Jakub Przybyl Phd Thesis to Be Submitted 28-8

Jakub Przybył

120909

Investigating the relationship between personality factors and the use of language learning strategies by Polish university students

Badanie związku między czynnikami osobowościowymi a użyciem strategii uczenia się języków obcych przez polskich studentów

Praca doktorska napisana na Wydziale Anglistyki Uniwersytetu im. Adama Mickiewicza w Poznaniu pod kierunkiem prof. zw. dr. hab. Mirosława Pawlaka

Poznań, 2019 SUMMARY

The present dissertation explores the relationships between the personality traits of Polish university students learning English as a foreign language and their use of language learning strategies. It consists of three theoretical chapters, a chapter dedicated to the methodology of the research project, and one presenting the results and implications of the main study. The chapters are accompanied by a brief introduction and a section offering some final conclusions. Chapter One presents multiple perspectives on the investigation of human personality, relating to the assumptions of psychodynamic theories, the behaviourist approach and the social learning outlook, the phenomenological viewpoint, the cognitive stance, and, finally, the trait approach. Since the last-mentioned approach constitutes the theoretical framework for the methodology employed in the main study described in the present thesis, particular attention is dedicated to the Five-Factor Model of personality and the description of the ‘big’ five personality traits (Costa and McCrae 1992). Chapter One closes with a review of studies into the role of personality-related variables in foreign language development and aims to define the status quo. Chapters Two and Three address language learning strategies (LLS) as a scientific notion. Chapter Two provides a historical background and chronicles the essential landmarks on the map of strategy investigations. This is followed by a review of definitions of the strategy construct and an attempt to define the thematic scope of referring to LLS in the 21st century. Afterwards, various categorisations, classifications, and taxonomies of strategies are discussed. Particular emphasis is laid on Oxford’s (1990, 2011, 2017) classification systems of LLS as they explain the theoretical foundations of the adaptation of the research instrument for the measurement of strategy application, that is the SILL ver. 7.0 (Oxford 1990). The subsequent sections focus on linking the LLS concept to that of self-regulated language learning by providing a theoretical structure for the latter one, analysing relevant studies, and pointing to conclusions which legitimise the co-existence of the two constructs. The aspects of self-regulation under consideration correspond to the recognised strategy dimensions, that is learners’ motivation and goal orientation, affect, cognitive resources, and the interpersonal sphere.

ii Chapter Three is dedicated to the relationships between LLS and other learner characteristics. A number of variables mediating strategy use are discussed, including the foreign language itself along with cultural variables along and instructional settings as well as learner-specific attributes, such as language learning experience, gender, age, beliefs, aptitude, and motivation. Also, the nature of the complex relationship between LLS and foreign language proficiency is considered. The final section of Chapter Two offers reflection on the role of the language task and its impact on learners’ strategic choices, and thus concentrates on the microcontext of strategy use. Chapter Four sets the methodological framework and describes the rationale for conducting the main study. First, it presents the design and briefly discusses the results of two pilot studies. Capitalising on these results, it introduces a number of implications for the main study including the choice of the research instruments. Second, it provides a thorough description of the participants of the Main Study and the sampling scheme. This involves discussing the representativeness of the sample and the legitimacy of generalising research findings over the entire populations of students of Adam Mickiewicz Uni- versity in Poznań and WSB University in Poznań. Third, research questions are formulated and finally, the analytical procedures employed in order to provide relevant answers are elaborated on following the order of the research questions. Chapter Five presents the results of the main study. It addresses the four research questions framed in Chapter Four, and thus reports and reviews findings concerning the relationships between LLS and university students’ personality traits, links both constructs to the level of linguistic proficiency, and analyses clusters of learners in terms of their strategic and personality features. The outcomes are then referred to the results of earlier studies investigating LLS and personality traits in foreign language development. Also, an attempt is made to draw conclusions resulting from the revealed links between students’ personality traits and their strategy repertoire. Moreover, Chapter Five also discusses the limitations of the main study. The concluding part of the present dissertation addresses potential pedagogical implications as suggests directions for future research.

iii STRESZCZENIE

Niniejsza dysertacja poświęcona jest badaniu związków między cechami osobowości stu- dentów polskich uczelni, uczestniczących w zajęciach lektoratowych z języka angielskiego, a stosowaniem przez nich strategii uczenia się. W jej skład wchodzą trzy rozdziały teoretyczne, rozdział poświęcony metodologii projektu badawczego oraz ostatni, prezen- tujący wyniki i implikacje badania głównego. Towarzyszy im krótkie wprowadzenie oraz podsumowanie w postaci końcowych wniosków. W rozdziale pierwszym omówiono różne koncepcje osobowości, odnosząc się do założeń teorii psychodynamicznych, podejścia behawioralnego i teorii społecznego uczenia się, koncepcji fenomenologicznej, stanowiska kognitywistów oraz podejścia cecho- wego. Temu ostatniemu poświęcono szczególnie dużo uwagi, ponieważ postulowane przezeń Pięcioczynnikowy Model Osobowości oraz inwentarz pomiaru osobowości NEO-FFI (Costa i McCrae 1992) wykorzystano do pomiaru osobowości w badaniu głów- nym. Końcowa część rozdziału zawiera przegląd badań na temat roli zmiennych osobo- wościowych w procesie rozwoju języka obcego oraz próbę ustalenia status quo. W rozdziałach drugim i trzecim przedstawiona zostaje charakterystyka konceptu strategii uczenia się języków obcych. Przedstawiony rys historyczny uwzględnia przeło- mowe odkrycia i koncepcje w badaniach nad strategiami. Kolejne sekcje obejmują prze- gląd definicji oraz próbę zdefiniowania współczesnego rozumienia zakresu tematycznego pojęcia "strategie uczenia się języków obcych". W dalszej części rozdziału przedstawione zostały kategoryzacje, klasyfikacje i taksonomie strategii ze szczególnym uwzględnie- niem tych autorstwa Oxford (1990, 2011, 2017), stanowiących teoretyczne podwaliny dla narzędzia badawczego wykorzystanego w niniejszej pracy do pomiaru częstotliwości ko- rzystania ze strategii, tj. SILL ver. 7.0 (Oxford 1990). Kolejne sekcje skupiają się na po- wiązaniu konceptu strategii z modelem samoregulacji w nauce języka obcego poprzez przybliżenie założeń dotyczących tego ostatniego, analizę poświęconych mu badań oraz uzasadnienie, dlaczego oba konstrukty można pojmować jako komplementarne, a nie - substytucyjne. Omawiane aspekty samoregulacji korespondują z powszechnie uznanymi wymiarami strategii, tj. motywacyjnym i zorientowanym na cel, afektywnym, dotyczą- cym zasobów poznawczych oraz sfery interpersonalnej. Rozdział trzeci poświęcony jest kontekstowi, w jakim ma miejsce korzystanie ze strategii uczenia się języków obcych. Opisane są interakcje strategii z szeregiem

iv zmiennych mediujących ich użycie, począwszy od samego języka obcego oraz jego sze- roko pojętej kultury i tendencji w dydaktyce na terenie danego kraju, a skończywszy na atrybutach nierozerwalnie związanych z samym uczącym się, tj. doświadczeniem w nauce języka obcego, płcią, wiekiem, przekonaniami dotyczącymi uczenia się języka obcego oraz motywacją do nauki. W następnej części rozdziału znajdują się rozważania na temat skomplikowanego związku strategii uczenia się z poziomem opanowania języka obcego. Rozdział zamyka sekcja poświęcona wpływowi specyficznego zadania języko- wego na wybór strategii uczenia się, a więc skoncentrowana na mikrokontekście funk- cjonowania strategii. Rozdział czwarty ma charakter metodologiczny i stanowi swoistego rodzaju opis wykorzystanych w pracy procedur badawczych wraz z uzasadnieniem ich zastosowania. Po przedstawieniu założeń i, pokrótce, wyników dwóch badań pilotażowych, omawia szereg implikacji płynących z nich dla badania głównego. Rozdział zawiera również szczegółowy opis uczestników badania oraz doboru próby, odnosząc się przy tym do kwestii reprezentatywności próby badawczej i zasadności dokonywania generalizacji wy- ników na populacje studentów, uczących się języka angielskiego na lektoratach na Uni- wersytecie im. Adama Mickiewicza w Poznaniu oraz na Wyższej Szkole Bankowej w Poznaniu. W dalszej części sformułowane zostają pytania badawcze oraz opisane proce- dury analityczne, mające na celu udzielenie na nie odpowiedzi, w kolejności odpowiada- jącej pytaniom badawczym. W rozdziale piątym przedstawiono wyniki badania głównego. W oparciu o dane empiryczne odpowiedziano na cztery postawione w poprzednim rozdziale pytania badawcze, dotyczące związku strategii uczenia się języków obcych z osobowością uczą- cego się, związku obu tych zmiennych z poziomem biegłości językowej oraz możliwością wyodrębnienia profili uczestników zajęć lektoratowych w oparciu o ich cechy osobowo- ści oraz wykorzystywany przez nich repertuar strategii. W dyskusji odniesiono uzyskane wyniki do zaprezentowanych wcześniej rezultatów badań nad strategiami i osobowością, opisanych w literaturze. W rozdziale zamieszczono ponadto refleksję dotyczącą ograni- czeń badania głównego. W końcowej części pracy podjęta została próba wyciągnięcia wniosków oraz przedstawienia implikacji pedagogicznych z punktu widzenia procesów uczenia się i nauczania języka obcego. Zawarto również sugestie dalszych badań w przedmiotowym ob- szarze.

v Poznań, dnia ......

OŚWIADCZENIE

Ja, niżej podpisany Jakub Przybył, student Wydziału Anglistyki Uniwersytetu im. Adama Mickiewicza w Poznaniu oświadczam, że przedkładaną pracę dyplomową pt.: „Badanie związku między czynnikami osobowościowymi a użyciem strategii uczenia się języków obcych przez polskich studentów” napisałem samodzielnie. Oznacza to, że przy pisaniu pracy, poza niezbędnymi konsultacjami, nie korzystałem z pomocy innych osób, a w szczególności nie zlecałem opracowania rozprawy lub jej części innym osobom, ani nie odpisywałem tej rozprawy lub jej części od innych osób. Oświadczam również, że eg- zemplarz pracy dyplomowej w formie wydruku komputerowego jest zgodny z egzempla- rzem pracy dyplomowej w formie elektronicznej. Jednocześnie przyjmuję do wiadomości, że przypisanie sobie, w pracy dyplomowej, autorstwa istotnego fragmentu lub innych elementów cudzego utworu lub ustalenia naukowego stanowi podstawę stwierdzenia nieważności postępowania w sprawie nada- nia tytułu naukowego.

[ ]* - wyrażam zgodę na udostępnianie mojej pracy w czytelni Archiwum UAM [ ]* - wyrażam zgodę na udostępnianie mojej pracy w zakresie koniecznym do ochrony mojego prawa do autorstwa lub praw osób trzecich

*Należy wpisać TAK w przypadku wyrażenia zgody na udostępnianie pracy w czytelni Archiwum UAM, NIE w przypadku braku zgody. Niewypełnienie pola oznacza brak zgody na udostępnianie pracy.

(czytelny podpis studenta)

vi Table of contents

TABLE OF CONTENTS ...... VII

LIST OF TABLES ...... XI

LIST OF FIGURES ...... XIII

INTRODUCTION ...... 1

CHAPTER 1 : PERSONALITY AS AN INDIVIDUAL VARIABLE ...... 9

1.1. INTRODUCTION ...... 9 1.2. THEORETICAL FRAMEWORKS FOR PERSONALITY INVESTIGATIONS ...... 10 1.2.1. Psychodynamic theories ...... 12 1.2.2. Behaviourist and social learning perspective ...... 16 1.2.3. Phenomenological approach ...... 18 1.2.4. Cognitive approach ...... 20 1.2.5. The trait approach ...... 22 1.3. THE FIVE FACTOR MODEL OF PERSONALITY ...... 24 1.3.1. Historical perspective ...... 25 1.3.2. Major theoretical assumptions of the Five Factor Model ...... 26 1.3.3. The ‘Big’ Five personality traits ...... 29 Extraversion ...... 30 Openness to Experience ...... 34 Agreeableness ...... 36 Conscientiousness ...... 37 Neuroticism ...... 40 1.4. PERSONALITY IN FOREIGN LANGUAGE DEVELOPMENT ...... 41 1.5. CONCLUSION ...... 57

CHAPTER 2 : THE CONCEPT OF LANGUAGE LEARNING STRATEGIES .. 60

2.1. INTRODUCTION ...... 60 2.2. HISTORICAL PERSPECTIVE ...... 61 2.2.1. Early landmarks in strategy investigations ...... 62 2.2.2. Good language learner studies ...... 64

vii 2.3. DEFINING LLS ...... 66 2.2.1. Early definitions of LLS ...... 67 2.2.2. The evolution of the concept in the 1990s and strategy features ...... 70 2.2.3. Mapping the scope of LLS in the 21st century ...... 72 2.4. CATEGORIES OF LLS ...... 77 2.4.1. Early LLS labels ...... 78 2.4.2. Rubin’s (1981) classification of LLS ...... 78 2.4.3. O’Malley and Chamot’s (1985) classification of LLS ...... 79 2.4.4. Oxford ‘s (1986, 1989/1990) early taxonomies of LLS ...... 81 2.4.5. Purpura’s (1998) classification of LLS ...... 86 2.4.6. Macaro’s (2001) tentative classification of strategy categories ...... 87 2.4.7. Schmidt and Watanabe’s (2001) classification of LLS ...... 88 2.4.8. Woodrow’s (2006b) classification of LLS ...... 89 2.4.9. Tragant and Victori’s (2012) typology of LLS ...... 91 2.4.10. Griffith’s (2013) taxonomy of LLS ...... 92 2.4.11. Cohen’s (1998, 2014) typologies of LLS ...... 92 2.4.12. Oxford’s (2011, 2017) classification of LLS ...... 94 2.5. LLS AND THE CONCEPT OF SELF-REGULATION ...... 98 2.5.1. Theoretical assumptions and models of self-regulated learning (SRL) ...... 99 2.5.2. Selected studies investigating SRL ...... 102 2.5.3. Self-regulated language learning (SRLL) ...... 106 2.5.4. Investigations of SRLL ...... 111 2.6. CONCLUSION ...... 118

CHAPTER 3 : VARIABLES MEDIATING LLS USE ...... 120

3.1. INTRODUCTION ...... 120 3.2. THE IMPACT OF THE LANGUAGE BEING LEARNT ON THE CHOICE OF LLS ...... 122

3.3. THE IMPACT OF THE DURATION OF LANGUAGE LEARNING EXPERIENCE ON THE

CHOICE OF LLS ...... 125 3.4. THE IMPACT OF CULTURAL VARIABLES ON THEIR CHOICE OF LLS ...... 128 3.5. THE IMPACT OF AGE ON LEARNERS’ CHOICE OF LLS ...... 130 3.6. THE IMPACT OF GENDER ON LEARNERS’ CHOICE OF LLS ...... 132 3.7. THE IMPACT OF LEARNERS’ BELIEFS ON THEIR USE OF LLS ...... 135

viii 3.8. THE RELATIONSHIPS BETWEEN LLS AND LEARNING STYLES ...... 136 3.9. THE RELATIONSHIPS BETWEEN LLS AND APTITUDE ...... 138 3.10. THE IMPACT OF MOTIVATIONAL VARIABLES ON THE CHOICE OF LLS ...... 139 3.11. LLS AND TARGET LANGUAGE ATTAINMENT ...... 142 3.12. LLS IN MICROCONTEXT: THE IMPACT OF THE LANGUAGE TASK ...... 151 3.13. CONCLUSION ...... 155

CHAPTER 4 : METHODOLOGY OF THE RESEARCH PROJECT ...... 157

4.1. INTRODUCTION ...... 157 4.2. PILOT STUDY ONE ...... 158 4.2.1. Participants and data collection ...... 159 4.2.2. Measures ...... 160 4.2.3. Analytical procedures ...... 164 4.2.4. Selected findings and implications for the main study ...... 165 4.3. PILOT STUDY TWO ...... 169 4.3.1. Participants and data collection ...... 170 4.3.2. Measures ...... 171 4.3.3. Analytical procedures ...... 173 4.3.4. Selected findings and implications for the main study ...... 174 4.4. THE DESIGN OF THE MAIN STUDY ...... 175 4.4.1. Research questions ...... 176 4.4.2. Participants ...... 177 4.4.3. Data collection and research instruments ...... 188 NEO-FFI ...... 189 SILL ver. 7.0 ...... 195 Semi-structured interviews ...... 199 4.4.4. Analytical procedures ...... 200 Testing data sets normality ...... 201 Descriptive statistics ...... 202 Identifying the link between strategy use and personality ...... 204 Investigating the link between strategy use, personality and attainment ...... 205 Correlations ...... 206

ix Regression model ...... 207 Clustering techniques ...... 208 Analysis of interviews ...... 210 4.5. CONCLUSION ...... 211

CHAPTER 5 : RESEARCH PROJECT ...... 213

5.1. INTRODUCTION ...... 213 5.2. STUDY RESULTS ...... 213 5.2.1. The relationship between LLS use and personality ...... 214 Students’ personality traits and strategy use ...... 214 Relationship between LLS use and personality ...... 221 Correlations between personality traits and specific LLS ...... 230 LLS use explained by stepwise regression of personality traits ...... 236 Insights from semi-structured interviews ...... 238 5.2.2. The relationships between language learners’ LLS use and attainment .... 239 5.2.3. Students’ personality traits and attainment ...... 242 5.2.4. Student profiles ...... 246 Results of two-step cluster analysis ...... 247 Results of k-means cluster analysis ...... 249 5.3. DISCUSSION ...... 252 5.3.1. Personality and LLS use ...... 253 Commentary on the investigated students’ strategic preferences ...... 253 The relationships between learners’ personality traits and LLS ...... 254 5.3.2. The relationships between learners’ use of LLS and language attainment 261 5.3.3. The relationships between learners’ personality and attainment ...... 263 5.3.4. Clusters of LLS users ...... 265 5.3.5. Limitations of the study ...... 266 5.4. CONCLUSION ...... 268

CONCLUSIONS, DIRECTIONS FOR FUTURE RESEARCH AND PEDAGOGICAL IMPLICATIONS ...... 272

REFERENCES ...... 283

APPENDICES ...... 339

x List of tables

Table 1. Major defence strategies recognised by psychoanalysts...... 14 Table 2. Stages of psychosocial development ...... 15 Table 3. The structure of personality as proposed by Rogers...... 19 Table 4. The origin of emotions in construct change ...... 21 Table 5. Historical outline of studies relevant to the development of the FFM ...... 26 Table 6. The ABCDs of Extraversion...... 33 Table 7. Linguistic manifestations of Openness to Experience...... 35 Table 8. Alternative arrangements of the facets of Conscientiousness...... 38 Table 9. Early investigations of the impact of personality on L2 development...... 42 Table 10. Investigations of psychological factors in language acquisition ...... 45 Table 11. Overview of strategy definitions by Griffiths and Oxford (2014)...... 75 Table 12. Examples of language learning strategies based on the SILL ver. 7.0 ...... 85 Table 18. Types of investigations of links between strategies and proficiency ...... 144 Table 14. Participants of Pilot Study 1...... 159 Table 15. Cronbach’s alpha values for strategy scales and personality dimensions. ... 162 Table 16. Additional questions introduced in the adaptation of SILL ver. 7.0...... 163 Table 17. Data normality for strategy scales and personality dimensions...... 164 Table 18. Descriptive statistics for Pilot Study One...... 165 Table 19. Participants of Pilot Study Two...... 170 Table 20. Spearman-Brown reliability statistics for TIPI in Pilot Study Two...... 172 Table 21. Cronbach’s alpha for strategy scales in Pilot Study Two...... 172 Table 22. K-S normality test results in Pilot Study Two...... 173 Table 23. Participants’ professional activity in the main study...... 184 Table 24. AMU students participating in interviews ...... 188 Table 25. WSB students participating in interviews ...... 188 Table 26. NEO-FFI scales...... 190 Table 27. Reliability of personality scales...... 194 Table 28. SILL ver. 7.0 scales...... 196 Table 29. Reliability of the adapted SILL ver. 7.0 scales...... 199 Table 30. LLS and personality scales tested for normality ...... 202

xi Table 31. Results of Levene’s test for normally-distributed strategy scales ...... 205 Table 32. Descriptive statistics for strategy scales...... 216 Table 33. Descriptive statistics for personality scales - raw scores...... 218 Table 34. Descriptive statistics for personality scales - women aged 18 and 19...... 219 Table 35. Descriptive statistics for personality scales - men aged 18 and 19...... 219 Table 36. Descriptive statistics for personality scales - women aged 20-29...... 220 Table 37. Descriptive statistics for personality scales - men aged 20-29...... 220 Table 38. Moderate correlations ...... 231 Table 39. Stepwise regression parameters ...... 237 Table 40. Use of cognitive and metacognitive strategies across levels of proficiency. 240 Table 41. Cluster distribution ...... 247 Table 42. Distribution of personality traits across clusters of learners ...... 248 Table 43. ANOVA for final clusters ...... 250 Table 44. Centroids for final clusters ...... 251

xii List of figures

Fig. 1. Strategy dimensions and categories in Oxford’s (2011, 2017) S2R model...... 95 Fig. 2. Participants’ age...... 183 Fig. 3. Participants’ level of competence in English...... 185 Fig. 4. Participants across departments...... 186 Fig. 5. Participants’ place of residence while receiving education...... 187 Fig. 6. Use of memory strategies across levels of Openness to Experience...... 222 Fig. 7. Use of memory strategies across levels of Conscientiousness...... 223 Fig. 8. Use of memory strategies across levels of Openness to Experience...... 224 Fig. 9. Compensation strategies use across levels of Openness to Experience...... 224 Fig. 10. Use of cognitive strategies across levels of Extraversion...... 225 Fig. 11. Use of cognitive strategies across levels of Openness to Experience...... 226 Fig. 12. Use of cognitive strategies across levels of Conscientiousness...... 226 Fig. 13. Use of cognitive strategies across levels of Openness to Experience...... 227 Fig. 14. Use of affective strategies across levels of Openness to Experience...... 227 Fig. 15. Use of social strategies across levels of Extraversion...... 228 Fig. 16. Use of social strategies across levels of Openness to Experience...... 228 Fig. 17. General use of LLS across levels of Extraversion...... 229 Fig. 18. General use of LLS across levels of Openness to Experience...... 229 Fig. 19. The distribution of residual parameters in the stepwise regression model...... 237 Fig. 20. General frequency of strategy use across proficiency levels...... 240 Fig. 21. Two-step cluster model summary...... 247

xiii Introduction

While it appears that applied linguists and foreign language educationalists have long been aware of the influence of learners’ individual characteristics on the process of foreign language development, the popular paradigm according to which the role of psychology should be limited to providing a certain theoretical background to a given methodology, as expressed by Belâev (1969: 11), has hindered in-depth investigations of the learner for several decades. Consequently, although early volumes dedicated to the study of foreign language learning were not entirely devoid of descriptions of learner-specific factors affecting the process of learning a foreign language, they tended to discuss the role of aptitude, intuition, or memory, and, to some degree, the contribution of attitude and motivation (cf. Belâev 1969; Jankowski 1973). The growing interest in the principles of person-centred education, deriving from the assumptions of humanistic psychology (Bugental 1964), gradually led to a paradigm shift, marked predominantly by the arrival of good language learner studies (Rubin 1975, Stern 1975). A number of studies carried out in the area of individual variation in foreign language development may be, at least in substance, classified as studies seeking to establish the qualities of good language learners. One example could be Biedroń’s (2011) study investigating near-nativeness as a function of cognitive and personality factors whose findings lend support to the claim that high proficiency in a foreign language is a matter of an optimal interplay between’ IDs rather than possessing a certain required, non-negotiable set of characteristics. It seems that the major factor which diminished the recognition for research into learners’ IDs for at least some time in the past was the conviction that their vastly endogenous character results in little scope for any modification or adjustment, and therefore allows for few, if any, practical implications for the field of foreign language pedagogy.

1 This conviction is, however, only true to a limited extent. According to Biedroń and Paw- lak (2016), the properties of learner IDs could be summarised under four major headings: heterogeneity, mutual dependence and proneness to environmental influences; pertinence to the stage of language development and specific skills; dynamics and changeability related to mutual interactions; existence in multiple combinations enabling learners to succeed in developing satisfactory communicative ability. Indeed, the complexity of IDs does constitute a certain methodological challenge for researchers. However, the actual premise that poses a motivational barrier for some researchers and discourages at least some ELT practitioners from following the findings in the field might be the alleged scarcity of pedagogical implications, as suggested by Mardsen and Kasprowicz (2017) or, perhaps, the impossibility to translate them directly into classroom practice. In the specific case of personality investigations, Biedroń and Pawlak (2016) proposed that practical implications could focus on enhancing the individualisation of the learning and teaching process, awareness raising, and, finally, diversify- ing classroom activities and tasks so as to adjust the language learning instruction and materials for the specific needs of particular learners. A vast body of research, including studies conducted by Lowie et al. (2017), Przybył (2017), or Jackson (2018), seems to have adopted that kind of approach. Also, given the abovementioned features of learners’ IDs and their commonly acknowledged interdependence (Ehrman and Oxford 1995; Dö- rnyei and Skehan 2003; Dörnyei 2005; Oxford 2011; Gregersen and MacIntyre 2014; Dörnyei and Ryan 2015; MacIntyre, Gregersen and Mercer 2016; Oxford 2017), it surely is legitimate to pursue investigations of the intricate relationships among language learner-specific variables. Among these, language learning strategies occupy a special place as their application has commonly been associated with foreign language proficiency and achievement, both in general and with reference to specific skills (Bialystok and Fröhlich 1978; Olivares‐Cuhat 2002; Altan 2003; Magogwe and Oliver 2007; Wong and Nunan 2011; Szyszka 2015), as well as associated with self-regulated language learning (Oxford 1999; Chularut and DeBacker 2004; Rose and Harbon 2013; Seker 2016). The co-dependence of learner variables in foreign language development has been subject to multiple investigations, including quantitative and qualitative studies of virtually all possible paired combinations of features. Moreover, a number of researchers have examined the relationships among several variables rather than just two of them. To quote just

2 one example, in Verhoeven and Vermeer’s (2002) study of the impact of personality traits on developing communicative competence investigations were conducted parallelly for two different languages. The present author’s interest in the correspondence between university students’ personality traits and their use of language learning strategies is far from random. Firstly, it originates from a firm conviction that through broadening knowledge of the participants of foreign language classrooms, instructors can individualise both the teaching and the learning process. Secondly, it pays tribute to over 40 years of research into language learning strategies. Although its usefulness has been questioned at times (cf. Dörnyei 2005; Tseng et al. 2006; Dörnyei and Ryan 2015), it remains invaluable from a language educator’s perspective as it broadens the scope of opportunities for language practice and provides the scaffolding for developing learner autonomy, as asserted by SLA researchers (Oxford 2003; Pawlak 2004, 2008a; Reinders 2010; Benson 2011; Pawlak 2011a). Thirdly, it rests on the premise that while personality traits are to a large degree genet- ically determined and cannot be altered, as asserted in recent studies (Widiger 2017), language learning strategies possess a number of features which make them a useful tool in foreign language development, including teachability, transferability, and facilitative- ness (Oxford 1990, 2011, 2017), all of which imply some flexibility and potential for adjustment in order to suit individual learners’ personal characteristics. Aiming to discuss the relationships between university students’ language learning strategies and personality traits, the present dissertation aspires to expand the scope of knowledge about language learners’ IDs by considering yet another combination of two learner variables and referring to a particular group of learners, that is, learners of General English at Polish universities. Whereas research into learning strategies has constituted a major area of studies dedicated to the language learner, the meaning of learners’ personality in their strategic choices has frequently been explored in studies investigating the links between strategies and personality types defined by the Myers-Briggs Types Indicator (MBTI) (Myers and Briggs 1976; Myers and McCaulley 1985) (Oxford and Nyikos 1989; Wakamoto 2000; Sharp 2004; Moreno and Bartlett 2016). Linking students’ use of language learning strategies to their personality traits investigated in large- scale studies could be perceived as an alternative to the parallel research based on MBTI as the trait approach offers a different framework for investigating human personality,

3 renowned for its psychometric excellence (Allport 1961; Costa and McCrae 1992; Widiger 2017). The choice of the above framework is explained in the in Chapter One, which also discusses alternative approaches to personality investigations. Not only is an attempt made to reflect on the development of consecutive personality theories, but the advantages of the trait approach from the perspective of the main study are emphasised. Also, the assumptions of the Five Factor Model (FFM) (Costa and McCrae 1985) are described as foundations underlying the construction of the NEO-FFI inventory (Costa and McCrae 1992) used for personality investigations in the present thesis. In particular, the ‘big’ five personality traits are accounted for in more detail. The following subsections aim to define and account for the underlying traits of each of the ‘big’ five dimensions of human personality. Relevant implications for language pedagogy are discussed either through elaborating on the psychological predictions of a given trait concerning its meaning for academic success or by referring to the traits ABCDs (affect, behaviours, cognition, desires) (Revelle and Wilt 2008). Since particular importance is granted to Openness to Experience in view of the results of the main study, its linguistic manifestations are also referred to. Finally, Chapter One also involves a section which outlines a selection of investigations dedicated to the role of personality in foreign language learning. At- tempts are thus made to present the status quo in studies into language learners’ personality characteristics as well as elucidate the role of the trait in the process of foreign language development. The complexity of the strategy construct is reflected the structure of the present dissertation as two theoretical chapters are dedicated to the notion. Attempts to discuss the history, understand the nature of LLS according to different theoretical approaches, and classify them are made in Chapter Two. A reflection on the origin of researchers’ interest in LLS and an explanation of the impact of good language learner studies (Rubin 1975; Stern 1975; Naiman et al. 1978) on strategy research are followed by an overview of definitions of LLS. Containing insights from over 40 years of research, it aims to present various perspectives in a dialectical manner. Contributions by different scholars are described with reference to both the very nature of LLS, which has been interpreted not only as behavioural (Politzer 1983) or cognitive (O’Malley and Chamot 1987), but also multidimensional, and thus incorporating affect (Oxford 1990) and motivation (Oxford 2011, 2017). Alternative systems of strategies are accounted for from multiple

4 perspectives. Some of them were developed on the foundation of empirical studies (Schmidt and Watanabe 2001) while others relied on theoretical models and insights from non-linguistic areas of research (O’Malley and Chamot 1985). The final part of Chapter Two seeks to reflect on the correspondence between LLS and self-regulation (SR). Generally believed to involve three main aspects, that is commitment to a standard, monitoring, and capacity for change, it can be analysed with respect to four categories, including motivational processes, goal-oriented behaviour, affective and cognitive processes, and interacting with others (Forgas et al. 2009). It could be argued that these four categories directly correspond to the four dimensions of Ox- ford’s (2017) model of strategic self-regulation (S2R), which assumes that language learners’ self-regulation encompasses cognitive, affective, motivational, and sociocultural-interactive domains. Section 2.4 thus aspires to capitalise on the similarities of the two constructs. At the same time, it argues to prove the benefits of an amalgamative approach towards the allegedly competing notions of LLS and self-regulated language learning (cf. Tseng et al. 2006). Recognising the importance of the context in which language learning takes place, Chapter Three discusses a number of variables which can be related to learners’ use of LLS. By adapting a variety of perspectives, it aims to account for learner-specific and situation-specific factors (Oxford and Nyikos 1989) as well as seeks to recognise the importance of individual and group variables as proposed by Takeuchi et al. (2007) and Pawlak (2011). First, attention is paid to language-specific variables, including the target language itself and the mother tongue of the learner, but also the instructional and cultural settings in which the learning process takes place. Insights are also provided concerning the role of learner-specific characteristics, such as age, gender, the length of the language learning experience, beliefs about language learning, language aptitude and motivational variables. Moreover, the links between LLS and learning styles and the meaning of language aptitude are also considered. An attempt is then made to account for the complex relationships between strategy use and attainment. Although no ultimate answer is provided in terms of causality of either of the two variables, a number of studies are described which confirm the essentiality of LLS for foreign language development. Finally, the concept of the language task is scrutinised and its impact on the strategy employment is discussed.

5 It is expected that the degree of scientific scrutiny applied in the study described in the present dissertation is sufficient to elaborate on the intricate relationships between the investigated learner variables, provide answers to all the research questions, and arrive at a number of pedagogical implications. Therefore, it is examined to what degree the employed research instruments, including Polish adaptations of NEO-FFI (Zawadzki et al. 2010) and SILL ver. 7.0 (Oxford 1990), fulfil the goodness criteria for psychometric tests (Hornowska 2007). Moreover, the numbers of participants in the main study from both Adam Mickiewicz University in Poznań and WSB University in Poznań were calculated so that the results of the study could be representative for entire populations of learners of English as a foreign language at both universities. Also, the main study was preceded with two pilot studies in order to optimise its rationale and examine the psychometric properties of the applied questionnaires. Firstly, this actually resulted in selecting NEO-FFI (Costa and McCrae 1992) as a tool for personality assessment, which proved to be more reliable than its considered alternative, TIPI (Gosling et al. 2003). Secondly, adapting the SILL ver. 7.0 (Oxford 1990) involved introducing some adjustments aiming to update the list of strategies and include those available to the investigated students, but not to their counterparts in the 1990s. Thirdly, the reliability of the tool was increased by expanding the problematic scale of affective strategies. This happened after addressing a panel of experts according to the procedure described by Brzeziński and Maruszewski (1978). Finally, in order to triangulate the data from obtained from the questionnaires, a series of semi-structured interviews was conducted with learners of English studying at both universities covered by the main study. Prior to providing answers to the research questions formulated in Chapter Four of the present thesis, the use of the above measures contributed to an extensive description of the investigated populations of university students in terms of their personality traits and strategic choices. The exploration of participants’ personality traits was conducted according to the criteria for personality assessment (Zawadzki et al. 2010), and thus also separately for specific age groups and genders. As regards the psychological interpretation of the calculated levels of personality traits, it included elements of both profile and functional aspects. Apart from being a prerequisite for analytical procedures, the description also allowed the present author to establish university students’ strategic preferences as well as identify strategies which they tend to dismiss. Some practical implications of these findings are addressed in the final section of the present thesis.

6 Four major areas are discussed in Chapter Five of the present thesis. To begin with, an attempt is made to elaborate on the relationships between the investigated university students’ ‘big’ five personality traits (Openness to Experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism), and the reported frequency of their use of language learning strategies (LLS) measured for six scales (memory, cognitive, compensation, metacognitive, affective, and social strategies) as well as their general frequency of strategy use (GULLS). They are explored in the analysis of differences in strategy use between data sets distinguished on the basis of the level of learners’ personality traits according to psychological interpretation (Zawadzki et al. 2010), correlations between the investigated scales, analysis of regression, and thematic analysis of interviews. Further investigations employ analogical procedures and aim to provide answers to Research Questions Two and Three concerning the relationships of the investigated IDs with students’ attainment1, operationalised as their diagnosed CEFR level. Analyses and are conducted parallelly for language learning strategies and personality. As far the quantitative analysis is concerned, each time a relationship between two variables is explored, its statistical significance is reported. At the same time, results obtained in statistical procedures are triangulated with findings from a series of semi-structured interviews with BA students of the two investigated universities. Afterwards, two types of clustering techniques, that is hierarchical and non-hierarchical clustering, are employed in order to answer Research Question Four concerning the possibility of creating language learner profiles on the basis of their personality traits and strategic preferences. In other words, an attempt is made to group the investigated language learners into relatively homogenous clusters of strategy users and describe them in terms of the levels of their personality traits. After describing the distinguished groups of learners in terms of their strategic and personal characteristics, an attempt is made to reflect on the type of relationship between the investigated variables incorporated as co-existing characteristics. Chapter Five fin- ishes with a brief consideration of the limitations of the main study and ideas which might contribute to improvements in future research. The conclusions of the present dissertation include a number of implications which result from the established correspondence between learners’ personality traits and

1 The use of the terms, language attainment and CEFR level is therefore interchangeable in the present thesis since attainment is treated as given (diagnosed by CEFR level) and constant at the time of conducting the study.

7 their use of language learning strategies. They include suggestions of possible methodological improvements referring to the use of specific instruments in the research project and certain suggestions for foreign language pedagogy. First of all, the potential benefits of both, cognitive and strategic training, are elaborated on in view of the findings granting a particular role among learner IDs to Openness to Experience and scientific evidence that the level of the trait can be stimulated (Jackson et al. 2012). Secondly, suggestions are made to advocate for the employment of certain specific procedures, such as guided discovery (Marton 1988) in classroom practice. Examples of possible classroom activities which can prove advantageous for students’ language development are also provided. Thirdly, some principles of fostering EFL teachers’ relationships with their students are discussed. Lastly, the present thesis is intended to encourage EFL teachers’ interest in language learners as human persons by providing evidence that personality and strategy investigations are, by all means, feasible and beneficial.

8 Chapter 1: Personality as an individual variable

1.1. Introduction

Personality explorations have constituted a valid area of research since the very early days of scientific and philosophical investigations. References to the views of ancient thinkers that can be found in contemporary handbooks of psychology include such notions as Plato’s division of the forces that drive human behaviour, i.e., reason, emotion, and ap- petite, and Aristotelian ideas of biologically-conditioned psyche, consisting of three composites: the nutritive faculty, the perceptual faculty, and the intellectual faculty (Ellis et al. 2009: 3). One of the most popular classification of personality types (Sanguine, Chol- eric, Melancholic, and Phlegmatic) dates back to 2AD, that is the time when Galen, influenced by Hippocrates’ theory of four humours, attempted to scrutinise the study of temperament (Loehlin and Martin 2018: 99). It could be argued that the interest in human personality was at least partly induced by the inherent belief that knowledge of an individual’s characteristics can make their behaviour more predictable and serve better mutual understanding among human beings. The view that personality determines or, at least, interacts with people’s actions, is not rare among psychologists; however, it is the wealth of existential domains which are likely to be influenced by personality, that is really impressive. According to Zawadzki (1970), personality is a modifier of development and a determinant of actions; it affects learning and people’s personal goals, at least partly accounts for people’s needs and desires, affects attitudes, and even influences the components of thinking. Language and personality are inseparable. The use of language is determined by the speaker’s personality, regardless whether it is a native language (Beukeboom et al. 2013), a second language (Dewaele and Furnham 2000), or a foreign language (Studenska 2011b). Contrariwise, any investigation of personality is bound to involve descriptions of characteristics, depends on the thoroughness of the descriptions, and is likely to tackle the problem of the multitude of personality adjectives. Indeed, even the wording of the definition of personality in layman’s terms is far from obvious. Interestingly, the variety of descriptions offered by online dictionaries of English, referring to “the type of person” (Cambridge Dictionary), “the combination of characteristics or qualities” (Oxford

9 Dictionary of English), “someone’s character” (Longman Dictionary of Contemporary English), or “the part of a person that makes them behave in a particular way in social situations” (MacMillan Dictionary), seems to reflect the lack of uniformity in how personality psychologists explain the construct. The plurality of views on personality which have shaped the contemporary perception of the phenomenon is considered in the present chapter. It starts with an attempt to map the area of investigating personality in the present dissertation. First, the distinction between the major approaches to human personality is drawn, second, a review of the most influential theories is offered, and, finally, the Five Factor Model (FFM) Costa and McCrae 1985) is discussed. The model is given particular attention since, for one thing, it underlies a vast body of research into personality in gen- era1, and for another thing, it underlies the construction of the research instruments described in the final chapters of the dissertation. Finally, an attempt is made to account for the status quo in personality studies, and to recapitulate on the most relevant findings so that they can serve as theoretical background for investigating the role of the construct in foreign language development.

1.2. Theoretical frameworks for personality investigations

Choosing a framework of investigation seems indispensable in personality studies; however, it needs to be stressed that no definite answer has been provided regarding the superiority of a particular approach or perspective over the years. It could be argued that the development of all psychology, including personality psychology, has been dialectic in nature, and thus it has involved the emergence of a number of thesis proposals, responded to a number of antitheses and finally integrated by syntheses of viewpoints (Sternberg and Sternberg 2012: 4-5). For example, researchers scrutinising defence mechanisms have attempted to provide scientific evidence for their existence or non-existence, but even more than 100 years after Freud’s revelations, there seems to be no consensus in this respect (Cramer 1991; Freud and Breuer 2004). It can be concluded that such factors as scientific progress resulting from the rise of new technologies, increasing availability of information, and the speed of its spread all work in favour of accessibility of knowledge and the opportunity to exchange views, but at the same time, no ultimate solutions have

10 been offered for the examination of key concepts, such as cognition, behaviour, wellness, or personality. Consensus is present among scholars with regard to the very existence of personality except for extreme situationists, who only account for the differences in human behaviour and attribute them to the dissimilarity in cultural or situational backgrounds (Ellis et al. 2009: 3). According to Ashton (2017: XV-XVI), personality studies usually adopt one of two perspectives, i.e., the idiographic perspective or the nomothetic perspective. The former is more common for in-depth investigations which are based on the general assumption of any individual’s uniqueness, such as, in the first place, case studies. The latter focuses more on personality characteristics that are universal, investigates a greater number of individuals, underlies the quest to describe tendencies rather than unique properties. The quantitative studies described in Chapters Four and Five of the present dissertation represent the nomothetic approach; however, the qualitative study conducted in order to triangulate the quantitative findings has allowed for more insights into the idiosyncratic characteristics of the investigated language learners and could thus be described as idiographic in nature. Apart from falling into either of the above categories, personality studies typically adopt a given approach, and embrace its methodology, which particularly relates to personality assessment. In other words, theoretical speculations tend to be cast into testable forms and studied empirically (Mischel and Mischel 1973: 3). Nowadays, the major approaches to the investigation of personality which are discussed in handbooks of psychology or more specialised literature on personality include the psychodynamic approach, the behaviourist approach, the phenomenological (sometimes referred to as humanistic) approach, the social-cognitive approach and the trait approach. Some theorists also consider psychobiological findings to constitute a separate approach (Chamorro-Premuzik 2011). However, although treating the human person as a biological entity is by all means valid from the scientific angle, a number of psychologists maintain that the scope to which biological factors have been considered in personality models is relatively limited as such (Magnusson and Törestad 1993: 430). Even though examples of scientific research into the biological foundations of different aspects the human psyche can be provided (Wundt 1883; Strelau 1983; Eysenck 1990), they conclude that the biological perspective is only used for selected research areas. It is necessary to comment here on the development of positive psychology and its impact on the area of second/foreign language development,

11 which has been explored in recent years (MacIntyre and Mercer 2014; MacIntyre et al. 2016). Although it is methodologically correct to distinguish between humanistic psychology and positive psychology (Waterman 2013, 2014) the latter has not developed a separate theory of personality, and hence, by considering it as a concept related to within- person traits, beliefs, attitudes, goals, and predictors of human behaviour (Hefferon and Boniwell 2011: 13), it actually accepts the foundations of the trait approach. It could be argued that the personality theories which have been developed since the beginning of the previous century are not simply equivalents of the above approaches, but the at same time, no approach could exist without a theoretical framework. This view is shared by Ellis et al.:

Underlying all clinical approaches is a theory of human personality. This theory can be tacit or explicit, but it is requisite. One cannot diagnose an infirmity without a model of health. Nor can one treat it without a conceptualization of its etiology. A personality theory provides the foundation for understanding how both healthy and dysfunctional personalities develop, and it makes the distinction between the two. In addition, such a theory provides a framework for the research and experimental testing of both itself and the clinical approaches based on it (Ellis et al. 2009: XVI).

A brief description of the major personality theories constitutes the following part of the present chapter. The consecutive subsections thus discuss the main assumptions of psychodynamic, behaviourist, humanistic/phenomenological, cognitive, and trait theories of personality.

1.2.1. Psychodynamic theories

According to Engler (1979: 28), Freud’s concept of personality derives from psychoanalysis. Indeed, psychodynamic assessment of personality has commonly employed free association and analysis of dream reports or aimed to free the float of attention, “which turns inward to the observer’s ruminations while remaining turned outward to the field of observation and waits to be impressed” (Erikson 1958: 65-67). At the same time, it should be pointed out that while Freud emphasised the importance of psychosexual concepts of personality development, the vast number of his followers shifted the focus towards more social concepts (Mischel and Mischel 1973: 9).

12 Freud (1923) claimed that human behaviour was influenced by emotions, largely formed on the basis of memories, some of which would later be expressed as neurotic symptoms; however, at the same time they would also be affected by resistance and repression, both of which could prevent conscious access to memories (Engler 1979: 38). Freud’s structural model of personality assumed the existence of id, ego and superego. The first was a representation of instinctual energy and operated on an unconscious level, the second one could be understood as “thinking, remembering, planning and executing the necessary actions to obtain the need-gratifying object” (and thus following the reality principle or, in other words, performing secondary processes) and the final one was responsible for internalising the values, norms or beliefs acquired in socialisation (Barone et al. 1998: 30-31). These constituents of human personality were, in the classically psychoanalytic view, governed by different principles, i.e., the pleasure principle operating for id and the reality principle constituting the drive for ego. Wish fulfilment was subject to mediation between the id and the outer world, which did not allow direct gratification of some behaviours, especially those that were not socially accepted (Mischel et al. 2008: 161-162). The father of psychoanalysis identified three distinct types of anxiety, i.e., objective anxiety (fear caused by external threats), neurotic anxiety (caused by the conflict between the id and the ego) and moral anxiety (resulting from the conflicts between the ego and the superego). He also suggested that ego developed defence mechanisms against each of the above types of anxiety, including repression, denial, displacement, rationalisation and sublimation, which were all supposed to serve two functions, i.e., protect the ego, as well minimise anxiety (Larsen et al. 2013: 234-239). Table 1 contains a description of defence mechanisms and connects them with particular threats against which they are used. In 1936 the list of defence mechanisms was extended by Anna Freud, who introduced the concepts of identification with the aggressor, and altruism (Freud 1992). The former one can be described as finding active ways in becoming the aggressor, while the latter may actually be regarded as overlapping with projective identification brought forward by Klein (Edgcumbe 2000: 12).

13 Table 1. Major defence strategies recognised by psychoanalysts (adapted from Barone et al. 1998: 32). Defence strategy Description of operation Denial blocking of threatening external stimuli from entering awareness Projection attribution of ego-dystonic material onto others Undoing attempts to dismiss unacceptable acts Reaction formation management of anxiety-provoking impulses by expressing opposite ones Rationalisation employing apparently logical responses to explain anxious behaviour Displacement substitution of a less threatening behaviour for a more anxiety-generating one Sublimation transformation of unacceptable impulses into socially acceptable activities Introjection internalisation of external values, beliefs or attitudes Identification wish to become similar to another, sometimes feared, person/ object Regression primitivisation of behaviour to an earlier stage of development Intellectualisation avoidance of anxiety-arousing situations by dismissing the emotional factor Repression maintenance of anxiety-provoking thoughts within the unconscious

Adler and Jung are commonly referred to as Sigmund Freud’s (Ellis et al. 2009). Structured as an autonomous system, the construct of personality was expanded by Jung (1921) in a number of ways, although it was probably the new interpretation of the unconscious, which suggested that personality was actually composed of the collective unconscious and the personal unconscious, that could be seen as paradigm shift. Castilho (2014: 15) interpreted the personal layer of the unconscious as the more superficial one, whereas she regarded the collective layer as universal or innate. Drapela (1995: 25-28) briefly summarised the Jungian model of human personality as a composite of four elements, including the ego, the self, and the twofold unconscious. Overall, eight typological constituents of human personality can be distinguished in Jung’s vision of the psyche on the basis of grouping individuals according to their attitudes, including introversion or extraversion, and, for each of the two attitudes, four modes of orientation, i.e., thinking, sensation, intuition and feeling (Sharp 1998: 10). Remembered as the founder of the school of individual psychology, Adler recognised the role of both nature and nurture in human development, but at the same made a claim that it was impossible to understand the role of either of these two factors without realising “what interpretation of them an individual has made and to what use she or he puts these aspects of life” (Bitter 2011: 422). Consequently, personality according to the Adlerian stance may be regarded as “the individual’s creative exercise or use of self with others in the world” (Bitter 2011: 422). The links between the individual and others are also notably present in Erikson’s concept of psychosocial development and his theory of the adult. In fact, Hoare (2002: 4) claimed that Erikson was the first illustrator of the “social world” in an individual’s “psychological apparatus”. On the other hand, Barone (1998: 40) summarised Erikson’s

14 stance as that of an ego psychologist, but also close to object relations theorists, one of the similarities being the development theory. Erikson views personality as dynamic and dependent on the interaction with a number of individuals and institutions, as he believed that using the strategies of ritualisation and ritualism, ego progresses through a sequence of eight stages of psychosocial development, each one resulting in crisis resolution and a new virtue as an added value (Erikson 1968: 92). Erikson’s stages of psychosocial development are listed in Table 2, which also includes information on relevant processes, names the stages and accounts for the resolved conflicts.

Table 2. Stages of psychosocial development (adapted from Erikson 1968: 92). Psychosocial crisis Age Resolution and gained virtue Trust vs. Mistrust acquisition of basic trust, the virtue of hope, and the nu- 1 minous idea of mother’s presence vs. idolism – unreal, perfect bond between mother and child Autonomy vs. Shame & Doubt 2 sense of autonomy, and the virtue of will vs. legalism Initiative vs. Guilt developing the sense of purpose and achieving authentic- 3-5 ity vs. impersonation Industry vs. Inferiority ego’s development, and the virtue of competence, ac- 6-puberty companied by formality, vs. formalism Identity vs. Role Confusion adolescence fidelity and ideology vs. totalism Intimacy vs. Isolation early adulthood love and affiliation vs. eliticism Generativity vs. Stagnation care and transmission of societal values to offspring vs. middle age authoritism Ego Integrity vs. Despair old age wisdom and integration vs. salietism

The psychodynamic view of personality is neither uniform, nor dismissed, especially in modern psychoanalysis. It has evolved over years and been pursued in modern approaches to the human self, such as the object relations theory, which attempts to account for people’s existence in two dimensions, the external and the internal one, and explore the relationships between them (Greenberg and Mitchell 1983: 12). Although psychoanalysts remain heavily criticised for providing insufficient scientific support of their claims, their theories have exerted influence on modern psychology, including personality psychology, and the popularity of the psychoanalytic approach is reflected in the popularity of the psychodynamic therapeutic treatment.

15 1.2.2. Behaviourist and social learning perspective

It is impossible to omit behavioural insights into the study of personality in the present thesis because of its profound interest in learning (Thorndike 1911, 1931), and its impact on the English teaching methodology consisting in, for instance, its fundamental importance for the audiolingual approach (Richards 2002: 20). As the name suggests, behaviourists showed primary interest in demonstrable behaviour which they approached as a series of observable events. Social behaviour theories of personality bring a shift in focus from the individual’s intrapsychic processes to environmental impact. The belief in the power of the stimulus gave rise to the development of the two core concepts of behaviourism, i.e., classical and operant conditioning (Romero 1995). At the same, a certain degree of variation exists between the extent of social determinism in the views of social behaviour theorists, which is reflected in the scope of shaping one’s own personality and, more broadly speaking, control over life that is left to the individual. While some behaviourists, such as Skinner, adopted extreme positions and their research derives much from Pavlov’s experiments on animals, others, such as Bandura (1977), suggested that learning, and particularly – social learning – takes place in everyday life and largely through observation. According to Skinner (1953), human behaviour was determined by purposes, incentives and goals, all of which are involved in the concept of operant conditioning. Epstein (1991: 362) provides an account of two key concepts brought forward by Skinner, operant behaviour, which was originally labelled as emitted, not elicited by a stimulus and, therefore, assumed to be in a way spontaneous, and determinism, explained by genetic endowments and environmental histories. Analysis of behaviour finally urged Skin- ner to put forward his own definition of personality, which he understood as:

“a locus, a point at which many genetic and environmental conditions come together in a joint effect. As such, he [an individual] remains unquestionably unique. No one else (unless he has an identical twin) has his genetic endowment, and without exception no one else has his personal history. Hence no one else will behave in precisely the same way. We refer to the fact that there is no one like him as a person when we speak of his identity” (Skinner 1974: 172).

Other behaviourists, such as Staats (1971, 1996), abandoned the extreme position in order to include the biological and evolutionary perspective as well as allow for a more systematic view of individual variation. Consequently, some of them contributed to the

16 development of the personality theory known as PB’s2 basic theory of personality whose main assumptions according to Staats (1996: 200-253) could be summarised as follows: a child has no inborn behavioural repertories, only sensory mechanisms and unorgan- ised response mechanisms; a child can develop sensory-motor, language-cognitive, and emotional-motivational repertoires through learning, which results in particular behaviour and characteristics; personal characteristics are provided by an individual’s basic behavioural repertoires (BBRs), unique for everyone given the uniqueness of everyone’s environment (and thus experience), and the overall number of BBRs; BBR is both a dependent and independent variable; mechanisms of behaviour are biologically-determined while learning provides content; behaviour can be treated as a demonstration of some BBRs, which are considered to be general, not situation-specific; abnormal behaviour can be approached as a demonstration of verbal or motoric repertoire rather than the result of unresolved past conflicts. Not only did the interest in learning persist among social learning theorists such as Bandura, but it was also incorporated into the construct of a person. For Bandura, “(c)onceptions of human behaviour in terms of unidirectional personal determinism are just unsatisfying as those espousing unidirectional environmental determinism” (Bandura 1986: 22-23). Instead, in his publication Social Foundations of Thought and Action he attempted to picture the human person as subject to the interplay between environmental factors, cognitive factors and other personal factors (Bandura 1986: 18). Social behaviourists also believed an individual to be possibly subject to modelling (a term sometimes interchangeably used with observational learning) if they fulfil four criteria. These would then include paying attention to relevant events, managing to retain the relevant material through a representational system, converting representation into appropriate actions of the primarily modelled behaviour, and finally, the appearance of sufficient incentive so that the modelled action can actually be performed (Grusec 1992: 782). It is necessary to recognise the findings of behaviourists and their followers regarding personality, not merely because of the length of the period of their dominance in

2 PB is the abbreviation which is commonly used with reference to psychological behaviourism.

17 psychological research, but also because of the impact that they exerted on the study of the person. To draw a parallel, linguists who neglect the language output in order to sub- scribe to the superiority of competence over performance, may not only fail to notice at least some processes regarding language change, but also fail to analyse it as a system accurately enough. Much in the same way, language educators who dismiss demonstrable behaviour as a source of knowledge about a person and neglect their environment or, broadly speaking, experience as the source of influence and individual variation, face the risk of not fulfilling their basic function to assist their learners in the process of language learning.

1.2.3. Phenomenological approach

Humanistic psychology developed as a response to both Freudian views of the person and the subconscious drives that the person would be governed by and behaviourists’ view of the person being utterly dependent on environmental conditioning. Instead, humanistic psychologists were willing to account for the possibly holistic picture of the human person, taking into consideration such concepts as cognition, emotion, feeling, will, morality, ethics, aesthetics, as well as intrapersonal, interpersonal and transpersonal relationships (Rennie 2007: 1). One of the two distinct representatives of the humanistic approach, Abraham Maslow (1999), saw humans as active agents equipped with free will. Known for the creation of the concept of the pyramid of needs (including physiological, safety, belongingness and love, esteem, and cognitive needs), Maslow (1999) believed that individuals’ actions could eventually compensate for the biological inheritance that they had been endowed with, and the unfavourable constitutional factors originating in the environment. He did not, however, provide a systematic description of personality that would support his theory of self-actualisation, which was later developed by contemporary psychologists as self-determination theory. Similarly, instead of developing a personality questionnaire, Maslow’s followers constructed the Personal Orientation Inventory, which measured the degree of people’s self-actualisation. The development of phenomenology as an alternative direction in psychology was closely linked not merely to an alternative outlook on such notions as the observer’s frame, or reference, but also the novice approach to the Freudian ego and the self. Smith

18 (1950: 519) distinguished between the phenomenal construct, the self, and the non-phenomenal, subjective construct of the ego, pointing that “(t)he distinction is that between a dynamic configuration of on-going processes, inferred from many facts of biography and behavior, and a phenomenal entity resulting from these processes and affecting them.” The traces of this such philosophy can also be found in Rogers’s works (1959, 1981, 2011), who, like Maslow, dedicated much of his studies to actualisation, and shared the humanist conviction that individuals can choose the best directions, even in their own therapy, thus depreciating the role of the almighty therapist introduced by psychoanalysts. According to his person-centred theory, individuals have an innate, gradually-developing need for positive regard (Rogers 1959). His view of the human person can be best summarised with his own words, i.e., “The good life is a process, not a state of being. It is a direction, not a destination” (Rogers, 1969: 186). The assumptions Rogers’s structural model of personality were discussed by Ewen (2003: 199). Its key constituents are shown in Table 3.

Table 3. The structure of personality as proposed by Rogers (adapted from Ewen 2003: 199-204). Key concept Insights experience everything available to an individual awareness at a given moment, i.e., thoughts, emotions, perceptions, and needs; only a small proportion of experience is conscious the organismic valuing process humans’ innate ability to value positively whatever is, in their view, actualising, whereas value negatively anything that is not self-concept conscious representation of experience self-actualisation tendency to satisfy the demands of self-concept conflicts between actualising and self-actualising tendencies parents’ unconditional positive regard vs. conditional positive regard, i.e., positive self-regard becomes dependent on satisfying introjected (most often imposed by parents) conditions of worth, which replace the organismic valuing process as an inner guide to behaviour (incongruence of actualising and self-actualising tendencies) defence concerns experiences which result in incongruence and results in change of behaviour or attitude personality development no distinction of specific personality development stages is drawn; the final stage in personality development is characterised by absence of conditions of worth (not satisfying the introjected standards of other people); manifests e.g. by the individual’s ability to form successful interpersonal relationships

19 To summarise the phenomenological perspective in Mischel’s words, it can be concluded that human behaviour is best explained by means of what a person can do with it (Mischel and Mischel 1973: 13). Although the phenomenological (and, broadly speaking, humanistic) approach imposes an inherent difficulty on researchers of human personality, as it requires them to accept that human behaviour is not predictable or control- lable, it remains the approach which is not only frequently adopted by psychotherapists. It is also increasing in popularity nowadays because of its association with positive psychology (Snyder and Lopez 2002: 7).

1.2.4. Cognitive approach

The cognitive view of the human person is deeply rooted in a philosophical framework. Rychlak (1968) created the construct of the image of man, whom he equipped with certain specific features, including self-concepts, phenomenal fields, life lines, and personal constructs. Warren et al. (1987: 107) linked environmental experience with an individual’s search for meaning and pointed to the role of active reflection, which may result in volitional action. Similar views had been expressed by Kelly (1963) who had developed the personal construct theory. It differs considerably from other personality theories since it refutes the existence of identical constructs for individuals. Also, it supports the claim that individuals resemble scientists in the way they handle construct change. Other major assumptions incorporated in Kelly’s (1991) paradigm involve: psychological channelization of an individual’s personal processes which serve event anticipation; construction corollary – anticipating events by means of constructing their replications;

individuality corollary – individual variation in terms of event construction; organisation corollary –individuals’ evolution is convenience-driven, but convenience concerns event anticipation; dichotomy corollary – the composition of an individual’s construction system of a finite number of dichotomous constructs; choice corollary – an individual’s choice of the one of the dichotomous alternatives which ensures a better systematic elaboration;

20 range corollary – equivalence of a construct for the anticipation of a limited number of events; experience corollary – the variation in an individual’s construction system caused by construing replications of events; modulation corollary – the limitation in the variety of an individual’s construction system caused by the permeability of the ranges of convenience; fragmentation corollary – the inferential incompatibility of construction systems employed by an individual; commonality corollary – the similarity in experience construction between individuals; sociality corollary – the scope of co-construing the construction processes among individuals belonging to the same society. It is not merely the existence of a vast number of personal constructs that makes the theory truly complex. Another factor which adds up to its intricacy is the dynamic nature of constructs, which impedes their validation. Moreover, Kelly (1963) emphasised the necessity to account for the role of emotion in construct change and asserted that when confronted with all their emotions, an individual needs to adapt to the development of these constructs. A procedure which consists of five steps is repeated every time a construct changes, and involves the phases of anticipation and hypothesis formation, invest- ment in the event, encounter with the event, confirmation or disconfirmation of hypothesis, and constructive revision. Table 4 includes a brief description of the emotional factor in construct change.

Table 4. The origin of emotions in construct change (based on Kelly 1963: 114). Emotion Cause Anxiety recognition of the occurrence of events outside the range of convenience and the construct system Hostility failure of an individual’s prediction of their construct system accompanied by attempts to enforce its recognition by others Threat and fear external changes of the core construct system Guilt, shame, embarrassment perception of the self in the core constructs Confidence, pride awareness of the validation of an individual’s core constructs Humour realising that more interpretations of the same construct are possible Love awareness of the concordance between our own and somebody else’s construct system

21 Schultz and Schultz (2005: 356) stressed that it was the individual’s interpretation of the events they experienced rather than the events themselves that mattered for their development. They also added that the theory had a potential for therapy on various grounds. More specifically, it could provide empirical measures of individuals’ psychological problems, shed new light on psychopathology by providing the scientist-experi- menter’s perspective, and, finally, possibly facilitate devising new ways of psychotherapy by equipping the client with a new, powerful tool – the tool of anticipation. Finally, the need for distance to any psychological theory of personality was advocated along with and the need for humbleness in trying to explore personality in a scientific manner.

1.2.5. The trait approach

One of the distinguishing features of the trait approach is its diligence in personality assessment and commitment to methodological excellence in measuring human personality traits. At the same time, the simplicity of reasoning of the trait approach is clearly its major asset confirmed by the simplicity and clarity of personality definitions. For instance, Guilford (1959: 5) claimed that “(a)n individual’s personality is, (...) his unique pattern of traits”. According to Chamorro-Premuzik (2011: 61), the main assumptions of the trait approach include:

a) consistency of behavioural, emotional and thought patterns; b) belief in the internal nature of traits (though sometimes subject to debate); c) limited number of personality traits (or dimensions) to be presented in a personality model; d) quest for psychometric robustness of the created instruments. One personality psychologist who stands out because of his laboriousness and some vital differences in his methodological assumptions is Gordon Allport, also known as the founder of the trait approach. Schultz (1981: 179-180) listed 11 techniques of inquiry that were employed by Allport, including constitutional and physiological diagnosis, socio-cultural setting, membership and role, personal documents and case studies, self-appraisal, conduct sampling, ratings, tests and scales, projection, depth analysis, and synoptic procedures. At the same time, one should not only associate Allport with his belief in the deterministic role of traits, whose meaning he emphasised as the key to

22 understanding different reactions to the same stimulus), but also the scholar who linked the study of personality with the interest in individual differences. Mischel et al. (2008: 49) credited Allport with actually founding the field of personality psychology and expanding the scope of investigating the human person as opposed to analysing its specific segments such as learning or memory without accounting for individual variation. In other words, Allport’s traits are dispositional and can be used to predict individual’s behaviour and performance. After decades of research, Allport (1961: 28) finally formulated the definition of personality stating that the concept should be understood as “the dynamic organization within the individual of those psychophysical systems that determine his characteristic and behavior and thought”. He considered traits to be the constituents of human personality, which gave rise to a popular view of human personality, i.e., a collection of traits. In Allport’s view (1961), traits have a number of distinct characteristics. They are a) real, not merely nominal; b) general, in particular more generalised than a habit; c) dynamic, or, at least, determinative in behaviour; d) established empirically; e) relatively independent from each other; f) not synonymous with values, virtues or any form of moral or social judgement; g) analysable from either the individual’s or the society’s perspective; h) not always consistent with an individual’s behaviour. Strelau (2008: 527) elaborated further on Allport’s division of traits into three categories, depending on their intensity in an individual, that is cardinal, or dominating (typically one per person), central (only a few of which, according to Allport, can exist in a person) and secondary ones, which are displayed relatively seldom. The researcher asserted that while comparisons between individuals are possible, they can pose serious problems due to the idiosyncratic (or, to use Allport’s language, idiographic) nature of individuals. Scheier and Carver (1992) stress that some traits which occupy the central position in one person may be secondary in another and have no presence in yet another one. Gross (2009: 732-733) takes a more extreme position and suggests that some traits are only characteristic of single individuals and even if the same trait label is used with reference to two individuals, its semantic interpretation can vary. The possibility of making personality-related comparisons among individuals remains one of the key areas of

23 disagreements between the advocates of the idiographic and approach and those who support the nomothetic stand. Some psychologists, such as Zuroff (1986) signalled that Gor- don Allport should not be considered a trait theorist because of his distinction between idiographic and nomothetic traits, as well as his commitment to approaching the human person with an emphasis on the recognition of their purposes or perceptions and respect for their individuality. Allport’s interest in individuals’ purposes is reflected in his assumption that the study of human personality cannot be separated from that of motivation. Commenting on an alternative hierarchical organisation of personality traits, Gray (1999: 564-565) linked specific behaviours with traits and views them as the ground for drawing more general conclusions about individuals. Basing on Eysenck’s (1982) model, he saw specific behaviours as demonstrations of surface traits, which could then be organised into central traits. For example, a person who becomes easily involved in debates in which they defend unpopular stands, but also frequently expresses their opposing views in letters to the editors of the newspapers that they he or she reads, is likely to be argu- mentative (surface trait level). If the person also displays pugnacious and competitive behaviours, their central trait is likely to be aggressiveness. Consequently, a trait theory of personality attempts to provide a thorough description of personality composed of as few central (possibly non-redundant) traits as possible. Huffman (2008: 462-463) listed three major criticisms of trait theories, which are the following: lack of explanation of the source of human personality traits and individual variation, insufficient specificity in explaining developmental tendencies of personality, and neglect of situational effects in personality assessment or attempts to predict people’s behaviour. At the same time, trait theorists attempt to minimise these drawbacks and new theories of personality emerge in a dialectic manner.

1.3. The Five Factor Model of personality

It is the trait approach that underlies the methodology employed in personality investigations described in the two final chapters of the present thesis. Consequently, it is the model of personality which was developed by trait theorists, which underlies the construction of the personality questionnaire used in the quantitative part of the research project and referred to in the qualitative part of it. In spite of being the subject to a great deal of debate,

24 the Five Factor Model (FFM) of personality (Costa and McCrae 1985) can be referred to as a personality model of unique strength and predominance in psychology. Widiger (2017) suggests that the FFM largely owes to its integrative character to the compatibility of its framework with the English lexicon used to describe traits, its adaptability for various personality theories, such as those of neurobiological, cognitive, or even psychodynamic origin), and a vast amount of empirical research which has been conducted in order to test its validity. The following subsections of the present chapter are dedicated to the presentation of its historical development of the model, describe its major theoretical assumptions, and attempt to thoroughly describe each of the ‘big’ five personality traits in terms of their heredity, development across human life, and their underlying characteristics.

1.3.1. Historical perspective

The development of the Five Factor Model of personality (Costa and McCrae 1985) is traditionally linked with the rise of the trait approach in personality psychology the 1930s. Wiggins and Trapnell (1997: 738) provide a brief outline of the development and use of a five-factor personality model, according to which the existence of five factors was first cited by Thurstone (1934), followed by a reference made by Allport and Odbert (1936). A number of different five-factor models have later emerged over the decades in the works of such psychologists as Cattell (1943), Cronbach (1956) Guilford (1959), Eysenck (1960), Tupes and Christal (1961), Goldberg (1990), and, finally, Costa and McCrae (1985, 1992, 1995). Table 5 provides a selection of studies which have been relevant to the development of a five-factor model of personality. Research into the validity and cultural universality of the Big Five model, as well as the stability of personality traits, has continued through the 1990s and in the 21st century, and selected findings are related to in later sections of the present paper.

25 Table 5. Historical outline of studies relevant to the development of the Five Factor Model. Study Findings Thurstone (1934) A study requesting participants to name 60 commonly used trait adjectives in describing people resulted in identifying five factors that account for the intercor- relations between the named traits. Allport and Odbert (1936) Examination of trait-descriptive terms resulted in the classification of lexical items under the categories of personal traits, temporary states, social evaluations, and metaphorical terms. Cattell (1943) 4,500 terms were arranged in 171 groups based on semantic similarity, from which 35 clusters were subsequently obtained to account for a “standard reduced personality sphere”, which was initially declared to have 12 underlying factors. Self-ratings, teammate ratings and staff-assessment ratings were used to provide data for 22 of the original 35 rating scales. Fiske (1949) The first study of consistency of primary factors confirmed a high degree of consistency of factors among different raters. Cronbach & Meehl (1955) Construct validity became part of the philosophy of science. Consequently, a stricter approach to test construction gave rise to strivings to excel personality tests contrary to the behaviourist perspective on the human person. Loevinger (1957) The monograph Objective tests as instruments of psychological theory became a milestone in the history of psychometrics and test construction. References to introversion and neuroticism were made, but test responses were constructed as instances of behaviour and dichotomous scales replaced Likert scales for fear of self-reports. Tupes and Christal (1958) The investigation of 20 bipolar rating scales resulted in the emergence of a five- factor model built of Extraversion, Agreeableness, Conscientiousness, Emotional Stability, and Culture, four of which appear in the Big Five model. Tupes and Christal (1961) A study of universality of the above five-factor model demonstrated a high stability of the five-factor solution across different samples and conditions. Norman (1963) The rationale and procedures for developing a taxonomy of personality traits was scrutinised. Attention was also paid to such key issues as the generalisability and using peer ratings in order to improve Cattell (1973) Criticisms of the five-factor model centred around underscoring correlation ma- trices and the mechanical rotation of factors through clusters of surface variables. Goldberg (since 1981) The search for universal dimensions in trait names and work on redeveloping a Costa and McCrae better instrument were resumed. (since 1985)

1.3.2. Major theoretical assumptions of the Five Factor Model

Supporting the views expressed by Hjelle and Ziegler (1976), Costa and McCrae (1996, 2017) refer to a set of essential premises concerning human nature, which, as they claim, underlie the Five Factor Model of personality (Costa and McCrae 1985). The model could also be described as resting on three major assumptions (Costa and McCrae 2017: 12-14): a) knowability, that is the belief that personality can be a valid scientific research area rather than merely a metaphysical entity, variability, that is the belief that human personalities are different, but nevertheless can be studied in an attempt to account for possible universals;

26 b) proactivity, that is people’s empowerment to behave in a certain way, juxtaposed with the kind of determinist view of individuals advocated by behaviourists; c) reality, that is the belief in the relationship between the manifested exemplifica- tions of human behaviour reported on questionnaires and the trait construct. Adopting the Five Factor Model of personality also means accepting that personality attributes are continuous dimensions rather than discrete or categorical types, hierarchically ordered (with multiple layers of specificity), and that they cannot be easily structured. It can be assumed that the Five Factor Model of the personality (Costa and McCrae 1985) is not merely a model, but also, in fact, a scientific concept, as it underlies a separate personality theory. This theory does not merely assume that a person’s individuality can be demonstrated in a variety of patterns of thoughts, feelings, and behaviours, but also recognises the following premises: a) the initial assumption that individuals’ plans, goals, and schedules are influenced by their personality traits (Murray and Kluckhohn 1953); b) the premise that individuals have a conscious view of themselves, which is both cognitive and affective in nature (McCrae and Costa 1988); c) the recognition of an individual’s perception of themselves, which is coherent from their point of view, and consistent with their personality traits (McCrae and Costa 1996); d) the assumption that personality traits are accountable for “consistency and coher- ence in affect, behaviour, cognition, and desire” (Wilt and Revelle 2017: 57)3; e) the belief in the endogenous origin of traits which are substantially heritable (Tel- legen et al. 1988), reflected in the claim that all the Big Five traits are biologically determined (McCrae and Costa 2008)3; f) the acceptance of the developmental nature of traits, which are fully developed by about the age of 30 (Costa and McCrae 1994); g) the recognition of the hierarchical organisation of traits, which consists in their specific order, that is from specific to broad dispositions, which can be eventually generalised as the ‘big’ five traits, i.e., Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness (Costa and McCrae 1995).

3 Evolutionary psychologists suggest that traits can be viewed as strategies to meet adaptive challenges in the social environment (Nettle 2005; Wilt and Revelle 2015, 2017).

27 According to Costa and McCrae (2017), adaptations to the environment are influenced by both personality traits and earlier adaptations. Consequently, adaptations which fail to successfully match cultural values or individuals’ goals can be interpreted as mal- adjustments and refer to personality disorders, and finally, adaptations vary throughout an individual’s life, and can be affected by both changes to the environment and an individual’s conscious decisions. They recognise that human actions may be subject to multiple determination, i.e., they satisfy various needs connected with various traits. At the same time, they are aware that human behaviour is subject to external influences, which results in the interplay between an individual’s personality and their environment, constructing the environment consistently with an individual’s personality traits and reci- procity, i.e., influencing some element of the environment that influences an individual. Costa and McCrae (2017) claim that the dynamic processes can be universal in nature (i.e., affect all individuals in a similar manner), but they can also be affected by individual variation among individuals; the expression of some traits can be shaped by the expression of others. The model recognises the duality principle, that is the assumption that traits and trait indicators constitute two entirely different categories (or levels) and that trait indicators are characteristic adaptations but also, at the same time, proxy measures of basic tendencies. As Costa and McCrae (2017: 25) put it, “situations, behaviors, and explanations are some of the ingredients of indicators of heritable traits”4. Hence, it is possible to distinguish between dispositional states and personality traits, and, at the same time, draw a parallel between personality states, which can be understood as short-term, con- crete and contextualised patterns of affect, behaviour, condition, and desire (ABCD) (Wilt and Revelle 2015), and personality traits, involving decontextualised ABCDs (Wilt and Revelle 2017: 68). Wilt and Revelle suggest that personality traits can be described in the same way as personality states, which can lead to a conclusion that no need exists to design a separate framework of investigation of personality states. McCrae (2010: 59), presented the taxonomy of individual difference variables which included the personality traits from the Five Factor Model shown as dispositions. These dispositions were thus Neuroticism, Extraversion, Openness to Experience, Agreeableness, and

4 The proponents of FFM do not offer a single explanation of the relationship between an individual’s traits and their behaviour. Along with the ABCD framework, they present the conceptualisation introduced by Yang et al. (2014), according to which traits can further be divided into situations in which trait-relevant behaviours take place and explanations for those behaviours.

28 Conscientiousness (NEOAC). Together with abilities they could be understood as manifestations of tendencies, and manifestations of psychological individual differences (IDs) and they correspond to the ‘big’ five personality traits, suggested as the constituents of human personality (McCrae and Costa 1985; Costa and McCrae 1992). To recap, a number of conclusions can be made about the features of the NEOAC dispositions, such as: hierarchical organisation of traits; independence of adult personality from upbringing, cultural norms or historical conditions; gradual stabilisation of personality traits in adult individuals; genetic influence on personality; the relationship between personality and age; the impact of one’s objective biography, which can be understood as an individual’s factual life story, or, to put it in a more detailed way, their real successes and failures, struggles, or redemptions in the course of their lives; the influence of brain pathology and pharmaceutical agents on personality; the transcultural character of personality traits. Personality can thus be viewed as a construct defined by the Five Factor Model, and as such, to a large degree, biologically determined, and identifiable across a vast number of cultures, and possible to be investigated in cross-cultural contexts. In order to account for cultural differences, adaptations of personality tests need to be applied which fulfil relevant goodness criteria for psychometric tests relating to cultural adjustment.

1.3.3. The ‘Big’ Five personality traits

The following subsections deal with the description of the dispositions which constitute the elements of the Five Factor Model (FFM), i.e., Extraversion, Openness to Experience, Agreeableness and Conscientiousness, and Neuroticism.

29 Extraversion

Like all the other Big Five personality traits, Extraversion has a major impact on an individual’s everyday existence, as well as their well-being. Wilt and Revelle (2017:57) define Extraversion as “the dimension of personality reflecting individual differences in the tendencies to experience and exhibit positive affect, assertive behaviour, decisive thinking, and desires for social attention.” They also list a number of characteristics of extraverts, such as being energetic, dominant, spontaneous and sociable, and juxtapose these with the typical characteristics of introverts, such as being lethargic, inhibited, reflective, and generally quiet. The term was introduced into psychology by Jung (1921), who described extraverts as more focused on the outer world, or, in other words, more engaged with it, whereas introverts as those focused more on their own inner mentality, or, to put it differently, drawn inward into thought. All major personality models, including the Myers-Briggs Type Indicator (MBTI), (Myers and Mc Caulley 1985), Cattell’s Sixteen Personality Factor Questionnaire (16 PF) (Cattell 1956), and Eysenck’s Personality In- ventory (Eysenck 1964), include Extraversion as a distinct personality domain. One of the predecessors of the Big Five Model, developed by Tupes and Christal, (1961) includes the dimension of surgency, which is the equivalent of extraversion, built of the underlying traits of talkativeness, assertiveness, as well as being energetic and adventurous. Costa and McCrae (1992) list six components which together create the dimension of Extraver- sion, that is warmth, gregariousness, assertiveness, activity, excitement seeking, and positive emotion. These characteristics are, according to Wilt and Revelle (2017), connected through some common processes. The founders of the Five Factor Model suggested that the common process be the disposition to engage in social behaviour (McCrae and Costa 1997), although it needs to be acknowledged that other explanations suggest positive emotionality (Tellegen 1985; Watson and Clark 1997) or a propensity to seek social attraction (Ashton et al. 2002). A number of theorists have attempted to account for the organisation of Extraver- sion as a higher order trait. Eysenck et al. (1970) introduced a framework according to which it is possible to explain Extraversion as a phenomenon that can be analysed at four different levels of abstraction, which comprise: specific responses, which appear on single occasions, such as flirting with a specific person at a given party;

30 habitual responses, such as behaving in a lively manner at different parties on various occasions; facet-level constructs, for instance gregariousness; the general (broad) trait of Extraversion. According to the above model, the constructs residing at lower levels (e.g. the flirting behaviour) are incorporated within only one higher level trait (e.g. Extraversion). This suggestion has not been supported by sufficient evidence, however, in terms of scientific scrutiny. One controversial issue in the interpretation of Extraversion is whether to include impulsivity as one of its lower-order features (Revelle 1997). While scholars have tried to account for the degree of relevance of both concepts, studies showed that impulsivity and sensation-seeking tend to correspond to the Conscientiousness scale, but sensation seeking corresponds more strongly to Extraversion (Whiteside and Lynam 2001). At the same time, the results of a study conducted by Quilty et al. (2014) indicate that sensation seeking is related, but not entirely explained by assertiveness or enthusiasm, both being aspects of Extraversion. Evolutionary studies of Extraversion confirm its hereditary character (Bouchard and Loehlin 2001). The existence of genetic polymorphism with regard to the trait (i.e., individual variation consisting in the degree of individuals’ Introversion and Extraver- sion) can be explained in a number of ways. For example, the potential benefits (or, to put it in environmental terms, potential fitness benefits) from Extraversion include a higher likelihood of developing interpersonal relationships and having more sexual partners (Nettle 2005), and a relatively better social status, which is realised by greater dominance. On the other hand, extraverts tend to expose themselves to more safety risks such as injuries, and are more likely to enter antagonistic competition (Schaller and Murray 2008). Introversion can thus be viewed as a stance of safety or protection. Lukaszewski and Roney (2011) also point out that more attractive individuals favour Extraversion in order to increase their chances of mating. A number of controversies exist over the biological mechanism which underlies Extraversion. Eysenck (1967a, 1967b) suggested that an individual’s level of Extraver- sion depends on the functioning of their ascending reticular activating system (ARAS), which he had studied before in the processes of attention and learning. The claim was rebutted by Stelmack (1990), who pointed to the fact that resting brain activity does not show a clear relationship with the level of Extraversion, as well as Matthews and Gilliand

31 (1999), who found no clear relationship between the levels of cortical arousal and the level of Extraversion. Gray (1970, 1981, 1982) proposed an alternative theory, known as the Reinforcement Sensitivity Theory (RST), assuming the existence of three neural systems determining human behaviour, including: the Behavioural Approach System (BAS), which was believed to be relevant to an individual’s impulsivity; the Behavioural Inhibition System (BIS), which was presumed to relate to an individual’s anxiety; the Fight-Flight System (FFS), regulating an individual’s responses to threats. Gray (1970, 1981, 1982) suggested that the traits of impulsivity and anxiety should be treated as primary traits that together account for the general trait of Extraver- sion. In other words, he compared an individual’s level of Extraversion in Eysenck’s theory to their level of impulsivity lessened by their level of anxiety. BAS was re-examined in later studies and subsequently linked to such concepts as pursuit of reward, increased risk-taking tendency, pleasure-seeking and search for incentives (Carver and White 1994; Corr 2008). The positive relationship between BAS and the level of Extraversion was also confirmed in later studies. It needs to be emphasised that the origin and the biological basis of Extraversion are not generally agreed upon. Alternative models, such as the Be- havioural Facilitation System framework, linking the level of an individual’s Extraver- sion to their dopaminergic functioning or, more generally, to reward processing (DeYoung 2010), are not scarce. Since the ABCD framework contributes to a better understanding and a more accurate description of extravert characteristics, attention will now be given to the affect, behaviours, cognitive patterns, and desires which are typical of extroverts. These are presented in Table 6, which is based on the summary offered by Wilt and Revelle (2017).

32 Table 6. The ABCDs of Extraversion. A - affect B - behaviours C - cognition D - desires - trait Extraversion - extraverts view their - extraverts are likely - extraverts demon- and positive affect behaviours as bold, to engage with posi- strate relatively (at either, momen- socially adept, and tive stimuli and tend higher expectations tary, situational or secure, e.g. they at- to avoid negative for happiness once trait level) show a tend parties more of- stimuli, but as far as they have achieved strong, positive rela- ten, enter a higher neutral situations are their goals, tionship, number of relation- concerned, they - Extraversion is con- - social perception of ships, consume more view them relatively nected with ap- extraverts reinforces alcohol, but at the more positively than proach motivation, the belief about their same time do more introverts i.e., they are more greater effectiveness exercise, - they exhibit rela- reward-oriented of pursuing goals - extraverts display tively more favoura- - extraverts are char- and succeeding in more assertive be- ble cognitions con- acterised by greater accomplishing aims haviours, but at the cerning social need for social con- same time show situations than intro- tact, intimacy, and more enthusiasm in verts, interdependence, interacting with - they tend to be rela- - extraverts strive for other individuals tively better multi- social status, and - they also talk more taskers, which they positive affect, as often, use more ges- owe to their ability well as a career in tures, or, at least, to restrict distrac- the social domain, present different tions, - they demand rela- communicational - introverts tend to tively more personal styles perform better in agency, tasks which require - they have a rela- persistence or sus- tively greater need tained attention of affiliation

As far as an individual’s objective biography is concerned, apart from actually entering a higher number of relationships, greater job satisfaction, and even better performance at work, an extravert is more likely to involve more positive emotions in a story, as well as centre the story around such key concepts as status, optimism, sociability, and activity. They are also more likely to share their memories with others, and while doing so, they have a tendency to emphasise interpersonal trust (McLean and Pasupathi 2006). Also, from the perspective of an individual’s self-esteem, extraverts tend to show a greater praise of themselves, which Crocker and Luhtanen (2003) refer to as a global evaluation of general worth as a person.

33 Openness to Experience

Openness to experience is a trait which has not been unanimously recognised as one of the pillars of human personality. Problems concerned the very naming of the trait, which changed from model to model, shifting from such notions as intellect and culture to those of imagination and unconventionality (Sutin 2017). The authors of the FFM define openness (term used interchangeably with “openness to experience”) as “the breadth, depth, and permeability of consciousness, and (...) the recurrent need to enlarge and examine experience” (McCrae and Costa 1997: 826). A brief analysis of the definition was conducted by Sutin (2017) who made an attempt to explain its key elements, as well as expands it by considering the motivational dimension of motivation: the breadth of openness refers to the range of an individual’s interests (e.g. the number and variability of one’s hobbies); the depth of openness refers to the density of associations regarding particular ideas held by individuals, which can also be referred to as divergent thinking skills; the permeability of openness refers to malleability of mental boundaries, or, broadly speaking, amount of differentiation in a wide spectrum of perspectives; the motivation to actively engage in search for variety and/or novelty of experiences. From the psychometric standpoint, Openness to experience is built of the following dimensions (Costa and McCrae 1992): O1: fantasy, which involves a wide imagination; O2: aesthetics, which manifests itself in a person’s appreciation of beauty; O3: feelings, represented in plethora of emotions that an individual experiences; O4: actions, which result from a person’s eagerness to try new things; O5: ideas, which dwell in a person’s intellectual curiosity; O6: values, which - as regards an open individual - are liberal. From the nomological perspective, openness to experience can be linked with a number of other human characteristics, including authoritarianism, need for cognition, need for closure, or emotional intelligence (Sutin 2017: 86). Given the focus of the dissertation, special attention should be paid to the relationship between openness to experience and need for cognition. Some vital characteristics of individuals who are highly open to experience include verbal fluency, sense of humour, and being expressive (Sneed et al. 1998), as well as greater cognitive flexibility, which is demonstrated by storing,

34 manipulating, and retrieving information with ease (Sutin 2017). Consequently, it is logical to expect that both openness to experience and people’s greater need for cognition positively correspond with their use of language learning strategies. Table 7 contains examples of language-related manifestations of Openness to Experience.

Table 7. Linguistic manifestations of Openness to Experience. Study Findings Bates and Shieles (2003) Open young adults tend to perform better on vocabulary and comprehension tasks Noftle and Robins (2007) Open adolescents achieve better results in verbal sections of language aptitude tests Ayotte et al. (2009) Open individuals are characterised with better executive functioning, which predis- poses them to perform better on verbal fluency tasks Sharp et al. (2010) Open older adults show better verbal measures of cognition, such as analogies and synonyms Ritchie et al. (2013) Open individuals demonstrate better reading comprehension skills

In terms of the causes of openness, researchers generally agree that it is at least partly explained by genetic factors, much in a similar way that an individual’s general cognitive ability is (Riemann et al. 1997). Studies confirm the hereditary character of the trait; however, estimates of the contribution of genetic factors to an individual’s level of openness vary considerably, ranging from slightly over 20% (Power and Pluess 2015) to nearly 50% (Larsen and Buss 2013). A considerable amount of research concerns the changeability of openness across individuals’ lifespan. It is commonly believed that adolescents and young adults tend to be relatively more open to experience than adults. How- ever, the fact is that openness to experience is already a differentiating factor in childhood (Soto and John 2014). From then on, it continues to increase in adolescence and young adulthood (Robins et al. 2001). Moreover, people’s educational choices can matter in the development of the trait trajectory. For instance, a study conducted on German university students indicated that the level of their openness to experience was significantly higher than their counterparts’ who chose vocational education (Lüdtke et al. 2011). It is still debatable when the level of openness starts declining, but many specialists agree that it is after reaching the age of 60 that the level of an individual’s openness to experience starts declining (Roberts and Mroczek 2008).

35 Agreeableness

Graziano and Tobin (2017) briefly described Agreeableness as a synthetic measurement of the motivation to maintain positive relationships with others, and, at the same time, emphasised that the label prioritises these over individual differences in social cognition and thought. Prior to that, Costa and McCrae (1992) had redefined the underlying facets of Agreeableness, among which they included altruism, compliance, modesty, straight- forwardness, tender-mindedness, and trust. According to another description provided by McCrae and Costa (2003), individuals who are highly detached tend to be critical, scep- tical, try to push limits, express hostility more directly, and are more often reluctant to agree whereas highly agreeable individuals are relatively more sympathetic, considerate, warm, compassionate, likeable, and generous. The underlying traits of Agreeableness are described by Graziano and Tobin (2017) in the following way: altruism may be demonstrated as defensive impression management; compliance decreases as a result of traumatic events, and is also reported to matter in succeeding in forming and maintaining a relationship and finding employment; modesty is related to self-estimated intelligence, which means that people who are modest tend to lower their personality scores; tough-mindedness (the opposite of tender-mindedness) is the strongest facet predicting right-wing authoritarianism as well as sexism; trust (more specifically – organisational trust) determines people’s assessment of workplace. It is also worth considering the relationships between Agreeableness and other semantically related characteristics, that is empathy, pro-social behaviour, and, to some degree, self-regulation. As far as empathy is concerned, high Agreeableness scorers demonstrate a better developed skill of perspective-taking while not being affected by personal distress in case of witnessing others’ misfortunes. They also behave more pro- socially (Graziano and Habashi 2015). People who are more agreeable tend to show better self-regulation abilities, especially with regard to frustration management in social situations. It has also been reported that a positive relationship between Agreeableness and Behavioural Inhibition System exists, whereas the relationship between the Behavioural Approach System and Agreeableness is negative (Graziano and Tobin 2017). Another related notion is that of effortful control, which can be understood as the ability to

36 suppress a dominant behaviour. It is said to be the common characteristic of individuals who display high Agreeableness. In school and university settings, fewer problems related to noncompliance or inattention are reported with children who are high in Agreea- bleness (Laursen et al. 2002). The trait can thus have a considerable impact on learners’ performance and participation in classroom activities. Similarly to other personality traits, Agreeableness can generally be measured by observation. Relevant information can come from observers such as spouses or various types of supervisors) or a range of introspective methods (with questionnaires being the most widely employed tool). This obviously car- ries some degree of risk of bias, specifically towards higher Agreeableness, since qualities such as friendliness or cooperativeness are considered to be more socially desirable than their opposites, such as argumentativeness, or antisocial behaviour (Bègue et al. 2015).

Conscientiousness

Conscientiousness can be briefly described as a wide range of constructs that reflect the propensity to be self-controlled, responsible to others, hard-working, orderly, and rule- abiding (Jackson and Roberts 2017: 133). It is not without reason that the definition covers a wide composition of underlying lower-order individual differences. Similarly to other constituents of the Five Factor Model, it is hierarchically organised, and reflects relatively stable, automatic patterns of thoughts, feelings, and behaviours. A number of researchers distinguish between the proactive and inhibitory component of Conscien- tiousness (Costa et al. 1991; Roberts et al. 2005; Jackson et al. 2010). Jackson and Roberts (2017) explain that the proactive component can be associated with the willingness to accomplish goals and attempts to perform well whereas the inhibitory component is related to responsibility, delayed gratification, and controlling one’s instincts or impulses. Table 8 contains an overview of the lower-order facets of Conscientiousness suggested by various models.

37 Table 8. Alternative arrangements of the facets of Conscientiousness. Authors Constituents Roberts (2005), industriousness and orderliness; empirical tests prove that this arrangement DeYoung et al. (2007), only caters for the proactive component of Conscientiousness

Roberts et al. (2005), achievement, self-control, and responsibility Jackson et al. (2010) Perugini and Gallucci (1997) orderliness, industriousness (constituents of the proactive component), re- Peabody and De Raad (2002) sponsibility, and self-control (constituents of the inhibitory component)

The problem encountered in comparing various organisations built of five or more components of Conscientiousness consists in the difficulties to replicate studies, as well as in accounting for the differences regarding the inclusion of different underlying factors by different authors. A comprehensive list of characteristics that are ascribed to conscientious individuals has been proposed by Jackson et al. (2009), who pursue the four-facet hierarchical organisation of Conscientiousness: people high in orderliness enjoy cleanliness and neatness, which can manifest in arranging belongings in an organised manner, but also drawing plans and attempting to stick to them; industrious individuals are basically hard-working and not easily discouraged by ob- stacles in achieving their goals, as well as able to cope with challenge and persevering; highly responsible people easily conform to rules and follow instructions, but at the same time they are more likely to keep promises or appointments; self-control is an inhibitory facet which determines an individual’s ability to refrain from immediate action, restrict one’s impulses, or resist the urge of immediate gratification; Numerous attempts have been made to trace the developmental pattern of Consci- entiousness. The trait is reported to increase steadily until the age of fifty, level off between the years of 50 and 70, and decrease in old age (Roberts and DelVecchio 2000; Lucas and Donnellan 2011). The belief in the stable character of Conscientiousness results in the assumption that an individual who scores high in Conscientiousness in childhood is expected to achieve a relatively higher score in adolescence and adulthood. These expectations are, to a certain degree, mediated by cultural differences (McCrae et al. 1999), which are also responsible for at least a part of variance in Conscientiousness. Some psychologists point to the existence of the so-called Conscientiousness paradox, as nationalities commonly regarded as highly conscientious, and also relatively better-off,

38 such as the Japanese, have a tendency to lower their self-reported Conscientiousness in comparison to other nationalities, a phenomenon which is influenced by respondents’ modesty and high demands concerning laboriousness (Chen et al. 2014; Meisenberg 2015). More detailed patterns concerning the development of the four constituting facets of Conscientiousness are presented below: self-control and responsibility are reported to increase in young adulthood and in later stages of life (Jackson et al. 2009); industriousness is subject to considerable fluctuations during young adulthood and generally increases during lifespan (Terracciano et al. 2010); orderliness does not significantly change as individuals mature (Jackson et al. 2009); impulse control is also partly responsible for the overall increase in Conscientiousness (Jackson and Roberts 2017). Conscientiousness plays a vital part as a predictor of people’s objective biog- raphies, as it is one of the best predictors of academic success (Bratko et al. 2006) as well as underlies students’ better grades (Poropat 2009). It also determines the level of students’ dedication to handling assignments and projects (Duckworth and Carlson 2013), corresponds with students’ motivation to succeed in school performance (Capara et al. 2011), and, finally, determines a successful transition from school to professional settings (Jackson and Roberts 2017). As can be seen, Conscientiousness can exert a considerable influence on learners’ behaviour, performance and attainment. It could also be expected that a high level of this trait is correlated with the employment of those learning strategies whose use requires relatively more effort and diligence. Moreover, the level of Conscientiousness might mediate between learners’ motivation and their involvement in performing specific tasks, as well as affect the efficiency of managing the learning process. In other words, highly conscientious learners might be more likely to display greater willingness to take charge of the learning process. Therefore, it could be expected that more conscientious learners would demonstrate greater responsibility and accountability as language learners, and their use of learning strategies could, potentially, be more frequent.

39 Neuroticism

For a variety of reasons, defining Neuroticism and describing its underlying traits does not constitute an easy task. Firstly, a number of findings refer to clinical settings, psychopathology, and areas of life such as physical or mental health disorders. To put it more bluntly, Neuroticism shows clear associations with major forms of psychopathology (Tackett and Lahey 2017). Secondly, the developmental patterns of Neuroticism are not obvious, and it is not possible to clearly attribute a certain amount of its variation to heredity and another amount to environment, since environmental influence is particularly dubious in the specific case of Neuroticism. Neuroticism is said to reflect individual differences in tendencies toward negative affect, including sadness, anxiety, and anger, as well as individual responses to threat, frustration, or loss (Widiger 2009). It is alternatively referred to as Negative Affectivity (NA) or Negative Emotionality (NE) (Rothbart et al. 2001). The FFM (Costa and McCrae 1985) assumes that existence of six underlying facets of Neuroticism, including anxiety, angry hostility, depression, self-consciousness, vulnerability, and impulsiveness. While Neuroticism is substantially hereditable, with around half its variance explained by genetic influences (Lahey 2009, Widiger 2009), environmental factors prove equally important. At the same time, the variation of Neuroticism can be attributed to either non- shared or shared environmental factors (the latter showing complex connections with genetic variation) (Tacket and Lahey 2017). The development of Neuroticism shows complex patterns, but it is possible to trace slight increases of rank-order stability among both young and adult individuals (Roberts and DelVecchio 2000). Also, the overall variability of an individual’s personality is associated with emotional lability, often considered to be a facet of Neuroticism, if not its synonym (Widiger 2009). Although manifestations of Neuroticism which could affect educational performance are scarcely described in literature, some studies which deal with general aspects of Neuroticism can shed some light on predicting academic performance: high Neuroticism scorers show preferences for insufficient forms of coping, such as avoidance or disengaging attention (Bredemeier et al. 2011);

40 the cognitive response of individuals high in Neuroticism to both negative and ambiguous stimuli is relatively higher than those lower in Neuroticism (Hirsh and Inzlicht 2008). Regarding the predictions of academic success, studies indicate that a high level of Neuroticism can result in learners’ low performance (Trapmann et al. 2007; Bhagat and Nayak 2014), give rise to situational fears or anxiety (Hakimi et al. 2011), and even impede learners’ self-actualisation (Duff et al. 2004). At the same time, the impact of Neuroticism on, especially, academic performance, is not free from controversy, as one of the underlying facets of Neuroticism, i.e., anxiety, is known to take a range of forms. Its two extreme forms, debilitative and facilitative anxiety were distinguished by Alpert and Haber (1960), who also designed an instrument to measure both, that is the Anxiety Achievement Test. A study conducted by Carrier et al. (1984) indicated that debilitative anxiety impaired students’ performance in educational activities, such as note-taking, and subsequent tests, whereas facilitative anxiety did not significantly correlate with students’ educational performance. The impact of facilitative anxiety remains dubious according to the results of a more recent study, dedicated specifically to the role of anxiety in learning English as a foreign language. Selvam et al. (2016) found that that 70% of the students they investigated had negative attitudes towards speaking in English (their L2) despite being diagnosed with facilitative anxiety in speaking English on the basis of how their scores from an adaptation of the Foreign Language Classroom Anxiety Scale (FLCAS) (Horwitz et al.1986). Neuroticism was also found to account for more than 10% of variance in exam marks and impair academic performance, and its influence could not be compensated by learners’ ability or aptitude (Chamorro-Premuzik and Furnham 2003).

1.4. Personality in foreign language development

The interest in language learners’ personality derives from humanistic psychology, whose five major postulates were brought forward by Bugental (1964: 23-24) and included treating the human person as superior to merely a sum of parts, existing in a human context, aware of their own existence and experience, making constant and conscious choices, and acting intentionally. Accepting the above assumptions in the field of foreign language development resulted in attaching increasingly more importance to the language learner

41 and gradually gave rise to an explosion of research in learners’ IDs. The interest in the learner gradually centred around more specific notions, such as their age, beliefs, language aptitude, personality, motivation, learning styles, and learning strategies. Before that, however; a major change in the perception of the human person had been brought forward by humanistic psychology and a shift away from the behaviouristic understanding of people’s intentions and actions were both echoed in foreign language education. Then, from the 1970s, the focus of researchers’ attention gradually relocated from efforts to produce the perfect, universal teaching method that would satisfy everyone to coming to terms with, embracing, and trying to find a way to deal with individual variation in language learning (Droździał-Szelest 2013)5. An overview of early studies into the impact of language learners’ personality on L2 development was provided by Leino (1972), who described the major studies conducted until 1970. Its summary is presented in Table 9 which lists the major conclusions of each study in the form of findings and/or hypotheses. Overall, the described studies tended to focus on the impact of certain specific personality traits on the language learning process.

Table 9. Early investigations of the impact of personality on L2 development. Ordering by author findings / hypotheses Dunkel (1947) Basing on a study investigating the meaning of compulsiveness in studying Latin, the author concluded that a certain level of the trait could actually facilitate learning the language, but once the critical level is surpassed, it becomes an impeding factor rather than a facilitator. Wittenborn et al. (1945) The investigated variables included not merely the habits of high and low achievers among university students, but also their three personality traits, i.e., ability to concentrate, anxiety, and self-confidence. Students’ overall approach to work and ability to concentrate were found to be valid predictors of educational attainment for freshmen. Plimsleur et al. (1962) The authors tested a number of traits, such as social conformity, flexibility, ambiguity tolerance, and tendency to frustrate, but failed to obtain clear results due to methodological limitations of the study (e.g. differences in participants’ age and achievement measures limited to written test results). Konttinen (1970) Significant correlations were found between Neuroticism, Social Extraversion, Im- pulsivity, and English Language Skills of university students. The study used Ey- senck’s theory of personality and thirty-two measures of foreign language skills.

5 Droździał-Szelest (2013) points out that even in the post-method era, practitioners in the ELT industry should not merely passively consume the knowledge of other people, but also be able to explain their choices and judgements of learning and teaching practices and rest on a theoretical background whose part the knowledge of the teaching methods is.

42 The 1970s where the decade when interest in the good language learner had already been propelled by the increasing popularity of both, humanistic psychology, and humanistic education (Rogers 1969). At the beginning of the decade, Taylor et al. (1971) examined the influence of 28 language learners’ sensitivity, conceptualised as empathy, tolerance to anxiety, intelligence, involvement in emotional experiences, and perception of emotional expression on their learning attainment. The researchers claimed to have proved the relevance of learners’ sensitivity to the correct pronunciation of Japanese utterances in spontaneous performance, while relating its other dimension, i.e., intuition (defined by tolerance to anxiety, intelligence, and psychological mindedness) to the pronunciation authenticity in repeated sentences. The researchers also pointed to the scarcity of psychological assessment of language learners. Leino (1972) examined the relationship between personality, which she considered to be a collection of temperament traits, and English school achievement of secondary school leavers. The study was conducted on 64 final grade secondary school students. Low correlations were observed between English school achievement variables and two of personality variables measured by Cattell’s 16PF, that is General Intelligence-Mental Defect, and Parmia-Threctia (the dichotomy of boldness corresponding with Extraversion and lack of self-control). The researcher concluded that the instructional method applied, meaning audiolingualism, failed to account for the importance of individual variation in second language learning. Oskarsson (1975) investigated the relationship between foreign language proficiency and various psychological variables, including dominance, deliberateness, sociability, emotionality, flexibility, perseverance, emotional stability, and independence, measured by a battery of tests containing ten self-assessment questions per trait. The participants of the study included 53 farm management students who, apart from taking psychological tests tapping the above factors, were assessed in terms of their reading, listening, speaking, and writing performance. The researcher found a negative correlation between high deliberateness and test achievement in multiple choice tasks (r = -.33), as well as significant positive relationships between reading comprehension skills and sociability and emotionality (positive), as well as with deliberateness (negative). The study can be viewed as very innovative regarding its area of focus, but at the same time, its results should be approached with caution due to a relatively low number of participants (only 20 students were tested with all the instruments employed in the study).

43 A number of studies investigating the role of personality in foreign language development could be referred to at least one aspect of Krashen’s (1981) theory of second language acquisition. The theory involves five main hypotheses, i.e., the acquisition- learning hypothesis, the monitor hypothesis, the natural order hypothesis, and the affective filter hypothesis. In fact, personality occupies a central position in Krashen’s affective filter and monitor hypotheses. More specifically, the distinction between monitor over- and underusers rests on the juxtaposition of their individual characteristics, i.e., the self-consciousness and introversion of overusers, and outgoing, uninhibited personality of underusers (Krashen 1981: 17). Consequently, relatively more outgoing individuals would be expected to demonstrate less hesitance in spoken style while their introvert, self-conscious opposites would typically display hesitate more often when speaking in a foreign language (Krashen 1981: 18). According to Krashen (1981: 23), personality factors are “interrelated with motivational factors” which would suggest that an outgoing and self-confident individual would be expected to have a lower affective filter and experience fewer difficulties during the intake processes of a foreign language. The filter was also believed to be lower if a foreign language speaker was empathetic and could thus more easily identify with speakers of the target language. At the same time, Krashen supported Schumann’s (1975) view, according to which “the natural factors that induce ego flexibility and lower inhibitions (assumed to relate to increased empathy) are those conditions that make the learner less anxious, make him feel accepted and make him form positive identifications with speakers of the target language.” (Schumann 1975: 227). Other psychological factors, viewed by Krashen as personality factors, included attitude towards the classroom and the teacher and analytic orientation6. The former factor was assumed to generate high input (Krashen and Seliger 1976), whereas the latter was believed to facilitate conscious learning and improve performance in tests (Krashen 1981: 24). To support his claims, Krashen (1981) referred to a number of empirical investigations. These could be grouped into six categories, including explorations confirming the relationship between various forms of anxiety and target language achievements, research addressing the positive relationship between self-image and second language achievement, investigations dedicated to the positive implications of having an outgoing personality, examinations of the role of empathy in foreign language development, explorations

6 The concept of learners’ analytic orientation is related to Krashen’s idea of analytic personality.

44 of the role of learners’ attitudes, and, finally, studies of the implications of having an analytic personality. A summary of relevant results is provided in Table 10.

Table 10. Investigations of psychological factors in language acquisition (adapted from Krashen 1981) ordering by author findings Investigations of the links between anxiety and TL achievement Wittenborn et al. (1945) relationships between low / high achievement and anxiety and the level of confidence Carroll (1962) weak negative correlation between anxiety and accomplishment in intensive foreign language courses Chastain (1975) significant, moderate, negative correlation between test anxiety and success in learning French (audiolingual instruction) and weak positive correlations found between anxiety and success in learning German and Spanish (“traditional” methods) Gardner et al. (1976) negative relationship between classroom anxiety and speech skills/grades in French courses Naiman et al. (1978) relevance of classroom anxiety and failure to fear of rejection and performance in an imitation test and listening comprehension tasks Investigations of the links between self-esteem and TL achievement Heyde (1977) a strong positive correlation between ESL learners’ self-esteem (understood as their evaluation of their own worth) and teachers’ evaluations of their oral performance Oller et al. (1977) a positive relationship between Chinese learners’ self-concept and their achievement in ESL Naiman et al. (1978) learners’ self-confidence regarded by teachers as a characteristic of good language learners Studies examining the role of foreign outgoing personalities Chastain (1975) positive relationship between performance in a foreign language and being outgoing Wesche (1977) relationship between role-playing (believed to reflect the willingness on the part of the student to engage in a dialogue or situation) and proficiency in listening comprehension and reading skills Naiman et al. (1978) no clear relationship between foreign language proficiency and Introversion/Extra- version Investigations of the role of foreign language learners’ empathy Guiora et al. (1972) individuals more aware of feelings are also more sensitive to details and demonstrate it when speaking a foreign language Naiman et al. (1978) no significant relationship found between empathy and listening comprehension / imitation skills Studies evaluating the impact of learners’ attitudes Gardner et al. (1976) learners’ assessment of their learning situation reflected in spoken performance and grades Bialystok learners’ subjective evaluation of their learning situation found to be a predictor of and Frohlich (1977) reading comprehension achievement Naiman et al. (1978) learners’ general attitude (i.e., their perception of their language situation) measured in interviews found to be the best predictor of their overall success in language learning Research into the meaning of learners’ analytic personality Oller (1976a) The following self-reported traits were found to facilitate the acquisition of English in a group of Chinese learners: - being logical, stubborn, generally happy; - being ‘teachable’ and friendly; - being clever, broad-minded, intellectual, and calm. Oller (1976b) Mexican Americans learning English who perceive themselves as logical, sensitive, and shy, performed better in a multiple choice cloze test.

45 Lalonde and Gardner (1984) put forward a model which aims to integrate personality variables into second language acquisition, focusing in particular on learners’ analytic orientation and seriousness. Their empirical findings based on investigating a group of 88 learners of French as a foreign language included the following significant relationships can be summarised as follows: self-esteem and organisation positively correlated with self-perception of proficiency; innovation negatively correlated with achievement; personality variables correlating with integrativeness included breadth of interest, organisation, responsibility, self-esteem and social desirability (positive relationship) as well as conformity and impulsivity (negative relationship). Ehrman and Oxford (1988) carried out a study of 78 adult learners and trainers of various languages, investigating the relationships between their personalities and other individual characteristics, such as gender, professional choices, cognitive styles, and learning strategies. The overall purpose of the study was to incorporate the concept of personality type into designing language training and counselling for the specific needs of language learners. The treatment of personality in the study was peculiar in two ways. Firstly, the employed model of personality was MBTI, and, secondly, personality was analytically combined with cognitive style to create an aggregated variable, the so-called “psychological type”. The researchers referred to personality traits as to “patterns of behaviour that characterize a person’s response to the environment” (Ehrman and Oxford (1988: 2). They listed several factors which they associated with a positive impact on affect language attainment, including a sense of humour, achievement orientation, out- goingness, impulsivity, assertiveness, risk taking, ambiguity and frustration tolerance, self-confidence, determination, and empathy. Griffiths (1991) compiled a report in which he included a number of conclusions from several studies of learners’ personalities. Since Eysenck’s Personality Questionnaire (EPQ; Eysenck and Eysenck 1975) was selected for the investigation of personality, the study can be classified as following the trait approach to personality research. The findings from the study could be summarised under five major headings: In terms of the employed methods of learning and teaching (discovery vs. receptive learning), extraverts performed better in discovery-driven learning, whereas introverts achieved more when taught in a more receptive manner as indicated by the mean values of post-test and retention test.

46 Regarding pairwork and groupwork, the efficiency of these classroom arrangement modes depended on whether students were paired/grouped with similar or different partners. Specifically, students paired with partners who differed considerably in terms of language anxiety achieved relatively better results in terms of oral test performance, showing solidarity, raising each other’s status, and spent more time asking for orientation, information, and confirmation. The possession of some personality traits might be the reason behind learners’ preference for some specific language tasks. For example, with respect to the range of necessary stimuli extraverts tended to expect greater variety in classroom activities and even prefer higher levels of noise in the classroom as optimal. Praising could work better for extraverts than for introverts as the former showed preference for reward while for the latter punishment and negative reinforcement seemed work more efficiently. Testing, due to differences in information retrieval and using episodic and semantic memory, can produce considerably different results for introverts and extraverts, especially when conducted in different time intervals. Carrell et al. (1996) investigated the relationships between 76 learners’ personality types, established by means of MBTI (Myers and Briggs 1976), and their language achievement, measured by test results illustrating learners’ performance in the areas of reading comprehension, vocabulary, grammar, and writing, as well as providing learners’ overall assessment (final tests). Firstly, grammar test performance correlated negatively with the Judgement scale, and positively with the Perception scale. Secondly, introverts outperformed extraverts in vocabulary tests. Correlations were also calculated for personality types as bipolar dimensions; however, it was only the above relationship between two scales, namely Introversion and Vocabulary Test Scores, which proved to be statistically significant. The problems with significance may have resulted from other factors, such as uneven distributions of gender (women considerably outnumbered men in the study and, at the same time, types varied significantly across genders), personality types (introverts were outnumbered by extraverts by 12%, but all the other categories showed greater discrepancies, the greatest of which amounted to 86% - in the case of the difference in the numbers of Sensing and Intuitive students tested). Also, different results for the continuous and rank mode of using MBTI scores could signal an overall lack of

47 validity of the tool, confirmed by changes in types for 5 out of 21 students chosen to test the reliability by means of the test-retest method. Onwuegbuzie et al. (1997) studied the factors which might underlie language learning anxiety, understood “a stable personality trait, among experienced language learners” (MacIntyre and Gardner 1991: 297). In a study of 210 students participating in tertiary-level foreign language courses, a battery of tests was used to investigate the factors which could induce language learning anxiety, including the Foreign Language Classroom Anxiety Scale (Horwitz et al. 1986), the Self-Perception Profile for College Students (Neeman and Harter 1986), the Social Interdependence Scale (Johnson and No- rem-Hebeisen 1979), the Academic Locus of Control Scale (Trice 1985), the Study Habits Inventory (Jones and Slate 1992), and a background demographic form. Using stepwise multiple regression, the authors of the study established that students’ gender, age, academic achievement, semester course load, travelling abroad experience, as well as prior learning experience, anticipated performance in foreign language class, learners’ self-perception, cooperativeness, and academic locus of control were predictive of more than 40% of variation of foreign language anxiety. The results of the study indicated that foreign language learning anxiety increased linearly as the length of studying at university increased, whereas no link was observed between anxiety levels and learners’ proficiency. In another study which addressed anxiety and other learner IDs, such as language aptitude, attitudes and motivation, integrativeness, field dependence and independence, language learning strategies, and learners’ self-confidence, Gardner et al. (1997) used factor analysis in order to identify the major correlates of language learning success. Again, a complex battery of self-reporting questionnaires was employed in order to obtain measures on 34 variables from 102 participants. It turned out that learners who had a high level of self-confidence, and lower level of anxiety, typically manifested a stronger desire to learn the foreign language. The level of anxiety also showed a significant, negative correlation with learners’ grades. The growing need to order and typify individual learner variables in foreign language learning was addressed in the study conducted by Onwuegbuzie et al. (2000), who considered the influence of cooperativeness, competitiveness, individualism, and locus of control on learners’ attainment. Personality variables were used as one of four groups of predictors of foreign language achievement in the study; however, the study also

48 incorporated measures of foreign language anxiety, perceived intellectual ability, perceived scholastic competence, and perceived self-worth, all four of which are classified as affective variables. The participants included 184 university students (mostly women), enrolled in foreign language courses in Spanish, French, German, or Japanese, who did not differ significantly from each other in terms of achievement. The battery of tests consisted of the Foreign Language Anxiety Scale (Horwitz 1986), the Self-Perception Profile (Neemann and Harter 1986), the Social Interdependence Scale (Johnson and Norem-He- beisen 1979), the Academic Locus of Control Scale (Trice 1985), and the Study Habits Inventory (Jones and Slate 1992). The analytical procedures involved calculating Pear- son’s product-moment coefficients, as well as using multiple regression in order to explain as much of achievement variance as possible. The results of the study indicated that two of variables investigated by the authors in the personality group, i.e., that is individualism, cooperativeness, significantly correlated with foreign language achievement; however, the strength of these correlations was weak. At the same time, 34% of the variance of language achievement was explained by five variables, including cooperativeness, classified as a personality variable. Although Onwuegbuzie et al. (2000) ensured proper use of statistical procedures, not defining personality as a construct, and referring to just a selection of variables resulted in lack of contingent results that could holistically account for the impact of personality on language learning. One of the first studies which examined the role of some personality factors in foreign language achievement within the framework of the trait approach to personality was conducted by Oya et al. (2004). The researchers investigated the links between personality and anxiety characteristics and how 73 Japanese learners of English performed in an oral test so as to account for the possible correlates of oral fluency, accuracy, complexity, as well as global impression of learners’ performance. Personality was considered as “the enduring, inner characteristics of individuals that organize their behavior” (Der- lega et al. 1991: 2). Oya et al. (2004) followed Eysenck’s expectations of an extravert to be sociable, fond of parties, excitement, spontaneous, and rather impulsive. Naturally, the possession of such characteristics was expected to exert an influence on the course of the language learning process. The inventories used in study included the Maudsley Person- ality Test (Jensen 1958), based on Eysenck’s trait approach (Eysenck 1959), the Spiel- berger State, the Trait Anxiety Inventory (Spielberger 1962), and a story-telling task, adapted from Wechsler’s Adult Intelligence Scale (Wechsler 1955). The performance on

49 the latter is evaluated in terms of fluency, accuracy, complexity, and global impression. On the one hand, no fluency, accuracy, or complexity measure significantly correlated with learners’ oral performance, but on the other hand, global impression of participants’ oral performance showed a significant, positive, moderate correlation with learners’ extraversion. Also, there was a significant, negative, weak correlation between learners’ state anxiety and clause rate accuracy. It was concluded that it is worth cultivating or encouraging those aspects of an extravert’s profile which could improve the overall impression of their oral performance (Oya et al. 2004: 850). Hurd (2006) investigated the personality self-perception of 500 learners of French in the specific context of distance learning at The Open University (UK). She assumed that distance learners should have a relatively high degree of self-regulation and attempted to link it with the personality traits reported by the learners. The instruments used in the research included two questionnaires and a set of interviews, both think-aloud verbal pro- tocols, and telephone interviews. The established rank of personality features which could facilitate distance language learning included 17 characteristics, the initial five of which (reported both in February and in June) included being motivated, enthusiastic, persistent, systematic, and self-confident. The set of the above adjectives was only slightly different in learners’ self-descriptions as language learners. While motivation, enthusiasm, and persistence continued to occupy positions 1-3, communicativeness and independent mindedness were reported as self-characteristics ranked 4 and 5 in the initial study. No- tably, learners did not pay considerable attention to the extra-/introversion distinction. In their thorough study of the effect of Extraversion on oral proficiency in a foreign language, Van Daele et al. (2006) described the relationships between 25 Dutch learners’ of French and English supertraits measured by the EPQ (Eysenck and Eysenck 1975). They chose the instrument because of their preference for including a single dimension of Psychoticism rather than dealing with Agreeableness and Conscientiousness separately, and their recognition of the reticulo-cortex arousal as the neurological explanation of Extroversion (Eysenck1967a). They also accepted Eysenck’s assumption that introverts outperform extraverts in formal settings because of greater mental concentration ability. However, by referring to an earlier study (Dewaele and Furnham 1999), they suggested that the area in which introverts perform better was explicit (learning) ability, while, conversely, they underperform in their use of communicative skills. Van Daele et al. (2006) were primarily interested in the effects of Extraversion on the level of oral

50 fluency, complexity and accuracy in learners L1 (Dutch), and in English, and French, learnt as foreign languages. It turned out that the measure of lexical complexity, Guiraud’s Index (Guiraud 1959), significantly correlated with the Extraversion variable - in other words, extraverts demonstrated a greater lexical complexity than introverts. At the same time, no correlations existed between the level of Extraversion and learners’ fluency, syntactic complexity or lexical and grammatical accuracy, which researchers attributed to the specificity of research settings (relaxing rather than stressful) and to the psychological characteristics of the investigated learners (relatively high average Extra- version scores). Ockey (2009) examined the effect of learners’ personalities on other test takers in a group oral exam, investigating the influence of one of the facets of Extraversion in the Five Factor Model of personality, namely, assertiveness. The study involved a group of 113 Japanese first-year students of English, whose personality measurement involved ad- ministering the NEO-Personality Inventory Revised (NEO-PI-R), Costa and McCrae 1992). Referring to Bachman’s (2002) views on oral test performance8, Ockey (2009) hypothesised that assessing candidates’ performance could be biased due to the personal characteristics of interactants9. According to the results of the study, assertive candidates received higher scores than expected once grouped with non-assertive candidates, and lower scores than expected once grouped with only assertive candidates. The researcher suggested that in group-oral testing “rater-training sessions should include guidance on how to evaluate a test taker in the context of the group in which the test taker is assessed and assign scores that are based not on a comparison of proficiencies of group members” (Ockey 2009: 161). Biedroń (2011) conducted an in-depth analysis of the relationships between the personality traits, measured by the Polish adaptation of NEO-FFI (Costa and McCrae 1992) (Zawadzki et al. 1998), and language learning aptitude, measured by The Modern Language Aptitude Test (MLAT), Carroll and Sapon, 2002), and the Language Ability Test (Test Zdolności Językowych, TZJ, Wojtowicz 2006). Her investigations of two groups of learners, 44 gifted, and 46 non-gifted students, cast new light on personality as a predictor of foreign language aptitude. On the one hand, no statistically significant differences in the levels of the ‘big’ five personality traits were detected between the two groups of participants, but on the other hand, a number of statistically significant correlations could be revealed between the level of personality traits, and the participants’

51 language aptitude in the group of 46 non-gifted learners. These correlations included a moderate relationship between the subcategory of TZJ referred to as Grammar and Vo- cabulary, and Openness to Experience and negative, moderate correlations between two MLAT constituents, number learning and paired associates, and the level of Neuroticism. Apart from that, multiple regression indicated that a number of personality traits were, in fact, valid predictors of foreign language aptitude. Openness to Experience was the strongest positive predictor, whereas Extraversion affected foreign language aptitude most negatively. The other personality traits exerted an influence on language aptitude with regard to specific components of MLAT and TZJ; it was positive for Conscientious- ness, and negative for Neuroticism. Hajimohammadi (2011) investigated the impact of Iranian female students’ personalities on their performance in a writing self-correction task. In a study of 120 pre- intermediate Iranian female EFL students, the researcher applied the Nelson English Lan- guage Test (NELT) (Fowler and Coe, 1977), and the Eysenck Personality Questionnaire (EPQ) (Eysenck and Eysenck 1975). The results of the study turned out to be significant with regard to comparisons across different types of correction, but no significant relationships between its efficiency and learners’ personality was revealed. Fayyaz (2011) investigated the relationships between Pakistani learners’ metacognitive listening skills and their personality traits in two phases, including a pilot study, and a main study using amended personality measurements. The researcher employed the Metacognitive Awareness Listening Questionnaire (MALQ, Vandergrift et al. 2006) and a Pakistani adaptation of NEO-FFI for personality measurement, aiming to provide insights into the affective domain of listening comprehension. Overall, total MALQ scores showed a significant negative correlation with Neuroticism scores while the relationship between the scores and Extroversion, Openness, and Conscientiousness showed significant positive relationships. In general, the study could be perceived as particularly useful in setting new directions of research in the area of personality variation in EFL learners. Khodadady and Zabihi (2011) explored the relationships between the school performance of 419 students majoring in English or Persian, their cultural and social background, as well as personality characteristics. Using their Grade Point Averages (GPAs), as well as various measures of learners’ social background, and the Persian adaptation of NEO-FFI (Costa and McCrae 1992). The results of the study indicated a significant difference in GPAs between English and Persian majors, as well as a significant difference

52 in income levels in their families. With regard to personality traits, openness to experience turned out to be a positive correlate of English learners’ GPAs, which could be attributed to the benefits of being more open to both the language and culture of the target language, but also to a wider repertoire of language learning strategies employed in the learning process. At the same time, the relationship between English learners’ Conscientiousness and their GPAs was not confirmed. It could be argued that the findings of the study provide support for the assumption that language learning is a specific area of education, in which it is openness to experience rather than Conscientiousness that can be viewed as a predictor of good school performance. Ockey (2011) re-examined the roles of self-consciousness and assertiveness in learners’ oral ability applying the latent variable approach. In his investigation of 360 Japanese first-year university students he used an adaptation of the NEO-PI-R (Costa and McCrae 1992) as well the L2 Group Oral Discussion Test, also employed in one of his previous studies (Ockey 2009). The researcher accepted Costa and McCrae’s (1992) definition of self-consciousness, according to which it can be understood as the degree to which individuals experience shame, embarrassment, and social anxiety once in a group of people. He also referred to the FFM creators’ notion of assertiveness (Costa and McCrae 1992), thus viewing assertive individuals as free of hesitation, and potential group leaders, also regarding conversation. Logically, assertiveness could be commonly believed to facilitate L2 learning and acquisition, whereas self-consciousness could be seen as an obstacle in L2 development. Importantly, both characteristics are primary facets in the FFM (Widiger 2017); however, self-consciousness refers to Neuroticism while assertiveness underlies Extraversion. The results of the study indicated that assertiveness was a significant explanatory variable of oral L2 performance; however, self-consciousness was not. Moreover, covariance analysis revealed that the subconstructs of L2 oral ability were also explained by assertiveness to some degree, (ẞ coefficients ranged from 0.13 for grammar to 0.18 for communication skills and fluency). Consequently, one can infer that non-assertive Japanese learners could be at a disadvantage in their attempts to learn English. Ożańska-Ponikwia and Dewaele (2012) investigated the link between immigrants’ personality traits and their L2 use. Their study, conducted on a group of 102 Polish adults, indicated that openness to experience and self-esteem constituted two significant predictors of the perceived L2 achievement. In fact, openness to experience was found to

53 explain more than 7% of the variance in self-reported L2 use. At the same time, participants’ self-perceived L2 proficiency positively correlated with their Agreeableness, Openness to Experience, and empathy, which is one of the characteristics underlying emotional intelligence. Although the settings of the study were not classroom-based, it is worth mentioning that again the vital role of Openness to Experience in learning a foreign language was confirmed. Ramírez-Esparza et al. (2012) investigated 700 immigrants’ socio-interactive behaviours and how they corresponded to their extrovert/introvert personalities, conceptualised as a set of classroom behaviours. Both concepts were also related to learners’ educational background, since learners had been split into two subcategories, that is low- education, and high-education learners. Learners’ socio-interactive behaviours were measured in five category clusters, including learner’s language learning tasks, resources, engagement, personality (only regarding Extra-/Introversion), and interaction. The results of the study proved that low-education learners displayed more introvert behaviours, and also, achieved significantly lower test scores. In particular, Introversion showed a significant, positive correlation with taking up the novice role, which was not the case once years of studying were taken into account (Ramírez-Esparza et al. 2012:566). Oz (2014) examined the relationships between pre-service 168 EFL teachers’ Big Five personality traits and their willingness to communicate, understood as “readiness to enter into discourse at a particular time with a specific person or persons, using an L2” (MacIntyre et al. 1998: 547). Two instruments were used in the study, i.e., the Willingness to Communicate Scale (WTCS) (McCroskey and Richmond 1987), and the International Personality Item Pool (IPIP) (Goldberg et al. 2006). The results of the study indicated a moderate correlation between the desire to talk to strangers and learners’ self-reported Extroversion, as well as statistically significant, positive correlations between Extrover- sion (a strong relationship), Agreeableness (a moderate relationship), and Openness to experience (a slight relationship) and WTC. Also, Agreeableness was found to moderately correlate with educational achievement. The results of the study differed significantly for male and female participants for two personality dimensions, that is Emotional Stability and Agreeableness. It could be inferred from the study that students who are open, sociable, and person-oriented, as well as friendly, curious, and creative, are more likely to communicate in English.

54 One of the most recent trends in second language acquisition and learning studies deals with the incorporation of social websites and virtual reality in L2 learning. Kao and Craigie (2014) investigated the impact of 164 Taiwanese learners’ personality traits and English usage on Facebook (EUF) on learners’ EFL achievement using an adaptation of the Big Five Inventory, the English Usage on Facebook Inventory for Language Learning (EUFILL, a self-created tool), and a General English achievement test. Extraversion turned out to moderately correlate with EUF, Conscientiousness showed a weak relationship with it, while for Neuroticism the relationship was weak and negative. Neurotics also performed significantly worse on the achievement test. Pourfeiz (2015) examined the relationships between 157 Turkish students’ of English self-reported global personality traits, measured by the IPIP (Goldberg 2006), and their attitudes towards foreign language learning, assessed by means of the Attitudes towards Foreign Language Learning inventory (A-FLL) (Vandewaetere and Desmet 2009). The researcher approached personality in a general manner, describing it as “a complex of all the attributes that characterize a unique individual” (Pourfeiz 2015: 467); however, his choice of the personality inventory indicated his preference for the trait approach. Not only did statistically significant, positive correlations exist between the ‘big’ five personality traits and attitudes towards foreign language learning, but also significant relationships were detected between the cognitive, affective/evaluative, and behavioural, and personality components of A-FLL and the five personality traits. Moreover, Open- ness to Experience was the strongest predictor of the cognitive component, Emotional Stability, Agreeableness and Openness to Experience accounted for a considerable proportion of variance of the behavioural/personality component. Finally, Conscientious- ness, Agreeableness, Emotional Stability, and Openness to Experience were all significant predictors of the affective/evaluative component - in particular, the lower the level of Emotional Stability was, the more challenging it became to feel positive about learning English as a foreign language, and the greater risk of language anxiety. The results of a qualitative study conducted by Coker and Mihai (2017) on a group of 10 learners of English studying in the USA are certainly an argument in favour of including personality not merely in investigating it as a variable in second/foreign language development, but also of familiarising language learners with its role in the language learning process. Specifically, the researchers pointed out that the investigated learners benefitted from being familiarised with information about their personalities

55 measured by the Enneagram (Daniels and Price 2009). The researchers expressed the belief that expanded person knowledge helped the language learners understand their reactions in different learning situations, but also equipped them with the ability to improve their study habits. It was also found that personality determined how participants interacted with each other and with their teacher. The Type 1s (Perfectionists) stayed focused on the goal, but also preferred secure situations and structured activities. The Type 2s (Helpers) were eager participants in group activities and enjoyed helping others. The Type 3 (an Achiever) pursued their aim to succeed in learning the language, but also demonstrated leadership skills. The Type 8 (a Leader) felt responsible for other members of the group and disapproved of any manifestations of laziness of other group members. In conclusion, the researchers emphasised that it should be vital for foreign language teachers to create an environment where each student could actually benefit from their personal characteristics. In a study of over 500 Polish adolescents and young adults, Piechurska-Kuciel (2018) investigated the role of Openness to Experience in learners’ willingness to communicate (WTC) in a foreign language. The level of the trait was measured by a 20-item constituent scale of the IPIP (Goldberg 1990) while learners’ WTC was measured by the Willingness to Communicate in the Classroom scale (MacIntyre et al. 2001), consisting of four scales dedicated to specific language skills. The results of the study confirmed the hypothesised positive relationship between both constructs, and a negative relationship between the level of openness to experience and the level of language anxiety, as indicated by the employed regression model. Also, the participants who were relatively more open, got significantly better final grades in EFL courses. It was also concluded that students relatively less open to experience might encounter more difficulties attempting to tackle linguistic idiosyncracies, communicative shifts, and ambiguity. Although some claims have been made that the area of second language acquisition and learning has failed to respond to the developments in personality psychology and it has been pointed out that a vast number of studies into the role of personality in second language development have only denoted weak or insignificant results, it is also commonly acknowledged that the importance of personality in learning a foreign language cannot be denied (Dörnyei and Ryan 2015: 33). Perhaps applied linguists are yet to recognise the potential of more complex models of human personality, such as “The New Big Five”, which takes into accounts not only learners’ dispositional traits, but also allows

56 for the general factor of human nature, the characteristic adaptations of the traits that individuals are equipped with, and the roles of integrative life narratives, as well as culture, understood as meaning systems and practices (McAdams and Pals 2006: 213). On the one hand, personality studies employing the MBTI (Myers and Mc Caulley 1985) as a research tool to investigate learners’ personal characteristics seem at least partly biased due to the necessity for participants to make choices between extremes. On the basis of the empirical evidence, conclusions can be made with reference to the role of the ‘big’ five personality traits in educational achievement (Dörnyei and Ryan 2015: 23-31): Conscientiousness and Openness to Experience appear to be the two best predictors of academic achievement. The impact of the latter trait is far from simple as it can facilitate critical thinking, but at the same time impede the ‘required’ acquisition of knowledge. Extraversion has been found to correlate negatively with attainment measures because of its inherent distractibility. Neuroticism is found to impede academic achievement since, in a sense, it includes anxiety.

1.5. Conclusion

Personality has been thoroughly investigated since the beginning of the previous century, and a considerable body of evidence suggests that it is a construct whose constituents can be identified from different angles. It might appear that contradictions in understanding the construct exist even within a given approach. For example, according to Allport (1961: 28), it is dynamic, whereas it is reported by Larsen and Buss (2013) to be organised and relatively enduring. On the other hand, researchers seem to agree that personality can be seen as an intrinsic psychological structure, described by adjectives referring to personality traits centred around a certain limited number of dimensions (Piechurska-Kuciel 2018: 192). The extent to which personality determines people’s actions is debatable; however, a considerable proportion of researchers believe that it underlies human behaviour, either in a deterministic manner, or, at least, by interacting with specific conditions (Ellis et al. 2009: 3).

57 It is logical to expect personality to affect virtually every domain of human existence, including learning styles, academic achievement (Komarrju et al. 2011), and language learning strategies (Oxford 1986, 1990, 2011, 2017). Moreover, learners’ personalities are believed to interact with each other within each individual learner’s microsystem (Oxford 2017: 107). As is the case of any other complex construct, scholars need a framework of investigation to approach personality systematically and account for its measurement. Over the decades, the psychodynamic approach, the behavioural approach, the humanistic/phenomenological approach, and the trait approach, have shaped the contemporary perception of personality. Within these approaches, numerous models have been developed so as to facilitate personality assessment. The advantages of the FFM (Costa and McCrae 2017) consist in its relative simplicity, and its capacity to describe individual, regardless of language or culture (Piechurska-Kuciel 2018). It does not diminish the role of situational variables in predicting human behaviour, either, since it was not designed to comply with just the trait approach or any other theoretical perspective (McCrae and Costa 2008: 159). On the other hand, it should be stressed that the use of the model, and particularly the NEO-FFI questionnaire (Costa and McCrae 1992), may produce a bias towards analysing personality in terms of its higher-order constituents rather than lower-order traits, and result in neglecting the role of such phenomena as language anxiety (Dörnyei and Ryan 2015: 32). However, the FFM model was constructed with due scientific scrutiny recognised by a vast body of researchers (cf. Widiger et al. 2017) and even those who manifest a certain dose of scepticism towards it acknowledge that it was developed in a psychometrically appropriate manner (Dörnyei and Ryan 2015: 19). It enables cross-cultural and longitudinal comparisons and recognises the genetic foundation of personality traits (Eysenck 1990; Ellis 2009). Its universal character has been confirmed across continents,10 and its popularity has expanded knowledge of the determinants of human behaviour in various areas of life, including foreign language development. Finally, the application of the instruments developed within the model (NEO- FFI and NEO-PI-R, Costa and McCrae 1992), is very likely to trigger reflection on the part of the learners. The author of the present dissertation strongly believes that such reflection is likely to increase learners’ person knowledge, which involves understanding one’s own strengths and weaknesses (Oxford 2011: 19), and, in effect, make learners more active, better self-regulated, and more positive about language learning. This, in turn, enhance creativity and facilitate the learning process (MacIntyre et al. 2016: 22).

58 The motivation behind the interest in learners’ personality traits is thus to a large extent generated by the conviction that self-awareness is a very desirable quality in learning a foreign language.

59 Chapter 2: The concept of language learning strategies

2.1. Introduction

Over forty years of investigating language learning strategies (LLS) has not merely expanded the knowledge about learners’ characteristics, thoughts, behaviours, and the desirable practices, which especially good language learners apply, but also made it more complex to adopt a view of what language learning strategies are, and how to best define or classify them. Moreover, some researchers have expressed reservations concerning the very notion of language learning strategies, pointing to their allegedly insufficient theoretical framework or postulating alternative ways of accounting for learners’ actions taken in order to learn (and acquire) a foreign language (Dörnyei 2005; Dörnyei and Ryan 2015). On the other hand, Ellis (1994: 473) includes LLS as one of three foundations of the framework for investigating individual learner differences and Droździał-Szelest (1997: 23) sees language learning strategies as a valid explanatory variable in investigating second language achievement, pointing to their relevance in defining the “previously unaccounted for variance”, which is certainly an argument for including LLS as one of individual differences (IDs) in empirical inquiries. Despite some the methodological shortcomings and theoretical discrepancies, a number of strategy researchers have persisted in conducting such empirical investigations, perhaps feeling obliged to strive to support learners and teachers with expert practical guidance (Gu 2012: 330). In an attempt to reflect on the progress made in strategy research, Oxford (2014) interviewed other strategy researchers and identified key themes in their reflections on investigating LLS. They included the focus on psychological aspects of language learners, including, first of all, their self-regulation, but also recognised the role of strategies in task accomplishment and self-reflection. Adopting a research paradigm seems to pose another major difficulty, considering the above areas of investigation, and a multitude of instruments that have been used in LLS studies. Moreover, attempts to reconcile qualitative and quantitative methods may prove challenging. While the former could potentially trigger more opportunities for learners’ reflections, they could, on the other hand, make it impossible to generalise research outcomes due to a limited, often unrepresentative number of participants. The

60 latter, which are more likely to ensure a proper number of participants, may at the same time entail the dangers of misinterpreting numeric values from questionnaires, or neglecting the idiosyncratic aspects of language learning. A number of studies suffer from limitations due to reliance on only one type of data, i.e., either qualitative, or quantitative. One way of overcoming this deficiency is applying a multi-approach and data triangulation, as suggested by Droździał-Szelest (1997: 113). Chapter Two of the present thesis attempts to address the afore-mentioned difficulties in accounting for the use LLS. It opens with a historical overview of approaches that aims to provide a broad context for analysing LLS. By accounting for the scope of findings and perspectives provided in the discussed studies, it is also intended as an argument for the legitimacy of the role of LLS in foreign language development. An overview of definitions of LLS is provided in section two of the present chapter. The third section of Chapter Two includes a review of taxonomies, classifications, and categorisations of LLS. It also briefly discusses the assumptions of Oxford’s (2017) criticism of strategy groupings. Finally, its last section addresses LLS through the prism of self-regulation (SR) and thus expands the macrocontext of strategy use. Through adopting an amalgamative perspective (Oxford 2017), it recognises the importance of SR and attempts to account for the construct through reviewing its psychological foundations, educational implications in general, and, finally, focusing on recent insights into self-regulated language learning (SRLL). The review of relevant literature involves providing the theoretical background of SR, SRL, and SRLL, and examples of empirical studies. The final part of the chapter is dedicated to the discussion over the future of research into LLS, including recent insights from Pawlak and Oxford (2018) and Rose (2018). An attempt is made to illustrate the current directions in strategy investigations with special emphasis on the developments after the introduction of the SRLL concept by Dörnyei (2005). Lastly, conclusions are summarised in the final section of the present chapter with regard to the trends and potential developments in strategy research in the future.

2.2. Historical perspective

The shift in the focus of EFL researchers from attempting to produce a perfect method of learning and teaching a foreign language to capitalising on learners’ individual

61 characteristics, which began in 1970s, was not merely a result of the increasing popularity of the principles of humanistic teaching, but also a logical stage after recognising the need for methodological guidance leading to learners’ self-direction. Researchers, such as Ellis and Sinclair (1989), viewed learner training as an inevitable component of language learning, while authors of EFL coursebooks, including the first coursebooks based on functional syllabi in the 1980s, incorporated language learning strategies as part of their courses. Abbs and Freebairn (1977, 1979, 1980, 1982a, 1982b) even decided to include the word ‘strategy’ in the titles of their coursebooks, and depicted the learning process sequence by naming the following parts of their language course Starting strategies, replaced by Opening strategies, Building strategies, Developing strategies, and Studying strategies. At the same time, empirical investigations (Bialystok and Fröhlich 1977, O’Malley and Chamot 1985, Oxford 1986, Oxford 1989) provided the foundations for the creation of paradigms for accounting for LLS.

2.2.1. Early landmarks in strategy investigations

In one of the pioneering works dedicated to the psycholinguistic aspects of foreign language development, Jakobovits (1970) expressed several beliefs about LLS; however, he did not propose a definition of the construct. Instead, he speculated that LLS could be both taught and learnt and suggested that relevant instruction could be based on the clues derived from the study of predictors of language learning success, referring to Carroll’s (1962, 1965) studies. Although Jakobovits’ approach to strategies was, to some extent, intuitive, he did express concerns about potential problems to be encountered while addressing some of Carroll’s attributes of good language learners and designing appropriate strategies. For example, Jakobovits (1970: 255) expressed criticism about a potential single strategy that would facilitate the development of phonemic coding ability and facilitate storing phonetic material in memory. The researcher also postulated that LLS be at least partly independent of teachers’ instructions and their use could potentially affect the efficiency of teaching methods. In other words, the need to account for strategies in learners’ environment was advocated. Also, a distinction was made between assimilatory and separative strategy use, which could be essential in considering the role of attitudinal factors affecting language transfer. Learners could thus either resort to their mother tongue

62 as a kind of reference structure (assimilatory strategies, potentially more likely if more similarity exists between L1 and L2) or rely on separative strategies if they felt that L1 and L2 were considerably different in a given aspect. According to Jakobovits (1970: 210), the use of either of these strategies would then only be altered, to a certain degree, by the teaching situation, and thus constitute an area of “individual differences in learning strategy”. It was also in the early 1970s hat Selinker (1972) decided to use the learning perspective when addressing the linguistic aspects of the psychology of second language learning and developed the theory of Interlanguage (IL). Considering the phenomena, which he viewed as the underlying elements of attempting meaningful performance in the target language (TL) by the learner, he listed language learning strategies and communication strategies as processes of vital importance to second language learning. Selinker (1972: 37) described LLS as “an identifiable approach by the learner to the material to be learned” and linked them to the culture of the learner’s country of origin, or, more specifically, their L1 country. He also presumed that strategies were consciously employed once a learner had realised that a particular aspect of L2 competence was missing and concluded that LLS affected “the surface structure of sentences underlying IL utterances” (1972: 10-11). For instance, a learner might use a strategy which would simplify the target language (such as using verbs in all aspects, assuming that all verbs can be used in all aspects), or a strategy consisting in looking for analogies and guessing (“copy the cue”). Also, a distinction was made between widespread LLS (e.g. cue-copying), and culturally- determined LLS (e.g. chanting). Additionally, Selinker et al. (1975) identified language transfer, overgeneralisation, and simplification as three central language learning strategies. Earlier, the notion of inference had been brought forward by Carton (1966), who viewed it as learners’ responses, relevant to prior learning history of the responding individual, involving an encounter with a novel attribute in a specific context, requiring conceptualisation. According to Carton (1966), prior experience is linked with the novel stimulus through the process of conceptualisation so that response can be formed. Carton (1966: 63-64) specifically described inferencing as “the aggregate of all behaviours calculated to achieve a major educational objective”. Inferencing was attributed a number of features shared by language learning strategies, such as the possibility of being employed outside the language classroom, being beneficial, if not inevitable, for learner’s

63 autonomy, and being subject to considerable variation among language learners. In the following decade, Bialystok and Fröhlich (1978) also referred to inferencing, which, along with other strategies, they considered as one of four factors that could influence language learning achievement. Importantly, the researchers reflected that the use of strategies could depend on both cognitive and affective variables. Apart from inferencing, they investigated monitoring and practising, which they regarded as any form of exercising the target language, and thus the most commonly used strategy. According to Bi- alystok and Fröhlich (1978: 329), monitoring could be described as “the conscious attention to the language code”, and hence, it could be argued that inferencing consists in attempts to derive meaning in spite of encountering unfamiliar linguistic items. They also suggested that the use of these strategies be modified by purpose and modality (written or spoken use of language), and purpose (formal practice or act of meaningful communication). In a study of lower-level high school learners of French, the researchers applied a self-devised questionnaire in order to quantify the use of the above strategies. At the same time, achievement was measured by means of a formal oral task, a formal written test, and a functional test. The statistically significant findings included a strong relationship between strategy use and correlations between achievement and both, aptitude, and strategy use.

2.2.2. Good language learner studies

With the onset of the humanistic approach, efforts to identify the characteristics which make some language learners more successful than others became more intense. For one thing, a number of researchers believed that expanding the scope of knowledge about learners’ inherent characteristics could contribute to a more harmonious development of the learner, and for another thing, it was expected that at least a part of the strategic repertoire of good language learners could be tailored to the specific needs and abilities of less successful learners. A number of researchers consider Rubin’s (1975) paper on good language learners to be “the initial spark for the language learning strategy field” (Oxford et al. 2014b: 31). According to Rubin (1975), good language learners possess the following characteristics: not being afraid to figure out meanings;

64 skilful management of information, communicative focus in language development; active search for opportunities to apply new knowledge; willingness to integrate with other speakers of the target language; ability to maintain motivation in the long run; drive to communicate despite language shortcomings; lack of inhibitions; attention to forms and patterns; awareness of connections between meaning and structure.

In another pioneering analysis of good language learners, Stern (1975) placed special emphasis on the ways in which learners tackled the three problems that they encountered as beginners, that is the disparity of language systems within the new target language, the “code-communication dilemma”, understood as the way in which a message is communicated by means of the imperfect linguistic system, and, finally, the choice on the part of the learner between adopting a more intuitive, or a more “rational” way of learning the target language. The researcher expressed a belief that in order to address each of the three dilemmas, language learners need to employ LLS. In spite of assuming that the language learning process should involve both conscious use of language learning strategies, and unconscious mental processes, Stern (1975) hypothesised that it was possible to isolate specific techniques or strategies in a process of prolonged observation. However; it turned out that drawing up a list of LLS to be observed might pose a serious problem, and observation might not be the best technique to employ in order to determine good language learners’ strategies. A number of researchers referred to the legitimacy of the paradigm to investigate good language learners’ strategies. For instance, Cook (2008) suggested it would be beneficial for language teachers and researchers to exploit the strategies used by GLL, and make students more independent through presenting them with a wide range of strategies, and training them in selected strategies. At the same time, he pointed to the necessity to bear in mind both the similarities and differences between content subject and foreign language learning strategies (Cook 2008: 119). Perhaps it is also needs to be pointed out that, according to Griffiths (2008), no single, universal set of good language learner strategies exists, since strategy use is influenced by a number of individual variables, and, at the same time, certain language learning strategies could be applied in order to develop a

65 number of specific language skills. Finally, Lesiak-Bielawska (2013: 29-30) observes that poor language learners’ strategies had received relatively little attention in studies of second language acquisition and learning and pointed out that some beliefs in this area might actually be speculative. At the same time, she supports Abraham and Vann’s (1987) opin- ion that ‘poor’ language learners are not fully able to reflect on the LLS which they apply or, perhaps, to word their thoughts. This, in turn, constitutes a major obstacle in strategy investigations, since a large number of studies rely on learners’ introspection. The relationships between LLS, learners’ level of proficiency, and achievement are analysed in section four of the present chapter, while methodological issues are addressed in Chapter Four.

2.3. Defining LLS

Interest in the already mentioned concepts, such as inferencing (Carton 1966, 1971), or Interlanguage (Selinker 1972), and noticing the need to distinguish between communication strategies and learning strategies by Tarone (1981, 1983) gave rise to definitional efforts to account for language learning strategies. In retrospect, it seems that consensus over defining LLS has not been reached, and efforts to fully account for the concept are still being made, as signalled by Oxford (2017). The major issues which underlie the lack of unanimity in addressing the notion of LLS include: a) the legitimacy of LLS as an ID7 and their uniqueness against ordinary learning activities, as expressed by Dörneyi and Ryan (2015), and b) the location (mental or behavioural domain), size, degree of abstractness, and contextual setting of LLS, as indicated by Macaro (2006). Focusing on the latter aspects, Macaro (2006) referred to the relationships between strategies (thus recognising the complexity of strategy chains or clusters), the compliance of LLS with individual learners’ goals, and use of LLS to deal with specific tasks. The following subsection offers an overview of definitions which aims to provide an outline of how researchers have discussed the strategy concept over the past decades with particular

7 Dörneyi and Ryan pointed to the ambiguity of strategies by indicating that a number of definitions, such as that by Cohen (1998: 4) describe strategies as features of the learning process rather than an attribute of the language learner.

66 emphasis on the indicated problematic areas. The following subsection aim to describe the evolution of the strategy concept and account for the status quo in strategy studies.

2.2.1. Early definitions of LLS

In the early years of interest in LLS, Rubin (1975: 43) described the construct as “the techniques or devices which a learner may use to acquire knowledge”, while Naiman et al. (1978: 4) regarded LLS as “more or less deliberate approaches, and more specific techniques, i.e., observable forms of language learning behaviour”. Finally, Tarone (1983: 67) viewed them as attempts on the part of the learner “to develop linguistic and sociolinguistic competence in the target language - to incorporate these into one’s interlanguage competence”. A somewhat more complex account of learning strategies in general was introduced by Carver (1984: 124-125), who defined them as “the overt or covert behaviour, conscious or unconscious”, which he believed to stem from “higher level categories”, such as learning styles, understood as “learners preferences for ways of organising his learning”. The researcher assumed that strategies derive from studying, or, more generally, working habits, and plans, and comprise a statement of objectives along with the assumed time span, list of necessary materials, and techniques, and, finally, some methods of evaluation. He also suggested that learning styles may be modified by both learner’s personality and situational factors, which, given their relation to LLS, would imply that learners’ personality exerts influence on strategic choices. Commenting on Carver’s views, it could be argued that much as it is necessary to recognise the need to see language learning strategies in a broader context, the assumption that learning strategies are limited to behaviours appears to be a simplification, neglecting a number of dimensions of a human person, such as affective variables or metacognition. LLS researchers frequently relied on the findings of educational psychologists. For example, O’Malley et al. (1985: 23) referred to Rigney’s (1978) and Dansereau’s (1985) explanations of learning strategies, and consequently, defined them as “any set of operations or steps used by a learner that will facilitate the acquisition, storage, retrieval, or use of information”. They also lent support to Bialystok’s (1983: 101) definition, according to which, LLS are “activities in which the learner may engage for the purpose of improving target language competence”. Chamot and O’Malley (1986) continued to

67 advocate for the cognitive foundation of LLS, claiming conscious and at the deliberate use of strategies constitutes declarative knowledge, which can become procedural knowledge through practice, but then it ceases to be to strategic and becomes autonomous. For a number of applied linguists and language educators, Rebecca Oxford remains the unquestionable figure of researching LLS. Originally employed by the U.S. Army Research Institute for the Behavioral and Social Sciences, Oxford stands out as the researcher who has introduced a great deal of scientific scrutiny and set the ground for psychometric investigations of LLS. Although Oxford’s views have evolved as her following research findings appear, her belief in the benefits of studying LLS and expectations of the potential of LLS as a predictor of foreign language performance and proficiency, as well as skill development has persisted. In one of her earliest papers on LLS, Oxford (1986) referred to Weinstein and Rogers’s (1985: 3) definition of learning strategies, according to which learning strategies are “cognitions or behaviours that a learner engages in during learning that are intended to influence the encoding process so as to facilitate the acquisition, retention, and retrieval of new knowledge. Moreover, she suggested that language learning strategies could be juxtaposed with teaching techniques and instructional strategies, the former ones depending on the learner’s choice and therefore being vital to learner autonomy. She also described LLS as “highly individualized and personalized” (Oxford 1986: 2) and recognised the necessity to account for strategies in context, in their relation to other IDs. Consequently, she also hypothesised that the use of strategies depended on a number of factors, including the level of the target language competence, the purpose of learning L2, and, obviously, the idiosyncrasies of L2. Wenden (1986) provided a number of more detailed descriptions of LLS. First, she identified language learner strategies as “language learning behaviours learners actively engage in to learn and regulate the learning of a second language” (1986: 6), thus classifying strategies as behaviours rather than mental actions, yet recognising the component of volition (learners acting as agents). She suggested that learning strategies involve strategic knowledge, which can be investigated when learners are asked to reflect on their learning process. Wenden (1986) also pointed out that efficient strategy use requires a certain amount of awareness of personal factors which facilitate and impede learning a foreign language, and depends on the recognition of one’s own competence as a language user/ Another description of LLS provided by Wenden states that they are “steps or mental operations used in learning or problem solving that require direct

68 analysis, transformation, or synthesis of learning materials in order to store, retrieve, and use knowledge” (Wenden 1986: 10). After studying 34 adult learners of English in semi- structured interviews, Wenden concluded that the kinds of strategies that learners use and favour reflect their beliefs and asserted that raising learners’ awareness of LLS can help them diagnose their language problems, reflect on achievement, and even handle feelings that accompany learning English as a foreign language. Learners should therefore be empowered to become in charge of their learning process. A considerable number of researchers were, at least initially, in favour of setting LLS in the cognitive domain of human activity. Chaudron (1988: 109) described LLS as “cognitive operations that learners apply both in the classroom and in other learning situations” and pointed to the paradox of the two settings. He observed that on the one hand, studies of LLS are essential for the understanding of the learner and equipping them with the right repertoire of tools to acquire knowledge and skills in L2, but on the other hand, they remain a phenomenon that is relatively difficult to investigate in the classroom. In- troduced by Chamot and O’Malley (1986), the Cognitive Academic Language Learning Approach incorporated strategy use as a means to help learners understand and remember important concepts. According to the definition provided by Chamot and O’Malley (1986: 15), LLS could be referred to as “conscious processes and techniques that facilitate the acquisition and retention of new skills and concepts”. Their application was thus likely to facilitate the learning process with reference to both, language, and content learning. The definitions of LLS coined by other researchers, such as Tarone (1981, 1983) and Bialystok (1983), emphasised the role of LLS in the communication process, and signalled the role of strategies in developing communicative competence. Tarone (1983: 67) considered LLS to be attempts on the part of the learner “to develop linguistic and sociolinguistic competence in the target language - to incorporate these into one’s interlanguage competence”. The researcher pointed to the semantic proximity of the two notions of language strategies and communication strategies, which could be distinguished from each other on the basis of the primary purpose of their use. If the latter consisted in communicating a meaning to the listener, a strategy should be classified as communication strategy; however, the distinction could become problematic if the motivation of the learner (“to learn or to communicate”) was unknown. Another attempt to distinguish between language learning strategies and communication strategies was made by Bialystok (1983), who actually claimed that a particular strategy could be used for different

69 purposes, and be characterised by different degree of control, and a certain level of consciousness. The researcher hypothesised that perhaps the origin of a given strategy could be used in differentiating between the two types of strategies, that is those more intrinsic could be classified as language learning strategies, and those more attached to the language could be considered as communication strategies. Perhaps referring to the definitions of communicative competence by Hymes (1972) and Canale and Swain (1980), it is possible to assume that the same tactic or learning behaviour can serve two purposes and thus be perceived as both learning and communication strategy. For instance, if a learner uses any type of circumlocution in interaction with another speaker of the target language, it is difficult to classify the circumlocution as either a learning strategy or a communication strategy. Similarly, if a learner enquires about an unknown word or expression and their interlocutor clarifies it, the situation could be interpreted as both language learning and spontaneous language use.

2.2.2. The evolution of the concept in the 1990s and strategy features

Oxford’s (1990) book Language learning strategies: What every teacher should know was a milestone in the study of LLS since it sets out a thorough framework for investigating them as an ID in a scientific manner. The work is deeply rooted in the principles of humanistic psychology, which is reflected in the references made by Oxford (1990:1- 2) to language learning strategies as “steps taken by the students to enhance their own learning”, “tools for active, self-directed involvement, which is essential for developing communicative competence”, and “the way students learn a wide range of subjects”. The echoes of the Communicative Approach (CA) are present in the main assumptions of the concept of language learning strategies, but it would be an incomplete picture to merely resort to the main ideas of CA in order to justify the framework for investigating LLS created by Oxford. First of all, Nyikos and Oxford (1993: 11) suggested expanding the strategic component of communicative competence, included in Canale and Swain’s (1980) model. Nyikos and Oxford (1993: 11) also pointed out that strategic competence, understood as the competence in using language learning strategies, “fosters competence in grammatical, discourse, sociolinguistic, and psycholinguistic areas”, both in classroom and naturalistic settings. Moreover, Oxford (1990: 5) also clarified that the term “input”

70 by explaining its relevance not merely to the language system that the learner is exposed to, but also to the “a variety of student and teacher characteristics, such as intelligence, sex, personality, general learning or teaching style, previous experience, motivation, attitudes, and so on”, along with a number of societal and institutional factors. All the above considerations are reflected in another Oxford’s (1990: 8) definition of LLS, according to which “learning strategies are specific actions taken by the learner to make learning easier, faster, more enjoyable, more self-directed, more effective, and more transferrable to new situations”. From the early years of interest in LLS, researchers have advocated the characteristics which facilitate their understanding as a scientific construct. These, according to Oxford and her co-workers (Oxford 1990; Nyikos and Oxford 1993) include: contribution to communicative competence and learners’ self-direction as the underlying goals of their training and use; expansion of teacher’s role in foreign language instruction as a requirement for their introduction and training; orientation towards problem-solving; learners’ agency; holistic treatment of the learner’s person, i.e., inclusion of various dimensions of language learning, not merely connected with cognition; variation in the scope and level of influence on learner’s performance and achievement (direct vs. indirect); variation in their visibility to the observer; multidimensionality, resulting from the fact that learning as such is multidimensional; synergism, which manifests in the advantages of using strategy chains; scientifically confirmed demonstrability. A number of the above characteristics were recognised by Ellis (1994: 532-533), including the use of reference to both general approaches and specific actions, problem orientation, awareness of LLS, the mental and behavioural dimension of LLS, and the variation in strategy use resulting to the variety of language learning tasks. At the same time, Ellis (1994) enumerated the following additional features of LLS: involvement of both linguistic and non-linguistic behaviour; potential for linguistic strategy use in both L1 and L2;

71 indirect and direct contribution of LLS to the learning process, consisting in either providing language learners with the data on L2, which they can process, or addressing specific areas of L2 competence). In an early paper, Cohen (1995) addressed the issue of conscious strategy use by referring to Schmidt’s (1994) distinction between focal and peripheral attention and concluded that if a learning behaviour was totally unconscious, that is if the learner couldn’t answer the question what they had been just doing, their behaviour should be treated as a process rather than a strategy. He also referred to Ellis’s claim (1994) that strategies that become proceduralised may not be accessible for verbal description by the learners. Co- hen (1995) specifically linked using LLS with the completion of certain tasks, which a well-functioning repertoire can foster. Unlike Oxford, Cohen (1995, 1998, 2003, 2014) drew a distinction between language learning strategies and language use strategies. Moreover, referring to the latter, he further distinguished between language performance strategies and communication strategies, proposing that “language performance strategies include cognitive processing strategies, strategies for solidifying newly acquired language patterns (...), and strategies for determining the amount of cognitive energy to ex- pend” (Cohen 1995: 5). Briefly referring to communication strategies, he considered them to be strategies that are used in communicating messages. At the same time, he acknowledged the existence of some overlap between the above categories.

2.2.3. Mapping the scope of LLS in the 21st century

After the dissonance in understanding the strategy concept became more evident, some researchers attempted to name the major problem areas in accounting for LLS (Cohen 1998, 2003; Griffiths 2004, 2013; Dörnyei 2005; Macaro 2006; Rubin 2008, Liang 2009; Pawlak 2011; Gu 2012; Griffiths and Oxford 2014; Oxford 2017, Pawlak and Oxford 2018). These areas could be summarised under the following headings: the conscious/subconscious dilemma of strategy use; the features of LLS, such as their location, and degree of abstractness; the contextual setting of strategies among other IDs, learning goals, and tasks; strategy chains and internal relationships between particular LLS; the relationships between LLS and plans and processes;

72 the correspondence between LLS and motivation; references to cognitive styles and learning styles; the inclusion of language skills in the strategy model. Addressing the problems of the size and scope of LLS, Cohen (2003: 280) referred to LLS as “conscious or semi-conscious thoughts and behaviours used by learners with the explicit goal of improving their knowledge and understanding of a target language”. Referring to Wenden’s (1986) definition of strategies, Griffiths (2004) pointed out that LLS could be observed directly, but they could also be inferred from language learners’ behaviour. According to Macaro (2006), the existence of a number of definitions of LLS does not facilitate finding the answer to the dilemma. The researcher indicated that situating the strategy concept within the framework of cognitive theory and lending support to the ACT model (adaptive control of thought) (Anderson 1982) would imply accepting the assumption that strategies are located in the brain; however, according to other standpoints, represented by Chamot and El-Dinary (1999), strategies cannot be restricted to the mental dimension since they actually include overt activities. Macaro (2006) concluded that the breadth of LLS definition corresponds with the scope of problems that one is likely to encounter making attempts to locate the area of human activity responsible for LLS, but also pointed out that the semantic proximity of such notions as strategies, operations, routines, processes, procedures, actions, tactics, techniques, plans, or steps seriously impedes the interpretation of strategies. He proposed a definition of strategies, according to which LLS could be considered as conscious steps or actions that originate from working memory, thus being controlled by the central executive (cf. Baddeley 1997). Macaro (2006: 325) also suggested that although some mental actions might be unconscious, those taken in order to accomplish a learning goal and evaluated by means of certain specific criteria can only be of conscious nature. It can be inferred from the framework that strategies need to be accounted for in context, which includes the goal they serve, as well as the particular task and situation, and they are consciously employed by language learners. These views are reflected in Macaro’s (2006: 325) three assumptions about the nature of LLS: The description of LLS needs to involve three aspects, that is the goal, the situational component, and that of mental action; The effectiveness of LLS depends on their use for particular tasks and the operation of various processes;

73 The constructs which cannot be used as synonyms of LLS include language learning processes, styles, plans, and actions which happen automatically. A number of other researchers have also referred to the dilemmas addressed by Macaro. For example, White (2008b: 9) described language learning strategies as “the operations or processes which are consciously selected and employed by the learner to learn the TL or facilitate a language task”, adding that language learners choose them from a variety of options, and their decisions are influenced by both the dynamic environment, and their goals in learning and / or using the target language. The researcher also emphasised the importance of exploring LLS in order to gain insights into independent learning, noticing considerable potential in the impact of the findings from strategy research on the enhancement of the language learning process, both inside and outside the language classroom. Liang (2009: 199) listed the following features of LLS, allowing for a greater dose of freedom in terms of what can qualify as a language learning strategy: LLS can be behavioural (observable) or mental (non-observable); LLS can vary from general approaches to specific actions or even techniques that learners adopt in order to learn the target language; LLS are consciously used; however, the success in their operation may require some unconsciously operating processes. According to Pineda (2010: 96), LLS could be defined as “conscious, selected behaviors, used for overcoming certain educational challenges which depend on the matter to be solved (...) used to memorize information, to synthesize it, or to use that information in speaking or writing” and “complex behaviours which may change according to the learning task, and (...) vary from one person to another”. Relying on Oxford’s considerations of LLS (1990), he considered strategies to be the key to learner’s self-direction, problem-oriented, teachable, transferable, and influenced by a number of other IDs. Oxford (2011, 2017) emphasised the role of volition in learners’ use of LLS, indicating that deliberateness could actually serve as the criterion to distinguish between strategies (consciously employed) and skills (automatically exercised and demonstrated). She also updated the list of characteristics of LLS, among which she mentioned the following (Oxford 2011: 14): consciousness in strategy use, involving awareness, attention, intention, and effort (as listed by Schmidt 1995);

74 facilitative character, which consists in making the learning process easier, more enjoyable, and/or more effective; potential to take the form of different tactics, work in different contexts, and be used for a variety of purposes; inclusion of a number of aspects of the learner as a human person, e.g. cognition, metacognition, affect, etc.; condition of joint use of a number of strategies (strategy chains) as a premise for successful self-regulation in strategy use; adaptability and transferability across a number of situational contexts. In an attempt to address the allegedly insufficient scientific rigour underlying the strategy concept (cf. Dörnyei and Skehan 2003; Dörnyei 2005; Macaro 2006), Griffiths (2013) provided a thorough description of the origins and the features of LLS and relevant notions, such as target language and speakers of other languages. Referring to an earlier paper (Griffiths 2008), she also summarised the key features of LLS, considering them to be active, conscious, chosen, purposeful, regulatory, and learning-focused. Basing on the above features, she proposed her own definition of LLS, according to which “(l)anguage learning strategies are activities consciously chosen by learners for the purpose of regulating their own language learning” (Griffiths 2013: 15). Griffiths and Oxford (2014) considered the status quo in strategy research in the 21st century, taking into account the lack of consensus across defining the strategy concept, the relationship between strategy use and learners’ proficiency, some theoretical underpinnings, categorisational issues, contextual relevance of LLS, their teachability, research methodology, and analysis techniques. They chose a set of definitions to demonstrate the evolution of the strategy concept, presented in Table 16.

Table 11. Overview of strategy definitions by Griffiths and Oxford (2014).

Paper Definition

the techniques or devices which a learner may use to acquire Rubin 1975 knowledge Rigney 1978, O’Malley et al. procedures which facilitate acquisition, retention, retrieval, and per- 1985 formance specific actions taken by the learner to make learning easier, faster, Oxford 1990 more enjoyable, more self-directed, more effective, and more transferable to new situations activities consciously chosen by learners for the purpose of regulating Griffiths 2013 their own language learning

75 Referring to the existence of multiple definitions of LLS, describing strategies as learning behaviours, tactics, or techniques, Griffiths and Oxford (2014) expressed their conviction that efforts to unify and consolidate the beliefs about the source and nature of language learning strategies should be maintained so as to account for a proper investigation of a scientific construct. Commenting on the nature of LLS, the researchers suggested that LLS be considered a cognitive construct, involving the processing and acting upon information. However, the processing is not limited to behaviouristic ‘machine-like’ production of output, since it involves generating rules, learning from errors, developing interlanguage, creating schemata, and employing metacognition. The researchers drew the following conclusion which is probably a very accurate comment on the co-existence of numerous theoretical frameworks explaining the functioning of LLS:

Strategies are theoretically multifaceted, and although the need for a sound theoretical base for the purpose of meaningful research is acknowledged, perhaps we should also be careful that attempts at clarification do not oversimplify and thereby reduce the richness and predictive potential of a phenomenon which is by its nature extremely complex. (Griffiths and Oxford 2014: 2).

In an attempt to account for the complexity of the strategy construct, Cohen (2014) advanced a definition of language learner strategies, a label under which he included both, language learning, and language use strategies. According to the definition, language learner strategies are “thoughts and actions, consciously chosen and operational- ized by language learners, to assist them in carrying out a multiplicity of tasks from the very onset of learning to the most advanced levels of target-language performance” (Co- hen 2014: 7). On the one hand, maintaining the learning and use divide distinguishes Cohen from other strategy researchers (cf. Griffiths 2003, 2008, 2013; Oxford 1990, 2011, 2017). On the other hand, Cohen’s (2014) claim that strategy use is conscious and his recognition of the meaning of context in strategy use across different proficiency levels are shared by a great number of strategy researchers who consider LLS as one of the pillars of learners’ self-regulation (Oxford 1990, 2011, 2017; Griffiths 2003, 2008, 2013). Recently, Oxford (2017) reviewed 33 definitions of LLS and identified the pre- vailing themes in the analysed accounts of other strategy researchers. She gathered her findings under nine major themes, including the diverse form of LLS, their purposeful- ness, the variety of purposes that LLS serve, consciousness, the way of strategy use, their contextuality, teachability, and the nomenclature of addressing LLS. According to the

76 most commonly shared views, LLS involve a mental and volitional component, they serve a certain purpose, and their use is intentional, they facilitate learning, and constitute a major component of learners’ self-regulation, and their use is conscious. All these apsects are referred to in the definition of LLS which attempts to reconcile the notion with the concept of self-regulation (see section 2.4.3 of the present chapter). As a final comment, on the one hand, it seems that there is no uniformity regarding the holistic treatment of strategies, the way in which they are used in terms of flexibility and co-dependence, their teachability, and their contextual setting. On the other hand, however, it could be argued that the lack of consensus in the above areas does not constitute a reason to worry, but instead can be used as a reference to determine the directions of future strategy research.

2.4. Categories of LLS

The classifications, categorisations, typologies, and taxonomies8 of LLS have evolved along with the development of the strategy concept. Initially, they tended to be more of ad hoc collections of labels, created in order to account for the strategies recognised in early strategy studies, such as practising, monitoring, and inferencing. With time, the scope of identified strategies became considerably wider, which led to the creation of more thorough classifications, typologies, and taxonomies.

8 Mapping the area of LLS poses a challenge not merely because of the scope of strategy research, but also due to a variety of co-existing approaches which researchers have adopted when attempting to distinguish between various strategies. Scholars have resorted to categorisations, consisting in recognising, differentiating, and understanding ideas which are grouped for specific purposes (Cohen and Lefebvre 2005), classifications, ordering entities into groups of classes on the basis of their similarity (Bailey 1994) in order to achieve within-group homogeneity and between-group heterogeneity, typologies, conceptually developed classifications, based on theories, intended to produce classificatory framework for ordering individual items (Eppler and Mengis 2011), and, finally, taxonomies, i.e., empirically derived groupings, aimed to introduce structure into a body of facts, and build a unified and homogeneous view of a particular domain (Chandra and Tumanyan 2005: 247).

77 2.4.1. Early LLS labels

Carton (1971) suggested a simple division of strategies relying on various types of cues used by foreign language learners in their attempts to handle unfamiliar linguistic data, that is: relying on intralingual cues, including those provided by the target language - used by relatively more advanced learners, and connected with a facilitative regularity of the target language; relying on interlingual cues, including the search for similarities between the mother tongue and the target language, e.g. noticing derivations or cognates; relying on extralingual, i.e., contextual cues, i.e., resting upon the representative function of language in general, and the possibility to make predictions and notice similarities between objects and events which constitute the existing reality. Taking into account the purpose and modality of LLS, Bialystok (1981) initially recognised three categories, i.e., practising, monitoring, and inferencing, which she later expanded into four by distinguishing between formal and functional practising. She also observed that learners are given the opportunity to practise both aspects of language, i.e., formal and functional, but at the same time pointed out that overall opportunities for functional practice occurred much more frequently outside the language classroom. By distinguishing two types of modalities, i.e., oral and written, along with two types of practice, i.e., formal and functional, Bialystok ultimately produced a four-cell classification of LLS, distinguishing between formal/oral LLS, formal/written LLS, functional/oral LLS, and functional/written LLS. This classification differs considerably from later classifications developed by other researchers not only because it very concise, but also it treats modality as a basis of distinction between LLS, and it reflects the different teaching and learning focus, i.e., that is formal vs. functional practice.

2.4.2. Rubin’s (1981) classification of LLS

Basing on her analysis of cognitive processes in learning a second language, Rubin (1981) grouped LLS into two major categories: strategies directly contributing to learning, including:

78 (a) asking for clarification or verification concerning either the form or the manner of using a problematic word or expression; (b) monitoring, which involves, in the first place, self-correction regarding various aspects of linguistic correctness; (c) memorisation, which learners resort to in their attempts to acquire new elements of the target language, e.g. using associations; (d) guessing and inductive inferencing, consisting in using linguistic clues to interpret meaning or ‘discover’ the rules of the target language; (e) deductive reasoning, comprising the search for target language patterns; (f) L2 practice, through which learners can experiment and make use of the acquired knowledge of the target language, and strategies indirectly contributing to learning, including: (a) looking for and creating opportunities for practice, such as engaging in conversation with native speakers; (b) compensating for missing chunks of language, for instance using synonyms, circumlocutions, cognates, or even gestures, collectively described as ‘production tricks’. In the course of time it became evident that Rubin’s (1981) classification of LLS mainly accounts for cognitive strategies. Overall, only two classes of strategies are listed under the indirect label. It could be argued from today’s perspective that they correspond to metacognitive and compensation strategies. One more exception that stands out is the inclusion of clarification/verification strategies in the direct category of LLS, as these strategies are interactive in nature and thus can be referred to the social dimension.

2.4.3. O’Malley and Chamot’s (1985) classification of LLS

Following Brown’s (1982) classification scheme of learning strategies, O’Malley and Chamot (1985) developed their own classification of LLS which encompassed three main categories: metacognitive strategies, cognitive strategies, and social mediation. They included any strategy requiring cooperation with peers in order to model a language element or receive feedback in the final category. In the first group, the following strategies were included:

79 advance organizers, defined by O’Malley et al. (1985: 33) as “comprehensive preview of the organising concept or principle in an anticipated learning activity”; directing attention to learning tasks; focusing on specific aspects of language (selecting); self-management, consisting in creating favourable learning conditions; advance preparation, involving both planning and rehearsing; self-monitoring, connected with strive for accuracy, and thus manifesting in self-correction and situational appropriateness; consciously scheduling production after reception; self-evaluation, referring to both, one’s progress and accuracy; self-reinforcement. Under the cognitive label, the following strategies were classified by O’Malley et al. (1985: 33-34): repetition, involving imitation of model input, and silent rehearsal; resourcing, consisting in the use of target language materials; directing physical response to physical actions as in giving orders; translation; grouping based on attributes; various types of note-taking; deduction with reference to target language rules; recombination, involving manipulating target language chunks in order to create larger sequences; imagery, or visualising new elements of the target language; auditory representation, involving the retention of target language sounds; key word strategy, requiring the use of a familiar word which somehow resembles the word to be learnt; using a newly learnt word in context; elaboration, consisting in relating new information to previously learnt concepts; transfer, involving the use of previously acquired knowledge to deal with a new task; inferencing, applied in order to handle new language items, predict content, or compensate for missing information; asking for clarification.

80 2.4.4. Oxford ‘s (1986, 1989/1990) early taxonomies of LLS

Oxford’s (1986) taxonomy of LLS attempted to describe learners’ efforts to develop all four language macroskills. It was descriptive rather than prescriptive in nature, available in two versions, i.e., simplified or expanded, and designed to be used by both profession- als (teachers, researchers, etc.), and L2 learners. Capitalising on Rubin’s (1981) work, Oxford (1986) divided LLS into two major categories, that is direct (primary) and indirect (support) strategies, which she explained in the following way: primary LLS are employed to operate directly on the learning materials, including: (a) L1 to L2 strategies, such as translation, interpretation, transfer, contrastive analysis, and analogy; (b) inferencing strategies, consisting in attempts to comprehend the meaning of L2 words of phrases; (c) emphasis or summary strategies, including note taking, various forms of outlining, summarising, highlighting, or using context-signalling devices; (d) clarification or verification strategies; (e) resourcing; consisting in skilful use of materials, such as dictionaries or recordings; (f) formal practice strategies, such as rule generation and/or revision, rule search and/or application, exercising, generalisation, deductive reasoning, analysis, practice, repetition, imitation, and noticing patterns; (g) functional practice strategies, e.g. recombination, naturalistic practice, L2 self- talk, or L2 games; (h) communication strategies, for instance attempts to maintain communication in L2 or making use of available information9; (i) mnemonic strategies, such as preparing lists, dealing with them, e.g. breaking lists into smaller lists, assigning attributes to parts of lists, using acronyms, loci, flashcards, relating L2 entities to situations or contexts, mechanical tricks, rhyming, using auditory associations, imagery, keyword techniques, elaboration,

9 Communicative strategies were excluded from the strategy taxonomy in later papers (Ehrman and Oxford 1990, Oxford 1990) as the name tended to be misused and the tactics under the label specifically referred to compensation strategies (cf. Tarone 1981, 1983).

81 physical response, phonological aids, rote learning, silent rehearsal, or learning a whole passage, and support LLS are employed in order to develop an appropriate attitude, handle various constraints in the learning process, such as tiredness, boredom, or distractions, etc. They facilitate learning in a number of ways, and involve in particular: (a) general study strategies, such as organising work and the learning environment; (b) planning and goal setting strategies, referring to both the language learning process and particular tasks; (c) attention-enhancing strategies, e.g. advance organisers, or focusing on certain types of input; (d) self-management strategies, involving learners’ monitoring, assessment, evaluation, estimation, diagnosis, prescription, and reinforcement; (e) social cooperation strategies; (f) creating practice opportunities; (g) cultural orientation towards the L2 society and history. (h) affective strategies, involving searching for incentives, reduction of anxiety, and ways of improving perseverance. In her taxonomy, Oxford (1986) also differentiated between cognitive and metacognitive LLS. The former LLS fall into the category of direct LLS, whereas the latter LLS can be all classified as indirect LLS. Cognitive strategies would thus involve manip- ulations and adjustments on the part of the learner with reference to the learning material in order to support retention or enhance learning, whereas metacognitive strategies would involve reflection on language learning, knowledge about it, and increase in the ability to regulate one’s own learning. Another distinction reflected by Oxford (1986) in her first LLS model refers to the inclusion of both syntactic and semantic LLS. Both of them represent direct learning strategies; however, syntactic LLS focus on function words, affixes, and some content words, whereas semantic strategies require the use of contextual cues which can be found in learners’ environments. A number of other strategies are recognised in the taxonomy, such as formal vs. functional practising (as suggested by Bialystok 1981), certain affective strategies, or some LLS specifically dedicated to processing texts in the target language. Moreover, as the researcher claims, the distinctions between the above categories

82 may become blurred in practice since learners may apply some strategies in either a direct or indirect manner. For example, social strategies are recognised as a separate group of LLS, but not classified as either direct or indirect strategies. In an attempt to simplify the taxonomy, Oxford (1989ab) introduced a classification which reflected the scales on the strategy inventory, SILL ver. 7.0, which included: cognitive strategies, whose main focus remains the target language itself, e.g. reasoning, analysis, note taking, as well as both functional and formal practice, and which can all be classified as direct LLS; memory strategies, which help the learner store and retrieve relevant information; compensation strategies, which enable language learners to deal with insufficient knowledge of the target language, such as inferencing or using circumlocution; metacognitive strategies, used by learners in order to plan and evaluate their learning, and therefore execute control of the learning process; affective strategies, which help learners in managing their emotions, and attitudes, and seriously influence learners’ motivation; social strategies, whose application requires the presence of other people in the learning process, and are often based on cooperation. Ehrman and Oxford (1990) classified memory, cognitive, and compensation strategies as direct strategies, and labelled metacognitive strategies, affective, and social strategies as indirect strategies. In her milestone publication, Oxford (1990) presented a more thorough version of the previously described taxonomy. Examples of strategy categories belonging to each of the six scales are listed in Table 17. Each of the subcategories (19 strategy sets) can be expanded into particular strategies, 62 in total. For example, the cognitive strategies of practising include repeating, formally practising with sounds and writing systems, recognising and using formulas and patterns, recombining, and practising naturalistically, while the memory strategy of applying images and sounds can take the form of using imagery, semantic mapping, using keywords, or representing sounds in memory. Of course, a certain dose of overlap can be expected. For instance, it may prove difficult to classify a strategy as structured reviewing (a memory strategy which is an example of reviewing well) or repeating, which is an example of practising, a category existing in the cognitive domain of LLS. According to Oxford (1990: 39), strategies work well when used simultaneously. For example, the efficiency of using memory strategies increases

83 when metacognitive and affective strategies are employed, e.g. paying attention or reducing anxiety. Additionally, the material to be learnt or memorised needs to be personally meaningful and significant to the learner. According to Oxford (1990), memory strategies are employed in order to help learners store and retrieve information which they need to communicate. They are also applied when information used in the communication process needs to be moved from the fact level to the skill level. This may take the form of matching different elements, e.g. assigning verbal labels to mental images of words or phrases. Cognitive strategies constitute a relatively heterogeneous set; however, as reported by Oxford (1990: 43), they all serve the purpose of manipulating or transforming the target language. Compensation strategies are vital whenever the language learner encounters difficulties caused by short- ages of knowledge, either trying to understand or communicate something. Metacognitive strategies provide language learners with the opportunity to order and coordinate the process of their own learning. They enable the language learner to avoid information overload through planning and organising the learning process as well as various tactics of maintaining focus. Also, since metacognitive strategies include seeking the opportunities for practice, their use considerably affects the amount of target language input that the language learner is exposed to. Affective strategies are connected with learners’ management of emotions, attitudes, motivations, and even the values which they believe in, as stated by Oxford (1990: 140). All of them influence one’s self-esteem as a language learner, the degree of inhibition which one experiences in learning or using the target language, or one’s tolerance for ambiguity. Oxford stressed the importance of teachers’ role in creating facilitative atmosphere in the classroom and through it, assisting the learner in their search for the sense of efficacy. Finally, social strategies are an inevitable ingredient in successful language learners’ repertoire, since not only do more proficient interlocutors provide learners with opportunities for practise, but they also serve as a source of input. Moreover, other L2 users can serve as instructors or experts in charge of correcting learners’ mistakes or errors and even attempt to consciously work on the areas which, in their view, need guidance or improvement. Table 12 illustrates Oxford’s (1990) division of LLS into subcategories.

84 Table 12. Examples of language learning strategies based on the SILL ver. 7.0 (Oxford 1990: 17-25).

Strategy scale Subcategories

Memory strategies Creating mental linkages Applying images and sounds Reviewing Employing action. e.g. through physical response Practising Receiving and sending messages

Cognitive strategies Analysing and reasoning Creating structure for input and output

Making guesses Compensation strategies Overcoming limitations in speaking and writing

Centring one’s learning Metacognitive strategies Arranging and planning learning Evaluating learning

Lowering one’s anxiety Affective strategies Self-encouragement Taking one’s ‘emotional temperature’

Asking questions Social strategies Cooperating with others Empathising with others

Oxford and Cohen (1992) addressed the issue of the subjectivity of strategy categories, which they attributed to a variety of epistemological beliefs and diverse inclinations of strategy researchers. They used three examples of strategies in an attempt to prove that the same strategy might be classified differently: asking questions for clarification or verification, which various researchers classified as either a social or a cognitive strategy; monitoring one’s errors, which was either classified as a cognitive or metacognitive strategy; seeking practice opportunities, which was either recognised as a metacognitive or social strategy. Oxford and Cohen (1992: 17) concluded that differences in strategy classifications at least partly resulted from different research goals in strategy research, which ranged from accounting for the characteristics of good language learners to investigating the learning process as such, and even validating ELT methods. Attempts to bring strategy categories into order were made in Oxford’s (2011, 2017), and Cohen’s (2014) later

85 works, which, on the one hand, seem to stem from similar basic convictions, but, on the other hand, reflect at least partly different methodological assumptions (later categorisations by Oxford and Cohen are discussed in the following subsections of the present chapter).

2.4.5. Purpura’s (1998) classification of LLS

Capitalising on the model of information processing introduced by Gagné et al. (1993)10, Purpura (1998) drew a taxonomy of cognitive and metacognitive LLS.11. Among cognitive strategies, he included: comprehending processes, e.g. analysing inductively and clarifying or verifying; storing or memory processes, e.g. associating, repeating or language transfer from L1 to L2; using or retrieval processes, analysing inductively, applying rules or practising naturalistically. At the same Purpura (1998) incorporated the following categories of LLS under the metacognitive label: assessment strategies, implemented by the learner before taking any specific action, therefore ordering the use of other strategies; monitoring, i.e., assessment strategies introduced as a learning action progresses; self-evaluating, involving a set of strategies employed after a specific learning action; self-testing, which can take place both before and after a learning action, and reflects learners’ interest in their progress.

Purpura’s (1998) taxonomy differs considerably from Oxford’s (1986, 1990) taxonomies. Since it is based on the model of information processing, it focuses on two groups of strategies, namely, cognitive and metacognitive LLS. Memory strategies are

10 According to the model, human learning consists of four stages, including identification, e.g. selecting and isolating information which is to be processed in later stages, comprehension, e.g. maintaining information in working memory in order to transform it into meaningful symbols or understand it, storing or memory , e.g. active elaboration, making associations, recognising patterns or relating information to prior knowledge, and using or retrieval of information (e.g. involving the use of long-term memory in creating linguistic output) (Gagné, Yekovich and Yekovich 1993). 11 The two categories correspond with the previously developed questionnaires of cognitive and metacognitive strategies developed by Purpura (1997).

86 not treated as a separate category; however, this does not seem to pose any methodological difficulties. On the other hand, although some cognitive strategies which are labelled “clarifying” and “verifying” may be applied in the presence of another speaker, the social component of LLS is not recognised as a criterion important enough to consider it as a base of creating a separate category of LLS. Similarly, compensation strategies are not recognised as a separate category, although one might expect them to be employed in comprehension processes, especially those involving negotiation of meaning. Finally, it appears that one of the weaknesses of the model is the lack of inclusion of affective variables. Focusing solely on the cognitive aspect of learning may pose a threat to treating the language learner holistically, since humans, unlike computers, not only process information, but also experience different types of emotions, which may condition adopting different attitudes to learning.

2.4.6. Macaro’s (2001) tentative classification of strategy categories

After providing an extensive list of examples of strategies employed by language learners performing routine learning activities12, Macaro (2001) considered the cognitive model of learning and suggested that LLS should be mapped on two continua based on the direct-indirect and conscious-subconscious dichotomies. The strategies which he viewed as relatively more indirect and subconscious included: linking L2 words to mental images; inferring meaning from close context; grouping vocabulary items into lexical categories; storing L2 chunks in short-term memory before analysing them; referring to gist while attempting to understand details of meaning. The strategies which were situated in the central part of the continuum included: reflecting on the similarity of L2 and L1 words; repetition of new words or phrases; memorising new language items; providing answers (in one’s mind) to questions asked to other students.

12 Three activities were scrutinised, including questioning sentences, understanding a written text, and a writing homework.

87 Finally, the following strategies were classified as relatively more direct: making mental associations with objects/ shapes of words in writing; asking for clarification; practising with peers (e.g. acting out dialogues); relaxation strategies employed in order to prevent being overwhelmed by a task. Although it seems legitimate to situate LLS on continua in order to account for their complexity, the list provided by Macaro (2001) resembles a set of practical examples rather than a classification or even a typology. The strategies named by Macaro (2001) are likely to be commonly employed by language learners, but at the same time, due to their level of concreteness, they cannot be used as general labels for specific tactics. In fact, the boundary between tactics and strategies appears to be blurred. Similarly, the distinction between natural and taught strategies appears a bit arbitrary, especially as the evidence for it comes from a single case of a specific learner. Since strategy use is modified by learners’ intrinsic features, such as personality traits or gender (Oxford 1989a; Nyikos and Oxford 1993; Griffiths 2015), categorising strategies on the basis of such evidence is prone to bias.

2.4.7. Schmidt and Watanabe’s (2001) classification of LLS

After analysing 23 strategies, Schmidt and Watanabe (2001) introduced a classification of LLS based on the results of explanatory factor analysis. Four factors appeared as the most essential variables explaining a great deal of the variance in strategy use as follows: strategies linked to study skills, consisting in a methodological allocation of resources for the purpose of task completion; cognitive strategies along with some metacognitive tactics of learning management; coping strategies, which constitute a challenging area since they do not show a high homogeneity; social strategies, whose use involves either interaction with classmates or the involvement of other target language users. The described model was subject to further analysis by Woodrow (2005, 2006b), who reported that despite the good data fit, a three-, rather than four-factor structure of the model appeared to represent a better fit. The researcher advocated the inclusion of the

88 cognitive, social, and metacognitive dimension in the model of LLS, thus dismissing the separate scale of general study skills. Also, she remained deeply critical about the possibility of generalising research findings based on questionnaires, and particularly those based on Likert-type scales, due to the lack of replicability of such studies in confirmatory factor analysis, which, as she concluded, results from the existence of multiple contextual influences in strategy use. The researcher also recommended alternative categorisations of LLS, derived from educational scientists, such as that introduced by Pintrich et al. (1991b).

2.4.8. Woodrow’s (2006b) classification of LLS

Woodrow (2006b) juxtaposed two broad categories of strategies with each other: deep and surface strategies, which, as she claimed, reflected two different approaches that learners could adopt towards the learning process13: deep strategies, which are oriented towards transforming knowledge; they involve a conscious attempt to comprehend the learning material, as well as establish a correspondence between the newly acquired knowledge and the existing knowledge and experience, draw conclusions on the basis of the available evidence, and assume that the learner remains critical by examining the logic of arguments; surface strategies, which are oriented towards reapplying information, and consist in recollecting and reproducing previously learnt ideas or information which have previously been passively accepted; they also consider the requirements of the task or assessment and require sticking to them strictly so that reflection on their purposeful- ness does not take place. Referring to the above categories, Aharony (2006: 852) commented that deep strategies corresponded to a holistic approach to studying, concentration on relevant meaning, as well as learners’ ability to personalise knowledge and relate it to the learner’s everyday experience. The researcher also expressed the belief that deep strategies involve the use of metacognitive skills, manifested in creating learning materials or spreading new ideas. Moreover, he suggested that adopting a “critical-inquisitive” perspective by

13 The distinction was introduced by Marton and Säljö (1976), and reflects the general beliefs about language learning held by Säljö (1979).

89 the learner may actually guide them to the discovery of their “inner self” and facilitate academic achievement. According to Aharony (2006), the reason for applying surface strategies consists in learners’ willingness to accomplish the learning task as quickly as possible. The researcher also listed other characteristics of surface LLS, such as studying the material in a linear manner, the reliance on rote learning rather than strategies facilitating comprehension. The use of metacognition is hence limited to planning minimal efforts so as to accomplish the learning task. On the basis of the results of a confirmatory factor analysis and an earlier study, Woodrow (2006ab) distinguished the following categories of LLS: Metacognitive strategies: (a) relating words; (b) guessing from context; (c) evaluating progress; (d) deciding what to learn; (e) asking the teacher to explain; (f) practising with non-native speakers. cognitive strategies: (a) organising the learnt content; (b) reflecting on feedback; (c) rereading materials; (d) reviewing for tests; (e) memorising words; (f) determining unfamiliar concepts; (g) maintaining the right pace of progress. social strategies, including: (a) cooperating with others; (b) discussing course contents with others. This tripartite division of LLS reflects the categorisation suggested by O’Malley et al. (1985). First of all, memory strategies are not treated as a separate category. Instead, they form a part of the cognitive dimension of LLS. Secondly, it is different from Ox- ford’s (1990, 2011, 2016) and Griffiths’s (2013) organisations of LLS with regard to the way it accounts for guessing from context. This strategy is approached as inferencing rather than compensating for missing knowledge since it belongs the metacognitive

90 category. Finally, the classification does not include an affective or motivational dimension, which might result in a failure to account for all the variables which determine the quality of learning or determine learning outcomes.

2.4.9. Tragant and Victori’s (2012) typology of LLS

Tragant and Victori (2012) introduced a number of innovations into the investigation of LLS, resulting from their earlier study (2006). They attempted an equal distribution of questions across different aspects of language learning, dedicated not only to different functions, but skill areas, and areas of linguistic competence, such as the study of grammar and vocabulary, reading strategies, writing strategies, and conversational skills (referring to both listening and speaking outside the language classroom). The strategies were grouped by Tragant et al. (2013) under the following categories: extended practice strategies, taking advantage of learning opportunities, and maximising exposure to the target language (with no distinction being made between classroom and outside-classroom activities); study strategies, comprising learning the formal aspects of language, e.g. through memorisation; analysis strategies, employed in reading and listening comprehension, referred to as higher-order strategies; avoidance strategies, whose name might be considered a bit of a misnomer, since they are meant to either stop or maintain a production/reception activity which is problematic for the language learner; reduction strategies, involving restricting communication processes by limiting their scope, but at the same time, making the communication process less effective; form-focused strategies, employed to develop linguistic competence and language awareness and encode or decode language. The categories of strategies listed above were also later classified as either skills- based, deep processing strategies, demonstrated in reading, listening, speaking, and writing in the target language, or language study strategies, associated with learning vocabulary and grammar, whose frequent manifestation was simply memorising vocabulary or grammar rules.

91 2.4.10. Griffith’s (2013) taxonomy of LLS

Admitting that strategy classification remains an area of controversy among strategy researchers, Griffiths (2013: 43) developed a simple taxonomy of LLS consisting of five major categories: Cognitive strategies, central to the taxonomy since directly connected to the material to be learnt, e.g. learning vocabulary encountered in a text; Affective strategies, employed in order to control feelings and emotions, e.g. attempting to relax or rewarding one’s good performance; Social strategies, applied in order to manage interaction with others, e.g. arranging to meet with other L2 user; Memory strategies, used in order to remember the target material, e.g. writing down selected words or phrases in order to memorise them; Metacognitive strategies, involving the control, management, and regulation of the learning process, e.g. checking requirements. The above taxonomy is functional in nature. To a large extent, it reflects Oxford’s (1990) classification of LLS. Although Griffiths (2013: 42) acknowledged that distinguishing between memory and cognitive strategies had been criticised by LoCastro (1994) and Cohen and Dörnyei (2002), who viewed the former as a subcategory of the latter, she did not exclude memory strategies from her taxonomy. Instead, she eliminated the group of compensation strategies, either because they tended to be referred to as communication strategies rather than learning strategies (cf. Ellis 1994) or because they could be employed by learners with multiple functions in mind.

2.4.11. Cohen’s (1998, 2014) typologies of LLS

Cohen’s (2014) map of strategy categories confirms that some crucial discrepancies among strategy researchers have never been eradicated. Firstly, the distinction between language learning and language use is still present, and secondly, communication strategies are discussed as a category of language use strategies14. In order to explain the

14 Cohen (2014: 16) admits that communication strategies may or may not have impact on learning.

92 distinction between language learning and language use strategies (both belonging to a larger category labelled language learner strategies), Cohen (2014) provided examples of both and typified language use strategies employed by language learners in order to improve their command of the target language. Referring to the first category of language learning strategies, Cohen (2014: 12) mentioned: identifying, distinguishing, and grouping target language chunks; repeated contact with the target language, including solitary rehearsal; memory strategies, e.g. repetition and mnemonics. These categories partly correspond to the groups in Oxford’s (1990, 2011) taxonomies, and focus on the cognitive and metacognitive aspects of LLS; however, the affective and social dimension of the learner seem to be unaccounted for. Describing language use strategies, Cohen (2014: 13-14) identified four subsets: retrieval strategies, consisting in reaching for language material stored in memory; rehearsal strategies, through the use of which learners engage in various types of practice; coping strategies, comprising compensatory strategies, allowing learners to compensate for missing knowledge, and cover strategies, through which learners feel empowered and in charge of their learning process; communication strategies, which are employed in order to convey meaningful messages in speech or writing, reported to be further divided into: (a) intralingual strategies, such as overgeneralisation of target language rules; (b) interlingual strategies, such as transfer from L1 to L2; (c) topic avoidance or abandonment; (d) message reduction; (e) code switching; (f) paraphrasing, for instance through the use of synonyms or circumlocutions. Cohen (2014) also indicated that LLS could be classified by skills, functions, proficiency levels, and, finally, specific cultures and languages, and distinguished a number of subcategories under the first two discriminants. Considering the possible categories of strategies with regard to skill development, Cohen (2014) did not suggest that strategies differ regarding the development of each of the four basic macro skills, but also assumed that learners could employ specific grammar, vocabulary, and translation strategies across the four skills. Commenting on the possible functional division of LLS, he referred to

93 Chamot’s (1987) and Oxford’s (1990, 2011) categories of strategies and agreed that strategies could be classified as metacognitive, cognitive, affective from the functional point of view with no need to distinguish between language learning and language use strategies. At the same time, citing the example of monitoring and evaluation strategies, commonly classified under the metacognitive label, Cohen (2014: 20) pointed out that certain strategy categories could overlap, and their use might depend on specific language tasks.

2.4.12. Oxford’s (2011, 2017) classification of LLS

The development of Oxford’s (2011, 2017) Strategic-Self Regulation (S2R) Model did not merely provide theoretical and contextual frameworks for the investigation of LLS, but also shed new light on their classification. While the assumptions and implications of the Model are discussed in subsection 2.4.3, dedicated to self-regulated language learning, the present subsection concentrates on the updated LLS classifications within the S2R framework. In general, Oxford (2017) warned that LLS ought not to be rigidly assigned to merely a single functional category since that could result in oversimplification of the not merely the roles LLS play in foreign language development, but also the strategy construct as such15. The researcher emphasised that similar caution should apply when distinguishing between LLS on the basis of their skill-dependence16 as not only are some traditional skill divisions viewed as irrelevant nowadays, but also the same strategy can serve different purposes even when considered for a single skill, and, may be accompanied by a set of other, less evident, strategies. Oxford (2017: 209) also expressed doubts concerning the legitimacy of efforts to come up with a complete taxonomy of strategies, refine strategy categories, or teach strategy classifications to students, calling for greater flexibility in strategy investigations so as to account for the complexity of the construct. One of the manifestations of flexibility in accounting for and categorising LLS has been Oxford’s (1990, 2011, 2017) refusal to divide them into language learning and language

15 The examples that Oxford (2017: 203-206) provided included using analysing to regulate emotions, manage motivation, or deal with sociocultural issues and attitudes, and thus going beyond its cognitive dimension, and qualifying summarising texts, reconceptualising words, or skipping examples in a text as either a cognitive or metacognitive strategies, depending on the researcher’s interpretation. 16 For example, reasoning is traditionally associated with developing receptive skills rather than productive skills (Oxford 2016: 207).

94 use strategies. Unlike Cohen (1998, 2014), Oxford (2017: 210) expressed a strong belief that the distinction between learning and use was rejected not merely by educational psychologists, but also not recognised by institutions responsible for educational supervision, such as the Council of Europe. The S2R Model assumes that the organisation of strategies could, in fact, be simplified to four main categories of strategies, including cognitive, motivational, social, and affective strategies, corresponding to three relevant meta- categories. Initially considered by Oxford (2011) to be a subcategory of affective strategies, motivational strategies were distinguished as a separate group in Oxford’s (2016) most recent taxonomy. This conceptual development is illustrated in Figure 1.

2011 2017

Fig. 1. Strategy dimensions and categories in Oxford’s (2011, 2017) S2R model.

The S2R model corresponds to the assumption of treating the language learner holistically, and consequently, the division of strategies is functional, so as to account for each aspect of the human person. Each dimension of strategies (the core of Fig. 1) is ‘supervised’ by a relevant meta- dimension built of eight strategies which empower the learner so that they become in charge of their learning process. As Oxford (2017: 218) put it, “(m)etastrategies, by virtue of their executive-control and management function, help the learner know whether and how to deploy a given strategy and aid in determining whether the strategy is working or has worked as intended”. According to Oxford (2017), they include:

95 paying attention; for example noticing the markers of politeness in L2; planning, e.g. arranging to study at home; organising learning and obtaining resources, e.g. using dictionaries to find additional information on encountered L2 structures; monitoring and evaluating, e.g. checking one’s own understanding while reading a piece of text. The organisation of the above categories does not differ across the four dimensions of the learner, while it does, considerably, with reference to strategy categories. Oxford (2017: 246) listed the following strategies in the cognitive domain. Overall, these strategies are not only believed to support memory processes, such as remembering and processing information, but also facilitate the application of newly acquired knowledge: using senses to understand and remember, e.g. analysing visual clues that accompany texts or recordings; activating knowledge, e.g. brainstorming specific vocabulary; reasoning, e.g. making deductions about L2 grammar rules; conceptualising with details; e.g. breaking complex words into smaller components; conceptualising broadly, e.g. drawing semantic maps; going beyond the immediate data, e.g. using headings or topic sentences in reading comprehension. Initially, Oxford (2011: 64) distinguished two general strategy sets in the affective domain. The first one was named activating supportive emotions, beliefs, and attitudes, while the second one involved motivational strategies, which were later considered a separate category of LLS. Later on, Oxford (2017) expanded the category by including five additional families of emotional self-regulation strategies, initially introduced by Gross (2014), in the affective domain. Oxford (2011: 33) stressed that activating supportive emotions might manifest themselves through providing oneself with encouragement, such as self-talk, before performing a language task. Ultimately, Oxford (2017: 307) listed the following strategy sets, believed to help learners deal with emotions and optimise their beliefs and attitudes, and, in the long run, maintaining their motivation to learn, under the affective label: selecting the situation to influence emotions, e.g. avoiding unpleasant classmates; modifying external situations to control emotions, e.g. using and manipulating objects so as to encourage positive emotions;

96 deploying attention to control emotions, e.g. using distractions to reduce language anxiety; changing cognitive appraisals of situations to shape emotions, e.g. modifying one’s own judgement of pace, progress, or achievements in learning L2; modulating emotional responses, e.g. attempting to relax; making meaning as a means of handling emotions, e.g. valuing positive experiences in learning a foreign language. Finally, the three strategies in the sociocultural-interactive domain include, according to Oxford (2017: 271): interacting to learn and communicate, e.g. studying for tests in pairs or groups; overcoming knowledge gaps in communicating, e.g. using circumlocutions in presen- tations; dealing with sociocultural contexts and identities, for instance figuring out the expected amount of silence in turn taking. In short, the above dimension of strategies corresponds to two categories previously distinguished by Oxford (1989bc, 1990), namely compensation, and social strategies. Their inclusion in a single dimension seems legitimate as compensating for missing elements of language frequently takes place during interaction with other L2 users. Fur- thermore, sociocultural-interactive strategies are expected to help assist learners in their attempts to make use of the target language in various contexts, but also facilitate the use of the language whenever the learners’ level of competence is not sufficient to allow satisfactory expression. Also, learners might resort to this category encountering temporary problems, such as not being able to remember a word or expression which they need to use. Finally, as Oxford (2011: 33) suggested, some learners actually like studying together, and hence might typically choose to prepare for a test in pairs, or even groups. Motivational LLS are the final functional category recognised by Oxford (2011, 2017: 263), who described them as entailing the learner’s self-regulation of motivation and volition and specified the following motivational strategy sets: self-consequating, e.g. introducing a system of punishments and rewards for oneself; using positive self-talk and positive self-image, e.g. as preparation for delivering a presentation in L2; using defensive pessimism, e.g. providing oneself with the explanation that it is impossible to do well;

97 enhancing learning, e.g. making learning a play or a game; controlling attributions, e.g. not accepting blame for uncontrollable factors that cause failure. To summarise, Oxford’s (2011, 2017) classifications of LLS constitute interesting alternatives not only to her earlier taxonomies (1986, 1989ab, 1990), but also to other systematic organisations of LLS. The employed nomenclature remains general enough to classify particular tactics into one of the labelled categories of strategies. Moreover, the taxonomy is set in a theoretical and systematic framework, which, on the one hand, reflects the contemporary expectation to recognise the complexity of the learner, but on the other hand, makes it possible to incorporate such concepts as strategy chains or task- specificity, and does not deprive strategy researchers of the option to analyse strategies across skills or other variables. Perhaps most importantly, the ‘old’ categories suggested by Oxford (1986, 1989ab, 1990) easily fit into the new labels, which certainly facilitates comparisons across strategy studies conducted at different times. It could be concluded that strategy researchers are aware of the lack of consensus regarding not merely the definition of LLS, but also their organisation across skills, functions, and other categories. Responding to this problem, Griffiths (2013) even recommended avoiding existing strategy classifications and insisted on grouping strategies according to analyses following the conducted research. This solution, although it is methodologically grounded, does not constitute a systematic solution, and it seems that a general expectation exists that researchers “devise a more consistent set of theoretically sound strategy categories that can be widely applied” (Griffiths and Oxford 2014: 7). More recently, a tendency has emerged among strategy researchers to avoid divisions of LLS into clear-cut categories. Supporters of the tendency include Griffiths (2013), Oxford (2017), and Cohen (2014), who seem to agree that rigid classification frameworks may impede a full understanding of the strategy construct, as they are likely to fail to account for multiple functions of LLS.

2.5. LLS and the concept of self-regulation

The present section of the dissertation constitutes an attempt to account for self-regulation (SR) processes which accompany language learners. First, a brief definitional description

98 of SR is provided from the perspective of the model of limited resources, and the distinction between SR and self-control (SC) is explained. Second, a number of psychological perspectives on the phenomenon are presented, which can be grouped into four categories, suggested by Forgas et al. (2009), including motivational processes, goal-oriented behaviour, affective and cognitive processes, and, finally, social and interpersonal processes. Third, an attempt is made to account for self-regulated learning (SRL) from a theoretical perspective, and some leading frameworks for SRL are discussed, followed by a review of studies into SRL. Finally, self-regulated language learning (SRLL) is discussed, which involves a presentation of SRLL models, and a recapitulation of selected research findings.

2.5.1. Theoretical assumptions and models of self-regulated learning (SRL)

One major reason for educational researchers’ interest in SR the claim that it positively affects academic performance. According to Baumeister and Alquist (2009:23): trait self-control significantly predicted grade-point average (GPA): (1) trait self-control significantly predicted grade-point average (GPA); (2) trait self-control was a better predictor of GPA than IQ; (3) individuals characterised with relatively high levels of SR could be expected to have fewer absences at school or university, choose more competitive programmes of studies, and dedicate more time to homework while spending less time watching TV. The present section aims to provide support for these assertions by addressing relevant investigations. Prior to that, however, theoretical background is provided in order to consider SRL as a scientific construct. Schunk and Zimmerman (1994: 309) defined self- regulation as “the process whereby students activate and sustain cognitions, behaviours, and affects, which are systematically oriented toward attainment of their goals”. Although the definition might not seem complex, approaching self-regulation remains a challenging task at the explanatory level. A number of models were created in order to account for the complexity of the notion. Boekaerts (1997) assumed that two types of self-regulation could be distinguished, cognitive SR and motivational SR. According to the researcher, cognitive self-

99 regulation involves the use of cognitive regulatory strategies on the part of the learner, such creating mental representations of learning goals, designing action plans, or monitoring one’s progress goal achievement. By contrast, motivational self-regulation involves the use of motivational regulatory strategies, such as creating mental representations of one’s own behavioural intentions, linking intentions to action plans, being persistent in implementing and pursuing these plans in spite of potential problems, as well as disengaging action plans and behavioural intentions. The cognitive regulatory strategies empower the learner to orchestrate the use of cognitive strategies, such as paying attention selectively, decoding, rehearsing, elaborating, structuring, thinking of questions, activating and applying rules, uptake, and repair in case of making mistakes, and finally, proceduralising skills. On the other hand, motivational strategies orchestrated by the motivational regulatory strategies described above, include creating the learning intention in the learner, managing one’s negative emotions and limiting the stressors in learning, applying prospective and retrospective attributions, or even limiting effort. Both types of strategies are connected to domain-specific knowledge, involving conceptual and procedural knowledge in the content domain, and metacognitive knowledge shaped by motivational beliefs. The researcher specifically insisted on considering metacognitive knowledge as a learner’s prior knowledge, which she viewed as one of the basic components of SR, allowing the learner to improve their comprehension, monitoring, and assessment of both, conceptual, and procedural knowledge of a given domain. Focusing particularly on goal orientation, Pintrich (2000) proposed a framework for SRL, which is based on several important assumptions. Firstly, according to the cognitive perspective on SRL, learners are believed to actively construct goals and meanings, as well as strategies employed to attain them. Secondly, they are also expected to be able to monitor, control and, moreover, regulate, at least some aspects of their own cognition, motivation, and behaviour. Thirdly, the framework presumes the existence of certain goals, criteria, or standards, which are considered as desirable by the learners. Finally, self-regulated activities are considered to be mediators between learners’ IDs and contextual characteristics. A comprehensive model based on the four above assumptions assumes that SRL can be analysed in four phases, including: forethought, planning, and activation; monitoring; control;

100 reaction and reflection. Four areas are considered at each of the above stages, including cognition, motivation/affect, behaviour, and context. Specifically, in phase one, goals are adopted, and tasks are planned along with the dedicated time and effort necessary for their accomplishment. Judgements are made about self-efficacy, task difficulty, and task value. phase two involves monitoring cognition, motivation, and affect, as well as any changes in the context of task conditions, and requires self-observation. Phase three consists in the selection of relevant cognitive and motivational strategies, and effort management, as well as introducing any necessary amendments to the pursued tasks. Finally, phase four comprises making judgements regarding learners’ own cognitive abilities, affective reactions, behaviour, task, and context for its completion, and forming relevant attributions. The universal character of the above model is certainly its advantage. Moreover, the model also constitutes a valid attempt to account for the interplay between the contextual characteristics of the task, and the IDs of the learner, thus allowing for the inclusion of learners’ strategic knowledge and learning strategy skills, as well as other variables that significantly impact learning and achievement, in considering SRL. Referring to the way that information is acquired and organised by learners, Weinstein et al. (2000) expressed the belief that learning strategies can be applied to various types of tasks, both simple ones, such as memorisation, and more complex ones, for instance those involving conceptual learning. Depending on the perceived complexity of the task, learners might resort to different strategies, such as rehearsal strategies, elaboration strategies, organisational strategies, affective strategies, and support strategies. More recent definitions of SRL remain consistent with the already discussed ones. For example, according Kostons et al. (2012: 3), SRL is “an active, constructive process in which learners plan, monitor, and control their own learning process” that “can occur at different levels, from learners controlling how they engage in studying a given task or whether they want to restudy it (...) to learners controlling what information they want to study”. In general, SRL can also be conceptualised in terms of three processes, namely goal setting, goal operating, and goal monitoring, as suggested by Burnette et al. (2013). Their meta-analysis indicated that implicit theories of SR could be valid predictors of distinct self-regulatory processes, and therefore, also of goal achievement, while incremental theories could serve as significant predictors of goal setting (operationalised as performance and learning goals), goal operating (both helpless- and mastery-oriented

101 strategies), and goal monitoring (expectations vs. negative emotions). Finally, it should be pointed out that the contemporary SR considerations, such as the one proposed by Zimmerman and Kitsantas (2014), have often considered SR referring to processes in the cognitive, emotional, and behavioural sphere, which are also the frequently addressed dimensions of SR in foreign language development (cf. Oxford 2011; Gregersen and MacIntyre 2014), Dörnyei and Ryan 2015, Oxford 2017).

2.5.2. Selected studies investigating SRL

Since the present thesis focuses on the relationships between personality traits and LLS employed by university students and given the fact that SR in learning develops relatively late, as pointed out by de Bruin and van Gog (2012: 249), the scope of the review that follows is limited to studies investigating SR of adults and adolescents17. At the same time, investigations of SR require references to earlier developmental stages, as evidence exists that SR largely depends on early attachment experiences, as indicated by Drake et al. (2014). A number of researchers have emphasised the particular role of goal orientation in SRL. Studies conducted by Pintrich and Van De Groot (1990), and Pintrich and Garcia (1991) indicated that students who developed mastery orientation used cognitive strategies, such as elaboration, and organisation, relatively more frequently, and were more likely to process information more deeply. Similarly, Wolters (1999) proved that goal orientation was the strongest predictor of cognitive strategy use (explaining 25% of variance) and regulatory strategy use (35% of variance in self-regulation) by conducting a study on a group of more than 400 American seventh- and eighth-graders. According to the results of a study of the impact of metacognitive monitoring on the comprehension and learning of texts, conducted in a group of 66 psychology students by Thiede et al. (2003), prompting SR by simply asking students to generate key words before being tested on reading comprehension resulted not merely in significantly better results, but also improved their SR during learning. More specifically, those students who

17 It needs to be granted, though, that de Bruin and van Gog (2012: 245) also claimed that “(w)hen metacognitive instruction takes into account a number of design principles, accuracy of metacognitive processes in children can equal that of adults”.

102 were required to generate key words were also able to select pieces to re-read more accurately. The best results concerning both SR and ultimate reading comprehension test scores were obtained by the group in which generating key words was delayed by reading other, irrelevant texts. The results of a longitudinal experiment designed in order to illustrate the relationships between prompting SR and learning, attrition, and self-regulatory processes, conducted by Sitzman and Ely (2010) certainly correspond to the assumption of SR train- ability referred to in the former subsection of the present Chapter. Involving a sample of 479 adults participating in a voluntary Microsoft Excel course consisting of four consecutive modules, each finishing with a multiple choice test, the study proved that prompting SR through asking questions (continuous or delayed prompts) could increase learning, reduce attrition, and overall, exert a more powerful influence on a subsequent self-regulatory activity than learning performance in a previous activity. Also, delayed and continuous prompting resulted in participants’ allocation of greater amounts of time on tasks. Finally, continuous prompts were efficient in maintaining a relatively high level of SR irrespectively of trainees’ performance in a previous module. Overall, attrition was significantly lower among trainees who were continuously reminded to reflect on their learning, and the effect of prompting SR on learning was exclusively mediated by the amount of time trainees spent on task. A study of 148 college graduates was designed by Strunk and Steele (2011) in order to investigate the relationships between SR, self-handicapping, and procrastination. The instruments used in the study included two scales of the Motivated Strategies for Learning Questionnaire, developed by Pintrich and De Groot (1990): the Self-regulation and the Self-handicapping subscales. SR was found to be a statistically significant predictor of academic procrastination, and, at the same time, it solely explained the variation in self-efficacy18. Although some overlap was spotted between SR and self-handicapping, the researchers suggested that both constructs could be applied in studies explaining procrastination. Finally, no significant differences were found across genders on any investigated variables. Pieschl et al. (2012) assessed the monitoring and SR in a group of 119 undergraduate college students. The participants were asked to complete a task-specific

18Strunk and Steele (2011: 987) therefore advocated for using the term self-efficacy for self-regulation, as suggested earlier by Klassen et al. (2008).

103 questionnaire, and their learning behaviour was observed in the hypertext learning environment on genetic fingerprinting by means of log-file analysis. The researchers concluded that although the investigated students’ behaviour implied the adaptation of efficient learning strategies for three tasks of various level of difficulty, their log files indicated better adaptation to task demands than the self-reported measures. Also, perhaps a certain degree of dissociation might exist between explicit reflection and direct indicators of monitoring and SR. Kostons et al. (2012) carried out two experiments in groups of 80 and 90 secondary students. Seeking to evaluate the effects of observing modelled self-assessment and task selection on the development of parallel skills, the researchers also found out that improvement in self-assessment and task selection skills could enhance the effectiveness of self-regulated learning. Participants learnt about heredity, did a pre-test and post-test on relevant problems, rated their mental effort according to a scale developed by Paas (1993), engaged in self-assessment, which was later confronted with their test scores in both, the pre-test and the post-test, and selected upcoming tasks, which differed in complexity and level of support. As regards self-assessment, it turn out that participants who were given a chance to observe self-assessment modelling examples and task-selection modelling examples displayed significantly better self-assessment accuracy. At the same time, no statistically significant differences were found between the two groups subject to the two different ways of modelling. The results of the study demonstrate that it can be inferred that both self-assessment, and task selection skills can be trained, either through modelling or through explanations and practice, and by training, self-regulated learning can be, indeed, enhanced. Zimmerman and Kitsantas (2014) set out to investigate the relationships between 507 high school students’ self-discipline, SR, and their academic achievement, operationalised as GPA and Virginia Standards of Learning (VSOL), categorising scores into three main levels: pass advanced, pass proficient, and fail. Their study involved the application of multi-method, multi-source array of questionnaires. Self-discipline was measured by the scales previously used by Duckworth and Seligman (2005, 2006) to assess the construct: the Eysenck Junior Questionnaire Impulsivity Subscale (Eysenck Jr. et al. 1984) and the Self-control Rating Scale (Tangney et al. 2004). SR was assessed by the Motivated Strategies for Learning Questionnaire (MLSQ) (Pintrich 1991b), the Self-Efficacy Learn- ing Form (SELF) (Zimmerman and Kitsantas 2007), and the Perceived Responsibility for

104 Learning Scale (PRLS) (Zimmerman and Kitsantas 2005). According to the results of hierarchical regression, the combined composite of SD and SR accounted for more than 69% of variance in GPA, and the inclusion of SD as an independent variable improved the prediction by 5%. When tested separately, SD only accounted for 10% of GPA variance. The two composites also accounted for 42% of VSOL variance. Also, confirmatory factor analysis proved that the two composites, although moderately correlated, loaded on two separate factors rather than one, common latent factor. Lauriola et al. (2015) investigated the relationships between SR and epistemic curiosity (EC), which they measured applying the Interest (I) and Deprivation (D) subscales, both of them consisting of 5 items on a four-point trait scale, developed by Litman and Mussel (2013). Two studies were conducted in groups of 151 Italians (group 1) and 218 Americans and 56 Germans (group 2). Seeking to trace the monitoring and control processes that accompany individuals in their attempts to achieve higher levels of knowledge and proficiency, the researchers also applied several instruments designed to measure SR: the Emotion Regulation Questionnaire (ERQ) (Balzarotti et al. 2010), the Elaboration on Potential Outcomes Scale (EPO) Nenkov et al. 2008), and a scale to assess participants’ risk-taking behaviour, RT-18 (De Haan et al. 2011). All questionnaires were administered in participants’ native languages. The results of the first study indicated that D-type EC significantly correlated with Expressive Suppression and Genera- tion/Evaluation constituents of Emotion Regulation, while no significant relationship was detected with regard to I-type EC. Instead, I-type EC correlated significantly with positive focus on results, which made the researchers conclude that I-type EC involves being op- timistic about making discoveries. In study two, I-type EC correlated positively with Fun Seeking, which was interpreted as a proof of the importance of searching for joy in learning. At the same time, impulse control and punishment sensitivity were found to significantly influence D-type EC, which was interpreted as evidence that moderately-intense emotional states are favourable for concentration and prolonged intellectual activity. Finally, insights into the kind of processing that is involved in self-regulated learning were provided by Winne (2011, 2018). Since it is generally accepted that it is deep processing that ensures greater knowledge gain and makes knowledge sustainable, the researcher decided to employ the levels-sensitive approach in his investigations of self- regulated learning. Winne’s (2018) model assumes that self-regulated learning consists of four phases, including surveying task conditions, setting goals and planning, engaging

105 in the task, and designing adaptations for subsequent tasks. The main implication from Winne’s considerations rests in the conclusion that self-regulated learning should not be automatically classified as deep processing. Instead, according to Winne (2018: 9), “it is processing more complex - deeper - information about a different topic, namely processes for learning”.

2.5.3. Self-regulated language learning (SRLL)

Although it is hard to deny that research into LLS has contributed to advances in foreign language education, as was already mentioned in the present Chapter, studies of LLS have received considerable criticism due to the alleged insufficient reliability and validity, and the purported definitional and classificational fuzziness of LLS as a scientific construct (cf. Skehan 1989; Dörnyei 2005; Woodrow 2005; Dörnyei and Ryan 2015). Apart from numerous efforts to bring order to the realm of LLS, which have produced even more definitions and classifications, attempts have been made to reshape the concept of a strategic language learner, and account for strategy use in the context of learner’s SR. One example is the creation of a framework for motivational strategies by Dörnyei (2001). According to Dörnyei (2001), the five major classes of self-motivating strategies include: commitment control strategies, assisting the learner in their pursuit of their original goal, such as thinking of positive outcomes, incentives, or rewards, or anticipating the consequences of failure in goal attainment; metacognitive control strategies, employed by the learner in order to handle concentration issues, such as procrastination or distractors, for instance getting rid of disturb- ing classmates; satiation control strategies, applied by the learner in order to fight boredom and add excitement to the learning task, for example changing the characters in a role-play; emotion control strategies, assisting the learner in managing their moods, emotions, for managing disruptive emotional states, such as engaging in breathing exercises, yoga, or resorting to relaxing visualisations;

106 environmental control strategies, triggered to dispose of unwelcome environmental influences and make the best possible use of positive environmental settings, like choosing an air-conditioned library to study rather than a hot room. Some researchers claimed that Dörnyei’s (2005) model of SR and the learning strategy use models should not be treated as incompatible since they allow to account for two various aspects of foreign language development. This point of view was supported by Gao et al. (2006), who concluded that the apparently contradictory models could, in fact, allow researchers to account for two aspects of language learning: SR could be compared to the initial driving forces while strategy research would then represent the outcome of their operation. Other specialists, such as Rose (2012), even contested that the model of motivation control could be subject to similar definitional fuzziness and insufficient theoretical base as the formerly criticised functional classifications of LLS, and argued that shifting from investigating LLS as such to investigating them as part of self- regulatory processes might be compared to throwing LLS out of the bathwater, implying that research in foreign language development would suffer if efforts to investigate LLS were to cease. It is therefore surprising that some papers which discuss the relationships between LLS and SR advocate for replacing the strategy framework with the SR framework. In his view, such shift of focus could only result in incomplete findings about language learning. Although it may seem obvious, it needs to be emphasised SRLL involves the development, use, and successful management and adaptation of LLS. In order to illustrate the connection, Oxford (2017: 112) referred to numerous examples of both theoretical and empirical investigations, such as the works of Zimmerman and Martinez-Pons’s (1986), elaborating on the use of self-regulated learning strategies by successful students, or Winne and Perry’s (2000), who investigated the way that self-regulated students addressed potential problems in the learning process. Introduced by Oxford (2011), the Stra- tegic Self-Regulation (S2R) Model for Language Learning constitutes a bridge between the studies of SR in psychology, including the domain-specific findings of educational psychologists referred to in the previous subsection of the present chapter, and the studies of language learning strategies, carried out by language educators and applied linguists, and discussed in sections 2.1, 2.2, and 2.3. Central to the model remains, however, the reworded definition of LLS in which Oxford (2017: 79) aimed to identify the prototypical factors of strategies:

107 L2 learning strategies are complex, dynamic thoughts and actions, selected and used by learners with some degree of consciousness in specific contexts in order to regulate multiple aspects of themselves (such as cognitive, emotional, and social) for the purpose of (a) accomplishing language tasks; (b) improving language performance or use; and/or (c) enhancing long-term proficiency. Strategies are mentally guided but may also have physical and therefore observable manifestations. Learners often use strategies flexibly and crea- tively; combine them in various ways, such as strategy clusters or strategy chains; and orchestrate them to meet learning needs. Strategies are teachable. Learners in their contexts decide which strategies to use. Appropriateness of strategies depends on multiple personal and contextual factors.

The relevance of strategies to SRL can specifically be inferred from the following qualities, elaborated on by Oxford (2011: 14) and discussed in section 2.3 the present chapter. The organisation of the Model is simplified, that is each strategy category is accompanied by a parallel set of metastrategies. Oxford (2017) emphasised that these sets do not vary between each other, but obviously, the strategies are dimension-specific. Ba- sically, the distinction between strategies and metastrategies consists in the principal function of the latter. Oxford (2011: 14) described them as “crucial mental processes or tools” employed by language learners in order to control and manage the use of strategies from each strategy dimension. Referred to as executive functions, metastrategies were later described by Oxford (2017: 218) as assisting the learner in choosing whether or not to use a particular strategy and evaluating its performance through paying attention, planning, organising learning and obtaining resources, and monitoring and evaluating. It was also explicitly asserted by Oxford (2011, 2017) that the metastrategies could only be successfully set in motion after a learner has accessed their metaknowledge. The latter concept was originally defined by Oxford (2011: 289) as “knowledge underlying the use of metastrategies that enables a learner to learn more effectively” Six types of metaknowledge were distinguished by Oxford (2011, 2017): person knowledge, involving the understanding of people’s characteristics of language learners, e.g. their preferred learning and cognitive style, elements of SWOT analysis, as well as learning goals; group or culture knowledge, i.e., learners’ awareness of the cultural norms and expectations in both, their L1, and L2 culture; task knowledge, consisting in the right understanding of the demands imposed by a given language learning task; whole-process knowledge, going beyond the above, and concerning learners’ interest in L2 learning success, which requires thinking of future aims and outcomes;

108 strategy knowledge, regarding learners’ familiarity with language learning strategies, and involving both direct, and indirect strategies; conditional knowledge, which orchestrates the application of all other five types of metaknowledge (that is enables the learner to use the previous types of knowledge). Equipped with the necessary metaknowlege, and making efficient use of LLS, the self-regulated language learner is thus likely to display the following qualities, listed by Oxford (2011: 15, 2017:115, 130): active participation in the learning process; focus on learning goals by controlling different aspects of learning; ability to regulate their own cognitive and affective states, as well as observable behaviour, and learning conditions; control of their learning beliefs; ability to link elements of declarative knowledge to procedural (automatic) knowledge; ability to choose between a number of strategies on the basis of their appropriateness regarding their goals, needs, context, learning styles, and other individual variables; resistance to overreliance on a single strategy in various contexts; awareness of the relationships between strategy use and performance and language learning success; self-efficacy, overall greater sense of agency and autonomy. To summarise, as advocated by Oxford (2011: 40-41): the S2R model attempts to synthesise the findings of three scientific traditions in which research into LLS has been conducted, i.e., the psychological, social-cognitive, and sociocultural tradition; it caters for an appropriate balance between the above dimensions, not showing preference of any of them; it widens the meta- dimension of strategic competence, thus not limiting it to metacognition (which is related to 1. and 2.); it states that metastrategies are accessible at different levels, i.e., the task level, and the whole-process level; it distinguishes between deep processing and surface strategies, postulating the special role of deep processing;

109 it has double utility, which consists in its potential for ordinary, and crisis-like situations for using LLS; it is not overwhelmingly complex, comprising 19 strategies in total, but at the same time allows for in-depth analysis of strategy use at the level of particular tactics, and presents strategy chains; it recognises the findings of neuroscience and links them to strategy use; it includes references to a number of techniques for strategy development and assessment, which certainly adds up to its practicality. The complex relationships between the constructs learner autonomy, agency and SRL certainly require an explanation. According to Oxford (2017: 125), while each of these constructs is volitional in nature, agency and autonomy are relatively general notions, whereas the focus in accounts of SR is more towards more specific processes. On the one hand, without the sense of agency, language learners are unlikely to employ any LLS or initiate any goal-oriented actions, but on the other hand, as pointed out by Wil- liams et al. (2015), strategic knowledge is one of the pillars of agency. Also, while a number of researchers, such as Wenden (1991), Cotterall (1995), Oxford (1999), Mi- chońska-Stadnik (2008), and Griffiths (2013) view LLS as an important element of learner autonomy. At the same time, it also remains true, as pointed out by Oxford (2003, 2008), that whereas autonomy results in strategy use, autonomy also capitalises on strategy use. In order to account for the relationships between learner autonomy, strategy use, and SRL Oxford (1999, 2011, 2017) employed Vygotsky’s (1978, 1986) theory. In Ox- ford’s (1999: 111) view, the aim of learning consists in the creation of “an independent, self-regulated, problem-solving individual”, and the process needs the support of “the more capable others”, whom Vygotsky metaphorically referred to as “scaffolding”, assisting learners’ movement through their zone of proximal development (ZPD), which is a representation of the difference between what a learner knows and can do unsupported and what they could achieve once provided with just enough of support. Of course, the movement can be enhanced by efficient strategy use.

110 2.5.4. Investigations of SRLL

On the basis of Zimmerman’s (2002) Self-Regulation Model, Studenska (2011a) developed the Foreign Language Learning Self-regulation Difficulty Inventory (FLLSDI) to measure SR difficulty reported with regard to three basic stages of learning: planning, performing, and evaluation. Conducted in a group of 380 Polish learners of English as a foreign language, Studenska’s (2011a) study was designed to investigate differences in SR difficulty in learning English as a foreign language across genders and educational levels (primary/ lower-secondary/ upper-secondary). Consistently with the results of principal component analysis, four aspects of SR difficulty were scrutinised, strictly corresponding to the four scales of the FLLSDI: difficulty in choosing goals, ways, and conditions for learning; difficulty in planning, organising, and implementing learning; difficulty in motivational and emotional control; difficulty in reflecting and making changes. The results of the study indicated that significant main effects were detected for gender regarding three dependent variables: general foreign language learning self-regulation difficulty (a combination of the four above scales), scale 3 variable, and scale 4 variable. Specifically, the female participants of the study were found to experience less difficulty in planning; however, when analysed separately for each educational level, the differences only proved significant for primary and lower secondary school students. Evaluation of difficulties in self-regulated learning behaviour also differed across genders as twelve types of behaviour were ranked more difficult by the male participants of the study. Men found devoting time for independent learning, looking for people and materials that could help, planning learning in distant future, using various ways of learning, putting learning plans into practice (scale 2), increasing one’s willingness to learn, continuing learning instead of engaging in more pleasant activities, avoiding distracters, taking advantage of one’s dispositions (scale 3), making changes to learning when necessary, identifying causes in one’s learning results (scale 4), and choosing the way of learning to obtain the results one wants to achieve (scale 5) significantly more difficult than women. Also, the impact of the educational level on planning, organising, and implementing learning proved to be significant proved to be significant - a growing, linear trend was observed from primary to upper secondary school.

111 Sampson (2012) made a number of suggestions on how to introduce SR into the language learning classroom after investigating the relationship between learners’ self- images, socially-possible self-images, and language learning motivation during a 15- week course of English. The participants of the study included 34 female Japanese students at the Faculty of International Communication. The design of the study was based on the framework by Dörnyei and Ushioda’s (2009) Motivational Self System, and thus involved the Ideal L2 self, the Ought-to L2 self, and L2 Learning Experience. Three major research questions were considered while the three-cycle action research took place: 1. Whether collecting information about learners’ visions of their L2-selves would facilitate the development of more motivating lessons, 2. What self-conception-building lesson elements might motivate students, and 3. How do the participants perceive the change in their L2-self-image. Data was collected from a free-writing exercise in which the participants described the Best Possible English-Self, learning journal entries written after completing tasks based on Nunan’s (2004) concept of task-based language teaching19, and after participating in activities expected to foster goal achievement strategies. Overall, after conducting qualitative analysis of participants’ expectations and reflections, Sampson (2012) came to the conclusion that the evolution of the L2-self enhanced a more self-centred approach to learning English. Also, numerous connections between the learners’ ought-to selves and refining of ideal-self-images were observed. Although the participants were representatives of a collectivist culture, they were able to provide comprehensive conceptualizations of the selves, and actively reflected on the effort required to meet their individual goals. A final conclusion could thus be drawn that ensuring a student-centred learning environment can improve learners’ focus on their learning outcomes in general. A qualitative study of SR in learning the Japanese written characters, kanji, was conducted by Rose and Harbon (2013). Twelve volunteers, students of kanji courses at university, participated in a series of 10 interviews throughout the academic year. The set of interview questions was constructed on the basis of the study carried out by Tseng et al. (2006) and modified after a pilot study run a year prior to the year of the main study. Most of topics which were selected for analysis corresponded to the five major classes of self-motivating strategies described by (Dörnyei 2001: 133); however, additional themes

19Sampson (2012: 321) specifically reported employing schema building, controlled practice, authentic listening, focus on linguistic elements, freer practice, and a pedagogical task.

112 emerged, including goals, procrastination, boredom and interest, stress, and, finally, the learning environment. Participants’ SRL was assessed by the interviewers in order to present a synthetic picture of extremely well self-regulated learners, typical cases, and extremely low cases of self-regulation. Instruction in the kanji classes was focused on character list learning rather than improving the overall reading process and classes did not assist students in developing their comprehension skills in reading authentic materials. Regarding commitment control and goal-setting challenges, goal attainment proved more difficult for relatively more advanced learners20, which was due to the degree of difficulty of goal, and the amount to learn. Moreover, the most advanced learners goal attainment linked to both, future education, and future employment, which added up to the overall pressure and experienced stress. On the other hand, the relatively less proficient learners concentrated on short-term, and more easily manageable goals, whose attainment was only assessed in the classroom. A limited proportion of advanced kanji learners attempted to handle commitment control through into regular self-study tasks, which resulted in clearer goal management and increased confidence21. Learners, especially the relatively more advanced ones, also reported experiencing numerous negative emotions while pursuing their kanji learning goals, including frustration and defeatism, and had to deal with negative self-thoughts. Some of the most advanced learners even considered overall goal abandonment and changing their future career plans because of the experienced problems. Difficulties were also reported with reference to satiation control, as both, increasing boredom, and decreasing satisfaction was reported by the more proficient participants of the study. Some students were successful in dealing with satiation issues by employing the strategy of self-rewarding after completing smaller portions of tasks whereas in one case it was simultaneous participation in learning kanji and physical exercise that helped the student overcome boredom. Also, boredom was successfully dealt with by altering environmental settings, which drove Rose and Harbon (2013) to conclude that environmental control as such might not be a separate strategy, but a means to an end. On the basis of the findings, it was advocated that kanji teachers should support students’ motivation not only through inclusion of more authentic materials in the classroom, but also through systematic support in self-regulation related problems, such as

20 The most advanced learners aimed to review and study between 1000 and 2000 kanji characters. 21Rose and Harbon (2013: 101) indicated that similar conclusions were reached by Bandura and Schunk (1981) and Dörnyei (2001).

113 breaking learning goals into manageable tasks, indicating connections between the material learnt and the Japanese culture and everyday reality, reflecting on progress, and assisting students in taking emotional control of the learning process. Although based on the traditional LLS paradigm and strategy classification reflected in Oxford’s (1990) SILL ver. 7.0, the case study conducted by Véliz (2012) did manage to provide a number of interesting insights on the phenomenon of SRLL, especially with regard to the interrelationships between motivation and strategy use. By conducting semi-structured interviews with two third-year students participating in teacher- training phonetics courses, the researcher aimed to verify whether both direct and indirect strategies are improved for pronunciation improvement, and investigated the strategies applied in pursuit of native-like accent, the extent to which the participants referred to formal instruction in the learning process, and, finally, the relevance of motivation in learners’ attempts to attain native-like accent. On the one hand, certain similarities were observed, such as increased use of indirect strategies, especially metacognitive strategies, which served the purposes of centring, arranging, and planning learning. Specifically, the investigated individuals reported participating in a preparatory course or doing ‘pronunciation research’, which encompassed activities such as monolingual, and pronunciation dictionary use, but also looking for contextual use of words and phrases on YouTube. On the other hand, differences were observed regarding social strategy use, which was, in fact, modified by learners’ personal characteristics in order to boost SRL. The learner who was more retracted and embarrassed at the thought of making contact with native speakers of English would not engage in practising the language actively, whereas the learner who was more open and outgoing tended to look for opportunities to engage in conversations. Adopting well-suited indirect strategies did not merely result in a more effective application of direct strategies, such as creating mental images of new phonetic utterances, or engaging in routine practice tasks, such as repetition of new words and phrases, but also had a positive effect on learners’ motivation. Veliz (2012: 214) concluded that a similar degree of autonomy was, in fact, developed by both learners through the use of different, well-adjusted strategies, and both investigated learners benefited considerably from their positive perception of their educational attainment. Several attempts have been made to develop an instrument for SR measurement, both for selected aspects of language learning, such as vocabulary, and for SRL in general. The Self-Regulating Capacity in English Language Learning (SRClang) was

114 developed and examined with regard to its psychometric properties by Liu and Lee (2015). Over 500 high school students from east Asia completed the questionnaire in order to provide data on the instrument item-model fit, response category use, appropriateness for item difficulty level, reliability, and dimensionality. Based on the Self-Regulating Capacity in Vocabulary Learning (SRVoc) scale (Tseng et al. 2006), Liu and Lee’s (2015) questionnaire originally consisted of 27 items, 3 of which were subsequently removed due to their unsatisfactory fit to the item model. Infit and outfit item-fit means square were adequate for each category of the instrument. However, the differences of the maximum and minimum values between person trait scores and item difficulties suggest that the scale might be not comprehensive enough to investigate extremely-well/poorly self- regulated learners. The employed reliability measure, Pearson separation reliability estimate, indicated that the instrument could be used so as tool to differentiate between individuals. Finally, as far as the dimensionality of the instrument is concerned, analysis of variance indicated that the model explained less than 60%, and therefore it could not be assumed that the universe of content provided in the answers to the questionnaire items actually constitutes a single dimension of self-regulation in learning a foreign language. In other words, creating an instrument to measure language learners’ overall SR remains one of the challenges still faced by SR researchers. Ziegler (2015) investigated the efficiency of using SRCVoc scale introduced by Tseng et al. (2006) in predicting learners’ motivational characteristics through the prism of SR. His study of convergent validity of the two constructs conducted on 572 Saxon learners of English (grades 4-9) implied that whereas the SRCVoc significantly predicted all the scales of SRL measured by the MSLQ (Pintrich et al. 1991), effect sizes ranged from small for the three scales of Test Anxiety, Control Beliefs, and Extrinsic Goal Ori- entation to medium for the scales of Academic Self-Efficacy, Intrinsic Goal Orientation, and Task Value. The study revealed that applying the SRCVoc did not produce overall satisfactory results with regard to measuring learners’ overall capacity for SRL. Alt- hough, according to Ziegler (2015: 646), the questionnaire can serve as a “measure of the strategic behaviors characteristic of the performance phase of the self-regulated learning cycle”, perhaps its appraisal as a reliable and valid and reliable measure for strategic vocabulary learning is a bit of a hasty conclusion, especially as neither of these two goodness criteria was psychometrically analysed in the paper.

115 Aiming to select the best model of writing strategies, Teng and Zhang (2016) developed and validated the Writing Strategies for Self-Regulated Learning Questionnaire (WSSRLQ) in order to provide insights into language learners’ cognitive and metacognitive processes, as well as their social and motivational behaviours. Their study was conducted in a group of 790 Chinese university students learning English as a foreign language, and involved three major phases, item generating, pilot study, and psychometric evaluation. The final version of the 7-point Likert scale questionnaire consists of 40 items that divided into 9 subscales, including Goal-Oriented Monitoring and Evaluat- ing(GME), Idea Planning (IP), Peer Learning(PL), Feeedback Handling (FH), Interest Enhancement (IE), Emotional Control (EC), Motivational Self-Talk (MST), Text Pro- cessing (TP), and Course Memory (CM). The results of confirmatory factor analysis indicated an acceptable model fit for the 40-item, 9-factor construct, with significant pa- rameter estimates, robust reliability of each of the subscale, and statistically significant correlations between the subscales. In particular, strong correlations were detected between Goal-oriented Monitoring and Evaluating strategies were strongly correlated with Peer Learning and Interest Enhancement, which does not only confirm that LLS work in chains, as suggested by Oxford (2003), but also suggests that learners aware of the need to monitoring their task goals are also more likely to regulate their social behavior and manage their intrinsic motivation to control their involvement in foreign language writing tasks. It was observed that the model assuming SR as a hierarchical construct explaining 9 SRL strategies produced the best fit. Also, the four strategies belonging to the realms of motivation and metacognition: Idea Planning, Goal-oriented Monitoring and evaluating, Motivational Self-Talk, and Interest Enhancement, had relatively the largest loadings on SR, which proves the essential importance of planning and metacognition in the self- regulated development of writing skills in a foreign language, and supports the sociocog- nitive view of SRL processes. Rose et al. (2018) conducted a systematic review of studies of LLS conducted after Tseng et al. (2006) appealed for replacing the traditional strategy paradigm with SR- based research. Designed to identify adaptations in strategy studies, the review also took into account the ways of theorising LLS, research instruments, and the methodologies underlying the analysed studies. Three major categories of studies were identified by Rose et al. (2018: 155): studies adopting SR and SRL as the construct essential to their research framework;

116 c) studies allowing contributions from SR, but still within the traditional strategy paradigm; d) studies aimed to develop new instruments, and investigating the link between strategic learning and yet other theories. One of the advantage of category 1 studies, advocated by Dörnyei (2005, 2015), consists in the fact that SR has already been a well-established construct in psychology, and adopting the paradigm would thus result in resorting to apparently more sophisticated or efficient tools of investigation framed within methodologies with a long-standing tradition. Type 2 studies, certainly recognising the tradition of strategy research, did not only manage to respond to methodological criticisms, but also showed promising areas for generating even more knowledge about the learner, such as investigating the context of strategy use and the importance of the language task. Also, studies such as that conducted by Ardasheva and Tretter’s (2013) have proved that even investigations based on Ox- ford’s (1990) SILL (ver. 7.0) can still be successfully adapted and become valid, population-specific ways of analysing strategic learning. Finally, there are many researchers who adopt Gao’s (2007) standpoint, according to which, strategy and SR researchers explore different parts of the same process, and may prove to be of complimentary, rather than substitute nature. This has led to the emergence of hybrid models, such as Oxford’s (2011, 2017) S2R Model’, which are still awaiting to be conceptualised and operationalised in research. In the concluding article a special volume of Studies in Second Language Learn- ing and Teaching, dedicated entirely to strategy research, Pawlak and Oxford (2018) discussed the potential directions of research into LLS, addressed the emerging methodological challenges of such research, and appealed for increasing its focus on pedagogical implications. It was stressed that both types of studies in LLS, those of quantitative nature, conducted on a macro scale in large groups of learners, and those investigating the strategy use by individuals or in small groups of learners, could contribute profoundly to a better comprehension of the intricate nature of LLS. However, as some critical issues have been reported over the years, a firm belief was expressed that future studies ought to not only focus on strategy use with regard to specific subsystems and skills of the target language, the target culture, and affective factors on the part of the language learner, but also respond to concerns. Expectations therefore exist that future research focuses on strategy use in learning the culture of the target language, differences in strategy

117 repertoires in the development of following foreign languages, and, finally, self-regulation in foreign language development. With regard to the latter, a conviction was expressed that looking into the well-established concept (LLS) from a new perspective (SR) could, in fact, shed new light on strategy use. Consequently, Pawlak and Oxford (2018) called for empirical studies that would provide insights into how strategy use contributes to the development of language learners’ SR, from the behavioural, emotional, and motivational perspective. At the same time, they criticised the idea of an overall replacement of the strategy research paradigm with the SR paradigm, clearly opting for amalgamation of research results rather than dismissing the plethora of findings due to their traditional methodological background.

2.6. Conclusion

The growing level of awareness about the complexity of the strategy construct has not discouraged researchers from investigating the use of LLS and relevant variables. On the one hand, it remains true that factors such as the lack of consensus concerning the very name of the strategy ID (language learning strategies vs. language learner strategies), the co-existence of a multitude of strategy definitions, and the partly contradictory insights about strategy features can all exacerbate the challenge of arriving at some ultimate conclusions concerning the role of LLS in foreign language development. On the other hand, the dissemination of insights into the ways of using LLS, the patterns of their use, and the influence of a number of IDs on strategy use have all influenced the way that LLS are investigated. Originally, mostly accounted for from the functional and quantitative perspective, LLS have become a construct that is investigated in relation not merely to skills, functions, aspects of language such as grammar or vocabulary, and other learner IDs, also in context of particular tasks, and as a pillar of learners’ self-regulation (cf. Oxford 2017; Oxford and Amerstorfer 2018). Perhaps the final conclusion which arises at the end of the present chapter will partly echo the view present in Oxford’s (2017) second book on LLS as part of learners’ self-regulation. It seems that given all the complexity and ambiguity which relate to the strategy construct, definite and arbitrary judgements should best be avoided when addressing the scope of work and efficiency of strategy research. A trend to investigate language learning strategies in context has clearly

118 emerged and it is a matter of time before new research results can be analysed and conclusions can be summarised and generalised. It is already clear, though, that the context for analysing LLS involves both the micro- and the macroenvironment of strategy use, the former consisting in the use of LLS with reference to particular language tasks, and the latter requiring the inclusion of a whole range of learners’ individual characteristics, to be discussed in Chapter Three.

119 Chapter 3: Variables mediating LLS use

3.1. Introduction

As emphasised by Cohen (1998, 2014), the strategic character of language learners’ conscious choices in the process of foreign language development could only pertain to the given learner who makes them. In an attempt to account for the idiographic nature of LLS use, the present thesis aims to account for strategy application not merely from the microcontext perspective of strategy use, elaborated on in the previous chapter, but also addresses the strategy user in a more systematic way through considering a number of possible influences of IDs of various origin and character. Since the very construct of LLS is an extremely broad one and defies a strict classification as either behavioural or cognitive, as asserted by Michońska-Stadnik (2008), it is necessary to capitalise on the insights provided from both perspectives. Although the factors influencing learners’ strategic choices could be categorised as individual, group, or contextual (cf. Oxford and Nyikos 1989; Cohen and Macaro 2007, Pawlak 2011b), the considerations of the interactions between each of the analysed IDs and strategy use in the first section of the present chapter do not reflect this organisation. Instead, they comply with the way of analysis suggested in the previous Chapter and attempt to elaborate on the complex network of relationships which build the macrocontext of strategy use. This involves the impact of the language being learnt, the language learning experience, IDs such as age, and gender. Moreover, strategy use can also be mediated by learners’ beliefs, learning styles or language aptitude. Certain implications arise from learning a TL in a specific L1 cultural background and instructional settings. Attention should also be paid to the importance of the micro-context of strategy use, that is of the language task in learners’ strategic choices. Finally, various strategies are used by learners who differ in terms of attainment. It could be argued that the context of strategy use can be analysed from at least two various perspectives, including the micro- perspective of certain specific strategies employed in order to handle a specific language task, and the macro- perspective. The latter one allows researchers to investigate the intricate relationships between LLS and other IDs, such as learning styles, motivation, language aptitude, or affective factors,

120 creating bundles of factors which certainly influence strategy use, as pointed out by Paw- lak and Oxford (2018). To use Griffith’s (2015) words,

We have learnt that strategies are important, and that successful learners have a large repertoire of different types of strategies that they use frequently. But strategies are not the whole answer, and the strategies that are chosen and which are effective depend on the context, the learning goal, and the learner’s own unique set of individual characteristics (Griffiths 2015: 432).

The interaction between LLS and other IDs has been addressed and investigated by strategy researchers since the early years of interest in strategy use. Rubin (1975) listed a number of IDs which she expected to influence the use of strategies, such as learner’s age, the context of using the target language, individual learning styles, or cultural differences. Both LLS and each of the categories mentioned by Rubin (1975) are influenced by other factors, such as cognitive styles, personality traits, past educational experience (which may actualise as transfer of training), general educational achievement, or experience in learning other foreign languages. Hence, in order to fully understand the choice and use of language learning strategies, it is inevitable to pay attention to other aspects of individual variation among language learners. Oxford and Nyikos (1989) provided a synthetic list of factors reported by strategy researchers as essential determinants of learners’ strategic choices and categorised them as either learner variables or situational variables. Later, Takeuchi et al. (2007) and Pawlak (2011) grouped the factors affecting strategy choices and use under three major labels, including individual, contextual, and group variables. The following list of variables mediating strategy use could thus be formulated: the language being learnt; language learning experience; national origin, culture, and instructional settings; age; gender; learners’ beliefs and attitudes22‘; learning styles; language aptitude

22 According to Oxford and Nyikos (1989), learners’ attitudes are one of a number of affective variables, which also include their personality. Since personality cannot really be analysed in terms of an affective variable, regardless of the framework chosen in order to investigate it, it is considered as a separate factor.

121 a number of motivational variables, including the dynamic changes in their motivational intensity, as well as their learning goals (also referred to as motivational orientation) and learners’ career orientation; learners’ attainment; specific requirements of the performed language tasks; the degree of learners’ self-awareness23; personality24. Adopting the above list of factors influencing strategy use as a starting point, the following subsection of the present chapter seeks to explore their connection to learners’ strategy choices and strategy use. The overview of theoretical standpoints is supplemented with examples of empirical investigations so as to provide evidence of how various IDs might impact learners’ selection and application of LLS. The final two points, 12 and 13, are not covered in the present chapter. Learners self-awareness is discussed in the previous chapter along with language learners’ self-regulation while personality is discussed in Chapter One. Empirical findings regarding the relationships between personality and LLS are subject of the main study and discussed in Chapter Five.

3.2. The impact of the language being learnt on the choice of LLS

It seems obvious that the specificity of the target language (TL) exerts a powerful influence on the way it is learnt, as voiced by Oxford and Nyikos (1989), Grenfell and Harris (1999), Takeuchi et al. (2007), Pawlak (2011), Grainger (2012), Oxford and Gkonou (2018). First of all, the availability of the foreign language to the learner can vary, even though globalisation and the Internet make virtually all languages more accessible. Some languages, such as English, enjoy a more privileged (or dominant) position, and hence, the learner does not have to strenuously look for L2 input. Also, as pointed out by Harmer

23 This notion could be associated with the concept of person-knowledge, (cf. Oxford 2011), but also with the concepts of agency, self-control, autonomy, and self-regulation (cf. Pawlak 2011, 2018; Dörnyei 2005, Dörnyei and Ryan 2015; Oxford 2017). The correspondence between LLS and self-regulation is discussed in the previous chapter (see sections 2.4 and 2.5). 24 Oxford and Nyikos (1989) distinguished between the so-called personality characteristics, or, long-term traits, and general personality type, measured by MBTI, and therefore it could be implied that the researcher did not favour any particular approach to investigating personality, since, on the one hand, the ‘long-term traits’ represent a construct present in the trait approach, whereas personality types investigated by means of MBTI are said to be of Jungian, and thus, psychodynamic, origin.

122 (2001: 8), it is possible to learn a foreign language in different settings, that is in countries where the language is spoken as a first language, in countries where the language is spoken as a second, and sometimes also, official language, and, finally, in places where the language is recognised as important, for instance considering its cultural or economic impact. Moreover, even in the specific case of learning English, essential differences exist regarding the accessibility of both educational and authentic materials comparing Amer- ican, British, or Australian, and other resources, and the availability of dedicated learning resources is asymmetric, with the British ELT industry enjoying the dominant position on a number of European markets. Secondly, if a language learner makes an independent choice of which foreign language they want to study, their choice may be motivated not merely by practical or instrumental reasons, but also by their fondness of a certain culture which they associate the language with. For instance, one could prefer German rather than English because of their penchant for orderliness, Spanish or Italian because the ap- peal of their prosodic features, and French or Latin because of their attraction to high culture. Consequently, learners willing to master different languages, may favour different strategies. For example those that require planning and good organisational skills, participation in authentic communication with L2 speakers, or increasing the input by watching cultural programmes or films in L2. Also, as indicated by Grenfell and Harris (1999: 30), progression in reaching attainment targets by language learners depends on the type and form of the language learnt, its style, degree of grammatical sophistication, the approached content, the familiarity of the TL, encompassing, among others, its relative degree of unpredictability and the possibility and/or need for improvisation, level of support possible to obtain from visual clues, cognates, gestures, or written clues. Clearly, these features can be accounted for in both interlingual and intra-lingual comparisons, and even related to the differences between learning a language as such and learning a language for specific purposes. Finally, it is inevitable to grant at least a certain role to the developmental stage of the L2 teaching methodology. In the specific case of English, its popularity, availability, and development of the ELT industry have all led to a situation in which the language learner can enjoy a plethora of resources, both online, and those more traditional ones. The evolution of methods and approaches has undoubtedly left a trace in today’s eclectic reality of the ELT business, making the amount of instructional content in ELT books greater, the range of existing learning techniques and strategies wider, and thus increasing the probability of each learner finding their own way of

123 learning English as a foreign language. Other languages studied as foreign languages might not have entered this stage of methodological development yet, and hence learning and teaching resources might be relatively scarce, the level of pedagogical competence of language authors and course instructors could be lower, and, consequently, the range of available methods, techniques, and strategies to learn these languages, could be much narrower. The examples of studies referred to below are provided to illustrate the influence of the (chosen) TL on learners’ strategy repertoire and usage. According to the results of two studies reported by O’Malley and Chamot (1990), learners of Spanish and Russian used cognitive strategies slightly less often than learners of English. Conducted in groups of 70 EFL high school students, 67 high-school learners of Spanish, and 34 college students of Russian, the comparison across the language being learnt revealed that the investigated learners of Spanish and Russian also used metacognitive strategies relatively less frequently, and hardly ever resorted to social or affective strategies. Some strategies, such as memorisation of longer pieces of text, were only applied by the learners of Russian, confirming earlier findings reported by Chamot (1987), according to which learners of Russian demonstrated greater overall scope of strategy use than Spanish students. The results also lent support to the outcomes of Politzer’s (1983) study, comparing the learning behaviours in a group of 90 university students learning French, German, or Spanish as a foreign language, according to which learners of Spanish tended to engage in relatively fewer behaviours which were expected by the researcher to boost their overall learning attainment25 in comparison to learners of French or German. Additionally, Politzer’s (1983) study also demonstrated that learners of German and Spanish were relatively more oriented towards social interaction and increasing their TL input. It needs to be pointed out that the identified differences in strategies could also have reflected differences in instructional practices. Schmidt and Watanabe (2001) conducted a comparative study of more than 2,000 students of Mandarin Chinese, Tagalog Filipino, French, Japanese, and Spanish. The analysis of variance in learners’ motivation and strategy use across target languages proved that the investigated learners of Filipino were the most frequent strategy users, followed by learners of Spanish, while learners of Mandarin Chinese used strategies significantly less frequently than learners of the remaining languages. Regarding the use of

25Politzer (1983) investigated three types of students’ self-reported behaviours, including general behaviours, classroom behaviours, and interactions with others.

124 particular strategy categories, the analysed learners of French and Filipino were the most frequent users of cognitive strategies, which were relatively last frequently used by learners of Japanese. Social strategies were relatively more often used by learners of Filipino and Spanish, while relatively less popular with learners of Chinese. Students of Filipino also outperformed all other students in the use of strategies consisting in an efficient use of study skills and coping strategies. Finally, from the motivational perspective, it turned out that the learners of Spanish were principally interested in satisfying the language requirement of the university curriculum, and thus, being characterised with relatively lower intrinsic motivation, they reported to be less willing to participate actively in activities which require increased effort. To a large extent, the description also matched the learners of French; however, not all of the characteristics were statistically significant for them. Being the rarest strategy users, the learners of Mandarin Chinese were at the same time the most self-confident, the greatest believers in their own aptitude, and the least anxious, which can be linked to the reported personal, ‘heritage’ link to Chinese. The learners of Japanese were similar with respect to their self-reported connection to the heritage language. At the same time, the investigated learners of Japanese also acknowledged the difficulty of learning kanji and had relatively low attainment-related expectations. This combination of characteristics had a number of implications for their learning behaviours, which generally reflected the traditional pedagogical focus on grammar, vocabulary, reading, and writing. Similar findings were revealed by Rose and Harbon (2013), who investigated the motivational strategies of kanji learners - a study that is given more detailed description in the section dedicated to self-regulation in foreign language development. Finally, Schmidt and Watanabe (2001) also observed that the learners of Filipino exhibited both high social motivation, which allowed them to embrace innovative activities, such as using authentic materials, and made their learning more goal-oriented in general. Overall, the research may serve as valid evidence that the TL itself certainly influences the choice and use of LLS.

3.3. The impact of the duration of language learning experience on the choice of LLS

According to Wilczyńska (1999: 123), each time an act of learning happens, an individual pays attention to a new concept, manages to remember and comprehend it, reflects on it,

125 and imagines it in their own, personalised context. What happens next? It is possible to anticipate a number of developments that might emerge while individuals gain language learning experience. On the one hand, increased length of exposure to both the target language itself and learning instruction in the second place, could be expected to boost the volume and diversity of learners’ strategy repertoire, as asserted by Oxford and Nyi- kos (1989), Ehrman and Oxford (1990), and Pawlak (2011a). On the other hand, however, it would be logical to link the length of the foreign language learning experience to an increase in the learner’s awareness of the efficiency of strategy use and, perhaps, limiting the range of the applied strategies to those that the learner regards as potent or useful26. As Wenden (1991) pointed out, during the learning process, learners acquire strategic knowledge, and make decisions which strategies they continue using, and which strategies they abandon, thus developing a resource file on strategic knowledge. Additionally, they also accumulate knowledge of linguistic tasks, regarding the resources that are required for task completion, and the possible ways in which a given task can be accom- plished. According to Wenden (1991), foreign language learners also develop a resource file on task knowledge, and even a resource file on the nature of communication. The threads of the relationships between strategy use and language learning experience are thus complex, as they are interwoven with the impact of learners’ increasing proficiency, and instruction in strategy use. specifically recommended that language learners should reflect on their learning experience and evaluate strategy use:

Every so often, you should evaluate the learning strategies you are using. Some of them may be no longer useful because you have learned new ones or because you have reached a level of proficiency where they no longer help and you need to develop new ones. For example, at lower levels you may need to look up some words in a dictionary or guess their meaning from context. At higher levels, however, you might be able to figure out their meaning based on the meaning of their roots, your knowledge of word formation and/or comparison with vocabulary that you already know. (Leaver et al. 2005: 60)

Much as capturing the learning trajectories of foreign language learners and accounting for the amendments that they make to their strategy repertoires as they become more experienced is fascinating, it remains a challenge for strategy researchers due to all the factors which certainly interact with the language (learning) experience. The

26 Interestingly, Williams et al. (2015) pointed out that language learners might develop ‘unhelpful’ strategies, for the very purpose of coping with difficult situations, such as mumbling answers in the classroom in order to avoid being reprimanded, or waving hands to show that they know the answer to a question, whereas the purpose of doing so might be to avoid being suspected of not knowing it.

126 challenge becomes even more daunting as language learning experience can be analysed as either TL-specific, or cumulative, and thus involving learning background in any foreign language, or even L1. According to Williams et al. (2015: 53), learners who have studied more than one foreign language are likely to compare experiences across languages, and form TL-specific self-concepts. A brief description of the findings from two studies aims to offer more empirically-grounded. After investigating a group of 275 Turkish students of English with Oxford’s (1990) SILL (ver. 7.0), Uztosun (2014) arrived at a number of interesting conclusions regarding the links between learners’ educational backgrounds, their language learning experience, and strategy use. The participants of the study belonged to three categories: 38% attended preparatory (‘year 0’) courses, 34% were first grade, and 28% were second grade students. Overall, the participants favoured compensation, metacognitive, and social strategies, which lent support to previous findings concerning the strategic preferences of Turkish students revealed by Yılmaz (2010). The most frequent strategies indicated by the investigated students included paying attention to the interlocutor, asking the interlocutor to slow down/repeat, and using a different word/phrase with a similar meaning. Social strategies were used significantly more often by the least experienced learners, which could imply that the ‘year 0’ students might require more assistance of ‘the significant others’ in learning a foreign language27. Finally, the relatively little dependence on cognitive strategies, which might be worrying, especially in a group of future English majors, could also indicate that Turkish learners need more task-specific training in strategy use. Conducted in a group of 107 BA university students of English, Pawlak and Kier- masz’s (2018) study provided a number of insights into the intrapersonal differences in strategy use with reference to L2 (English) and L3 (Spanish, German, French, Italian, Russian, Dutch, and Polish), and the possibility of strategy transfer across languages. In- volving the use of a Polish adaptation of SILL (Oxford 1990), and a series of semi-structured interviews, the study was designed to compare overall strategy use with regard to the above languages and identify the dominant patterns in L2 and L3 strategies. Overall, the reported strategy use was significantly more frequent among in L2 learning than in L3 learning. The differences were most striking for memory, metacognitive, and social

27 A similar conclusion was drawn by by Magogwe and Oliver (2007) with regards to a group of Batswana students learning English as a foreign language.

127 strategies, which were at the same time the only categories where strategy use could be interpreted as high for learning L2. Additionally, individual variation in strategy use was considerably higher in the case of L3, which confirms the essential meaning of individual learning experience in strategy use. The analysis of qualitative data confirmed the close relationship between learning experience and the level of proficiency. Differences in learners’ strategic choices reflected in the analysis of quantitative data were also attributed to the disparity of proficiency levels for L2 and L3. The strategies favoured in L3 were associated with lower-level strategies, such as looking for cognates in L1, relying on gestures, or creating associations with images or sounds. Hence, it could be argued that the study thus also signalled that the interaction between strategy use and language learning experience could be mediated by the level of linguistic competence in any of the analysed languages.

3.4. The impact of cultural variables on their choice of LLS

It is hard to deny the impact of learners’ national origin and culture when analysing the factors potentially influencing strategy choice. Firstly, educational models differ across borders, and hence learners are exposed to different kinds of instruction depending on the country where they receive education, as indicated by Pawlak (2011a). Secondly, some cultures favour more individualistic behaviours while others, such as a great deal of east- ern countries, train the individual mainly to become a useful part of a society, which could also result in some of their desires and needs being unattended to, and, possibly, some strategic choices being unrecognised, or, from a different perspective, never familiar to the learners. It could also be argued that the instructional materials that language learners are exposed to reflect not only a certain philosophy of language as such, or learning a foreign language, but also some underlying values. As pointed out by Canagarajah (1999), these values might stand in contrast to the learners’ own values and result in learners’ adaptation of strategies preventing them from feeling culturally or linguistically dominated, such as drawing pictures of their symbols or objects which they consider valuable on the margins of the course books. Also, as observed by Oxford (1989c), the understanding of what language learning consists in, differs across cultures, and consequently, so does the curriculum. The curriculum might, or might not, involve elements of

128 strategy instruction, and the degree of emphasis on learner autonomy in it might differ considerably. For example, in the Australian, American and Central and Western Euro- pean context, the conviction that strategy training could encourage the development of foreign language skills is generally rather firm, and has been expressed by a vast number of researchers, such as O’Malley and Chamot (1990), Droździał-Szelest (1993, 1997), Grenfell and Harris (1999), Taylor et al. (2006) or Griffiths (2013, 2018). Although some researchers, such as Hoff and Paige (2008), optimistically assumed that some LLS could not only be expected to be universal, but also teachable, and learners, once trained, could make use of them in various setting, regardless of the culture that they originate from, the odds are that in some countries language learners might never be exposed to them due to differences in instructional settings. In Poland, for instance, evaluation procedures might reflect some relatively old methodological assumptions, resulting in a bias in strategy use towards more traditional strategies, such as memorization or repetition, and a neglect of metacognitive, affective, social, and more sophisticated cognitive strategies, as suggested by Pawlak and Kiermasz (2018). Overall, researchers seem to agree that learners’ identities need to be accounted for in studies of LLS, and that LLS repertoire should be inclu- sive of strategies for cultural acclimatisation, identity-stretching, or managing power relationships in different learning situations, as proposed by Lee and Oxford (2008). At the same time, doubts have been raised whether learners’ L1 national culture and origin could be considered as strong predictors of strategy use, and, as indicated from Griffiths’s (2003), and Grainger’s (1997, 2012) studies, learners’ culture should not be interpreted as a barrier in language learning. Examples of relevant research findings are described below. Early studies addressing the impact of the learner’s nationality and cultural background, such as those conducted by Politzer (1983), Politzer and McGroarty (1985), and O’Malley et al. (1985), indicated that rote learning and mnemonic strategies rather than communication-oriented strategies tended to be favoured by oriental learners who were also found to be generally less responsive to training, but at the same time, their strategies seem to be more teacher-centred. The final claim was later supported by Usuki (2000), who identified psychological barriers which were partly to blame for the lack of success in effective adoption of language learning strategies by Japanese students, allegedly too passive in order to make use of some strategies.

129 Commenting on the results of his study investigating the use of strategies by 73 Jap- anese female college learners of English, Takeuchi (1993) linked the participants’ preference for some learning strategies, such as writing notes, avoiding word-by-word translation, analysing words or paying attention over other strategies, e.g. asking questions, using flashcards, looking for opportunities to use English, or writing down one’s feelings in a diary, to the learners’ cultural beliefs and shared values. Magogwe and Oliver (2007), who raised the issue of the cultural specificity of the investigated 480 Batswana learners’ of English strategic choices, reporting that compensation strategies were least frequently used among the six strategy categories of the SILL (Oxford 1990). Their finding was consistent across all the three examined levels of education, from primary school to university level. Finally, Grainger (2012) analysed the differences in the strategy use by nearly 200 learners of Japanese in the Australian context. Significant differences were found at item level for nearly 17% of all the investigated items, and it was concluded that native Australians were able to concentrate more and made conscious efforts to think in Japanese more than non- Australian learners with Asian backgrounds.

3.5. The impact of age on learners’ choice of LLS

The role of age in learners’ strategic choices remains debatable as, similarly to the already described other IDs, its influence is believed to be interwoven with the impact of other variables which tend to vary with age, such as language learning experience, level of proficiency, or learners’ beliefs28. Dakowska (2005) specifically reflected on five aspects of language learners’ development with reference to age: the state of their cognitive system, with implications for learners’ attention span, their ability to manage their learning process, use memory strategies, and the cognitive load they can successfully handle in terms of not merely linguistic parameters, such as vocabulary acquisition, but their ability to think in abstract terms, use symbols, or visualise concepts;

28 The combination of the two variables, age and learners’ beliefs, may be of critical importance, as some adult learners might not even make any efforts to learn, for example, the pronunciation of the target language, only due to their conviction of being “too old”, as reported by Stern (1992).

130 the communicative accomplishment, which is mediated by the degree to which the learner can adopt their interlocutor’s perspective, be goal-oriented, or process conversational feedback; the linguistic development, demonstrated in the length of utterances, produced in both L1, and L2, stage of syntactic or phonological acquisition, and metaknowledge; the social growth, involving the development of a sense of identity and cooperative behaviours, but also the ability to follow rules; the emotional development, involving, in the first place, the ability to control one’s emotions. Clearly, given the above premises, the development of a foreign language by those who already speak other languages differs considerably from the development that takes place in children who are only beginning to speak their L1, and, at the same time encounter another language. In general, there seems to be a clash of at least two contradictory views regarding age. On the one hand, the conviction that younger learners are somehow superior to older learners, seems to be quite common, supported with a lot of real-life examples, referred to by Griffiths (2008) as anecdotal evidence, but, at the same time, still quoted by her. On the other hand, at least with reference to LLS, a number of researchers tend to agree that older learners can choose from a wider spectrum of strategies and, as strategies are at least initially employed consciously, they can choose more ap- propriately. According to Oxford (1986), and Ehrman and Oxford (1989c), they are also more likely to use ‘more sophisticated’ strategies. Indeed, it was found by Rohwer and Bean (1973), and Pressley and Levin (1977) that learners become more skilled at applying certain strategies as they get older, or as they progress from one developmental stage to another. For example, according to the results of the study conducted by Pressley and Levin (1977) in a group of 95 seven year-olds and 90 eleven year-olds, older learners performed significantly better in using memorisation strategies. The procedure involved associating a Spanish word to an English word (the keyword) whose sound resembled a part of the Spanish word and remembering an image of the keyword and translation ref- erents interacting. Two more recent studies conducted by Griffiths (2003, 2013) have failed to confirm the existence of a simple, linear relationship between learners’ age and their use of LLS. The data for the first study was collected by Griffiths (2003) from 26 students of a private language school in Auckland, New Zealand, through the administration of the

131 SILL (ver. 7.0, Oxford 1990), and semi-structured interviews. The aim of the study was to examine whether older language learners (specifically more than 40 years old) were less successful at language development than younger learners, and, it that be the case, identify the factors behind it. The interviews revealed that although age did influence the investigated learners’ strategy use and development of language skills in English, it was primarily a matter of attitude and, partly, the difficulties that the learners had to face, which were not directly related to the learning process. The fact that two learners of the same age (41) progressed at very different rates and the frequency of their strategy use differed considerably was largely due to their beliefs - the less successful learner was convinced that she was not able to learn more efficiently because of her age and the over- whelming problems with her children, while the successful learner claimed that he put his whole heart in the learning process, and actively explored new strategies, including some very unconventional ones. The difficulties encountered by the 64 year-old learner were of cultural and social nature rather than due to his age. He reported resorting to some strategies, such as spending more time with his notebook instead of engaging in conversations with his classmates, which he developed as defence mechanisms to fight anxiety and self-respect problems. Overall, the findings were supported by Griffith’s (2013) later (2013) quantitative study of 348 learners of English as a foreign language aged 14-62, which also provided a number of interesting findings concerning the meaning of age in strategy application. The learners were divided into two groups, 172 younger learners (14-23 years old), and 176 older learners (24-64 years old). In general, younger learners tended to attend slightly more advanced classes than older learners, however, both groups were characterised by the same frequency of strategy use which was estimated at 3.2, a medium frequency according to Oxford (1990).

3.6. The impact of gender on learners’ choice of LLS

While the existence of gender-related differences with respect to learner language, language development processes, and strategy use has been generally acknowledged (Ehr- man and Oxford 1988, Oxford and Nyikos 1989, Green and Oxford 1995), the impact of gender as an ID is far from easy to measure due to its co-existence with other related factors, such as learners’ career choices, cultural background, or the culture of the

132 countries where the TL is spoken. Even the classification of gender as an ID is not as clear-cut as it might seem. Ellis (1994) considered sex, and, later (2008), gender, as a social factor influencing second language acquisition, while Morgan and Clarke (2011) considered gender to be a primary demographic domain being part of a broader concept, learner’s identity. It appears that, traditionally, female language learners would be expected to achieve better results in foreign language tests and become more proficient than their male counterparts in the same classrooms, employ more positive attitudes, and deal with language tasks differently from men (Gass and Varonis 1986). More specifically, it was found that women tended to engage in interaction mainly to gain TL input whereas men did that so as to provide output, which would naturally mean differences in strategy use. After tools for strategy assessment were developed, a number of studies have addressed the impact of gender in learners’ strategic choices, regarding both the overall range of the strategy repertoire, the general frequency of strategy use, and the favoured strategy categories across genders. Selected research findings are discussed below. Oxford et al. (1988) analysed the variation in strategy use across genders by considering the results of four empirical investigations, including the study conducted by Politzer (1983) on a group of 90 university students participating in French, Spanish, and German courses, the study conducted by Nyikos (1987) in a group of 135 first-term students of German, and the factor-analytic study of a group of 1200 university students learning French, Spanish, Italian, German, and Russian conducted by Oxford and Nyikos (1989). The researchers sought to explain the causes of differences in strategy use across genders and aimed to account for the interpretation of the impact of gender on strategy use as an independent variable. Overall, it was concluded from the analysis of the four studies that the female participants used a wider range of LLS with regard four investigated categories: general study, functional practice (understood as authentic language use), searching for and communicating meaning, and self-management strategies, which the researchers attributed to women’s overtly more social orientation, and their different (traditionally deemed better) use of verbal skills. Specifically, in Politzer’s (1983) study, females were reported to use social learning strategies more frequently than men. The finding was supported by the results of Green and Oxford’s (1995) study of Puerto-Rican students, although the frequencies of strategy use for both genders could classified within the same interval (medium strategy use) when interpreted according to Oxford’s (1990:

133 291) guidelines. Green and Oxford’s (1995) study also revealed that women significantly outperformed men in the use of memory, metacognitive, affective, and social strategies, while men tended to use TV and watch films in order to learn English29 . The female learners’ significantly more frequent use of conversation/input elicitation strategies was interpreted by Oxford et al. (1988) by referring to the findings of earlier research on gender differences in conversation. It was also hypothesised that women’s preferences of LLS could also be affected by their potentially greater desire for social approval, greater compliance with existing social norms, and the allegedly greater verbal ability. Overall, perhaps apart from the last conclusion, the beliefs about gender differences in strategy use have not altered considerably until present. More recent papers dedicated to the impact of gender on the variation in strategy use, such as the paper by Nyikos (2008) tend to represent the view that women generally demonstrate a more frequent use of strategies for authentic language use, communicating meaning, self-management, as well as social, affective LLS, and strategy use in general. Some studies have not confirmed the relationship between gender and strategy use. For instance, in her study aimed to examine the patterns of strategy use, Griffiths (2003) set out to identify the variation in reported strategy use regarding a number of IDs, including gender. Conducted on a group of 348 mostly Asian participants in English courses at different levels (A1-C2) in an Australian language school, the research was based on the use of the SILL (ver. 7.0, Oxford 1990), the study did not reveal significant differences across genders, either in general, or with respect to any particular level. In a more recent study, conducted on a group of 120 undergraduate Turkish learners of Eng- lish Hakan et al. (2015) found significant differences in the frequency of strategy use between the male and female participants only with regard to the compensation strategy use of the SILL (ver. 7.0) (Oxford 1990), and it was in favour of male students.

29 Green and Oxford (1995) related this finding to the tendencies which they had observed in TV programmes, which were, at they claimed, dominated by programmes preferred by men, e.g. broadcasts of sport events or action films.

134 3.7. The impact of learners’ beliefs on their use of LLS

According to Dörnyei (2015: 186-187), since their introduction into the L2 literature by Horwitz (1985), language learning beliefs have gained considerable prominence in studies of foreign language development because of their anticipated impact on both learning behaviours and learners’ performance. It could be argued that some beliefs, such as those about language learning and beliefs about self, underlie self-regulation processes in foreign language development. For instance, learners who feel more empowered and possess a stronger conviction of their agency in the language learning process could simply be expected to be more active in their search for LLS, as well as in the process of selecting the right LLS for a particular task. Benson and Lor (1999: 462) referred to the role of learners in that respect by describing them as “cognitive resources on which students draw to make sense of and cope with specific content and contexts of learning”. From the practical point of view, language learning beliefs can determine learners’ choices of a specific language course that one decides to attend or be the reason for leaving the country if an individual wants to immerse in the L2 environment. Therefore, learners’ beliefs do constitute an important determinant of strategic choices, and a particularly intricate one, at the same time. Commenting on the nature of learners’ beliefs, Pelligrino (1998) warned that learners’ perceptions of language learning abroad, combined with cross-cultural issues, could negatively affect their learning due to the prioritisation of learning abroad over studying in the classroom in one’s own country. At the same time, a number of positive consequences in the strategy use of such learners could be anticipated. For instance, one could expect a rise in the frequency of social strategy use and, possibly, compensation strategy use. Yang (1999) traced a relationship between the use of LLS and learners’ beliefs and discovered that self-efficacy beliefs correlated with the use of LLS, especially functional practice strategies. Attempting to summarise the findings about the relationships between beliefs and strategies, White (2008) concluded that successful learners were able to develop insights into their beliefs about language learning, their abilities, and, as a result, use LLS effectively.

135 3.8. The relationships between LLS and learning styles

According to Dörnyei and Ryan (2015: 106), learning styles (LS) can be described as the different ways in which people learn, chosen to make learning more effective. While there are certainly methodological criticisms concerning the very core of the concept of LS, such as those voiced by Coffield et al. (2005), the potential promise that foreign language instruction could be matched to learners’ preferences, is alluring, as pointed out by Grif- fiths (2012) and Dörnyei and Ryan (2015). Oxford (2003: 16) appealed for more studies investigating the relationship between styles and strategies, expressing the belief that “(t)he more that teachers know about their students’ style preferences, the more effectively they can orient their L2 instruction, as well as the strategy teaching that can be interwoven into language instruction, matched to those style preferences”. The multitude of the categorical labels that have emerged over the years of research into LS have certainly added up to the complexity of the LS notion. Dörnyei and Ryan (2015) adapted the map of learning styles family introduced by Coffield et al. (2004), and distinguished between the following categories of LS: physiologically-based LS, including the visual, auditory, kinaesthetic, and tactile modalities, introduced by Dunn et al. (1975); LS based on learners’ cognitive structure, introduced by Riding and Cheema (1991); LS based on personality types introduced by Myers and Briggs (1976); ‘Flexibly stable’ learning preferences, first investigated by Kolb (1984); LS understood as approaches, strategies, or orientations, initially investigated by Entwistle (1990). In an attempt to illustrate the distinction between LS and LLS, Oxford (1989b) referred to LS as general problem-solving tendencies pointed out that they could directly affect specific learning behaviours, including strategy use. A similar distinction was made by Carson and Longhini (2002), who inferred from their diary study in immersion settings that LS were generally more consistent than LLS. On the other hand, the relationship between styles and strategies was approached from a different angle by Harris (2001), who hypothesised that only through the use of certain language learning strategies can some learners discover how to make the most of their learning style. Examples of studies that have confirmed the relationship between the two variables are discussed below:

136 In a qualitative study, Stevick (1989) conducted a series of interviews with seven randomly selected foreign language learners, and categorised them as an intuitive learner, a formal learner, an informal learner, an imaginative learner, an active learner, a deliberate learner, and a self-aware learner. Each of the learners was eager to reflect on their learning styles and preferences, which was interpreted as a quality of good language learners. Stevick (1989) found a number of differences in the use of metacognitive strategies among seven learners who showed various learning style preferences. In general, he inferred that the tools to learn the target language were adjusted by the learners to fit the modality they favoured. For example, some of the learners preferred engaging in social conversations with a list of linguistic aims in mind, while others devised their one own phonetic transcription, and one specific learner was ‘me- chanically’ oriented, which was demonstrated in his fondness of drills and production of charts. In general, the study confirmed the correspondence between the learners’ general perceptual and learning preferences and the LLS which they employed. It was also observed that a successfully built language style relied on learners’ strengths rather than weaknesses. An analysis of an advanced learner’s diary was conducted by Ma and Oxford (2014) in order to account for the interaction between LS and LLS for used in listening and speaking in English. The insights of the diarist-researcher entered in the learning diary for 85 days amounted to a corpus of 18 entries, and more than 13 thousand words in total. Reflections were entered in the forms of in-class notes, involving the use of key words, and retrospective descriptions of learning sessions in class or plans for future learning. It was inferred that for a reflective, introverted, visual, and environmentally- sensitive learner, the tailor-made strategy repertoire was efficient if it included note- taking in lectures (a form of surviving in the mainly auditory learning environment), establishing communicative goals and attempting to accomplish them through stretching one’s introvert style, as well as making efforts to rely more on intuition. Overall, it was concluded that any language learner should at least attempt to account for the advantages and disadvantages of their own learning style through the application of metacognitive strategies. A belief was also express according to which learners who engage in that kind of reflection can become in charge of their own self-regulation.

137 3.9. The relationships between LLS and aptitude

Describing the relationships between strategy use and aptitude poses a difficult task since the measurement of aptitude and relevant variables, such as intelligence, has been subject viewed as taboo. Although instruments designed to measure language aptitude include items that imply the use of strategies, such as inferencing, conceptualising narrowly or conceptualising broadly, the very notion of language aptitude remains debatable. For example, Richards (2015: 741) defined language aptitude as merely “(t)he theory that some people have a special aptitude for second language learning, and are better able to reor- ganize grammatical relations, master pronunciation, and learn words more easily”. On the other hand, the tradition of exploring the role of language aptitude has been well established for the simple reason that it has traditionally been recognised as a strong predictor of academic success in general, as pointed out by Dörnyei (2005), Ranta (2008), Gregersen and MacIntyre (2014) and Dörnyei and Ryan (2015). Empirical evidence for these claims had been asserted, for example, by Ehrman and Oxford (1995) who found significant correlations between language aptitude and the rate of language learning and attainment. Similarly to LLS, aptitude could be approached from the specific task perspective, which was suggested by Robinson (2002) and Ranta (2008), who were in favour of measuring the effect of aptitude on learners’ performance in specific instructional activities rather than in general. According to Gregresen and MacIntyre (2014), aptitude could also strictly be analysed with regard to the ability of learners’ noticing the gap between the correct and the incorrect, a concept introduced by Robinson (2002), or memory for contingent speech, investigated by Ranta (2008). Furthermore, Gregersen and Mac- Intyre (2014) also lent support to Skehan’s (2002) views, according to which aptitude could be analysed regarding four stages of second language acquisition, including noticing, patterning, controlling, and lexicalising. The direct link between the stages and strategy use seems quite obvious, as noticing requires paying attention (a cognitive strategy) while patterning involves analysing, processing, and making generalisations, for instance through conceptualising broadly (a cognitive strategy, again). Lexicalising, in turn, involves gathering a certain collection of expressions that can be accessed quickly, which could easily be used whenever compensation strategies are employed to support communicative processes in the face of difficulties with retrieving TL elements.

138 Dörnyei and Ryan (2015) pointed out to new developments in assessing language aptitude. For example, introduced by Grigorenko et al. (2000), the Cognitive Ability for Novelty in Acquisition of Language (CANAL-FT) involves five knowledge acquisition processes in learning an artificial language, which, similarly to the above stages of acquisition proposed by Skehan (2002), require the learner to employ certain learning strategies. Dörnyei and Ryan (2015) described the CANAL-FT test in terms five tasks: 1. Learning meanings of neologisms from context, 2. Understanding the meaning of pas- sages, 3. Continuous paired-associate learning, 4. Sentential inference, and 5. Learning language rules. The five processes involve selective encoding, through selecting relevant information, likely to require learners to employ cognitive strategies, such as paying attention, accidental encoding through the use of context, likely to require learners to conceptualise broadly, selective comparison, consisting in selecting old information for future tasks in order to enhance learning, which might require learners to use metacognitive strategies, such as planning and orchestrating the learning process, selective transfer, consisting in applying inferred rules in new context, and selective combination, relying on synthesising relevant pieces of information. Although the battery was not introduced on a larger scale, due to its unavailability to a wider body of researchers, its categories, and the underlying processes imply the relationship between aptitude and strategy use, and therefore confirm the intuitively expected correlation with aptitude and strategy use.

3.10. The impact of motivational variables on the choice of LLS

The existence of a vast number of reasons why individuals learn foreign languages is certainly reflected in their selection of LLS. Obviously, these reasons may be related to other, previously discussed, variables affecting strategy choices. As pointed out by Gregersen and MacIntyre (2014: 107), although the importance of motivation in foreign language learning is commonly acknowledged, multiple perspectives exist on how it works and where it originates from. The definitions of motivation offered by applied linguists are perhaps not mutually exclusive, but they tend to stress different aspects of the notion. For example, Ellis (2008: 972) referred to motivation as “the effort that learners put in learning an L2 as a result of their need to desire to learn it”, while Gass and Selinker (2008: 520) described it as “(t)he characteristic that provides the incentive for learning”.

139 Williams et al. (2015: 101) warned that it might be misleading to link motivation as an ID with the core verb, motivate, from which it derives, as this could result in a bias in interpreting motivation as merely an external entity. Indeed, the results of early research into the impact of motivation on learning achievement, conducted by McGroarty (1987) indicated that lack of motivation could impede progress in language learning and negatively affect educational achievement even if learners were provided with state-of-the-art instruction. It could thus be valuable to consider the characteristics of a motivated language learner, which, according to Gardner (2013: 443) include having a goal, making efforts and being willing to achieve it, adopting favourable attitudes pursuing it, being persistent, focused and attentive, making attributions about success and failure. It appears that each of these characteristics requires the use of certain strategies. For example, pursuing an educational objective requires the use of metacognitive strategies, such as planning or organising one’s learning, being willing to achieve one’s educational goal can be fostered with the use of affective strategies, and finally, making efforts can take place by application of virtually any learning strategy. Elaborating on good language learners’ characteristics, Griffiths (2008) described their strategic behaviour as goal-oriented and purposeful, which certainly links to their being motivated. What seems much more challenging for applied linguists is accounting for the sources of motivation. While it is not the focus of the present thesis, it should be noted that adopting a certain theoretical perspective, such as the social psychological approach (Gardner 2010), or the dynamic perspective of motivation, introduced by Dörnyei and Otto (1998), certainly influences not only the beliefs that a given researcher holds about motivation, but also its impact on strategy use. For the purpose of the present considerations, Gardner’s (2013: 443) list of the characteristics of a motivated learner is used as the base of linking motivation and language learning strategy use. The other perspective is closely connected with Dörnyei’s (2005) concept of self-regulation (SR), discussed in section 3.2 of the present chapter. It could be argued that the relationships between strategies and motivation are mutual since, as worded by Harris (2001: 16), learners “are more likely to persist if they feel the outcome of their learning is not predetermined and they have some control over it” and“(s)trategies can play an important part in giving them that sense of control and changing their perceptions of themselves”. As asserted by Macaro (2006), motivation might stem from learners’ use of metacognitive strategies of self-assessment and reflecting on one’s progress, hence making the interaction between motivation and strategies

140 more complex. Recently, Oxford (2017) even specifically distinguished a separate category of motivational LLS that learners employ to maintain enthusiasm and interest in improving their foreign language skills. Although the dual nature of the relationship was recognised in the early days of strategy research (cf. Oxford and Nyikos 1989), the present chapter considers LLS as a dependent variable influence by other learner IDs, and consequently, the perspective of motivational impact on strategy use has been adopted. The study conducted by Oxford and Nyikos (1989) in a group of 1,200 American university students learning five different foreign languages confirmed the intuitive assumption of the positive relationship between the level of learners’ motivation and the frequency of strategy use. The applied research instruments involved a 121-question version of the SILL (Oxford 1990), and a background questionnaire which was intended to provide information about participants’ gender, language learning experience, the type of language course (elective vs. obligatory), and self-perceptions of motivation and language proficiency. Overall, the learners who perceived themselves as relatively more motivated used LLS significantly more often. Motivation also had the most powerful impact on the strategic factors which had emerged from confirmatory factor analysis, statistically significant for four out of the five factors which emerged in the confirmatory factor analysis, including: (a) formal, rule-related practice strategies, involving the use of knowledge about language structures, making cross-linguistic comparisons, generating and revising rules, or analysing words; (b) functional practice strategies, involving various forms of practising outside the classroom, such as seeing films in the TL, making contact with native speakers, and imitating them, initiating conversations in TL, or using authentic materials; (c) general study strategies, consisting in planning and organising the learning process, ignoring distractions, and time management; (d) Conversational input elicitation strategies, including guessing meaning from context, or asking the interlocutor to speak slowly or act as a pronunciation role-model. Conducted from the foreign language classroom perspective, Pawlak’s (2012) mixed-method study of the dynamic aspects of motivation provided a number of insights into the impact of learners’ levels of motivation on their use of LLS. Evidence came from

141 a longitudinal study of 28 Polish upper secondary school students who were learning English as part of their curriculum. Data was obtained from observation during regular lessons, a 42-item, 6-point Likert-scale questionnaire designed to assess motivation, based on earlier instruments designed by Ryan (2005), Taguchi et al. (2009), and Csizér and Kormos (2009), and a series of 11 interviews with the investigated learners. With regard to strategy use, it was observed that the learners who recognised the role of Eng- lish, had relatively more specific future goals, were at the same time eager to dedicate more time and effort to boost their competence. This involved a considerable number of metacognitive, metasocial, and metamotivational LLS at work as the learners tended to increase exposure to English through participation in additional lessons, searching for additional resources, using the Internet reading books, or listening to English music. Con- versely, the students whose motivation and educational achievement in English were rather low displayed little interest in making the learning of English a more self-directed, voluntary, and, perhaps, most importantly, enjoyable experience. It was also found that students who planned in long run and who were prone to the impact of their ideal language self were also much more active, in their quest for well-suited, tailor-made strategies, both enjoyable and efficient, and certainly going far beyond the scope of school requirements. Interestingly, the qualities of these learners match the label of good self-regulation - a notion which will be discussed in the next subsection of the present chapter.

3.11. LLS and target language attainment

The interest in LLS is obviously connected with the beliefs of strategy enthusiasts that the use of LLS can boost learners’ performance and attainment30, and their conviction that both performance and attainment can be improved by strategy training31. Already

30 Language attainment is probably the most comprehensive term that refers to learners’ achievements, and learning or improving skills, which could be operationalised as performance in specific tests, as suggested by Pawlak (2008b), or quantified by self-assessment or CEFR level. Obviously, the term itself denotes a certain degree of ambiguity, which can impede investigating the relationships between language learning strategies and language attainment. 31 Strategy researchers’ belief in the benefits of strategy use led to the arrival of a number of guidebooks for both language learners and course instructors to accelerate the acquisition of the target language, which were later assessed in terms of efficiency. To mention just one example, Cohen (2005) conducted a comprehensive study of a group of language learners and course instructors in order to assess the impact of using the Language Instructors’ Guide (LIG, Cohen et al. 2005) and provide instruction for strategy training and use.

142 referred to in the present dissertation, Rubin’s (1975) and Stern’s (1975) papers have been used as points of reference for findings presented in later studies. In general, two perspectives could be distinguished in studies of the relationships between language learning strategies and learning achievement. On the one hand, a number of researchers have tended to link the use of language learning strategies to learners’ performance or attainment, measured by specific achievement or performance tests, or evaluated according to certain criteria. In perspective, these attempts appear to have aimed to prove the overall usefulness of language learning strategies, and their meaning as an explanatory variable in accounting for educational achievement in learning a foreign language. Studies within this paradigm are referred to as paradigm one studies in later parts of the present subsection. On the other hand, a number of researchers have investigated the variation in the use of LLS by learners at different levels of proficiency with the use of specific language learning strategies. Studies within this paradigm are referred to as paradigm two studies in later parts of the present subsection. It could thus be argued that a thorough description of the relationships between learners’ proficiency and their strategies should take into account both perspectives, since, as signalled by Halbach (2000), Kamper et al. (2003), and Pawlak (2008b), the correlational nature of the relationship between strategy use and learners’ proficiency impedes making any assumptions on the role of strategies in boost- ing learners’ proficiency. Oxford (2003) attempted to explain the ambiguous nature of the relationship through a reference to its spiral (mutual) nature, lending support to Vygot- sky’s (1978, 1986) views concerning the spiral nature of learning. According to a number of theoretical and empirical paradigm one studies, frequent language learning strategy users are typically more proficient learners. Starting with Rubin’s (1975) and Stern’s (1975) observations, research into strategies has resulted into the discovery of a plethora of relationships between strategy use and proficiency. Refer- ring to earlier findings by O’Malley and Chamot (1990), Oxford (1989a, 1996), and Wenden and Rubin (1987), Oxford (1999) concluded that successful language learning is typically associated with the use of a larger number of strategies, their wider range, greater effectiveness in strategy use, and a better fit to particular task demands. Table 18 provides more detailed information on the procedures employed in investigating the relationships between strategy use and proficiency level obtained in studies employing the SILL questionnaire (Oxford 1986, 1990).

143 Table 13. Types of investigations of links between strategies and proficiency (based on Oxford 1999).

Procedure Aim of application

Determining what percentage of language proficiency can be pre- Multiple regression, R2 dicted by the frequency of strategy use:

Calculating the strength and direction of the relationship between the Correlation, r frequency of strategy use and measures of proficiency Percentages of variance of language proficiency determined by the Analyses of variance frequency of strategy use (ANOVA) or strategy use and other varia- (ANOVA and MANOVA) bles (MANOVA)

The findings of selected studies using the techniques listed in Table 3 are briefly summarised below. In order to avoid information overload, the selection is limited to studies investigating learning English as a foreign or second language, and thus it does not include any references to the strategies employed by native speakers of English learning other languages. The findings are as follows: after conducting a study in a group of 141 adult learners of English as a second language, Phillips (1991) concluded that the mid-proficient learners demonstrated higher strategy use than low-proficiency or high-proficiency learners; in a study of 78 Japanese female college students carried out by Takeuchi (1993), 58% of variance in learners’ proficiency, measured by CELT test scores, was explained by eight LLS; in a study of 374 Puerto Rican university students, Green and Oxford (1995) found linear relationships between ELSAT test scores and four out of six strategy scales, i.e., compensatory, cognitive, metacognitive, and social strategies; a study conducted by Oxford and Ehrman (1995) on 262 adult learners of English employed in the US Foreign Service Institute indicated a weak, but significant correlation between the use of cognitive strategies and learners’ speaking proficiency, operationalised as end-of-term ratings; in a study of 305 South African speakers of Afrikaans, conducted by Dreyer and Ox- ford (1996), 46% of all variance was explained by the use of strategies measured by SILL ver. 7.0, with metacognitive strategies being the best predictor of learners’ proficiency, followed by social and affective strategies; the correlation coefficient calculated in the canonical correlation analysis amounted to 0.73;

144 in her comprehensive investigation of patterns of strategy use, Griffiths (2003) identified significant relationships between the applied strategies and course level, managing to select nineteen “plus” strategies, distinguishing advanced learners from others. Conducted on a group of 384 students of various nationalities, aged 14-64, attending courses in a language school, the study sought to trace the patterns of strategy use across different proficiency levels, identified by Oxford Placement Test (OPT, Allan 1995), and an oral interview by an experienced language teacher; the learners participated in a strategy awareness raising exercise or a course in study skills, which were both expected to orient learners more towards strategy use. The study confirmed that advanced learners used strategies significantly more often than elementary students. ANOVA showed that more than 10% of variance in learners’ proficiency was determined by their use of these nineteen strategies (p <0.05). The eight critical groups of strategies involved A. Interaction with others, B. Vocabulary learning strategies, C. Reading strategies, D. Ambiguity tolerance strategies, E. Strat- egies related to language systems, F. Strategies employed to manage feelings and emotions, G. Managing one’s own learning, and H. Strategies allowing learners to utilise available resources; investigating the impact of lexical inferencing strategies on second language reading proficiency, Parel (2004) found that reading comprehension skills in English as a foreign language may be improved by the acquisition of the meanings of the most productive affixes of English, accompanied with instruction of principles of morpholog- ical analysis; studying the learning strategies used by Batswana learners of English in order to handle the problem of learners’ relatively low fluency in English and their poor performance in English examinations in tertiary education, Magogwe and Oliver (2007) supported the claim that the relationship between proficiency and strategy use may largely depend on the type of strategy employed; they did not find any statistically significant relationships between learners’ proficiency and their use of LLS, but at the same time, observed that learners’ willingness to use metacognitive strategies increased considerably in secondary and tertiary education; a study of 77 full-time Slovenian students of the University of Ljubljana conducted by Jurkovič (2010) indicated that little efficiency could be expected from explicit strategy training; at the same time, the researcher suggested that one of the reasons

145 for it could have been the heterogeneous nature of the group: Learners varied considerably regarding their initial competence, and strategy training did not account for the differences in strategies used by lower level and higher level students. Habók and Magyar (2018) conducted a study on a group of 868 Hungarian lower secondary school students (aged 11 and 14) in order to examine the relationships between LLS, learners’ proficiency level, and their foreign language marks.; the study confirmed that moderate, significant correlation existed between overall strategy use and learners’ marks in foreign languages; however, the results showed no clear tendencies for the investigated scales of strategies. Although the results of the above studies are characterised by a certain degree of inconsistency, the general conclusion which arises from this overview is close to that made by Oxford and Ehrman (1995: 69), according to whom successful language learners manage a wide repertoire of learning strategies which they are able to apply accordingly to the particular task demands, the learning material, the individual’s personality, as well as their proficiency level, and their stage of learning. The body of paradigm two research dedicated to the investigation of the role of learners’ proficiency in their choice of their strategic repertoire seems somewhat limited in comparison to the plethora of studies conducted according to paradigm one. Some researchers that have followed the second paradigm have simply investigated the frequencies of overall strategy use or the use of certain categories of strategies (such as the scales of SILL ver. 7.0) among learners grouped according to their proficiency level, while others, such as Gregersen et al. (2001) or Lai (2009), have attempted to trace patterns in strategy use. In the case of paradigm two studies, an attempt can be made to analyse strategies as a variable explained by learners’ level of proficiency. A number of interesting perspectives on the use of strategies by Chilean students, who varied in terms of both, target language proficiency and attainment in their English studies, were introduced in a description of a pilot study conducted by Gregersen et al. (2001). The researchers investigated the differences in strategy use between good and poor learners (as indicated by a panel of professors) and between first-year and fourth- year students. They reported the following findings: as regards memory strategies, good learners outperformed poor learners in their use in the first year; however, the situation was reversed in the fourth year; the researchers concluded that memory strategies could be more naturally used by lower level

146 learners in order to help them expand their vocabulary range, whereas among more proficient learners, emphasis is more frequently put on acquisition rather than learning. the use of cognitive strategies remains consistently more frequent among good learners than among poor learners throughout the period of studies; in fact, the difference in cognitive strategy use in the fourth year exceeded the difference in the first year; the frequency of using compensation strategies did not vary among good and poor students in the first year, however, these strategies were relatively more frequently used by poor students in year four; the authors of the study hypothesised that exces- sive use of compensation strategies could possibly hinder the process of acquiring a foreign language; good language learners outperformed poor language learners in the use of metacognitive strategies and the efficiency gap in the fourth year exceeds the one between first year students; the scholars concluded that strategies such as planning, organising, or evaluation, are crucial to acquiring a foreign language; interesting revelations were obtained regarding learners’ use of affective strategies, which, as it turned out, were more frequently used by poor learners than by good learners; the researchers suggested that the need to use affective strategies may only arise when learners are experiencing problems or feel uncomfortable, but also pointed to the wording of the SILL questions in the affective scale; the use of social strategies remains consistently higher among good students than among poor students, and the tendency does not change over the time of studies. In order to investigate the patterns of strategy use, Lai (2009) conducted a study on a group of 418 students participating in the Freshman English for Non-Majors (FENM) program at Tunghai University in Taiwan. The researcher concluded that learners’ level of proficiency in English had a significant impact on their choice of strategies. First of all, more proficient learners used language learners significantly more often than less proficient users. Secondly, more proficient users showed preference for different groups of strategies than less proficient users (metacognitive and cognitive strategies being the domain of proficient students, and memory strategies being the domain of less proficient users). The researcher also inferred that the distinguishing strategies of good (and thus, proficient) language learners were arranging and planning one’s learning, and making

147 good use of one’s reasoning and analytical skills, as well as practising pronunciation and speaking. A study of 140 Turkish students of English conducted by Yılmaz (2010) indicated that the students who were classified as “good” on the basis of their cumulative average of grades in English subjects significantly outperformed poor students with regard to affective strategy use. He also concluded good students were able to encourage themselves to store and retrieve information more efficiently and lower their anxiety. Attempting to explain the differences in strategy application by good and poor Iranian students, Gerami and Baighlou (2011) examined a group of 200 students from two universities and measured their declared strategy application according to self-reported frequencies indicated in SILL ver. 7.0. These students were divided into three groups according to their TOEFL scores, i.e., high-level students (top 27% of scores), mid-level, and low level (bottom 27% of scores). Not only did the study indicate considerable differences in general strategy use in favour of good students, but also, it provided interesting insights into the patterns of strategy application. According to the results, high level students preferred metacognitive strategies, and deep cognitive strategies. By contrast, poor students preferred surface structure cognitive strategies, e.g. saying or writing new words in English several times, or looking for similar words in their native language. The results of the study confirmed the utmost importance of planning and organising one’s own learning. A study conducted by Salahshour et al. (2013) on a group of 60 high school students confirmed the particular role of metacognitive strategies, which were used significantly more often high-proficiency learners than low-proficiency learners. In an attempt to elicit the strategies of effective language learners, Wong and Nunan (2011) studied the strategies of 110 Hong Kong undergraduate students, 33 of whom were ‘poor’ learners, graded ‘E’ or ‘F’ on the Hong Kong Examinations and As- sessment Authority “Use of English Examination”, and 77 of whom were ‘more effective’ learners, graded ‘A’. The five most popular strategies among the ‘more effective’ students included: learning by watching/listening to native speakers; learning English words by seeing them; watching TV in English; learning in class by conversation; learning new words willingly.

148 At the same time, the strategies favoured by the ‘less effective’ students involved: listening to error correction done by the teacher learning English words by seeing them; using the teacher’s help to discuss one’s own interest; having one’s own textbook; learning new English words by doing something. Overall, merely one strategy proved to be favoured by both categories of students - learning English words by seeing them. It could be executed in a number of ways, which could differ considerably in the two analysed categories of learners. All the strategies favoured by the ‘more efficient’ seem to require relatively more involvement than those employed more frequently by the ‘less efficient’ learners. Those employed by the ‘less efficient’ learners also appear to be relatively more passive, perhaps with the exception of learning by doing. It might thus be concluded that successful learners are more at ease, more in charge, and more communicatively oriented than their unsuccessful counterparts, as signalled in earlier research conducted by Nunan (1997). Researching the relationships between strategy use and the proficiency level of first year Thai university students, Kunasaraphan (2015) was interested in finding patterns of the use of both, direct, and indirect strategies, as well as their dependence on participants’ level. Unlike previous studies of Thai students, such as those conducted by Kham- khien (2010) and Anugkakul and Yordchim (2014), the author found that Thai students used indirect strategies relatively more frequently than direct strategies. The linear relationship between the general use of language learning strategy and learner’s proficiency level was confirmed. The author also reported a significantly higher frequency of metacognitive strategy use by students whose proficiency level was higher (learners’ level of proficiency had previously been determined by means of a Thai SSRUIC test, and an interview, as either basic, intermediate, or advanced). On the other hand, lower-level students used memory strategies significantly more often than higher level students. Kuna- saraphan (2015) expressed a belief that it was possible to promote strategies typically chosen by high-achievement students among low-achievement students, and therefore influence their attainment. The particular role of self-regulation and orchestration of one’s own learning, along with the essential meaning of efforts on the part of the learners made in order to increase exposure to English were inferred to be of vital importance by Lee and Heinz

149 (2016). In their study of participants of translation and interpretation courses at a Korean university, they attempted to elicit the strategies employed by advanced learners. Alt- hough the researchers emphasised that the participants of the study were not typical learners of English as a foreign language and hence, their study results could not be generalised, their findings match the overall picture of high-proficiency learners. Finally, numerous researchers attempted to capture the complex relationships between LLS and proficiency by allowing more variables in their studies. For example, in a study already referred to in the present section, Habók and Magyar (2018) set out to analyse the differences between relatively more and less proficient learners of English in terms of strategy preferences. Although their results were significant for some strategy categories, showing an overall preference for metacognitive strategies, and a shift from memory and affective strategies to cognitive strategies, the results remain ambiguous as the learners were not merely different in terms of proficiency, but also in terms of age. Consequently, it is not clear whether the observed differences could be explained by learners’ increasing proficiency in English, their cognitive development connected with the process of growing up, or, perhaps, both. The study referred to in the previous paragraph constitutes a good example of how challenging it might be to capture the complex relationships between LLS and language attainment, learners’ proficiency, and educational achievement. First of all, some researchers do attempt to distinguish between the three categories. Of course, it is possible to operationalise each of the three as different variables. For instance, one could quantify language attainment as end-of-term evaluation of learners’ performance, learners’ proficiency as their CEFR level, and educational achievement as grade/mark average, but it remains true that each of them can constitute a measure of progress that learners make in foreign language development. Secondly, the age factor poses a serious challenge in any analysis of the developmental patterns in strategy use. The ceteris paribus condition is hard to meet in comparative research investigating the strategies of learners who differ in their level of proficiency, especially when juxtaposing groups of students from primary or secondary schools. It needs to be remembered that age constitutes an essential individual characteristic, which can seriously affect learners’ performance and attainment, as indicated by Singleton and Leśniewska (2012), and hence, any comparison across groups of learners whose ages are different, especially those at different developmental stages, could be biased. Another difficulty consists in the already mentioned correlational nature

150 of the relationship between LLS and target language attainment. Addressing the issue of investigating the links between LLS and target language attainment, Pawlak (2011: 25- 26) suggested that the cause-and-effect relationship between strategies should be embraced rather than feared, and recommended more specific investigations of strategy efficiency, allowing for more in-depth analysis of successful strategy use, accounting for fit to a particular language task, or the influence of other IDs, such as learning style. These are referred to as microcontext and macrocontext of strategy use, and referred to in the final subsection of the present chapter, and in Chapter Three.

3.12. LLS in microcontext: The impact of the language task

While it appears quite obvious that the type and specifics of a language task to be performed by the language learner exert an influence on their strategy choice, the nature of the influence remains complex. More specifically, learners making strategic choices may take into account immediate language tasks, but also forthcoming language tasks, and even anticipated language tasks or situations in general. For example, a learner might rely on contextual clues when dealing with unknown vocabulary in a reading comprehension task in a progress test, but at the same time, they could frequently check the meaning of unknown words or phrases in an online dictionary when doing the same task at home. The interaction between LLS and the language task has been already referred to in the present dissertation. In brief, it is generally believed that the overall purpose of LLS is to facilitate the completion of specific language tasks, as reported by Oxford (1990, 2011, 2016). Some strategy researchers, such as White (2008b) or Cohen (2014), have even further emphasised the meaning of the language task by including it in the definition of LLS. Others, such as Bialystok and Fröhlich (1978), and Cohen (1995) have stressed the necessity to refer to specific language tasks when judging the efficiency of LLS. Fi- nally, it is commonly believed that the variety of language tasks exerts a direct influence on the variation in strategy use, as indicated, for example by Ellis (1994), and Pineda (2010). The belief expressed by Oxford et al. (2004) that research into task-based strategy investigations should be complimentary rather than substitute to research into strategy use as such appears to be relatively common. For example, Rubin (2008) perceived

151 effective learning as dependent on the use of language learning strategies in terms of accomplishing specific language tasks rather than claimed that learners should be aware of the existence of certain language learning strategies as such. Also, it appears that the overall usefulness of some strategies is greater that of other strategies with respect to tasks requiring the use of particular modalities. For example, Bialystok (1981) suggested that self-monitoring might be more frequently used in writing tasks rather than in speaking tasks. As indicated by Oxford (1989c, 1990), strategies can be transferred across tasks. In Anderson’s (2000) view, this can happen provided there is a sufficient degree of similarity between the tasks, such as the necessity to employ certain specific language patterns. According to Cohen (2003), successful learners select and adjust their LLS as they progress in performing a task. For instance, in reading comprehension tasks, they might approach the task initially using the strategies for finding the main idea of a text, such as reviewing headings and subheadings quickly in order to comprehend the organisation of a written passage, but they are likely to replace them with strategies for making inferences from the text, such as noticing the key adjectives in order to identify the writer’s attitude and distinguish between facts and opinions. Good language learners can also apply different strategies for summarising the text, such as making marginal notes, or converting detailed points of a summary into more general points. Among a plethora of various factors which modify the use of LLS, the task which the learner has to accomplish is one that is frequently external to the learner (assignment from a language course instructor or requirement created by the conditions of learning the language in a foreign environment), and goes beyond the level of skills. For example, writing a postcard differs considerably from writing a report ordered by a superior. As it was put by Oxford et al. (2004: 2), “task-based strategy assessment seeks to anchor strategy use within the context of a particular language task, thus allowing for a more detailed, more contextualized analysis of L2 strategy use”. In their in-depth analysis of the meaning of the language task in the application of various types of strategies, Oxford et al. (2004) used a number of perspectives on the L2 task, including: the perspective of a duty to fulfil;

152 the perspective of a segment of curriculum, which, in turn, is reflected as a stage in the organisation of the teaching and learning process, as well as a checkpoint in the assessment of L2 outcomes; the perspectives of the application of tasks as communicative activities or accuracy- oriented activities. Obviously, performing a certain task may be connected with more than one perspective, and similarly, a number of task categories can be distinguished, which show a considerable overlap, e.g. role-plays, simulations, problem-solving, decision-making, sharing experiences. A number of dimensions of tasks exist, all of which have the in-built component of external origin. Oxford et al. (2004: 10-15) draw attention to the following: goals (convergent or divergent), related to systematic elements of the target language or a language function; the level of risk, which determines the level of stress which a language learner may encounter performing it; timing, which can be defined or not; input, which can take different forms, and vary in terms of comprehensibility; materials and settings, which constitute the input, and vary with reference to their relevance and suitability; output, which varies in terms of complexity, and reflects the level of understanding the task, the operating cognitive processes which underlie performance, and learners’ aptitude; the level of skill competences (any of the four macro language skills or their combination which results in an integrated use of language skills), and the key aspect of communicative competence addressed by the task; complexity, resulting from the involvement of certain cognitive processes, and the degree of it familiarity to the learner; the expected mode of interaction (e.g. verbal or non-verbal); planning, whose dose may differ depending on the learner’s use of metacognitive strategies, and the place on the continuum between accuracy and spontaneity which the learner relates to the completion of the task; amount of strategic control on the part of the learner, which strategy researchers believe can facilitate the completion of a number of foreign language tasks;

153 a number of intrinsic learner-specific factors, such as the learner’s personality traits, motivation, their dominating learning style, accompanying language anxiety, or preferences connected with the role which they adopt in tasks involving group work; teacher-related factors, such as teaching beliefs and preference for certain classroom techniques. All the above factors are connected with the degree of difficulty, which Skehan (1996) considered with respect to three aspects, including linguistic complexity, cognitive complexity, and communicative stress. Oxford et al. (2004) also referred to Pica (1993: 17), who listed four conditions to be met so that a task enables learners to develop their strategic competence through negotiation of meaning. These include: information split between interlocutors; the need to supply the missing information through communication; proximity of interactants’ goals; only one possible outcome possible to reach through the completion of interactants’ goals. Some more recent findings referring to the importance of the language task in strategy investigations are listed below: Elgort and Warren (2014) suggested that specifically, in vocabulary tasks, learners should be able to employ more explicit task strategies, such as mnemonics or other higher-order cognitive strategies, whereas they linked fluent language use with fast and accurate processing of word forms and the retrieval of their meanings. Uhrig’s (2015) conclusions from his case study of learners performing a series of reading comprehension tasks supported Cohen’s (2003) and Dörnyei’s (2005) assumptions about the influence of students’ learning style on their choice of strategies with relevance to particular tasks. According to Varasteh et al. (2016), learners who attribute relatively higher value to the performed language task are also relatively more frequent deep strategy users and, hence, achieve greater self-regulation. Although the importance of the language task in determining learners’ strategic choices is quite obvious, and the framework for investigating task-based language learning and teaching has long been in existence (cf. Ellis 2003), strategy researchers seem to prefer the functional approach, which corresponds with the early taxonomy of LLS developed by Oxford (1990). The relations between LLS and language tasks are, indeed,

154 complex, due to the involvement of a number of variables in their mutual interaction. As shown above, strategies can be accounted for with reference not merely to their functions, but also to the four skills, and even language areas, such as grammar or vocabulary. Per- haps the need for a new framework for investigating LLS in microcontext should thus be emphasised.

3.13. Conclusion

Although it is acknowledged that language learning strategies are a key to learner autonomy and a valuable element of foreign language education, as asserted by Droździał- Szelest (2004: 32), accounting for them still happens to be seen as a challenge by members of the L2 community, including learners, teachers, and scholars. As far as L2 language learners and teachers are concerned, according to Pawlak and Mystkowska- Wiertelak (2015: 5), the difficulty in recognising the role of LLS might be due to the fact that promoting autonomy in language learning has not become a priority for the teaching community, and consequently, a large proportion of learners limit themselves to performing obligatory or quasi-obligatory tasks, such as homework assignments or revising for tests. Droździał-Szelest (2004: 41) pointed out that learner autonomy partly consists in knowing one’s own strengths and weaknesses as a language learner. This sort of person knowledge is the first element of metaknowledge, as defined by Oxford (2011, 2017), which requires a complete focus on the individual. Therefore, knowing that the development of learner autonomy can only succeed through introducing solutions which take into account the curriculum, and tailor-made strategies of language learners, as asserted by Pawlak (2015), it seems necessary continue the strivings of strategy researchers to provide insights into as many aspects of the language learner as possible. The present chapter has attempted to illustrate the intricate relationships between LLS and a number of other learner characteristics. Its final section, which contains a review of selected studies into LLS in recent years, certainly supports the conviction that strategy research should continue, the paradigm does not have be exchanged, and, perhaps most importantly, the breadth of perspectives into the choice and use of strategies constitutes much more of an advantage than a drawback, as pointed out by Pawlak and Oxford (2018). It would thus be unwise to abandon strategy research or opt for an overall change

155 in the research paradigm, as new insights which contribute to a more holistic picture of the strategy user, continue to appear as results of studies largely relying on the traditional methodology, including the use of the SILL (Oxford 1990). It is a matter of time until the directions of future research, such as indicated by Pawlak and Oxford (2018), including studies of the impact of learners’ emotions on strategy use, or cross-linguistic influences in strategy application, become niches occupied by strategy researchers.

156 Chapter 4: Methodology of the research project

4.1. Introduction

The need to investigate language learning strategies became obvious after the emergence of the humanistic approach, first, in psychology, and, following that, in education, and foreign language pedagogy. A research orientation was born which emphasised the need to approach individuals holistically, recognised their uniqueness, and emphasised the complexity of human experience. Its foundations were specified by Bugental (1964: 23- 25), according to whom: human beings cannot be reduced to the components of which they are built; human beings exist both in their unique context and in the universe; human beings are aware and conscious, which is also true with regard to their interaction with other human beings and self-awareness; human beings make choices and therefore bear responsibility; human beings act intentionally, set goals, and can anticipate the consequences of their actions. In fact, all the above points correspond to Oxford’s (1990, 2011, 2017) views on the features of language learning strategies, which she asserted to be conscious, facilitative, holistic, scalable, and transferrable. Moreover, the belief in human consciousness entitles the strategy researcher to employ self-report questionnaires and interviews as research tools, since they rely on introspection. At the same time, some methodological precautions need to be taken so as to ensure the fulfilment of the goodness criteria for psychometric tools (Hornowska 2007), and scientific rigour needs to be introduced in order to guarantee the external validity of the obtained study results (Aronson et al. 2010: 38-39). In general, the purpose of the research project described in the present dissertation consisted in broadening and disseminating knowledge about the language learner, with a particular focus on the interaction between their personality traits and their use of language learning strategies (LLS). The following subsections include the descriptions of procedures applied in the pilot study, the main quantitative study, and the series of interviews conducted in order to triangulate the data obtained from questionnaires. The

157 chapter also contains a description of the participants of the study, which consists of two major parts, that is a section dedicated to the statistics accounting for participants’ individual characteristics, and a section that explains some characteristics shared by young adults which are vital for their development as such, but also cannot be neglected when analysing any process of learning. Particular emphasis is placed on a detailed description of the research instruments used in pilot studies and in the main study. Both adaptations of questionnaires are analysed in terms of their reliability and validity, and a rationale is provided for the use of a semi-structured interview in the qualitative part of the main study. Furthermore, the description of analytical procedures employed in each part of the research projects is provided, and, finally, a section addressing the specificity of the software applied for performing statistical analyses and calculations is explained. Additionally, the procedures of data collection at all stages are also explained in the present chapter.

4.2. Pilot Study One

Pilot Study One was conducted in December 2013 in order to assess the potential of using the Polish adaptation of NEO-FFI (Zawadzki et al. 2010) along with the Polish adaptation of SILL ver. 7.0 (Oxford 1990) for investigating the relationships between language learners’ personality traits and their choices of language learning strategies32. The research tools were thus verified in terms of their fulfilment of the goodness criteria applied in psychometric research, including objectivity, standardisation, reliability, validity, nor- malisation and accuracy of adaptation (Hornowska 2007: 25-29). It was also expected that the results of Pilot Study One would at least partly reveal the relationships existing between the characteristics of the investigated individuals and their choice of particular language learning strategies (LLS). Furthermore, Pilot Study One also allowed a number of observations of how participants differed in using LLS across genders, employment status, or type of received education, i.e., students’ participation in the full-time or extra- mural programme.

32 Both instruments are introduced in section 4.2.2 and discussed in detail in section 4.4.3.

158 4.2.1. Participants and data collection

The participants of the study were 50 students of Poznań School of Banking (today: WSB University in Poznań), 50% of whom were participating in a full-time programme, while the other 50% were extra-mural students. Table 14 contains descriptive statistics referring to participants’ type of education, employment, gender, age, and type of received education.

Table 14. Participants of Pilot Study 1. type of edu- employ- gen- no. of re- min. max. mean std. de- vari- cation ment der spondents age age age viation ance extra-mural unemployed female 3 20 40 27.00 11.27 127.00 programme employed female 19 21 43 27.42 7.19 51.70 male 3 25 33 28.00 4.36 19.00 full-time unemployed female 13 20 23 20.54 0.88 0.77 programme male 6 20 22 21.00 0.63 0.40 employed female 5 20 21 20.60 0.55 0.30 all partici- 49 20 43 24.12 6.122 37.49 pants

As can be seen from Table 14, women outnumbered men, which was due to the specificity of most fields of studies at WSB University. A vast majority of extra-mural students were employed whereas only about 20% of full-time students were. Age variance was much higher among extra-mural students, which might have entailed differences in the amount of exposure to English in the investigated students’ past education, and differences in the applied strategies reflecting various educational backgrounds. As far as the investigated learners’ personality traits are concerned, a high age variance results in the necessity to interpret the raw values of personality trait levels with caution as the same raw scores can imply different levels for different age groups (Nye et al. 2016: 473). The participants of Pilot Study One were asked to complete pen-and-paper questionnaires of Polish adaptations of SILL ver. 7.0 (Oxford 1990), and NEO-FFI (Zawadzki et al. 2010) during their English classes. 49 completed questionnaires were returned, 25 by extra-mural students in their third year, and 24 by full-time students in their second year33. They had been familiarised with the purpose of the study, encouraged to provide

33 According to the curriculum which applied in 2013, full-time students received 240 hours of instruction in English in total for two years while part-time students only received 120 hours of instruction for two and a half years. 50% of classes were General English courses and the other 50% - Business English courses.

159 true answers, and told that no correct or expected answers existed. Participation in the study was both voluntary and unsupervised, and questionnaires were collected by the author of the present dissertation.

4.2.2. Measures

Polish adaptations of two questionnaires, NEO-FFI (Zawadzki et al. 2010) and SILL ver. 7.0 (Oxford 1990) were used in Pilot Study One. Whereas a standardised version NEO- FFI was used, the adaptation of SILL included amendments which were expected to reflect technological progress so as to account for strategies that were available in the second decade of 21st century, yet unavailable back in the 1990s. Hence, two items were amended so that the questionnaire would reflect the 21st century universe of content with regard to LLS. The amendments in the wording of both questions resulted from the necessity to account for technological changes rather than cultural differences between American and Polish settings. When SILL ver. 7.0 was constructed, watching a film in English meant either watching it on TV or using a copy from a video rental, or going to the cinema. From the 1990s, there has been a growing trend to watch films online. It can be assumed that the exposure to the target language which one can experience through watching films can take place no matter where it happens. Technological advancement has also resulted in the increasing popularity of online correspondence, such as emails or instant messengers, to mention just two forms of writing which were not available when SILL ver. 7.0 was constructed. Therefore, the wording of question 17 was changed in a similar manner, i.e., the content of item 17 was subject to a generalisation so as to include new forms of correspondence in the universe of content. Thus, the following modifications were introduced: Question 15, whose original wording was: “I watch English language TV shows spoken in English or go to movies spoken in English” was altered in order to account for a greater variety of English input to which learners could expose themselves, predominantly online. After rephrasing, the question became “I watch English TV programmes and films”;

When the questionnaires were administered, the students were attending A1 courses in Business English, but they had already completed from a General English course before.

160 Question 17, whose original wording was “I write notes, messages, letters, or reports in English” was also redesigned in order to allow for a wider number of options available, such as text messages, emails, blogs, etc., and hence became “I write notes, messages, letters, or other forms in English”. As the questionnaire was translated into Polish, the wording of the amended questions was consulted with a Polish philologist. Both changes can be viewed as adaptation in the sense that Hornowska (2007: 29-30) described as the process of adjusting the original version of the questionnaire to the specificity of the local culture. However, another reason behind the implemented modifications, was the temporal aspect34. In terms of analysis, first, the internal reliability of the research instruments was assessed. For that purpose, Cronbach’s alpha coefficients (α) were calculated for each personality domain (Neuroticism, Extraversion Openness to Experience, Agreeableness, and Conscientiousness), and for each language learning strategy scale (Memory, Cogni- tive, Compensation, Metacognitive, Affective, and Social strategies). Based on the mean intercorrelation between the elements which constitute a scale, Cronbach’s alpha can be interpreted as the degree of variance to which the observed outcome explains the hypothetical scale of real outcomes consisting of all potential items or, to put it in a different way, the extent to which scores obtained by a person are the same if the person is re- examined by the same test on different occasions (Groth-Marnat 2003: 31). The value of the α coefficient was also calculated for the strategy construct as such, understood as the mean value of the frequency of strategy use, and named GULLS - General Use of Lan- guage Learning Strategies. The values of Cronbach’s alpha for all scales can be seen in Table 15.

34 Specifically, the Polish word trawestacja would be ideal to describe the type of adaptation, as indicated by Prof. Hornowska in personal communication (06 April 2016). However, it does not have an English equivalent that would be accurate enough. The term ‘travesty’ is a pejorative one and mostly used with reference to literary imitation or misrepresentation. Hence, “adaptation” is the most accurate expression which explains the type of modification of the original questionnaire.

161 Table 15. Cronbach’s alpha values for strategy scales and personality dimensions.

Scale No. of items Cronbach’s alpha

Memory strategies 9 0.63 Cognitive strategies 14 0.80 Metacognitive strategies 9 0.86 Compensation strategies 6 0.53 Social strategies 6 0.74 Affective strategies 6 0.27 GULLS 50 0.87 Neuroticism 12 0.86 Extraversion 12 0.64 Openness to Experience 12 0.50 Agreeableness 12 0.79 Conscientiousness 12 0.72

A number of comments need to be made with reference to the above table. First of all, three strategy scales, that is cognitive, metacognitive, and social learning strategy scales, showed alpha values that are acceptable in psychometric research (α > 0.7, as formulated by Kline 2000: 13). The alpha value for memory strategy scale was just slightly lower than 0.7 and it could be expected that the value may increase slightly in the main research as a consequence of the rise in the number of participants of the study. At the same time, two of the six scales of SILL ver. 7.0, that is the compensation strategy scale, and the affective strategy scale, showed low internal validity, with Cronbach’s alpha values amounting to 0.53 and 0.27 respectively (calculations based on 46 valid cases, amounting to 93.9% of all participants). In order to increase the reliability of the two scales, additional questions were introduced. A panel of judges were asked to rate the modified research instrument (Lawshe 1975; Brzeziński and Maruszewski 1978). The judges were all experienced EFL teachers employed at AMU Language School. They were asked in an email whether the added questions contributed to an increase in the level of knowledge about a given scale of language learning strategies (scales C - compensation strategies, and E - affective strategies). Table 16 presents the questions introduced into the Polish adaptation of SILL ver. 7.0, which were subject to the evaluation by the com- petent judges.

162 Table 16. Additional questions introduced in the adaptation of SILL ver. 7.0. Ques- Scale Question tion compensation I refrain from using online dictionaries or translators while com- C7 strategies municating on the Internet compensation C 8 I draw a word which I cannot remember in English strategies affective strat- E 7 I use emoticons when I write in English egies affective strat- I share my feelings about learning or using English on an Internet E 8 egies blog or on social portals affective strat- E 9 I try to perceive my English teacher in a positive way egies

The following ratio can be calculated to assess the content validity of added items (Law- she 1975: 567): �� − N/2 CVR = N/2 where: CVR - content validity ratio, ne - number of experts who perceive a given item as essential, N - number of all experts According to Lawshe (1975), any item which is perceived to be “essential” by more than half of the panellists has some degree of content validity. In practice, it means that if any of the questions introduced into the questionnaire is rated as essential by more than 50% of the judges, including it in the questionnaire has at least some chance of accounting for compensation and affective language strategies better. Since all the introduced questions got more than 50% of positive ratings, they were subsequently included in the adapted inventory used in the main study. Cronbach’s alpha coefficients showed satisfactory values for all personality dimensions except for Openness to Experience. The value of the coefficient calculated in a study conducted to validate the reliability of NEO-FFI in comparison to NEO-PI-R amounted to 0.73 (Zawadzki et al. 2010: 29) whereas the value of the coefficient calculated in Pilot Study One amounted to 0.5. Excluding question 1 from the Openness scale (“I do not like wasting time dreaming”) would result in an improvement in the value of Cronbach’s alpha to 0.6. However, it needs to be remembered that NEO-FFI is a standardised psychological test and any modifications would affect the overall structure of the tool and result in an uneven split of questionnaire into the ‘big’ five personality scales. Instead of modifying the questionnaire in the main study, a decision was thus made to

163 retain all questions from the original Openness scale and compare Cronbach’s alpha values of the original scale and the modified scale if item 1 was deleted in the main study.

4.2.3. Analytical procedures

The pilot study met the formal requirements imposed by the authors of NEO-FFI (Costa and McCrae 1992) since all missing data items accounted for far less than 1/6 of total data input items and no cases with three or more gaps per personality scale were re- ported35. Data sets were tested for normality with Kolmogorov-Smirnov test. The research procedure involved testing each language learning strategy scale, and personality dimension. Each time, two hypotheses were tested. H0 assumed that data distribution did not differ significantly from normal while H1 assumed that data distribution was significantly different from normal. Table 17 provides information on data distribution.

Table 17. Data normality for strategy scales and personality dimensions.

data distribution scale

Memory strategies (mean) normally distributed Cognitive strategies (mean) normally distributed Metacognitive strategies (mean) not normally distributed Compensation strategies (mean) normally distributed Social strategies (mean) not normally distributed Affective strategies (mean) normally distributed GULLS (mean) not normally distributed Neuroticism normally distributed Extraversion normally distributed Openness to experience normally distributed Agreeableness normally distributed Conscientiousness normally distributed

As can be seen from the above table, data sets tended to have normal distribution. Distributions which varied significantly from what could be seen as normal included metacognitive strategies, and social strategies. Also, GULLS as a scale proved to have data distribution which was significantly different from normal (p > 0.05). Table 21 is the basis for later decisions to use parametric or non-parametric coefficients in calculating

35 Missing values were removed from analysis automatically by the statistical software, SPSS ver. 25.

164 correlations between strategy scales and levels of personality traits in the final chapter of the present dissertation. Pearson correlation coefficient was used if the distributions of both correlated data sets do not significantly vary from normal while Spearman’s rank coefficient is used if the distribution of at least one of the correlated data sets varies significantly from normal. For the purpose of comparing the use of strategies across genders, employment status, or type of received education, t-tests were used for most scales and dimensions, whose data distributions did not vary significantly from normal, and the Mann-Whitney U-test for the GULLS, as well as for scales of metacognitive strategies and social strategies, whose data distributions varied significantly from normal as indicated in Table 4.

4.2.4. Selected findings and implications for the main study

Although Pilot Study One was mainly conducted in order to verify the rationale for the main study, with particular focus on the examination of the data collection instruments to be used in the main study, findings concerning the frequency of LLS use and the level of personality traits were also collected and analysed. For that purpose, descriptive statistics for each scale of LLS and the levels of the ‘big’ five personality traits were calculated. They are presented in Table 18.

Table 18. Descriptive statistics for Pilot Study One.

Scale mean std. deviation variance Memory strategies 2.58 0.53 0.279 Cognitive strategies 2.83 0.56 0.316 Compensation strategies 3.10 0.57 0.319 Metacognitive strategies 3.13 0.68 0.467 Affective strategies 2.49 0.54 0.291 Social strategies 3.04 0.70 0.488 GULLS 2.83 0.42 0.176 Neuroticism 22.92 8.85 78.41 Extraversion 30.65 5.25 27.61 Openness to Experience 25.90 5.00 24.97 Agreeableness 30.24 6.41 41.06 Conscientiousness 32.80 5.39 29.04

On the whole, according to the SILL Profile of Results (Oxford 1990: 300), the frequency of strategy use in the investigated group could be described as medium, with

165 values ranging from 2.49 (Affective strategies) to 3.13 (Metacognitive strategies). It was thus evident that the investigated students relatively seldom used any affective language learning strategies. Instead, they favoured metacognitive and compensation strategies, which could potentially be signs of planning and orchestrating the learning process on the one hand, and not being overwhelmed by insufficient vocabulary, potentially both in the classroom and in real life communication, on the other. The investigated group of students also used social learning strategies relatively frequently in comparison to other strategy scales. Direct strategies such as memory or cognitive strategies were not parts of the most frequently used strategy repertoire, which could logically be linked to the fact that the learners’ level of proficiency was relatively low (A1+ according to CEFR). Regarding the personality characteristics of the participants, no high (stens 7-10) or low (stens-1-3) levels of personality traits were observed for the investigated sample of students considering the mean scores of personality traits. When confronted with the norms for particular ages and genders, the mean values all fall into the average category, which, according to the manual description (Zawadzki et al. 2010) could be summarised in the following way: in terms of Neuroticism, the participants of Pilot Study One were well-balanced, and occasionally experience anger, grief or feel guilty; in terms of Extraversion, the participants of Pilot Study One were fairly sociable, displaying preference for both, entering into social interactions with others on some occasions, but showing preference for privacy and intimacy on other occasions; in terms of Openness to experience, the participants of Pilot Study One proved to be balanced in their readiness to explore new ideas and reliance on traditional values; in terms of Agreeableness, the participants of Pilot Study One were typically nice and friendly, but sometimes exhibited competitive behaviours; in terms of Conscientiousness, the participants of Pilot Study One did show some organisation of their goals and lifetime ambitions, but, on the other hand, were not likely to strive for them at all price. One of the aims of Pilot Study One consisted in examining the role of categorical variables, such as gender, type of received education, or employment status, as well as other variables which might exert influence on either learners’ use of language learning strategies, or the levels of their personality traits. Regarding the scales whose data distributions varied significantly from normal, the results of the Mann-Whitney U-tests

166 indicated that no statistically significant differences existed between the investigated men and women regarding the frequency of using LLS in general (p > 0.05), or in the frequency of using metacognitive or social strategies (p > 0.05). For the scales whose data distribution did not significantly vary from normal, the results of the independent samples t-tests indicated that the female and male participants did not significantly differ in the use of the remaining strategy scales, i.e., including memory, cognitive, compensation, and affective strategies. The lack of statistically significant differences across genders was rather surprising since the results of a number of earlier studies indicated that women were more frequent strategy users than men (Green and Oxford 1995; Dreyer and Oxford 1996; Nyikos 2008; Yılmaz 2010). It was inferred that since very few men participated in the study, and thus the number of men may not have been enough to reveal the differences. The conclusion for the final study was therefore to ensure a greater proportion of male participants. Another data split for analytical purposes involved grouping learners according to the type of received education. For that purpose, comparisons were conducted for full- time and extra-mural students. Tests were applied analogically to the previous data split (t-tests and Mann-Whitney test for independent samples were used for normal and non- normal sets of data respectively). With respect to personality, extra-mural students scored significantly higher on the scale of Conscientiousness (p < 0.05). At the same time, the differences in the frequency of using language learning strategy between full-time and extra-mural students were not statistically significant. These findings constitute the first reason why participation in the main study was limited to full-time students. Afterwards, learners’ age was used as a criterion for the data split. Learners were divided into two categories, young adults (19-30), and learners in their mid-adulthood (30-60). Interestingly, three statistically significant relationships were observed. Firstly, learners in the middle adulthood stage showed noticeably higher levels of Conscientious- ness than their younger counterparts, which can probably be related to the fact that people who decide to continue their education after reaching the age of 30 need more determination and effort to succeed due to other than educational involvements in their lives, including professional and family-related duties. The proportion of these people in the student population was rather low (about 25%) and also, in general, those who decide to take up studies in middle adulthood are typically truly determined to sacrifice a considerable dose of effort and contribute a lot of hard work to accomplishing their educational

167 objectives. As could be expected, middle-aged adults exceeded young adults in the use of memory strategies, which might be viewed as traditional and more mechanical than other language learning strategies. A similar trend was observed in the use of metacognitive strategies, which means that older learners might also be better organisers, and experience fewer problems being in charge of the language learning process. The greatest differences were demonstrated in using flashcards, analysing learners’ own mistakes, and planning to learn the language (all three favoured by older students). Students in middle adulthood were also more often able to specify their language learning aims. The last observation could be linked to stronger intrinsic motivation among older students. All the described tendencies in the use of particular tactics were statistically significant (t-test for independent samples, p <0.05). It could be concluded that while participants’ age did not affect the possibility of using both research instruments applied in Pilot Study One, it certainly exerted influence on the variation in strategy use. The obtained results constitute the final premise on the basis of which participation in the main study was limited to full-time students so as to exclude the impact of age from analysis and concentrate on the relationships between personality and use of language learning strategies. To sum up, in terms of the fulfilment of the goodness criteria by the research instruments applied in Pilot Study One, no concerns arose regarding their as participation in the study was voluntary, unsupervised and neither rewarded with any extra credits nor even recommended. Both instruments were standardised and normalised as they were administered according to the guidelines by their authors (Costa and McCrae 1992; Ox- ford 1990), and the manuals for both of them also included instructions for interpreting raw scores so that no two same scores could serve as a basis for different interpretations. Steps were taken to improve the reliability of the problematic categories of the SILL through addressing a panel of experts and expanding the scales. Also, attempts were made to render a good quality of both the translation of the original SILL ver. 7.0 and the wording of the added questions, so as to ensure a good quality of the final version of the adaptation. Information on the validity of the research instruments is presented in the subchapter dedicated to the main study.

168 4.3. Pilot Study Two

Pilot Study Two was conducted in July 2014 in order to revisit the design rationale of the main study, with a major emphasis on the personality assessment tool. In particular, the Polish adaptation of Ten-Item Personality Inventory (TIPI) (Gosling et al. 2003) was used as an alternative measurement of personality to NEO-FFI, which was employed in Pilot Study One. Both TIPI and NEO-FFI are based on the Five Factor Model of personality traits, according to which human personality can is composed of five pillars, that is Open- ness to experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism. Recommended for the purposes of quick survey research, TIPI was adapted for Poles in 2014 (TIPI-PL; Sorokowska et al.2014). Its main advantages do not only consist in its brevity and free-of-charge implementation. The other advantage of the questionnaire consisted in having at least some potential to be used for scientific purposes as its Cronbach’s alpha ratios ranged between 0.44 (Openness to Experience, paper-and-pencil version) to 0.83 (Emotional Stability, Internet version) with regard to the Polish adaptation (Soro- kowska et al. 2014: 6), and similar values had been quoted in previous studies with regard to other adaptations (Muck et al. 2007; Renau et al. 2013). Although it cannot be viewed as a valid tool in clinical diagnosis, it is recommended for research, particularly among respondents with concentration problems, aimed to assess the personalities of many people, intended to be used as a sub-part in large-scale surveys, or pilot studies. Another suggested use of TIPI is for cross-cultural research, as the instrument is available in a number of language versions and even if it is not, translation by a specialist does not take a long time (Sorokowska et al. 2014). Apart from validating an alternative personality questionnaire, the study was conducted in order to investigate cultural differences in language learning strategy use, and, more specifically, to examine the impact of Polish and Italian teenage students’ cultural background on their strategic choices. It was expected that learners’ origin would affect their strategy repertoire, and possibly help detect tendencies in strategy use which would be characteristic of Polish and Italian learners. Finally, the aim of the study was also to provide insights into the relationships between learners’ personality traits and the frequency of using each of strategy scales distinguished by Oxford (1990), and identify which research tool, TIPI-PL (Sorokowska et al. 2014) or NEO-FFI (Costa and McCrae 1992) would constitute a more appropriate tool to interpret these relationships.

169 4.3.1. Participants and data collection

The participants of Pilot Study Two included 194 students, including 138 Polish and 56 Italian learners of English aged 13-18. The Polish learners were students of two types of schools, a lower secondary school and an upper secondary school in Rawicz, a small town 100km away from Poznań. In September 2014 the students of the upper secondary school participated in a lecture on language learning strategies by the author of the present dissertation, which was expected to encourage their participation in the study. During the following week, they were asked to complete adaptations of the above-mentioned questionnaires in their English classes. The questionnaires were collected by the teachers who were in charge of those classes. The Italian participants attended two types of secondary schools, scuola di primo grado and scuola di secondo grado, for students aged 11-13 and 14-15, with an optional extension for students aged 16-18. At the time of data collection, they were attending a summer language school in Bath, in the UK. The questionnaires were thus handed to the students during their English classes, and the students were supervised by their holiday teachers while they were filling in the questionnaires. Also, all of the Italians participating in the study had already been learning English before coming to Britain and were thus given a placement test on arrival. A separate part of the placement procedure was an interview, whose result was incorporated into the final result of the placement test (20% of the total test score). Levels ranged from 1 (complete beginners) to 6 (advanced students). No students were assigned to group 1. Table 19 presents descriptive statistics characterising the participants of Pilot Study Two.

Table 19. Participants of Pilot Study Two.

national- mini- maxi- std. de- vari- kurto- N range mean skewness ity mum mum viation ance sis

Italian 56 4 13 17 14.54 1.32 1.74 0.58 -0.57.

Polish 138 5 13 18 15.57 1.79 3.19 -0.07 -1.28

As can be seen from the above Table, Polish students were on average one year older than their Italian counterparts. The one-year difference in the age range for both nationalities was reflected in the values of both standard variation and considerably higher age variance for Poles. Also, the positive skewness observed in the group of Italian learners reflected the fact that within the group of summer language school attendants, more

170 learners were closer to the maximum age of 17 than to the minimum age of 13. At the same time, age values for Polish learners were almost evenly distributed, with the skewness value amounting to -0.07.

4.3.2. Measures

As mentioned earlier, one of the purposes of conducting Pilot Study 2 was to examine whether an alternative tool to NEO-FFI could be used for personality measurement in the main study. An attempt was made to assess the reliability of TIPI PL. Adaptation of SILL ver. 7.0 was also subject to reliability analysis. Out of the four methods of testing the reliability of a research instrument (Hornowska 2007: 49-69), the split-half test and Cronbach’s alpha were employed. The test-retest method was abandoned for two major reasons. First of all, the interval between the test and the retest would not be longer than two weeks in case of the Italian students of the British language school (the length of their stay in Britain was determined by the length of their English course). Secondly, a number of factors, including remembering the answers, the effect of learning, and changes in research settings, could influence the assumed parallel between the test and the retest (Nowakowska 1975: 38), which is particularly true for any pedagogical research. There- fore, it was decided to use the Spearman-Brown formula to calculate reliability for TIPI- PL. The formula is particularly recommended for two-item scales (Eisinga et al. 2013). Table 20 shows the values of Spearman-Brown formula for two-item personality scales in TIPI. The results show that three personality scales manifested extremely low split- half reliability values. However, it must be remembered that the inventory is designed to estimate personality dimensions (very broad constructs) by means of only two questions and the questions themselves are not merely mutually juxtaposed, but also relate to broader semantic categories so as to account for a more extensive part of the given personality dimension (Gosling et al. 2013: 516).

171 Table 20. Spearman-Brown reliability statistics for TIPI in Pilot Study Two.

scale items reliability values

Openness to experi- +5/-10 0.182 ence Conscientiousness +3/-8 0.441 Extraversion +1/-6 0.426 Agreeableness +7/-2 0.165 Emotional Stability +9/-4 0.205

Since all the scales of language learning strategies included six or more items, Cronbach’s alpha was calculated for each of them to evaluate the reliability of the adaptation of SILL ver. 7.0 as a research tool. Calculated in this way, the reliability coefficient refers to the internal consistency of a research instrument and thus also validates the homogeneity of a construct (Hornowska 2007: 55). Cronbach’s alpha values for strategy scales are shown in Table 21.

Table 21. Cronbach’s alpha for strategy scales in Pilot Study Two. reliability for nationalities no. of reliability for all scale items participants Italians Poles Memory strategies 9 0.602 0.591 0.617 Cognitive strategies 14 0.775 0.785 0.776 Compensation strat- 6 0.549 0.528 0.603 egies Metacognitive strat- 9 0.803 0.818 0.800 egies Affective strategies 6 0.528 0.514 0.509 Social strategies 6 0.698 0.692 0.704

As can be seen from the number of items in each of the scales, no modifications were made to the original scales36. In general, the alpha values were satisfactory for cognitive, metacognitive, and social strategies, and slightly below the acceptable level for memory strategies. As far as affective and compensation strategies are concerned, alpha values were poor, but they do not differ considerably from the results obtained in Pilot Study One (Table 2). Therefore, the decision to expand the two strategy scales of compensation and affective strategies, made on the basis of Pilot Study One, was not altered after conducting Pilot Study Two.

36 The panel of expert judges was being assembled at that time and strategy scales had not been expanded at the time when Pilot Study Two was being conducted.

172 4.3.3. Analytical procedures

The analytical procedures applied in Pilot Study Two reflected those used in Pilot Study One. First, data distribution was tested for normality within personality and strategy scales as well for the GULLS scale, calculated as the mean value of the frequency of overall strategy use. Kolmogorov-Smirnov test was run separately for two data subsets, that is Polish and Italian participants of the study. Each time, H0 assumed that data distribution was not significantly different from normal whereas H1 assumed that data distribution differed significantly from normal. H0 was accepted if p values exceeded 0.05. A summary of relevant findings can be found in Table 22.

Table 22. K-S normality test results in Pilot Study Two.

Data distribution Scale Italians Poles Memory strategies (mean) normally distributed not normally distributed Cognitive strategies (mean) normally distributed normally distributed Metacognitive strategies (mean) normally distributed normally distributed Compensation strategies (mean) normally distributed normally distributed Social strategies (mean) normally distributed normally distributed Affective strategies (mean) not normally distributed normally distributed GULLS (mean) not normally distributed normally distributed Emotional Stability not normally distributed not normally distributed Extraversion normally distributed not normally distributed Openness to Experience not normally distributed not normally distributed Agreeableness normally distributed not normally distributed Conscientiousness not normally distributed not normally distributed

Since the above results indicated that some of the data sets were normally distributed whereas the other distributions differed significantly from normal (p < 0.05), both parametric tests and their non-parametric equivalents were used in calculating the results of Pilot Study Two. Firstly, data sets of strategies were tested for differences either by Student’s t-test (for normal distributions), or by Mann-Whitney U test (for distributions which varied significantly from normal). Secondly, two types of correlation coefficients were computed, that is the Pearson correlation coefficient for any pair of parametric sets and Spearman’s rank coefficient for pairs where at least one distribution differed significantly from normal.

173 4.3.4. Selected findings and implications for the main study

The results of the Mann-Whitney independent samples U-test indicated that the Polish participants of the study used language learning strategies significantly less frequently than their Italian counterparts (GULLS mean values differed significantly; p < 0.05). The greatest differences were observed in the use of affective and social strategies. With regard to the affective scale, the Italian participants outperformed the Polish participants, too, and the results of the Mann-Whitney test for independent samples confirmed the statistically significant character of the difference (p < 0.05). Similarly, the Italian students were also significantly more frequent users of affective strategies (Mann-Whitney test, p < 0.05). It could generally be concluded from the study that the Polish participants were more passive strategy users. For example, they paid attention when something was said in English or revised material more frequently than they tried to find ways to be a good learner of English. Also, they did not recognise the usefulness of the tactic of rewarding themselves for good performance - they used it significantly less frequently than the Ital- ian students did. These tendencies could, at least partly, stem from the teaching methodology that the Polish learners had been exposed to, as well as the curriculum. No statistically significant correlations existed between the level of the investigated Polish lower secondary school students’ personality traits and the frequency of using any of the strategy scales, and only two moderate correlations were detected in the group of the investigated Polish upper secondary school students. The frequency of using affective language learning strategies was found to moderately correlate with the level of learners’ Openness to Experience (p < 0.05, ρ = 0.325), and Agreeableness (p < 0.05, ρ = 0.344). It was also observed that both categories of the investigated Polish secondary school students were either frequent strategy users or infrequent strategy users in general. In other words, the investigated students either used a number of different language learning strategies relatively frequently or they used all of them rarely. Therefore, it was difficult to trace any preferences of particular scales in strategy use among the investigated students. To summarise, one of the aims of Pilot Study Two was to determine whether the adaptations of two inventories, TIPI-PL and SILL ver. 7.0, could be used to measure the constructs that they are expected to measure and, with regard to the adaptation of SILL ver. 7.0, whether the split into six categories of language learning strategies is

174 psychometrically grounded. Overall, it could be inferred that the Polish adaptation of SILL ver. 7.0 (Oxford 1990) offers good prospects of strategy measurement with respect to most of the original strategy scales. Although space for improvement could be noticed as regards compensation and affective strategies, the modifications introduced on the basis of Pilot Study One, consisting in expanding the scales, was expected to result in enhancing a better reliability in both cases. On the other hand, the overall impression one could form about TIPI-PL was that the instrument left much to be desired in terms of its reliability, understood both as the measure of integrity, and homogeneity of human personality scales. Perhaps in consequence, hardly any relationships could be traced regarding the correspondence between the investigated students’ personality traits and their strategy use. Therefore, a decision was made to dismiss it, and employ the Polish adaptation of NEO-FFI (Zawadzki et al. 2010) in the main study. The research process which illustrates the transition from Pilot Study One to the main study can be seen in Figure 2.

Fig. 2.Transition from Pilot Study One to the main study.

4.4. The design of the main study

In order to comply with the contemporary standards of methodological data triangulation (Droździał-Szelest 1997; Pawlak 2010; Oxford 2017), the main study was designed as a mixed-method study. It was expected that its quantitative part would provide information

175 about the investigated learners’ overall use of LLS, as well the interaction between strategy use and specific learner IDs, including, most importantly from the point of view of the present thesis, their personality traits. As pointed out by Oxford (2017: 424), “(l)ook- ing at such background factors and at more detailed strategy information for any given individual or group of students would provide a more meaningful picture”. Through the use of statistical procedures it was thus intended to identify clusters of learners with reference to strategy use and other relevant IDs, and trace patterns in strategy use. At the same time, qualitative data obtained from semi-structured interviews was expected to provide contextual insights into strategy use, involving both the macro-context, such as the impact of the learners’ cultural background, and the micro-context of strategy use with reference to particular tasks, language subsystems, such as grammar or vocabulary, or the flexibility and multidimensionality of particular strategies. As stressed by Oxford (2017: 424), only by gathering data from both types of studies does research into LLS stand the chance of accounting for the complexity of the strategy construct.

4.4.1. Research questions

Over the last forty years, a great deal of research has been conducted in the field of language learning strategies, focusing especially on the between strategy use and attainment. At the same time, although numerous empirical investigations were dedicated to investigating the factors which underlie strategy choice, the number of studies investigating the relationships between learners’ personality and their language learning strategies remains relatively low, even though Oxford (2011: 40) stated that the choice of strategies might be connected with learners’ intrinsic characteristics, such as Extroversion/Introversion. Despite the fact that some studies address the issue of relationships between language learning strategies and personality, their results might be questionable due to insufficient attention given to the study of personality as an ID, an unclear framework for investigating the variation of personality traits, or insufficient consideration given to the fulfilment of the goodness criteria by the applied psychometric tests, as exemplified in the final part of Chapter One. In view of these considerations, the present dissertation aims to answer the following research questions:

176 1. Is there a statistically-significant relationship between the level of language learners’ ‘big’ five personality traits and their use of language learning strategies measured by the reported frequency of using memory, cognitive, compensation, metacognitive, affective, and social strategies, and the general frequency of strategy use (GULLS)? 2. What is the relationship between attainment and the use of LLs, both with respect to the six categories and overall? 3. What is the relationship between attainment and learners’ ‘big’ five personality traits? 4. Is it possible to group the investigated language learners into relatively homogenous clusters of strategy users and describe them in terms of the levels of their ‘big’ five personality traits? To summarise, the study was designed to examine how learners’ personality characteristics affected their strategic choices, as well as to provide insights how such hypothesised interaction translates into language proficiency. It was expected that conclusions could be drawn both with respect to learners’ general learning habits (insights from quantitative analysis) and in the context of specific skills and tasks (qualitative analysis).

4.4.2. Participants

The choice of the investigated population of learners in the main study was subject to a conscientious procedure since, apart from answering the above research questions, the study was also conducted in order to expand the existing knowledge about the use of language learning strategies by students from two universities in Poznań. For that reason, special attention was paid to sampling and ensuring the fulfilment of goodness criteria by the applied research tools in order to achieve satisfactory representativeness of both samples tested in the study. Attempts were made to conduct statistical inference for the investigated population of students, and to provide insights into language learning strategy repertoire, as well as to draw pedagogical conclusions. The description of the participants of the main study was designed to provide three types of factual information: first, a general characterisation of young adults is provided, as the author of the present dissertation believes that it is likely for studies dedicated to investigating

177 language learners to benefit from the general findings provided by educational and developmental psychologists; second, the population of learners for which the results of the study are later generalised is characterised in terms of the learning conditions imposed by the institutional settings of foreign language education at Adam Mickiewicz University and WSB Uni- versity in Poznań, the two universities where the study was conducted; third, a thorough exploration of the characteristics of sample of students who took part in both the quantitative and the qualitative part of the study is conducted in terms of such variables as age, gender, employment status, place of residence during a major period of receiving education, chosen programme and faculty of studies, and level of language proficiency37. Additionally, the characteristics of the 19 participants of the qualitative part of the main study are provided. All participants of the study were young adults according to Brzezińska’s (2005: 9) classification38, and thus were characterised by similar developmental patterns of skills and abilities, which are crucial to the process of learning. According to Brzezińska (2005), these include dialectic, meta systemic, and relativistic thinking, increased learning potential in areas of interest, ease of identifying and solving practical problems, considerable agility of short-term and long-term memory, good orientation in action settings, increase in IQ, and a maximum level of liquid intelligence. Moreover, as indicated by Anthis and La Voie’s (2006), individuals’ readiness to accept change in young adulthood may foster human development in multiple domains including learning habits. Finally, as asserted by Newman and Newman (2017), early adulthood is a period of time most distinct for people’s sensitivity and willingness to learn, which, according to Brzezińska and co-op- erates (2016), ensures individuals’ proper cognitive functioning. According to Arnett (2000: 470-473), the subjective indicators of becoming an adult involve assuming responsibility for oneself, making independent decisions, planning one’s own activity, creating plans for the future, and the ability to become involved in activities of one’s own choice, while objective indicators of becoming an adult

37 Although it is tempting to group individual differences into categories such as affective, cognitive, sociocultural, and other types of factors (cf. Cook 2008), recently a tendency has emerged to resist the need to classify these, as they may belong to more than one domain (e.g. age may be perceived as a biological or cognitive factor) as signalled by Pawlak (2012). 38Brzezińska (2005: 9) admits that while there is no controversy regarding the lower range of age (18), the upper range of the tended to vary in relevant literature from 30 to 35.

178 comprise such aspects as professional and financial independence, and living away from one’s family of origin. These are reflected in accomplishing such developmental tasks as settling down, entering relationships, getting married, raising children, entering the labour market, assuming social responsibility, and searching for social groups of similar values or characteristics (Brzezińska et al. 2016: 302). It is justified to expect that some of the mentioned qualities of young adults exert a direct and powerful influence on a number of different aspects of their learning, including language learning. For instance, assuming responsibility for oneself can be associated with greater independence and making the most of one’s autonomy as a language learner while the ability and willingness to plan constitute the core of metacognition. Moreover, creating plans for future may involve both career and educational choices, both considerably affecting the choice of language learning strategies, and greater involvement in activities of one’s own choice strongly corresponds with intrinsic motivation and volitional processes. Given the recognition of the role of the context of learning a foreign language (Oxford 2017; Pawlak and Oxford 2018), it is vital to consider the learners whose characteristics are investigated in the present dissertation in their specific educational environment. Therefore, one needs to recognise the rules which govern foreign language education at both universities considered. It needs to be pointed out that the description of the regulations and institutional framework for learning a foreign language at Adam Mic- kiewicz University (AMU) concerns learning English rather than learning any other language, as certain differences exist, mostly for practical reasons39. Learning a foreign language is obligatory at AMU and a vast majority of students need to pass the B2 certificate exam, conducted by AMU Language School, in order to graduate. Students may be ex- empted from this obligation if they hold one of a number of certificates specified in the regulations issued by the Minister of Science and Higher Education. Also, the requirement is not mandatory for the students of two other faculties, Faculty of Modern Lan- guages and Literatures, and Faculty of English, students of English at the Faculty of Fine Arts and Pedagogy in Kalisz, and students of Elementary Education and English at two faculties, the Faculty of the Fine Arts and Pedagogy in Kalisz, who become advanced

39 English is chosen by a vast majority of AMU students, and required as a foreign language by some Faculties or Institutes (e.g. AMU Institute of Psychology, Faculty of Mathematics and Informatics), which influences the organisation of other foreign language courses, e.g. certificate groups need to be combined with other types of groups at the same level, groups are created across different faculties, teachers and learners are under greater financial pressure, etc.)

179 users of a chosen language, and obtain a university diploma which certifies the level of their linguistic competence. All these students attend courses in foreign languages (30 hours per semester), which they are enrolled on the basis of their placement test scores. The total number of hours of instruction for the students targeted as participants of the present study, however, varies depending on the result of their placement test. The test is taken online and only accessible to first-year students who are about to start their tertiary education in English and it consists of Reading Comprehension and Use of English tasks, and a part testing functional language skills. For technical reasons, Listening Comprehen- sion and Writing skills are not tested in the placement test. In the Reading Comprehension task, students are asked to read short fragments of texts and choose the right answers to the questions based on the contents of the texts, whereas in the Use of English tasks, students choose one of four answers (the grammatical and lexical focus of these questions is evenly balanced). The test also includes a section which checks students’ ability to respond in a correct way in everyday real-life situations, whose main focus is functional language. Students who are diagnosed as A2 learners of a foreign language enrol in a 120- hour course, those diagnosed as B1 learners attend a 90-hour course, B2 learners attend a 60-hour course, and learners whose competence exceeds B2 are only entitled for a 30- hour preparatory course which familiarises them with the types of content and tasks which they can expect in their certificate exam. The number of hours of instruction does not exceed 30 in one semester, with the exception of psychology students, who still participate in a five-year programme of studies. The number of hours of instruction remains the same as for students of any other main subject, but they have English classes twice a week, which means that their courses last for the maximum of two semesters. Students of bioinformatics constitute another exception, as they participate in a joint programme, co- organised by two universities, Adam Mickiewicz University, and Poznań University of Technology. Their placement is organised by their teachers, and they participate in 120- hour course of English for Bioinformatics (30 hours per term). Unlike psychology students, they do not take the certificate exam. The procedures described above differ considerably from that present at the Poz- nań University of Banking. All students, irrespectively of their initial level of competence, need to attend a 240-hour foreign language course. They are also subject to much stricter regulations concerning exams. First of all, they take an exam in a foreign language

180 after each of four initial terms. Exam formats differ across levels. A vast majority of exam tasks directly reflect the tasks that candidates can expect at Cambridge BEC exams (BEC Preliminary, BEC Vantage, and BEC Higher). However, some tasks, especially those used in lower level exams, are simplified tasks, which were designed in order to prepare students for other, more complex tasks, once they have progressed to a higher level of linguistic competence. Full-time students are enrolled in specific courses on the basis of their secondary school final examination results in English. They attend 120-hour courses in their first year of studies (60hours per semester, English classes held twice a week), which are followed by similar 120-hour courses in their second year of studies. The only difference between first and second year students is that lower level learners attend Gen- eral English courses in the first year and Business English courses, while higher level learners attend Business English courses from the very outset. Much as one of the sampling criteria was to ensure a high homogeneity of participants in terms of age, similar homogeneity could not be ensured in terms of the investigated learners’ level of proficiency in English. Participants were students of BA and BS studies, whose level of English was diagnosed as A2, B1, or B2 at Adam Mickiewicz University, and A1+, A2, A2+, and B1 at WSB University in Poznań. For analytical purposes, students of the latter university were grouped into three levels: a) A2 (including students attending A1+ and A2 English lessons, but having passed school leaving exam in English); b) B1 (formally attending A2+ classes, but diagnosed to be able to work with B1 level coursebooks); c) B2 (formally attending B1 classes but using upper-intermediate and “upper-intermediate to advanced” coursebooks). In order to calculate the minimum sample size for each group of students, a formula was adapted from Brzeziński (2004: 247)40. According to the formula, minimal sample sizes were calculated for two populations:

40 () () � = () � = () , where where the following symbols represent: n – minimal sample size; P – estimated fraction (assumed to be 0.5 if there is no information); e – acceptable sampling error, amounting to 0.04; Z – value resulting from the assumed confidence interval; for the assumed confidence interval equal to 0.95, Z = 1.96; N – population size.

181 a) the population of Adam Mickiewicz University full-time BA students learning General English in classes taught by AMU Language School lecturers (13 out of 15 faculties41), which amounted to ca. 7500 students in the academic year 2014/2015. Given that the value of estimated fraction amounted to 0.5, the value of acceptable sampling error amounted to 0.04, and the Z value calculated for the confidence interval of 0.95 amounted to 1.96, the n value amounted to 557 students. The questionnaires were distributed to 700 students in order to limit the anticipated impact of returned empty sheets or randomly answered questions. 622 questionnaires were qualified for further analysis. The population was subsequently limited to 612 so as that analysis could be conducted on a relatively homogeneous group of students in terms of age, and after elimination of missing data which could not be replaced by means of any statistical procedure. b) the population of WSB University in Poznań full-time BA/BS students, in the academic year 2015/16. These students attended courses in General (year one) and Business English (year 2), but they received twice more instruction in one term (that is 60 hours for every student). Similarly to AMU students, WSB students are also obliged to take final tests, but the they are tested at the end of each term. The exam format is close to Cambridge BEC exams for appropriate levels. The minimum number of participants calculated in order to generalise the results of the study stood at 130, while the number of participants of the study amounted to 134. To sum up, the study is representative for the following population of full-time AMU students, who: a) study English as a foreign language at one of 13 faculties, excluding the Faculty of English, where English is studied as a main subject, and the Faculty of Pedagogy and Fine Arts in Kalisz; b) study General English as part of their BA curriculum and students of the Institute of Psychology who still participate in a five-year programme of studies;

41 Students of the Faculty of English and students of the Faculty of Fine Arts and Pedagogy attend English classes taught by faculty staff. Moreover, the organisation of these classes differs from the organisation of the classes taught by AMU Language School (most students learn English in more than one subject and the instruction is not limited to learning the language, but also covers learning about the language).

182 c) the population of full-time students at WSB University in Poznań, who study either General or Business English in their first and second year of studies as part of their BA/BS. The total number of participants of the main study amounted to 742. In terms of gender, the majority of participants, over 69%, were female, which reflected general trend in the populations of both universities. 19 students were selected for the qualitative part of the study, 16 of whom were female, and 3 of whom were male. The distribution of participants’ age is shown in Fig. 3.

Fig. 2. Participants’ age.

Only 8.3% of the investigated students were usually or always involved in voluntary work, and more than a half had never done apprentice of any kind. Students’ activity was considerably greater as far as part-time work was concerned - more than 60%

183 admitted that they had done part-time work at least sometimes. This is probably due to the fact that students put financial motivation for work first, and they may be forced to work due to insufficient levels of other sources of income. On the other hand, not even one in five students had ever worked as a full-time employee. This can be both a sign of changing tendencies in the labour market (permanent employment is gradually being replaced by temporary employment within projects), and a sign of young adults’ position in the labour market which forces them to accept employment in any form rather than allows them to choose favourable conditions. Detailed information on participants’ professional activity can be found in Table 23.

Table 23. Participants’ professional activity in the main study.

Voluntary Appren- Part-time Full-time

% of answers work tice work work

never 42.8 55.3 17.4 62.2

rarely 29.1 19.5 21.0 13.5

sometimes 19.7 15.8 35.7 6.5

usually 4.2 4.7 17.7 6.8

(almost) always 4.1 4.7 8.1 11.1

Figure 3 illustrates participants’ levels of competence in English in terms of the CEFR. As can be seen, A1+ participants only constituted a tiny minority of the entire sample. This results from the fact that English is only taught at this level only at WSB University and a vast majority of students have been learning this language since primary education.

184

Fig. 3. Participants’ level of competence in English.

The distribution of the participants of the main study across university faculties is shown in Figure 4. As can be seen, the number of students from the Faculty of Finance and Banking is considerably higher than the number of students of any other faculty. This is due to the organisational structure of WSB University in Poznań, which functions as a one-faculty university, whereas AMU is divided into 15 faculties, 13 of which were included in the present study. The numbers of participants from particular faculties result from the procedure which was employed to select respondents. In order to ensure random selection of respondents, stratified sampling was applied, and English teachers working at AMU Language School were chosen at random as a stratum, which resulted in the specific numbers of participants from different faculties.

185

Fig. 4. Participants across departments.

Finally, students were grouped according to their residence while receiving education. Figure 5 shows the shares of participants from different places of residence.

186

Fig. 5. Participants’ place of residence while receiving education.

The participants of the qualitative part of the main study were all volunteers, students of Adam Mickiewicz University in Poznań and WSB University in Poznań (formerly Poznań University of Banking) whose personality traits and LLS had already been investigated in the quantitative part of the study. Nineteen students were selected for the interview, including 17 females and 2 males. Participants were selected from the group of volunteers on the basis of the level of their personality traits. In order to connect strategy use to the level of learners’ personality traits, learners characterised by extremely high or low trait levels (stens 7-10 or 1-3 for any of the ‘big’ five traits) were preferred as participants. Relevant data about participants concerning their gender, level of competence in English according to CEFR, and level of personality traits can be found in Tables 24 and 25. The letters l, m, and h describe the levels of personality traits: l - low (stens 1- 3), m- medium (stens 4-6), and h - high (stens 7-10). The names of the participants are not revealed.

187 Table 24. AMU students participating in interviews Partici- 1 2 3 4 5 6 7 8 9 pant Gender M F F F F F F M F CEFR B2 C1 C1 B1 B1 B1 B1 B1 B1 N l l m h l h m L m E m l l l h m m M h O h h h m h m m H h A l m m m m l h L m C h m m h m m h M h

Table 25. WSB students participating in interviews Participant 1 2 3 4 5 6 7 8 9 10 11 Gender F M F F F F F F F F F CEFR B2 B2 B2 B2 B2 B2 B1 A2 A2 A2 B2 N m h m m m m m l m M m E l m l m h l m h m M l O h m h l h m m m l L h A m h l m m m h m m H m C m h m h m m h m l M m

4.4.3. Data collection and research instruments

In the quantitative part of the main study, participants were asked to complete two adaptations of the psychometric tests described in the previous chapter, i.e., the Polish adaptation of SILL ver. 7.0 (Oxford 1990, translated and amended by the author of the present dissertation), and the adaptation of NEO-FFI (Zawadzki et al. 2010). Data collection took place in participants’ English classes so as to maximise the number of returned questionnaires. Participation in the study was encouraged, but voluntary. The study was conducted after the official consents of the Head of AMU Language School, and the Dean of the Faculty of Banking at the WSB University, had been granted. Questionnaires were deliv- ered to AMU and WSB English teachers who supervised data collection in their classrooms. The choice of research instruments in the main study was determined by analysing the degree of meeting the goodness criteria by the instruments used in pilot studies. Due to the doubts concerning the internal validity of TIPI (Sorokowska et al. 2014) and described in subchapter 4.3 (Pilot Study Two), the instrument was abandoned and instead, NEO-FFI (Costa and McCrae 1992) was used in the main study. Since two of the six

188 original scales of SILL ver. 7.0, i.e., the compensation scale, and the affective scale, showed low reliability values in Pilot Study One (see section 4.2), an attempt was made to improve the instrument by expanding the scales. The amendments made to the questionnaire were assessed by a panel of experts (see section 4.2.2 above). The following sections, 4.4.3.1 and 4.4.3.2, discuss the research instruments applied in the main study which were used with the purpose of exploring the relationships between language learning strategies and learners’ personalities. Section 4.4.3.3 discusses the design and schedule of the interview which was used in the qualitative part of the research in order to triangulate the data obtained in the quantitative part.

NEO-FFI

Recommended by the Polish Psychological Association for both research and clinical purposes (Zawadzki et. al 2010), NEO-FFI is an inventory which allows a researcher to measure an individual’s personality in five bipolar dimensions corresponding with the Five Factor Model (FFM) of personality redeveloped by Costa and McCrae (1985). Also known as the NEOAC/OCEAN model, it postulates the existence of five factors underlying human personality, that is Openness to Experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism, and subscribes to the trait approach to personality, described in the final part of subchapter 1.2. Discussed in the second part of Chapter One, the FFM has two major advantages over alternative models. Firstly, it is said to measure different personality traits without the danger of their overlapping, and secondly, the traits display a considerable level of consistency in interviews, descriptions and when observed (Schacter et al. 2012: 496). In order to measure the level of personality traits according to the Five Factor Model, the NEO Personality Inventory (NEO-PI-R) was created and updated by Costa and McCrae (1985; 1992). Its revised version (1992) contains 240 items, and requires basic comprehension skills in participants’ mother tongue. In particular, form S was designed for self-reports, and it is appropriate for use with adults, including university students. Costa and McCrae’s (1992) NEO-FFI is a shorter version of NEO-PI-R and includes 60 items, which can be divided into five personality scales, i.e., Openness to Experience, Conscientiousness, Extraversion, Agreeableness, and Neuroti- cism, each one measured by 12 questions. Respondents choose their answers on a five-

189 point Likert scale by indicating to what degree they agree or disagree with a particular statement (1 = strongly agree; 5 = strongly disagree) At the end of the questionnaire, respondents also confirm whether they have answered all questions and whether their answers were truthful. The Polish adaptation is, in fact, a translation of the original. The only difference between the Polish and the English version consists in the fact that respondents write their answers on the answer sheet of the Polish version whereas answers to the English version are provided on a separate sheet. The scales of the questionnaire are described in Table 26, based on Zawadzki et al. (2010: 64-65).

Table 26. NEO-FFI scales.

Personality trait Definition Examples of manifestations a personality trait which describes an indi- excitement aroused by reading poetry or vidual’s tendency to search for and value Openness contemplating a work of art, experiences, as well as tolerance of new to Experience interest in the nature of the universe and concepts and cognitive curiosity human nature

a personality trait which describes the degree of an individual’s organisation, perse- clearly set objectives, verance, and their motivation to achieve Conscientiousness working systematically, goals, or, in other words, an individual’s being perceived as a reliable worker attitude to work

a personality trait which characterises the quality and amount of social interaction, as well as the level of an individual’s activity willingness to be surrounded by people, Extraversion and their ability to experience positive being cheerful and full of life emotions

a personality trait which describes an individual’s attitude to other people, or, in considered to be cold and shrewd, other words, their interpersonal orientation Agreeableness being able to manipulate others in order resulting in altruism / antagonism in to get what one wants thoughts, feelings, and actions

a personality trait which reflects one’s ability of emotional adaptation or lack of feeling tense or anxious, emotional stability, expressed through ex- feeling helpless, Neuroticism periencing negative emotions, such as fear, feeling dependent on others in solving confusion, dissatisfaction, anger, feeling of problems guilt, and sensitivity to psychological stress

As a psychological test, NEO-FFI needs to fulfil six goodness criteria, i.e., it needs to be objective, standardised, reliable, valid, interpreted according to norms, and adapted to the specific characteristics of a local population (Hornowska 2007: 25-32). Each of the above criteria was scrutinised and relevant conclusions are described in the following

190 paragraphs. First of all, the test can also be called “standardised”, as it meets the following criteria: general rules for testing are included in the manual, which enables the researcher to conduct the test in the same manner every time there is a need to repeat it; the instructions specify that the questionnaire can be used in group research, define the time limit, marking scheme, as well as suggest conducting the study in the pen- and-paper mode; the key defines how to calculate, code, check, and register test scores, norms presented in the manual refer to clearly defined populations (men and women belonging to specific age groups) and can be used. The Polish adaptation of NEO-FFI (Zawadzki et al. 2010) is also accompanied by norms depending on respondents’ age and gender. In order to optimise the discrimination power of the norms, they are presented as sten scores. Sten norms are presented for both genders across five age groups in appendix C of the adaptation. Second of all, both the English version of NEO-FFI (Costa and McCrae 1992) and its Polish adaptation (Zawadzki et al. 2010) include a key that needs to be used to calculate the scores for each of the five personality dimensions. The tests are therefore objective as the risk of interpreting the same result in a different way two different researchers is reduced to a minimum (“blind” diagnosis). Another measure which accounts for the degree to which a psychometric test fulfils the goodness criteria is its validity. According to Messick (1994: 6), “(v)alidity is an overall evaluative judgment of the degree to which empirical evidence and theoretical rationales support the adequacy and appropriateness of interpretations and actions based on test scores or other modes of assessment”. Apart from face validity, i.e., an estimate whether a test measures what it subjectively appears to measure to the person answering the test (Anastasi and Urbina 1997: 164), three main constituents of validity tend to be distinguished throughout most of the second half of the previous century: content validity, i.e., “the degree to which elements of an assessment instrument are relevant to and representative of the targeted construct for a particular assessment purpose” (Haynes et al. 1995: 238); criterion validity, which comprises the concurrent and predictive validity of a research instrument; according to the Standards for Educational & Psychological Tests, concurrent validity reflects only the status quo at a particular time, whereas predictive

191 validity can serve as a comparison between the measure and the outcome assessed at a later time (1985); construct validity, which Guion (1980: 390) recognised by judging whether a set of scores was free from contaminating sources of variance. At the end of the century, however, the validity of a psychometric instrument was more frequently viewed as an integral, unified concept, with six distinguishable aspects, i.e., its content, substantive processes, score structure, generalisability, external relationships, and testing consequences (Messick 1989, 1994, 1998). In order to verify the content validity of personality measurement with the Polish adaptation of NEO-FFI, its authors examined the very existence, invariance, and biological character of the five personality dimensions. After a detailed assessment which involved calculating content validity ratios, they arrived at the following conclusion: all the personality traits measured by the Big Five model comply with the existence and universality criteria; Agreeableness and Extraversion meet the invariability criterion, but only partly comply with the assumption of biological determination; Openness to Experience and Conscientiousness meet the assumption of heredity, but are only represented in the structure of biological traits of personality; Neuroticism scale fulfils all the above criteria, though higher validity can be obtained if measured by means of Eysenck’s EPQ personality inventory (Zawadzki et al. 2010: 42-56). As far as invariability of the traits measured by NEO-FFI is concerned, it is reported that the stability of all the ‘big’ five traits over age is greater when calculated in longitudinal studies than in self-reports of respondents’ perceptions of their own personality at different stages of their lives (Fleeson and Heckhausen 1997). On the other hand, it was inferred by Krahé et al. (2008) that exposing individuals to a situation in which a relatively high level of Extraversion might be expected, or to an extraverted person serv- ing as a stimulus, might actually affect their Extraversion reports. Obviously, further research is being conducted in order to achieve even greater accuracy, consistency and validity in measuring personality with NEO-FFI. For example, the number of factors have been considered in factor analysis (Wakabayashi 2014), and attempts have been made to produce a less time-consuming version (Körner et al. 2015), as indicated in Chapter 1 of the present thesis.

192 Specifically, for the purposes of evaluating the validity of the Polish adaptation of NEO-FFI (Zawadzki et al. 2010) in the present study, two procedures were applied. First, bivariate correlations between items were calculated with total scale scores, and the significance of the two-tailed test was examined. Items for which the significance level did not exceed the required significance level were recognised as valid. The required significance level was established by applying the Bonferroni correction for multiple comparisons and thus set at p = 0.0041. Second, the values of the correlations were confronted with those from the table of critical values for the Pearson correlation (p < 0.05; df = 2) and it was assumed that if the value of the correlation coefficient exceeded the critical value, the given position was valid. At both stages, the analysis of validity proved that each position of the adapted NEO-FFI was a valid element of the questionnaire. Also, for the non-parametric rank coefficient (Spearman’s r) was calculated in the place of Pearson correlation coefficient for data sets whose distributions varied significantly from normal. Detailed results for the significance test for the bivariate correlations of position and whole-scale values can be found in Appendix A. Critical r value for 749 (number of participants in the main study) was calculated and amounted to rc = 0. 105. As can be seen in Appendix A, all the listed r and r values exceed the critical value. Frequently referred to as the accuracy of measurement, the reliability of a psychometric test provides information on the degree to which the obtained test results reflect the level of the measured trait (Hornowska 2007: 28). Earlier, Guilford (1950: 474) described the reliability of psychometric tests as “the proportion of their variance that is true variance” and emphasised that reliability referred to both the population measured and the applied instrument itself. Another dimension of reliability was mentioned by Góralski (1976: 273), who referred to the reliability of a psychometric test as to the degree to which the result of applying the test to measure a sample remains constant. Ferguson and Takane (2002: 493) identified four methods of estimating the reliability of a construct: the test-retest method, consisting in using the same research tool twice (with a time interval in between42) in the same investigated sample of a population, and then calculating the correlation coefficient between the results obtained in both measurements;

42 The length of the time interval is not clearly defined in any commonly guidelines, since the impact of the carryover and reactivity effects remain debatable. For example, Murphy and Davidshofer (2005: 124) argue that both effects can either be treated as natural aspects of the attribute being measured, but they can also be recognised as sources of inflated or deflated test scores when the tests are retaken.

193 parallel-forms reliability, which involves using different variations of the same research tool, and then calculating the correlation coefficient between the obtained scores (this method also poses an additional demand to produce two parallel versions of a tool and form the criteria of their similarity); split-half method, which can be applied whenever the investigated test can be divided into half, and correlation coefficient can be calculated for scores obtained for the results obtained for both halves of the test (which usually takes the form of calculating the correlation coefficient between the results obtained for even and odd positions of a tool); methods related to establishing the internal consistency of a psychometric test, employing statistical techniques relying on the relationship between a test and its items, which are described below. One procedure which belongs to the final group of methods consists in calculating Cronbach’s alpha, the coefficient used in order to assess the reliability of all quantitative research tools, is one of the most popular measures of the internal consistency of a psychometric test. It is calculated according to the formula calculated by Cronbach (1970: 161 after Kuder and Richardson 1937). As far as the reliability of the test is concerned, the homogeneity of the Polish adaptation is acceptable, though not as high as that of the original version. Cronbach’s alpha values for personality scales range from 0.68 for Openness to Experience and Agreeableness to 0.82 for Conscientiousness (sample size: 2041 respondents) (Zawadzki et al. 2010). The results are only slightly lower than those reported by the authors of the original questionnaire, who provided values of internal consistency in the range from 0.72 to 0.88 in the Adolescent and Adult samples, with a median of 0.82 (McCrae and Costa 2007: 118). The specific values of Cronbach’s alpha calculated for the five personality scales for the investigated population of students are provided in Table 27.

Table 27. Reliability of personality scales.

Scale Cronbach’s alpha Openness to Experience 0.683 Conscientiousness 0.721 Extraversion 0.799 Agreeableness 0.750 Neuroticism 0.858

194 As can be seen, all the coefficient values are good or acceptable43, with only one value slightly lower than 0.7, i.e., the coefficient value for the trait of Openness to Experience. It needs to be remembered that Openness to Experience is the particular trait whose per- manence has been challenged or even questioned, since its level can be modified, for example through cognitive training (Jackson et al. 2012). Overall, the questionnaire is recommended for respondents of either gender aged over 15. According to the authors of the Polish adaptation, it only requires basic ability to read and write on the part of the respondent, and therefore is not demographically biased. Also, in view of the authors of the original questionnaire, the scales demonstrate a relatively low bias caused by social desirability. It needs to be emphasised that although the introduction to the questionnaire indicates that there are no good or bad answers, it is theoretically possible to predict (and, thus, fake) one’s scores, at least those regarding Extraversion, Conscientiousness, and Neuroticism (Furnham 1997). It is recommended that respondents are not time-limited, nor influenced by other respondents. In fact, the authors of the Polish adaptation recommend granting a certain level of intimacy to respondents.

SILL ver. 7.0

SILL ver. 7.0 is one of a number of questionnaires developed by Oxford in order to measure the frequency of language learning strategies use among foreign and/or second language learners (Oxford 1990). Despite criticisms, researchers have continued using adaptations of the questionnaire in order to identify and explore language learning strategies (Peacock and Ho 2003; Griffiths 2008; Liang 2009; Pineda 2010; Kayaoğlu 2013; Cas- tillo and Cordova 2014). Version 7.0 consists of 50 questions on a Likert-scale ranging from 1 (never or almost never) to 5 (almost always or always). Three of the scales listed above, i.e., memory, social, and affective strategies, appear in SILL ver. 7.0. New scales are combinations of the original scales and new questions were added as a consequence

43 Cortina (1993) suggests the following interpretation of Cronbach’s alpha value: α < 0.5 - unacceptable internal consistency; 0.5 ≤ α < 0.6 - poor internal consistency; 0.6 ≤ α < 0.7 - questionable internal consistency; 0.7 ≤ α < 0.8 - acceptable internal consistency; 0.8 ≤ α < 0.9 - good internal consistency; 0.9 ≤ α - excellent internal consistency.

195 of further factor analyses performed by Oxford and associates. The six scales which SILL ver. 7.0 is built of originally appear in Oxford and Crookall’s (1989) theoretical article (1989), which also defines each of the scales. They are presented in Table 28. For each scale the number of items in the Polish adaptation of the questionnaire is provided before and after modifications based on the results of Pilot Study One (see section 4.2.1).

Table 28. SILL ver. 7.0 scales. Description No. No. of of items Scale items (amended) (original) memory deal with storing and retrieving information 6 6 strategies involve manipulation / transformation of the language in some di- 14 14 rect way, e.g. through reasoning, analysis, note taking, and both cognitive functional practice in naturalistic settings and formal practice with strategies sounds and structures,

compensa- handle lack of / insufficient knowledge and thus include guessing 6 7 tion strate- while listening or reading, using synonyms or circumlocution gies while speaking or writing metacogni- aim to enable the learner to control the learning process through 9 9 tive strate- centring, arranging, planning, and evaluating their learning gies social strat- include interactions with other people which facilitate language 6 6 egies learning affective help learners manage their emotions, attitudes, and seriously influ- 6 9 strategies ence learners’ motivation

Since SILL ver. 7.0 (Oxford 1990) is, from the psychometric point of view, a test that accounts for the description of individual differences between language learners (Stu- art-Hamilton 2007: 214), the degree to which it fulfils the goodness criteria can also be analysed. Oxford (1986) considered SILL to be a psychometric test and set out to determine both the reliability and validity of the tool before developing the final version of the questionnaire. In terms of objectivity, the inventory is most likely to produce the same results for any two raters (“blind” diagnosis test), since it is also equipped with a simple marking scheme (“Appendix C, Working for Answering and Scoring”) (Oxford 1990: 297-300). Hence, scoring does not depend on the person who conducts the test. Also, the scoring procedure is described in a simple manner in order to minimise the risk of mis- calculating the score.

196 As regards standardisation, the test is accompanied by a set of instructions which appears in the initial part of the questionnaire. Specifically, for the purpose of both pilot studies, and the main study, instructions were provided in Polish so as to avoid ambiguity. Basing on the instructions, one can infer that the questionnaire can be used for both individual and group studies. Consequently, it can be used both as a self-assessment tool for learners who are willing to reflect on the way that they learn English as a foreign language, and for comparisons between learners, and across groups. Time limit is not specified; however, it is said that it takes no more than 30 minutes to complete the questionnaire. Obviously, it is not possible to discuss the results of one’s test by referring to norms, since norms as such do not exist in the case of SILL. However, the section “SILL Profile of Results” offers a key to understanding one’s scores, their graphic representation, and an explanation of what the averages mean. In terms of verifying the face validity of the instrument, Oxford (1989a: 292) investigated the classificatory agreement between two independent raters “blindly” matching the SILL items with strategies with a taxonomy to address face validity, whose value amounted to 0.95 when calculated by means of a coefficient in the quoted study, and even to 0.98 in earlier versions (Oxford 1986). As far as the content validity of the questionnaire is concerned, its connection with success in language learning or with measures of L2 proficiency and other similar instruments have been assessed by various researchers. A number of studies can be quoted to confirm the relationship. One example is the study conducted by Green and Oxford (1995), where 78% of variance in language proficiency was explained by the SILL. Although the validity of measuring language learning strategies by SILL has been challenged (Skehan 1991, Dörnyei 2005, 2009), it is an undeniable fact that the correlations of strategy use with language proficiency range from 0.30 to 0.73 and 21-56% of variance in proficiency is explained by strategy use reported on the SILL (Oxford 2013). Also, the confirmatory factor analysis conducted by Hsiao and Ox- ford (2002) proved that the 50-item SILL showed most consistent fit with learners’ strategy use. Similarly to the case of NEO-FFI, two procedures were employed to evaluate the validity of the SILL ver. 7.0: calculating bivariate correlations of the values of each position of the questionnaire with total scale scores in order to investigate the significance of the two-tailed test, and second, comparing the values of the correlations with critical values for the Pearson correlation (p < 0.05; df = 2). The analysis proved that all the items

197 of the SILL were valid. Detailed results of the significance test can be found in Specifi- cally, for the purposes of evaluating the validity of the Polish adaptation of NEO-FFI (Zawadzki et al. 2010) in the present study, two procedures were applied. Detailed results for the significance test for the bivariate correlations of position and whole-scale values can be found in Appendix B of the present thesis. Critical r and value for 749 (number of participants in the main study) are also calculated and provided in the final table in Ap- pendix B. As can be seen in Appendix B, all the listed r and r values exceed the critical values. With reference to the reliability of the test, Oxford measured the coefficient for the internal reliability of the whole inventory, which was each time very high (e.g. amounting to 0.95 for a 483-person field sample, 0.95 on average, or 0.96, specifically for a university sample of 1200 students (Oxford 1986, 1993; Oxford and Burry-Stock 1995). Cronbach’s alphas for particular strategy scales were first calculated in a study of 534 undergraduate students in Taiwan and amounted to 0.75 for the memory scale, 0.84 for the cognitive scale, 0.69 for the compensation scale, 0.86 for the metacognitive scale, 0.68 for the affective scale, and 0.78 for the social scale (Hsiao and Oxford 2002). Castillo and Córdova (2014) investigated a group of 1 283 Mexican students, of which 266 students were analysed separately in an exploration of successful language learners. Specific reliability coefficient for particular scales amounted to 0.8 for metacognitive strategies, 0.74 for social strategies, 0.78 for cognitive strategies, 0.76 for memory strategies, and 0.66 for affective strategies. It can be observed that the two scales which are characterised by relatively low reliability coefficients include compensation and affective language learning strategies. These two scales also proved to manifest low reliability values in Park’s study (2011), amounting to 0.43 (compensation strategies) and 0.45 (affective strategies) in contrast to other strategy scales, whose reliability coefficients range from 0.69 to 0.8. Table 29 contains the values of Cronbach’s alphas for the six strategy scales calculated with the data from the main study in two versions, i.e., for the original strategy scales, and for strategy scales expanded on the premises of Pilot Study One results. For scales which were not modified only the value of the original scale has been entered.

198 Table 29. Reliability of the adapted SILL ver. 7.0 scales. Cronbach’s alpha - Cronbach’s alpha Scale original scale - expanded scale Memory strategies 0.610 N/A Cognitive strategies 0.803 N/A Compensation strategies 0.625 0.631 Metacognitive strategies 0.830 N/A Social strategies 0.704 N/A Affective strategies 0.408 0.500

As can be seen in Table 33, Cronbach’s alphas improved after expanding the scales of compensation strategies and affective strategies. The values of the coefficient increased from 0.625 to 0.631, and 0.408 to 0.500 respectively. The reliability of the scale would further increase to 0.523 if question 2, I encourage myself to speak English even when I am afraid of making a mistake, was removed. One possible explanation behind such results consists in the fact that the statement is a complex one. Aimed to tap learners’ readiness to speak English in spite of being afraid of making a mistake, it may be ambiguous, as learners who do not use this strategy could be rare strategy users not because they avoid using affective strategies, but because their confidence as language users is high. Moreover, they might gain that confidence because of using other affective strategies, such as rewarding themselves, or having positive associations with (learning) Eng- lish. A conflict between question 2 and the rest of the scale, and a potential lack of internal consistency might therefore be one of the reasons for the low reliability of the affective scale.

Semi-structured interviews

It is generally accepted that the dependability of research can be enhanced by different kinds of data triangulation, as such approach caters for the demands of the positivistic view by employing quantitative methods, but at the same time can meet the criteria imposed by the naturalistic tradition through employing qualitative methods (Główka 2011: 298). Therefore, in order to ensure better dependability of the findings, a series of semi- structured interviews were conducted with a selected group of participants in the quantitative main study. The choice of the interview form was determined by the necessity to

199 allow for a wide range of topics including the emergence of unexpected themes (Richards 2009: 186). Also, the use of interviews was intended to make the study much more contextualised, which was attempted through engaging participants in conversations, and let- ting them interpret their own experience (Schultze and Avital 2011: 1). They were interviewed in Polish, by the author of the study, in July and August 2016. The interviews were all conducted in Polish and digitally audio-recorded and transcribed by the author of the present dissertation. The interview followed a semi-structured schedule and consisted of four major areas, that is introduction, applied language learning strategies, learner’s personality, conclusion. Details of each section can be found in Appendix C.

4.4.4. Analytical procedures

The present section contains a description of analytical procedures employed to handle quantitative analysis (findings from the applications of NEO FFI and the adaptation of SILL ver. 7.0) and qualitative analysis (thematic analysis of semi-structured interviews). Databases were created in Microsoft Excel and SPSS so as to secure data and allow for comparisons in cases of uncertainty. Personality data was recoded according to the instructions from the manual. No data was not standardised since SPSS creates standardised values as separate values ad hoc. Calculations were performed in SPSS versions 23 and 24. Commonly used in social sciences, and, particularly, in psychology and psychometrics, SPSS software has also gradually been gaining popularity among applied linguists and researchers in the domains of second and foreign language learning. Among its many advantages, SPSS does not require the user to enter commands, although, on the other hand, it has a built-in function that enables the user to create new commands should the need arise. The process of reading and transforming variables is relatively simple and intuitive. The software allows for nine kinds of variables, including numeric and string variables, as well as offering three different options of measurement, i.e., nominal measurement, ordinal measurement, and scale measurement. With regard to the qualitative part of the main study, recordings of semi-structured interviews were listened to so that the author of the present dissertation, who was also the transcriber, could become familiar with their content. Afterwards, preliminary tables of contents were created with key words which were expected to facilitate navigation in the

200 data files. Second, a sketch of each interview was performed with the use of the Microsoft Word processor. Finally, the sketches were used so as to add any missing details and achieve a maximum accuracy of the transcripts. The final transcripts were subject to thematic analysis described in section 4.4.4.8.

Testing data sets normality

In order to choose a proper set of analytic tools, a data set needs to be tested for normality. In other words, it needs to be verified whether the distribution of each data set further examined in a study is significantly different from normal distribution. For that reason, two tests can be used, Kolmogorov-Smirnow test or Shapiro-Wilk test. In both cases, H0 assumes that the data sets do not significantly differ from normal whereas H1 assumes that data sets are significantly different from normal (at an assumed significance level; p = 0.05). The result of the test (p value greater or lesser than 0.05 or 0.01, depending on the assumption) determines choosing a set of statistical procedures. In other words, if data sets do not significantly differ from normal, they can be analysed with parametric tests, while if their distributions differ from normal distributions significantly, non-parametric tests must be employed. This consequence is of vital importance as parametric tests are believed to be more powerful since they derive from standardised scores and enable researchers to compare sub-populations with a whole population (Cohen et al. 2005: 318). Specifically, for the data sets of personality traits and data characteristics, the Shapiro-Wilk test was used for the six strategy scales, the mean strategy use (GULLS), and the ‘big’ five personality scales. For each data set, the null hypothesis assumed that data distribution does not significantly vary from the normal distribution, whereas the alternative distribution is that data distribution is significantly different from normal (p < 0.05). Results are summarised in Table 30. The symbols used in the table refer to the following: - df indicate the number of degrees of freedom; - p value indicates whether the distribution is significantly different from normal (p < 0.05) or it does not significantly differ from normal (p > 0.05).

201 Table 30. LLS and personality scales tested for normality Shapiro-Wilk test results Distribution significance normality (p value) Scale Statistic df Memory strate- 0.993 540 0.014 gies different from normal Cognitive strat- 0.995 540 0.064 egies normal Compensation 0.993 540 0.010 strategies different from normal Metacognitive 0.995 540 0.053 strategies normal Affective strat- 0.987 540 0.000 egies different from normal Sociocultural 0.992 540 0.004 strategies different from normal GULLS – sum 0.998 540 0.859 normal GULLS – mean 0.998 540 0.854 normal Openness to ex- 0.992 540 0.007 perience different from normal Conscientious- 0.995 540 0.112 ness normal Extraversion 0.995 540 0.072 normal Agreeableness 0.987 540 0.000 different from normal Neuroticism 0.995 540 0.092 normal

As can be seen, a number of data sets for both strategy and personality scales show both types of distribution. Hence, different tests need to be employed in order to identify whether two data sets are significantly different from each other (t-tests for parametric data and Mann-Whitney U test for non-parametric data), in analysis of variance (ANOVA for parametric data and Kruskal-Wallis test for non-parametric data), and correlations (Pearson r coefficient for parametric data and Spearman’s rank coefficient, ρ, for non- parametric data). These procedures are described in the following subsections.

Descriptive statistics

In order to provide insights into the characteristics of the investigated variables, it is necessary to describe them in terms of basic statistical parameters, such as frequencies, per- centile values, or percentages, measures of dispersion (including standard deviation, variance, range), measures of central tendency (mean, median, mode), and distribution properties, such as skewness or kurtosis. For personality scales, calculating the results for

202 individuals consisted in adding the relevant scores for each scale, according to the key attached to the manual (Zawadzki et al. 2010), as well as entering recoded values for reverted items. In other words, items which tested individuals’ level of Emotional Stabil- ity had to be recoded so as to account for their level of Neuroticism, items that accounted for the level of Introversion had to be recoded in order to describe the level of Extraver- sion, items which indicated respondents’ Antagonism needed to be recoded so as to reflect their level of Agreeableness, items which reflected respondents’ Closed-Mindedness needed to be recoded in order to reflect their level of Openness to Experience, and, finally, items testing participants’ Impulsivity were recoded so as to account for their level of Conscientiousness. Mean values were calculated for all the participants of the study with regard to both the frequency of strategy use and the level of personality traits. For strategy scales, they were supplemented with the following measures: values of the standard deviation, which represented the square root of the arithmetic mean of the squared deviations of the measurements from their mean; values of the median, i.e., the point on the scale such that the frequency above or below the point is 50 per cent of the total frequency; values of the mode, that is the most frequently occurring value of the investigated variable. For personality scales, the descriptive statistics were calculated as follows: jointly, in a manner reflecting the calculation of descriptive statistics for strategy scales; across age groups and genders, so as to indicate whether the calculated trait levels were low, medium, or high since norms for any population differ across different age groups as well as across genders (Costa and McCrae 1992). Sten values for men and women aged 18-19 and 20-29 are presented in Appendix D of the present thesis. These comparisons are essential for making further trait comparisons across genders as the same raw score could not only be interpreted differently across genders, but also for participants varying with respect to age.

203 Identifying the link between strategy use and personality

In order to answer Research Question 1 and investigate the relationship between the level of language learners’ ‘big’ five personality traits and their use of language learning strategies, students were divided into three categories according to the level of their personality traits according to the norms presented in Appendix D of the present dissertation. Fol- lowing the interpretation suggested by Zawadzki et al. (2010), scores in stens 1-3 were interpreted as relatively low, scores in stens 4-6 were interpreted as medium, while scores belonging to stens 7-10 were marked as high. Two parallel procedures were employed in order to investigate whether the three groups of students (characterised by low, medium, and high level of a given trait) differed with regard to strategy use. For data sets whose distributions were not significantly different from normal, including cognitive strategies, metacognitive strategies, and GULLS, one-way analysis of variance (ANOVA) was applied. This type of analysis was chosen to test whether two or more means differed significantly, on condition that the investigated data distributions do not significantly vary from normal distributions (Cohen et al. 2005). Larson-Hall (2010: 389) defined ANOVA as a model including at least one categorical variable and enabling the researcher to observe whether groups defined by the independent variable (or variables) perform differently on the dependent measure. Brown (1988: 176-177) specifically described one-way ANOVA as the analysis applied to compare means of one interval-scale variable and one nominal-scale independent variable, asserting that the assumption of homogeneity of variances for the compared data sets should be met in order to conduct further analysis of variance. In order to examine that, Levene’s test was applied. The null hypothesis H0 assumed that the variances of the investigated data sets of GULLS, cognitive, and metacognitive strategies, were significantly different from each other across three data sets for each personality trait, while the alternative hypothesis, H1, assumed that they were not. Detailed information on the results of Levene’s test is provided in Table 31. Homogeneous variances were marked as H while non-homogenous variances were marked as non-H. On the basis of the test, Tukey’s post-hoc test was run for all cases where homogeneous variance was observed in order to investigate the differences in the frequency of strategy use across the participants’ groups varying with regard to the level of personality traits, and the significance level was adjusted by means of the Bonferroni correction to a = 0.01. As variances of GULLS were not homogenous

204 across learners whose levels of Conscientiousness were low, medium, and high, the impact of the trait on strategy use was investigated by robust tests for equality of variances.

Table 31. Results of Levene’s test for normally-distributed strategy scales Neuroti- Extraver- Openness Conscientious- strategy/factor cism sion to Experience Agreeableness ness GULLS H H H H n-H cognitive strategies H H H H H metacognitive strategies H H H H H

Should the data sets vary significantly from normal, the Kruskal-Wallis test needs to be applied in order to account for the differences between multiple sets of data. Brown (1988: 174-175) described the Kruskal-Wallis test as an extension of the Mann-Whitney U test that could be applied to compare more than two means if all groups are random samples from their respective populations and the assumption of their independence among samples is met. The test was applied for the scales of memory, compensation, affective, and social strategies, whose distributions were all significantly different from normal. The significance value for the test was each time adjusted by the Bonferroni correction for multiple tests which resulted in increasing the level of significance to a = 0.01.

Investigating the link between strategy use, personality and attainment

In order to answer Research Question 2, the Kruskal-Wallis was applied so as to identify any differences in the level of competence in terms of medians and distribution, in the groups of rare, frequent, and average strategy users. The test was selected due to the categorical character of both the dependent and the independent variable. The significance value for the test was each time adjusted by the Bonferroni correction for multiple tests which resulted in increasing the level of significance to a = 0.007. The groups of strategy users were created on the basis of their SILL scores according to Oxford’s (1990: 300) interpretation of results, with respect to both general use of strategies (GULLS) and each of the six investigated strategy categories. The frequencies of strategy use were respectively marked as - 1, for rare strategy users (scores lower than 2.5)

205 - 2, for participants whose strategy use was of medium frequency (scores between 2.5 and 3.4) - 3, for frequent strategy users (scores higher than 3.5) By analogy, in order to answer Research Question three, the Kruskal-Wallis was applied so as to identify any differences in the level of achievement in terms of medians and distribution, in the groups of participants characterised by low, medium, and high levels of personality traits according to the psychological interpretation provided by Zawadzki et al. (2010). The significance value for the test was each time adjusted by the Bonferroni correction for multiple tests which resulted in increasing the level of significance to a = 0.01. The values of trait levels were assigned to participants according to the interpretation of their sten scores (see Appendix D).

Correlations

According to the distinction made in Table 39 (see section 4.4.4.1) between parametric and non-parametric data, two basic measures of correlation between two single sets of data were applied: Pearson’s correlation coefficient for two sets of parametric data, marked as r44 was used to calculate correlations between the levels of three personality traits, Neuroti- cism, Extraversion, and Conscientiousness, and the frequency of using LLS in general (GULLS), and the frequencies of using strategies belonging to the cognitive dimension in order to answer research question one; Spearman’s rank coefficient, marked as ρ, was calculated for data sets whose distributions can differ significantly from normal45, including any correlations with memory, compensation, affective, and social on the one hand, and the levels of Open- ness to Experience, and Agreeableness, on the other hand, in order to answer research question one;

44 The coefficient is calculated according to the following formula (Guiford 1950: 157): rxy = (∑▒��)/�� , where rxy represents the correlation between X and Y, x stands for the deviation of any X score from the mean in test X, y stands for the deviation of the corresponding Y score from the mean in test Y, and σxv and σy represent the standard deviations of the distributions of X and Y scores. 45 The coefficient is calculated according to the following formula (McNemar 1962: 203): ρ = 1 - (6ƩD^2)/(N(N^2-1)) , where ρ represents Spearman rank-difference correlation coefficient, D - the difference between an individual’s two ranks (for two traits), and N - the number of observations

206 Spearman’s rank coefficient was applied due to the categorical character of the variables investigated in Research Questions 2 and 3. The following distinction, frequently applied in psychometric studies, was adapted from Bedyńska and Brzezicka (2007: 96) in order to interpret the obtained results: - coefficients below 0.3 indicate weak correlations, not sufficient to confirm a linear relationship; - correlations between 0.3 and 0.5 indicate moderate correlations; - correlations between 0.5 and 0.7 indicate strong correlations; - correlations between 0.7 and 1 point to very strong correlations.

Regression model

Larson-Hall (2010: 400) defined a regression model as an equation modelling a relationship between two or more variables by means of a regression equation. Using the SPSS software, one can assess how well a model matches the actual performance of variables, and assess the extent to which the change of a given variable, X, explains the value of the other variable, Y. In the main study, a regression model was used in order to determine the degree to which personality variables could explain the use of learning strategies. Ferguson and Takane (2002: 149-150) pointed out that researchers should carefully select an adequate model of regression, but at the same time claimed that the assumption of linearity was met by a major number of variables used in pedagogical and psychological research. By selecting the stepwise method of regression, one can ensure that only significant predictors remain in the final regression model. In other words, as Larson-Hall (2010: 186) puts it, stepwise regression accounts for all of the area where the explanatory variables overlap with the explained (or response) variable, leaving the model determination solely to statistical criteria. The model thus allows the researcher to elicit the predictor that contributes most to predicting the outcome variable and add it to the regression model if the impact is statistically significant (p < 0.05). The procedure is subsequently repeated for all predictors in the model, and finally, insignificant predictors are removed from the model. In terms of the output, the SPSS software supplies the researcher performing

207 stepwise regression with descriptive statistics, a correlation matrix, a table of entered and removed variables, a summary of what percentage of the variance of the explained variable is accounted for by the explanatory variables, and finally, information on how much of the variance of the response variable is explained by all the explanatory variables entered into the regression equation. The total impact of all the explanatory variables on the explained variable is represented by R2, while the relative impact of each predictor is represented by the value of the standardised b coefficient. The distribution of residual parameters in the regression model should not significantly differ from normal distribution so as to exclude their influence on the explained variable. When it comes to the data sets analysed in the present dissertation, the stepwise regression procedure was applied for the purpose of investigating the relationships between students’ use of language learning strategies and their personality traits. The initial set of explanatory variables involved all of the ‘big’ five personality traits, while GULLS was treated as dependent variable in the regression model. The values of R2 were calculated to assess the overall extent to which the variance of strategy use can be explained by learners’ personality traits, and b coefficients were calculated so as to assess the relative impact of each of the ‘big’ five personality traits. Finally, residual values were examined so as to exclude the existence of any regular pattern that could affect learners’ choice of language learning strategies.

Clustering techniques

The general aim of any cluster analysis is to classify data and conduct discriminative analysis. Landau and Everitt (2004: 311-312) differentiate between two types of clustering techniques, that is: agglomerative hierarchical techniques, which consist in combining individuals who are closest and afterwards combining them with other individuals, who are closest to them, applying either complete or average linkage for defining inter-group distance; k-means clustering, which divides a set of data into k clusters (k being a number established by the investigator), using the measure of Euclidean distance, with the overall aim to minimise the variability within clusters while maximising it between clusters.

208 Both techniques result in the emergence of clusters which show similar characteristics and were used for the same reason, that is, to answer the final research question. The SPSS software enables an assessment of the model of clustering created as a result of using either of the above clustering techniques by showing its adequacy on a graph. As far as using clustering in the analysis of learner variables is concerned, results of clustering can be referred to theoretical assumptions, but also, conclusions can be drawn with reference to individuals belonging to the obtained clusters. Specifically, the personality traits of frequent strategy users or those favouring a given strategy scale can be juxtaposed with the personality traits of rare strategy users or learners abstaining from the use of a particular strategy scale. The aim of the clustering procedure consisted in grouping language learners according to their personality characteristics and strategy use. For that purpose, Ward’s method was employed to minimise the total within-cluster variance and Euclidean distance was used as a measure of distance between cases which incorporated different types of data. For each participant the data involved the level of personality traits being significant factors in strategy variance, and the general frequency of strategy use (GULLS). In the following, two-step clustering procedure, which combines the features of hierarchical and non-hierarchical techniques, pre-clusters were initially formed from participants’ characteristics. Afterwards, the pre-clusters were used as cases in the sub-stage of hierarchical analysis. For the purposes of two-step grouping, students were divided into three groups, rare strategy users, average strategy users, and frequent strategy users. A categorical variable had been created prior to conducting cluster analysis - learners were divided into three groups: - rare strategy users, i.e., 1/3 of the entire population of the investigated students who used LLS least frequently, - frequent strategy users, i.e., 1/3 of the entire population of the investigated students who used LLS most frequently, - average strategy users, i.e., students who did not belong to either of the above groups and whose use of language learning strategies was neither frequent nor rare. The final stage of quantitative analysis involved conducting non-hierarchical clustering with the use of k-means procedure. It involved determining a fixed number of clusters, which was set as three, basing on the previous findings from hierarchical analysis.

209 Respondents were clustered according to the level of their personality traits as well as the frequencies of using each of the six strategy scales. The maximum number of iterations was set as 10, and cluster membership was saved as a separate variable in order to expand the range of characteristics for each respondent. Due to discrepancies in the range of values between personality scales and strategy scales (the former ranging from 0 to 40 as sums, the latter ranging from 0 to 5 as mean values), data was standardised and new, standardised variables were created. Again, learners were analysed in terms of the differences in their levels of personality traits and the frequency of using language learning strategies (both in general and with respect to particular scales) and the differences between clusters were tested for significance. On the basis of the findings of each part of the cluster analysis, learner profiles were created which were subsequently used in the discussion of the impact of learners’ personality traits on their strategic choices and learning behaviour.

Analysis of interviews

Both the interview schedule and the analytical procedure followed the guidelines for qualitative studies employing thematic analysis put forward by Braun and Clarke (2006), according to which, the following stages should be part of analytical procedures: becoming familiar with the data, which includes preliminary marking ideas for coding and transcription of verbal data; generating initial codes; searching for themes; reviewing themes; defining and naming themes; producing the report. The schedule focused on providing insights into the language learning process from the learner’s perspective. It involved questions corresponding to three strategy domains (i.e., cognitive, affective, and sociocultural-interactive domain; Oxford 2011), and provided space for reflecting on the metadimensions (planning and organising for cognition, affect, and social interaction). As the interviews were semi-structured, it was

210 expected that additional themes, not accounted for by the research questions, might appear in thematic analysis. The interview schedule is

4.5. Conclusion

Chapter Three of the present dissertation was intended to account for the rationale, design, and methods of investigation applied in the empirical study of the relationships between the level of the investigated learners’ personality traits and frequency and types of the language learning strategies which they employed in the process of developing their competence in English as a foreign language. The original idea behind the research project consisted in developing a methodology that would cater for the advantages of both the idiographic approach and the nomothetic approach towards investigating personality and individual differences in language learning strategy use. Consequently, the main study was planned as a mixed-method study. The tools of study were also piloted twice in order to optimise the battery of instruments applied in it. Moreover, particular attention was paid to selecting a representative sample of students so that results of the study could be generalised across the populations of at least the two universities where the study was conducted. The instruments employed in the quantitative part of the main study have a long- established tradition in personality investigations and studies exploring language learning strategies. With regard to the adaptation of the SILL ver.7.0 (Oxford 1990), it only requires respondents to have literacy and basic reading comprehension skills, and hence, it helps to minimise the risk of unreliable answers or unmotivated respondents described by Dörnyei (2003a). The second instrument, the NEO-FFI (Costa and McCrae 1992), formally requires participants to be 15 years old, and the study to be psychologically supervised, but does not require participants to possess any more advanced skills than the former instrument. A vital part of the chapter was dedicated to the investigation of how both instruments complied with the goodness criteria for psychometric tests (Hornowska 2007). Finally, the semi-structured interview, which was designed in order to triangulate the data gathered in the quantitative part of the study, followed a schedule developed according to a methodology suggested by experienced qualitative researchers (Braun and Clarke 2006), and steps were taken in order to scrutinise analytical procedures. It was

211 expected that given the advantage of the intimate familiarity with respondents (Delamont et al. 2010), no effects such as social desirability bias or self-deception should occur. Two checklists were used as points of reference to assess the rationale for the study, one of them specifically composed for studies into foreign language development by Dörnyei (2003), and the other one provided for psychometric research by Hornowska (2007). In spite of the conscientious planning and administration of the study, it was impossible to resolve all the problems encountered in the development of the main study. Specifically concerns might arise about the insufficient reliability of one of the SILL scales of affective learning strategies. In spite of the steps taken to improve its internal validity, it remained relatively low, which could be clearly related to Oxford’s (2017) recent comments on conducting large-scale quantitative studies investigating learners’ general use of strategies, and her encouragement to dedicate more studies to the issue of affect in foreign language development. However, in general, it was hoped by the author of the present dissertation that the developed rationale that could provide valid answers to the four research questions posed in the first section of the present chapter.

212 Chapter 5: Research project

5.1. Introduction

The present chapter aims to provide answers to the four research questions posed in Chap- ter Four. With that purpose, the results of quantitative analysis involving the procedures described in Chapter Four are presented and supplied with the insights from the thematic analysis of interviews in section 5.2. The section provides a general characteristics of participants in terms of their personality traits and strategic preferences and offers a description of correspondence between the two constructs. Moreover, it attempts to establish the nature of the relationships between learners’ repertoire of LLS, their personality traits and their level of attainment. It also aims to construct profiles of language learners at university on the basis of their personal characteristics and reported use of strategies. Afterwards, references to the results are made in the discussion section 5.3, which attempts to interpret, harmonise and systemise the findings with reference to each research question. The section also aims to elaborate on the weaknesses and limitations of the main study along with outlining the possible improvements for future research into the relationships between LLS and personality.

5.2. Study results

The present section is divided into two major parts. The first one attempts to provide a thorough description of the investigated population in terms of the levels of participants’ personality traits and their use of LLS operationalised as reported frequency. Its second part addresses RQ1 and contains the research findings concerning the relationship between learners’ level of personality traits and their use of LLS. They are based on three types of calculations described in Chapter Four, including tests for differences across data sets isolated on the basis of the levels of the investigated personality traits, correlations between LLS scales and personality scales, and between individual strategies and levels of personality traits, and finally, the model of stepwise regression. The following subsections investigate the relationships between LLS and learners’ level of proficiency in

213 English, and personality traits and learners’ level of proficiency in English respectively, relying on the results of one-way ANOVA tests and its non-parametric equivalent, the Kruskal-Wallis test. Afterwards, learner profiles are discussed on the basis of two independent clustering techniques: 2-step clustering and k-means analysis. Finally, the analysis of interviews provided support for interpreting the relationships between the analysed learner variables.

5.2.1. The relationship between LLS use and personality

Aiming to provide the answer to the first research question, the present subsection provides the results of three analytical procedures which were applied in order to measure the relationship between learners’ level of personality traits and their use of LLS. Prior to that, an extensive characterisation of the investigated population is provided with reference to the levels of learners’ personality traits and their use of LLS with reference to overall frequency, the popularity of specific strategy categories, and insights related to certain specific strategies. It is followed by a section intended to account for the correspondence between the level of the ‘big’ five personality traits and learners’ use of LLS measured as reported frequencies which relates the psychological interpretation of the level of each of the five traits to the reported frequencies of strategy use. The subsequent sections aim to provide supplementary information based on the analysis of correlation and regression.

Students’ personality traits and strategy use

Any attempt to account for the links between learners’ personality traits and their use of language learning strategies requires a preliminary characterisation of the involved variables in terms their central tendency and dispersion. Table 32 contains the descriptive statistics for six strategy scales, as well as for the mean frequency of strategy use strategy use, calculated for the data sets from the main study. Specifically, for each category, the values of the mean, the median, and the mode are provided, followed by the value of standard deviation, and mean values for each of the three quartiles. The final column

214 presents GULLS values, that is participants’ general use of language learning strategies, which can serve as the overall measure of the frequency of using LLS. According to the SILL Profile of Results (Oxford 1990: 300), 13.6% of all participants could be classified as rare strategy users, whose SILL averages did not exceed the value 2.5, a vast majority of 75.6% of all participants were characterised by medium strategy use, with average scores ranging from 2.5 to 3.4, and 10.8% of all participants were frequent strategy users, whose SILL scores exceeded the value 3.5. Also, it was observed that the investigated students showed a preference for two categories of LLS, compensation and social strategies, whose mean frequencies both exceeded 3.00. Both types of strategies from the cognitive dimension, cognitive and metacognitive strategies, were applied moderately frequently (mean values approximating 3.00). The investigated students were not particularly fond of two specific categories of LLS, memory, and affective strategies, whose mean frequency values only slightly exceeded the value 2.5, being the lower band of the interval of mean (average) strategy use according to Oxford’s (1990: 300) SILL Profile of Results. According to the instructional guidelines of the Profile, the aggregate values of frequencies regarding both the general use of language learning strategies (GULLS) and particular strategy categories (memory, cognitive, compensation, metacognitive, affective, and social) fell within the medium range of strategy use. For three scales the proportions of frequent strategy users were slightly lower than the proportions of rare strategy users with the differences amounting to 2.8% with regard to GULLS, 0.5% with regard to cognitive strategies, and 1.4% with regard to metacognitive strategies. The contrast was much starker with reference to memory strategies, which were frequently used by only 5.2% of the investigated population, but rarely used by more than 39%, and affective strategies, which were rarely used by more than 40% of the investigated population, but frequently used by only about 4%. Conversely, the groups of frequent strategy users were considerably more numerous than the groups of rare strategy users with regard to the categories of compensation strategies and social strategies, with differences amounting to 30.3% and 8.7% of the investigated population respectively. Two values of the standard deviation (SD) were relatively high, the SD of the social scale (nearly 0.75), and the SD of the metacognitive scale (more than 0.73), which indicated that the investigated students differed most in the use of these two categories of strategies.

215 Table 32. Descriptive statistics for strategy scales. Strategy Memory Cognitive Compen- Metacogni- Affective Social GULLS scales sation tive Valid 749 749 749 749 749 746 749 cases Missing 0 0 0 0 0 3 0 cases Mean 2.5528 2.9268 3.2453 2.9092 2.5438 3.0299 2.8444 Median 2.5556 2.9286 3.1667 2.8889 2.5556 3.0000 2.8519 Mode 2.89 multiple 3.17 3.00 2.56 3.00 multiple values values Std. devia- 0.58225 0.61427 0.64060 0.73057 0.52395 0.7446 0.47967 tion Percentiles 25 2.1111 2.5000 2.8333 2.4444 2.1111 2.5000 50 2.5556 2.9286 3.1667 2.8889 2.5556 3.8519 75 2.8889 3.2857 3.6667 3.4444 2.8889 3.1667 % of rare 39.7 19.6 9.1 23.8 41.3 21.1 13.6 strategy us- ers46 % of fre- 5.2 19.1 39.4 22.4 4.1 29.9 10.8 quent strategy users47

Within the two favoured categories of strategies, three specific strategies were used particularly frequently and the relevant average values could be described as high according to Oxford’s (1990: 300) SILL Profile of Results:

When I do not understand, I ask the speaker to slow down, repeat, or clarify what was said (social strategy; mean value 3.97); When I cannot think of the correct expression to say or write, I find a different way to express the idea; for example I use a synonym or describe the idea (compensation strategy; mean value: 3.96); When I cannot understand all the words I read or hear, I guess the general meaning by using any clue I can find, for example, clues from the context or situation (compensation strategy; mean value: 3.56). Other strategies whose frequencies of use were high included one cognitive strategy and one affective strategy: I say or write new English words several times (cognitive strategy; mean value: 3.56);

46 Oxford’s (1990: 300) key to understanding the averages assumes mean frequencies below 2.4 to be low. 47 Oxford’s (1990: 300) key to understanding the averages assumes mean frequencies above 3.5 to be high.

216 I try to have a positive image of my English teacher (affective strategy; mean value 4.03)48. No metacognitive strategies exceeded the lower band of the high strategy use interval. However, the mean values of three specific strategy frequencies were relatively high:

I notice my English mistakes and use that information to help me do better (mean value: 3.30); I pay attention when someone is speaking English (mean value: 3.33); I think about my progress in English (mean value: 3.27). As far as memory strategy are concerned, no specific value exceeded the lower band of high frequency of strategy use according to Oxford’s (1990: 300) SILL Profile of Results. The most frequently applied strategy was Remembering English words or phrases by remembering their location on the page, on the board, or on a street sign (mean frequency: 3.12) and it was the only strategy whose frequency of use exceeded the value 3.0. The strategies whose frequencies of use could be described as low according to Oxford (1990: 300) included three belonging to the affective domain, two affiliated with memory use, and three strategies which each belonged to a separate category. They are listed below with a mark in brackets indicating the category (‘A’ for affective, M for memory, Mcg for metacognitive, ‘C’ for cognitive, and ‘Cps’ for compensation): I keep a private diary or journal where I write my feelings about language learning (A; mean value: 1.19) I talk to someone I trust about my attitudes and feelings concerning the language learning process (A; mean value: 1.93); I share my feelings about learning and/or using English by writing a blog or using social media (A; mean value: 1.45); I use rhyming to remember the sound of the new word in my mind (M; mean value: 1.68); I initiate conversations in the new language (C; mean value: 2.65); I draw the meaning of an English word that I cannot remember (Cps; mean value: 1.45); I arrange my schedule to study and practice the new language consistently, not just when there is the pressure of a test (Mcg; mean value: 2.40).

48 The question was introduced after consulting a panel of experts, see Chapter Four, section 4.1.2)

217 The strategies enlisted above, which were the least popular among the investigated learners, mainly belonged to the category of indirect strategies, and hence, it could be inferred that students learning English as a foreign language might experience problems coordinating the process of language learning. However, no significant differences were observed between the use of direct and indirect LLS with regard to the overall frequency of their application. The mean values for the investigated population, calculated as aggregated means of memory, cognitive, and compensation strategies for the former category and metacognitive, affective, and social strategies for the latter, amounted to 2.83 and 2.80 respectively. The values of the SD were also similar, reaching 0.49 in the first case, and 0.54 in the second one. The results indicated a good balance between learners’ mental processing of the target language and management of the learning process. Overall, it could be concluded that the investigated learners exhibited a preference for language learning strategies which allowed them to overcome L2 deficiencies and handle potential problems that they could encounter when communicating in the target language. As indicated in the previous Chapter (section 4.4.4.2), in order to be interpretable, descriptive statistics for personality scales need to be calculated across genders and within specific age ranges. Therefore, as already explained, descriptive statistics for raw personality scores were calculated so as to reflect the procedure for strategy scales, but a valid interpretation of personality scores required data splitting across genders and the pre- scribed age intervals (Zawadzki et al. 2010). Table 33 contains descriptive statistics for raw personality values of the ‘big’ five personality traits while further tables contain data intended for psychological interpretation.

Table 33. Descriptive statistics for personality scales - raw scores. personality scale Neuroti- Extraver- Openness to Agreeableness Conscientious- cism sion Experience ness Valid 727 726 721 724 721 cases Missing 22 23 28 25 28 cases Mean 23.66 28.19 28.22 28.43 29.37 Median 24 28 28 29 30 Mode 21 30 24 30 30 Std. deviation 8.77 7.09 6.52 6.35 7.68

Tables 34-37 present descriptive statistics for the levels of the ‘big’ five personality traits for women aged 18 and 19, men aged 18 and 18, women aged 20-29, and men

218 aged 20-29. The final column in each table is the standard ten (sten) score that an individual would achieve if their trait intensity was equal to the mean value for the specific group (e.g. interpreted for a man, aged 20-29). It was expected that the mean values should correspond with the hypothetical values of stens 5 or 6, i.e., the middle of the sten scale, as it is typically true for data sets whose distributions do not considerably differ from normal distributions, as asserted by Hornowska (2007). Interpretation of the results is provided after the final table. Regarding the group of women aged 18 and 19, all mean values for Openness to Experience and Extraversion would be interpreted as sten 5 values, and the mean values for Conscientiousness, Agreeableness, and Neuroticism would be interpreted sten 6 values. Relevant data is presented in Table 34.

Table 34. Descriptive statistics for personality scales - women aged 18 and 19.

Personality trait Mean Std. Deviation Sten Openness to Experience 29.3735 6.62904 5 Conscientiousness 28.8795 7.26401 6 Extraversion 28.4699 6.93583 5 Agreeableness 29.4643 6.06480 6 Neuroticism 25.7381 8.54629 6

As far as the male participants of the main study aged 18 and 19 are concerned, the mean scores of personality trait levels also correspond with the ranges of central stens, 5 and 6. Interpreted for a single individual, the mean values for Openness to Experience, Extraversion, and Agreeableness would be diagnosed as sten 5 values, whereas the mean values for Conscientiousness and Neuroticism, would be diagnosed as sten 6 values. De- tailed information is provided in Table 35.

Table 35. Descriptive statistics for personality scales - men aged 18 and 19.

Personality trait Mean Std. Deviation Sten Openness to Experience 27.0000 7.52190 5 Conscientiousness 27.8718 9.06462 6 Extraversion 26.3590 6.47866 5 Agreeableness 25.7692 6.66673 5 Neuroticism 24.2564 9.31811 6

219 With reference to the female participants aged 20-29, the mean scores of personality trait levels read as individual values would be interpreted as stens 5 and 6, too. More specifically, the mean values for Openness to Experience, Conscientiousness, and Extro- version would be diagnosed as sten 5 values, while the mean values for Agreeableness and Neuroticism would be diagnosed as sten 6 values. Table 36 contains relevant information.

Table 36. Descriptive statistics for personality scales - women aged 20-29.

Personality trait Mean Std. Deviation Sten Openness to Experience 28.4123 6.37886 6 Conscientiousness 30.5546 7.60806 6 Extraversion 28.6143 6.87577 6 Agreeableness 29.0864 6.34975 5 Neuroticism 24.6253 8.51600 5

Finally, the values of trait intensity for men aged 20-29 could also be referred to mid sten scores of an individual. The levels of the particular ‘big’ traits would correspond to sten 6 for Openness to Experience, Conscientiousness, and Extraversion, and sten 5 for Agreeableness and Neuroticism. Table 37 contains detailed data.

Table 37. Descriptive statistics for personality scales - men aged 20-29.

Personality trait Mean Std. Deviation Sten Openness to Experience 27.6726 6.50540 6 Conscientiousness 28.2178 7.37041 5 Extraversion 27.8393 7.53117 5 Agreeableness 27.6667 6.26071 5 Neuroticism 21.1116 8.54275 6

Overall, when confronted with norms for Polish men and women in appropriate age groups, all the mean values of the levels of the ‘big’ five personality traits would represent sten 5 or sten 6 values. This means that no mean value for any personality trait lies in the extreme, upper (7-10) or lower (1-3) range of the sten scale, and thus the trait intensity can be rated as reflecting the levels for the majority of the Polish population. In brief, the mean personality scores for all the participants could be interpreted according to the brief interpretation introduced by Zawadzki et al. (2010):

220 regarding Neuroticism, the majority of the investigated students were generally calm and well-balanced, but sometimes experienced feelings such as anger, grief, or guilt; in terms of the level of Extraversion, participants’ level of social activity was fairly average, with signs of preference for both establishing contacts with others on the one hand and keeping a certain dose of privacy on the other hand; the average level of Agreeableness indicated that the investigated individuals generally displayed cooperative orientation, but at the same time, did sometimes enjoy ri- valry; the mid sten value of the level of Conscientiousness could be interpreted as indicating having clear targets in life on the one hand, but no orientation towards reaching them at all price on the other hand; the average level of Openness to Experience could be understood as a marker of being fairly practical and down to earth, but not deprived of unconventional interests or pursuits, and consequently, maintaining a balance between tradition and novelty. To recapitulate, the investigated population of students could be characterised as average in terms of strategy use, with respect to both the general frequency of strategy use and the frequency of applying particular categories of strategies. No clear indicators of overt psychopathologies were detected in the analysis of the investigated individuals’ personality traits. However, numerous individuals’ scores were interpreted as high or low according to the norms provided by Zawadzki et al. (2010). Since the analysed sample also included both rare and frequent strategy users within the meaning of Oxford’s (1990: 300) SILL Profile of Results, it was expected that the differences in the levels of their personality traits and the differences in the levels of proficiency would at least partly correspond to the frequencies of strategy use. Specific results of the analysis of that correspondence are provided in the following sections of the present dissertation in an attempt to answer all of its four research questions.

Relationship between LLS use and personality

In order to provide answers to the first research question, three types of analysis were conducted as mentioned in the previous Chapter (section 4.4.4.3). The present section describes the results of these analyses. First, the existence of differences in the frequency

221 of strategy use are accounted for across groups of students whose levels of the ‘big’ five personality traits are psychologically interpreted as low, marked as 1 (stens 1-3), medium, marked as 2 (stens 4-6), and high, marked as 3 (stens 7-10). Second, correlations between the frequencies of using strategy categories and students’ personality trait levels are analysed. Third, correlations are presented regarding the relationship between participants’ level of personality traits and specific LLS. Finally, the results of stepwise regression are discussed in order to explain the investigated learners’ employment of LLS through the prism of their personality characteristics. As far as the frequency of using memory strategies is concerned, the results of the Kruskal-Wallis test confirmed its statistically significant relationship with respondents’ levels of Openness to Experience and Conscientiousness (p < 0.01). Specifically, pairwise comparisons of ranks49 for the level of Openness to Experience indicated that participants whose level of Openness was diagnosed as high used memory strategies significantly more often than participants whose level of Openness was diagnosed as medium or low. A parallel observation could be drawn with regard to the level of Conscientiousness, as the investigated individuals who were diagnosed as highly conscientious used memory strategies significantly more often than those whose level of Conscientiousness was diagnosed as average or low. Figures 6 and 7 reflect the above findings.

Average ranks: Memory strategies

450 Level of Openness to Experience: 1-low, 2-medium, 3-high 400 428.06 350 360.15 300 344.91 250 200 150 100 50 0 1 2 3

Fig. 6. Use of memory strategies across psychologically interpreted levels of Openness to Experience.

49 The use of average ranks rather than means was required by the nature of the Kruskal-Wallis as the non- parametric equivalent of ANOVA.

222 Average ranks: Memory strategies Level of Conscientiousness 1-low, 2-medium, 3-high 450 400 428.46 350 356.03 300 316.43 250 200 150 100 50 0 1 2 3

Fig. 7. Use of memory strategies across psychologically interpreted levels of Conscientiousness.

Regarding the relationship between personality traits on the frequency of using cognitive strategies, statistically significant findings were limited to the relationships between cognitive strategies and the level of Openness to Experience. According to the results of One-Way ANOVA, the frequencies of using cognitive strategies differed significantly across groups of participants created on the basis of their level of Openness to Experience (p < 0.01). Since the condition of homogeneity of variances in the three groups was met, post-hoc tests were run in order to verify the statistical significance of the differences in the frequency of using cognitive strategies between participants characterised by low, medium, and high level of Openness. The results of Tukey’s test confirmed the statistical significance of the differences between groups 1 and 2, groups 1 and 3, and groups 2 and 3 (p < 0.01) and indicated a positive relationship between the level of the trait and the frequency of using cognitive strategies. The mean frequencies of using cognitive strategies by students characterised by various levels of Openness are presented in Fig 8.

223 Mean values: Use of cognitive strategies Level of Openness to Experience 1-low, 2-medium, 3-high 3,30 3,20 3.22 3,10 3,00 2,90 2,80 2.87 2,70 2.70 2,60 2,50 2,40 1 2 3

Fig. 8. Use of memory strategies across psychologically interpreted levels of Openness to Experience.

Regarding the frequency of use of compensation strategies, one statistically significant relationship was found. According to the results of the Kruskal-Wallis test for independent samples, the participants whose level of Openness to Experience was diagnosed as high (3) applied compensation strategies considerably more often than those whose level of the trait was either low (1) or medium (2) (p < 0.01). The values of average ranks of the frequencies of using compensation strategies across groups of students are presented in Fig. 9.

Average ranks: Compensation strategies Level of Openness to Experience: 1-low, 2-medium, 3-high 500,00 450,00 400,00 438.19 350,00 360.95 300,00 329.23 250,00 200,00 150,00 100,00 50,00 0,00 1 2 3

Fig. 9. Compensation strategies use across psychologically interpreted levels of Openness to Experience.

224 Students’ use of metacognitive strategies was subject to influences of as many as three out of the ‘big’ five personality traits. The results of one-way ANOVA indicated that the frequency of their use varied across groups of students distinguished on the basis of their levels of Extraversion, Openness to Experience, and Conscientiousness (p < 0.01). Since the variances in the analysed groups were homogeneous, post-hoc Tukey’s tests were run in order to account for specific differences across groups. According to their results, participants who were characterised by a high level of Extraversion were also significantly more frequent users of metacognitive strategies than participants whose level of Extraversion was diagnosed as low. Similarly, the individuals whose level of Consci- entiousness was the highest applied metacognitive strategies significantly more often than those whose level of Conscientiousness was the lowest. With respect to Openness to Ex- perience, significant differences were observed between the group diagnosed as the most open and the two remaining groups. The mean frequencies of using metacognitive strategies are shown in Fig. 10, 11, and 12.

Mean values: Use of metacognitive strategies Level of Extroversion: 1-low, 2-medium, 3-high 3,10

3,00 3.05

2,90 2.88 2,80

2,70 2.71

2,60

2,50 1 2 3

Fig. 10. Use of cognitive strategies across psychologically interpreted levels of Extraversion.

225 Mean values: Use of metacognitive strategies Level of Openness to Experience: 1-low, 2-medium, 3-high 3,20

3,10 3.14

3,00

2,90 2.87 2,80

2,70 2.70 2,60

2,50

2,40 1 2 3

Fig. 11. Use of cognitive strategies across psychologically interpreted levels of Openness to Experience.

Mean values: Use of metacognitive strategies Level of Conscientiousness: 1-low, 2-medium, 3-high 3,10

3,00 3.04

2,90 2.88 2,80

2,70 2.71 2,60

2,50 1 2 3

Fig. 12. Use of cognitive strategies across psychologically interpreted levels of Conscientiousness.

According to the results of the Kruskal-Wallis test, the frequencies of using compensation strategies differed significantly across groups of participants formed on the basis of their level of Openness to Experience (p < 0.01). Pairwise comparisons of average ranks indicated that the students whose level of Openness was diagnosed as high (3) were also more frequent users of compensation strategies than students whose level of Open- ness was diagnosed as medium (2) or low (1). These findings are presented in Fig. 13.

226 Average ranks: Compensation strategies

500,00 Level of Openness to Experience: 1-low, 2-medium, 3-high

400,00 438.19

360.95 300,00 329.23

200,00

100,00

0,00 1 2 3

Fig. 13. Use of cognitive strategies across psychologically interpreted levels of Openness to Experience.

The differences across groups of participants formed according to their level of Openness to Experience were also statistically significant with respect to the frequency of using affective strategies (p < 0.01), as indicated by the results of the Kruskal-Wallis test for independent samples. More specifically, the results of pairwise comparisons confirmed that the investigated students whose levels of Openness were diagnosed as high applied affective language learning strategies significantly more often than the students whose level of Openness was established as medium or low.

Average ranks: Affective strategies Level of Openness to Experience: 1-low, 2-medium, 3-high 500,00 450,00 400,00 429.16 350,00 366.67 300,00 325.65 250,00 200,00 150,00 100,00 50,00 0,00 1 2 3

Fig. 14. Use of affective strategies across psychologically interpreted levels of Openness to Experience.

Findings concerning the use of social strategies were related to two out of the ‘big’ five personality traits: Extraversion and Openness to Experience. With reference to both

227 traits, the results of the Kruskal-Wallis test revealed significant differences across the groups which varied in terms of their psychologically interpreted level (p < 0.01). Re- garding the former trait, pairwise comparisons of average ranks indicated that the participants whose level of Extraversion was high (3) used social strategies more often than the participants who belonged to the two other groups, characterised by medium (2) or low (1) level of Extraversion (p < 0.01). Regarding the latter trait, all comparisons across groups proved to be statistically significant (p < 0.01) and it was inferred that the respondents whose level of Openness to Experience was the highest (3) were also most frequent users of social strategies, and those who were diagnosed as conventional and traditional (1) were found to be rare users of social strategies. Fig. 15 presents the graphical representation of the comparisons of social strategy use across groups varying in the level of Extraversion while Fig. 16 does so for groups varying in the level of Openness to Expe- rience.

Average ranks: Social strategies Level of Extroversion: 1-low, 2-medium, 3-high 450,00 400,00 426.21 350,00 364.64 300,00 324.14 250,00 200,00 150,00 100,00 50,00 0,00 1 2 3

Fig. 15. Use of social strategies across psychologically interpreted levels of Extraversion.

Average ranks: Social strategies Level of Openness to Experience: 1-low, 2-medium, 3-high 500,00

400,00 456.03

300,00 360.99 298.18 200,00

100,00

0,00 1 2 3

Fig. 16. Use of social strategies across psychologically interpreted levels of Openness to Experience.

228 Finally, one-way ANOVA was conducted in order to investigate the differences in mean frequencies of general use of LLS across groups of learners varying in terms of the level of each of the ‘big’ five personality traits. According to its results, the only statistically significant differences referred to the groups created on the basis of the psychologically interpreted levels of Extraversion and Openness to Experience (p < 0.01). As far as the level of Extraversion was concerned, there were significantly more frequent strategy users among the participants who exhibited a high level of Extraversion (3) than those who exhibited low level of Extraversion (1). Regarding the level of Openness, significant differences in the mean frequency of LLS use existed between group 1 (the least open participants) and 2(participants characterised by medium Openness), groups 1 and 3 (participants whose level of Openness was diagnosed as high), and groups 2 and 3. The graphical representations of these findings are provided in Fig. 17 and 18, which both present the mean values of GULLS across the described categories of students.

Mean GULLS Level of Extroversion: 1-low, 2-medium, 3-high 3,05 3,00 3.02 2,95 2,90 2,85 2.89 2,80 2,75 2.81 2,70 2,65 1 2 3

Fig. 17. General use of LLS across psychologically interpreted levels of Extraversion.

Mean GULLS Level of Openness to Experience: 1-low, 2-medium, 3-high 3,20 3,10 3.13 3,00 2,90 2,80 2.88 2,70 2.75 2,60 2,50 1 2 3

Fig. 18. General use of LLS across psychologically interpreted levels of Openness to Experience.

229 Overall, significant differences in the frequency of strategy use were observed among groups of students who varied in terms of three personality traits: Conscientious- ness, Extraversion, and Openness to Experience. The impact of the last trait on the use of LLS seems to be the most pronounced as its level proved to affect the frequency of using each single category of LLS, and the overall frequency of strategy use. The level of Ex- traversion played a vital part in the frequency of participants’ use of social strategies, and LLS in general. A high level of Conscientiousness was found to be the feature which characterised frequent memory strategy users. The frequency of using the complex category of metacognitive strategies was significantly affected by each of the three mentioned traits. Therefore, in brief, the procedure which involved testing the existence of significant differences across groups of participants distinguished on the basis of the level of personality traits definitely provided a positive answer to Research Question 1.

Correlations between personality traits and specific LLS

The analysis of correspondence between learners’ levels of personality traits and strategy application revealed that the level of three personality traits: Openness to Experience, Extraversion, and Conscientiousness, significantly correlated with the frequencies of participants’ use of strategy scales and specific strategies measured by the adaptation of the SILL50. As far as strategy categories are concerned, moderate, significant correlations were observed between participants’ level of Openness to Experience and the frequency of their general use of LLS (r = 0.31, p < 0.01), and between participants’ level of Open- ness to Experience and the frequency of their use of metacognitive strategies (r = 0.32, p < 0.01)51. In terms of effect sizes, it could be inferred that the proportions of shared variances amounted to 9.9% and 10.5% respectively. With regard to specific SILL items, one moderate correlation was found. The values of Item 15 from the cognitive category, I watch English language TV shows spoken in English or go to movies spoken in English

50 As parametric and non-parametric coefficients needed to be calculated for different sets, r is used to represent Pearson’s correlation coefficient while r (rho) stands for Spearman’s correlation of ranks. 51 In order to avoid Type I errors in multiple comparisons, the Bonferroni correction was introduced and the required level of statistical significance was calculated for the set of five personality dimensions and established at α/5 = 0.01.

230 significantly correlated with the level of Openness to Experience (r = 0.301, p < 0.01)52, and shared the fraction of shared variances amounted to 9%. Table 38 summarises the above description of moderate correlational relationships.

Table 38. Moderate correlations

shared

Variable 1 Variable 2 r value variance

Openness to Experience GULLS 0.31 9.9%

Openness to Experience metacognitive strategies 0.32 10.5%

Openness to Experience SILL item 15 0.30 9%

A vast number of weak, but nonetheless, statistically significant correlations, existed between the investigated learners’ levels of Openness to Experience and the frequency of applying certain specific LLS belonging to all of the six categories forming the SILL ver. 7.0. To start with, the level of Openness significantly correlated with the reported use of some memory strategies, including SILL items 1, I think of relationships between what I already know and new things I learn in English (r = 0.13, p < 0.01), 2, I use new English words in a sentence so I can remember them (r = 0.17, p < 0.01), 3, I connect the sound of a new English word and an image or picture of the word to help remember the word (r = 0.10, p < 0.01), and 4, I remember a new English word by making a mental picture of a situation in which the word might be used (r = 0.10, p < 0.01). Furthermore, the level of the trait affected the frequency of using six specific cognitive strategies, that is SILL items 11, I try to talk like native English speakers (r = 0.24, p < 0.01), 12, I practice the sounds of English (r = 0.16, p < 0.01), 13, I use the English words I know in different ways (r = 0.13, p < 0.01), 14, I start conversations in English (r = 0.20, p < 0.01), 16, I read for pleasure in English (r = 0.28, p < 0.01), and 17, I write notes, messages, letters, or reports in English (r = 0.22, p < 0.01). It also demonstrated a number of statistically significant correlational relationships with five specific compensation strategies, SILL items 24, To understand unfamiliar English words, I make guesses (r = 0.22, p < 0.01), 27, I read English without looking up every new word (r = 0.21, p < 0.01) 28, I try to guess what the other person will say next

52 Spearman’s rank coefficient was calculated in place of Pearson’s coefficient due to the ordinal character of the data representing the frequency of using a single strategy.

231 in English (r = 0.10, p < 0.01), 29, If I can’ t think of an English word, I use a word or phrase that means the same thing (r = 0.18, p < 0.01), and 30, When I communicate in English on the Internet, I prevent myself from using online dictionaries or translators (r = 0.13, p < 0.01). The highest number of significant relationships between the level of Openness to Experience and strategy use was found regarding the metacognitive category of LLS and concerned items 32, I try to find as many ways as I can to use my English (r = 0.23, p < 0.01), 33, I notice my English mistakes and use that information to help me do better (r = 0.21, p < 0.01), 34, I pay attention when someone is speaking English (r = 0.19, p < 0.01), 35, I try to find out how to be a better learner of English (r = 0.12, p < 0.01), 37, I look for people I can talk to in English (r = 0.17, p < 0.01), and 38, I look for opportunities to read as much as possible in English (r = 0.22, p < 0.01). Students’ use of five specific affective strategies was also modified by their level of Openness. Correlations between the level of the trait and the frequency of students’ use of particular strategies proved to be statistically significant for the adapted SILL items 41, I try to relax whenever I feel afraid of using English (r = 0.10, p < 0.01) 42, I encourage myself to speak English even when I am afraid of making a mistake (r = 0.20, p < 0.01), 46, I talk to someone else about how I feel when I am learning English (r = 0.13, p < 0.01), 47, I use emoticons when I write in English (r = 0.14, p < 0.01), and 48, I share my feelings about learning/using English through a blog or internet portals (r = 0.16, p < 0.01). Finally, the degree of participants’ employment of five out of six social strategies also corresponded to their level of Openness. The relationship was confirmed for the adapted SILL items 50, If I do not understand something in English, I ask the other person to slow down or say it again (r = 0.13, p < 0.01), 51, I ask English speakers to correct me when I talk (r = 0.18, p < 0.01), 53, I ask for help from English speakers (r = 0.12, p < 0.01), 54, I ask questions in English (r = 0.29, p < 0.01) and 55, I try to learn about the culture of English speakers (r = 0.26, p < 0.01). The level of Conscientiousness demonstrated a positive relationship with the frequency of using metacognitive strategies (r = 0.20, p < 0.01)53, and language learning strategies in general (r = 0.15, p < 0.01). The level of the trait also positively correlated with the frequency of using five out of nine

53 Pearson’s correlation coefficient was calculated as both scales were normally distributed, as indicated in the previous Chapter (Table 17, section 4.4.2.3, Chapter Four). The level of statistical significance was set at α/5 = 0.01 after the Bonferroni correction for multiple comparisons had been introduced.

232 specific metacognitive strategies contained in the applied adaptation of the SILL ver. 7.0, including items 33, I notice my English mistakes and use that information to help me do better (r = 0.18, p < 0.01), 35, I notice my English mistakes and use that information to help me do better (r = 0.12, p < 0.01), 36, I plan my schedule so I will have enough time to study English (r = 0.23, p < 0.01), 39, I have clear goals for improving my English skills (r = 0.17, p < 0.01), and 40, I think about my progress in learning English (r = 0.13, p = 0.01). Moreover, it was observed that the level of Conscientiousness significantly correlated with participants’ use of certain specific memory strategies, including items 6, I use flashcards to remember new English words (r = 0.19, p < 0.01), 8, I review English lessons often (r = 0.22, p < 0.01), and 9, I remember new English words or phrases by remembering their location on the page, on the board, or on a street sign (r = 0.13, p < 0.01). Correlations of the levels of the trait were also observed with learners’ use of two cognitive strategies, items 10, I say or write new English words several times (r = 0.18, p < 0.01), and 18, I first skim an English passage (read over the passage quickly) then go back and read carefully (r = 0.10, p < 0.01) of the adapted SILL ver. 7.0, and the frequency of applying one compensation strategy, that is item 31, I draw the representation of what the word which I have forgotten (r = 0.10, p < 0.01). Also, the degree of participants’ Conscientiousness proved to influence their preference of two affective strategies, and one social strategy, that is items 43, I give myself a reward or treat when I do well in English (r = 0.11, p < 0.01), 48, I share my feelings about learning/using English through a blog or internet portals, and 52, I practice English with other students. The level of Extraversion was another factor which affected the frequency of participants’ use of certain specific LLS across all of the six SILL categories. The three items belonging to the memory category included item 1 (r = 0.15, p < 0.01), I think of relationships between what I already know and new things I learn in English, 2, I use new English words in a sentence so I can remember them (r = 0.10, p < 0.01), and 5, I use rhymes to remember new English words (r = 0.10, p < 0.01). Three other strategies whose frequencies of employment were correlated with the level of the trait were part of the cognitive scale: item 13, I use the English words I know in different ways (r = 0.11, p < 0.01)., item 14, I start conversations in English (r = 0.17, p < 0.01)., and item 23, I make summaries of information that I hear or read in English (r = 0.13, p < 0.01). Additionally, the degree of Extraversion was related to the frequency of applying two SILL compensation strategies, that is items 27, When I can’ t think of a word during a conversation in

233 English, I use gestures (r = 0.11, p < 0.01), and 28, I try to guess what the other person will say next in English (r = 0.15, p < 0.01). The use of nearly all metacognitive strategies also corresponded to the investigated students’ level of Extraversion. More specifically, the relationships were significant for SILL items 32, I try to find as many ways as I can to use my English, (r = 0.12, p < 0.01), 34, I pay attention when someone is speaking English (r = 0.10, p < 0.01), 35, I try to find out how to be a better learner of English (r = 0.13, p < 0.01), 37, I look for people I can talk to in English (r = 0.19, p < 0.01), 38, I have clear goals for improving my English skills (r = 0.15, p < 0.01), and 39, I think about my progress in learning English (r = 0.12, p < 0.01). Regarding the affective scale, three items significantly correlated with the level of the investigated students’ Extraver- sion: item 41, I encourage myself to speak English even when I am afraid of making a mistake (r = 0.16, p < 0.01), item 42, I give myself a reward or treat when I do well in English (r = 0.11, p < 0.01), and item 46, I use emoticons when I write in English (r = 0.19, p < 0.01). Finally, the application of social strategies demonstrated a statistical relationship with the level of the trait with respect to five out of six particular strategies, including the adapted SILL items 49, If I do not understand something in English, I ask the other person to slow down or say it again (r = 0.11, p < 0.01), 50, I ask English speakers to correct me when I talk (r = 0.15, p < 0.01), 51, I practice English with other students (r = 0.15, p < 0.01) 52, I ask for help from English speakers (r = 0.10, p < 0.01), and 53, I ask questions in English (r = 0.19, p < 0.01). Overall, the correlation of the level of Extraversion with participants’ frequency of using the category of social strategies was also statistically significant and amounted to r = 0.20 (p < 0.01). Although the degree of respondents’ Agreeableness did not considerably affect any of the six categorical constructs included in the adapted version of the SILL 7.0, significant correlations were observed between the level of the trait and the frequencies of using nine specific LLS, including two memory strategies, three cognitive strategies, three affective strategies, and one strategy from the social domain. More specifically, positive correlations were observed for items 8, I review English lessons often (r = 0.17, p < 0.01), 9, I remember new English words or phrases by remembering their location on the page, on the board, or on a street sign (r = 0.10, p < 0.01), 10, I say or write new English words several times (r = 0.13, p < 0.01), 44, I talk to someone else about how I feel when I am learning English (r = 0.11, p < 0.01), 49, I try to have positive feelings about my English teacher (r = 0.20, p < 0.01), and 50, If I do not understand something

234 in English, I ask the other person to slow down or say it again (r = 0.10, p < 0.01), while negative correlations were observed for items 15, I watch films or other programmes in English (r = -0.10, p < 0.01), 16, I read for pleasure in English (r = -0.11, p < 0.01), and 45, I write down my feelings in a language learning diary (r = -0.11, p < 0.01). Similarly to the level of Agreeableness, participants’ level of Neuroticism did not significantly correlate with the frequency of their use of an entire category of strategies. However, statistically significant correlations of the level of the trait were found with certain specific strategies from the cognitive, compensation, affective, and social categories. They included positive relationships for items 25, When I can’ t think of a word during a conversation in English, I use gestures (r = 0.12, p < 0.01) 26, I make up new words if I do not know the right ones in English (r = 0.10, p < 0.01), 44, I notice if I am tense or nervous when I am studying or using English (r = 0.23, p < 0.01) and 46, I talk to someone else about how I feel when I am learning English (r = 0.16, p < 0.01) and negative relationships for items 13, I use the English words I know in different ways (r = -0.10, p < 0.01), 14, I start conversations in English (r = -0.13, p < 0.01), 22, I try not to translate word-for-word (r = -0.12, p < 0.01), 30, When I communicate in English on the Internet, I prevent myself from using online dictionaries or translators (r = -0.13, p < 0.01) 42, I encourage myself to speak English even when I am afraid of making a mistake (r = -0.23, p < 0.01), and 54, I ask questions in English (r = -0.14, p < 0.01). Overall, most statistically significant correlations with the frequency of using both whole categories of LLS and specific individual strategies were demonstrated with participants’ level of Openness to Experience followed by the level of Extraversion and Con- scientiousness. These relationships were all positive, although they were never strong, and proved to be moderate for only two strategy categories and one particular strategy with regard to Openness. The level of Neuroticism did not correlate with the frequency of using any particular category of strategies, but it demonstrated mostly negative, weak relationships with the use of certain individual strategies. The level of Agreeableness did not demonstrate a clear influence on strategy use as correlations appeared for the lowest number of particular strategies and no specific category of strategies as such. Moreover, the relationship between strategy use and Agreeableness was not clearly positive or negative as comparable numbers of correlation coefficients reached values above and below zero.

235 LLS use explained by stepwise regression of personality traits

Another procedure employed in investigating the relationships between students’ use of language learning strategies and their personality traits was stepwise regression. The initial set of explanatory variables involved all of the ‘big’ five personality traits, while GULLS was treated as dependent variable in the regression model since the distribution of its values was not significantly different from normal (a preliminary condition to be met in regression analysis). Stepwise regression revealed that four out of five personality traits significantly contributed to the explanation of GULLS values. They included Open- ness to Experience, Extraversion, Conscientiousness, and Neuroticism. The only trait which was not included in the model is Agreeableness. As indicated before, the level of the trait did not significantly correlate with the frequency of use of any of the six strategy scales, or using language learning strategies in general (GULLS), and weak, significant correlations were found with the lowest number of specific LLS in comparison to the other traits. Table 39 contains the values of R, the correlation coefficient, R2, the determination coefficient, explaining how much the value of the dependent variable is influenced by the value of the explanatory variable, and the standardised b values (st. b), that is the relative measures of impact of each of the explanatory variables involved in the model. As can be seen, Openness to Experience proved to be the greatest predictor of strategy use in the model, accounting for nearly 11% of GULLS variance, and demonstrating the greatest impact on the use of strategies in comparison to other traits (b = 0.32). After incorporating the level of Extraversion in the model, the proportion of explained variance increased to 14%. The impact of Extraversion was relatively lower than the impact of Openness, as shown by the value standardised b, amounting to = 0.21. The levels of Con- scientiousness and Neuroticism also proved to affect the frequency of participants’ use of LLS, but their contribution to the percentage of the explained variance of GULLS proved to be considerably lower, as each of the two variables resulted in a 1% increase in the value of the determination coefficient, from 14% to the final value of 16%. Consequently, the values of the b coefficients calculated for Conscientiousness and Neuroticism only amounted to 11%. Finally, after including four personality traits, the model cumulatively explained 16% of GULLS variance - a proportion which cannot be ignored in discussing the determinants of strategy use.

236 Table 39. Stepwise regression parameters Explanatory variables R R2 bst. 1. Openness to Experience 0.33 0.11 0.32 2. Openness to Experience, 0.37 0.14 0.21 Extraversion 3. Openness to experience, Extraversion, Conscientious- 0.38 0.15 0.11 ness 4. Openness to Experience, Extraversion, 0.4 0.16 0.11 Conscientiousness, Neuroticism

The results of the stepwise regression confirmed the earlier findings from the sections of the present chapter, discussing the results of investigations of variance in groups of participants varying in terms of personality traits, and analysis of correlational relationships between personality traits and the frequency of strategy use. Openness to Expe- rience was found to be the greatest predictor of strategy use, while other personality traits, except for agreeableness, did influence strategy use, but their role was less obvious, or, perhaps, modified by other factors. The results suggested that a high level of Openness to Experience could facilitate strategy use, the primary reasons for that most probably being intellectual curiosity and preference for variety - in this case, a variety of strategies. More- over, the distribution of residual values, shown in Fig. 19, was not significantly different from normal (mean value = 0 and standard deviation = 1), which confirmed the good fit of the model.

Fig. 19. The distribution of residual parameters in the stepwise regression model.

237 Insights from semi-structured interviews

In terms of the interviewed learners’ strategy repertoires, a number of observations could be made. To start with, memory strategies were frequently referred to as a tool that learners used less willingly and only if they thought that their use was inevitable and they could not be replaced with a more efficient alternative. Some statements which exemplify this view are as follows: I only memorise things when I learn for school, I only memorise business words which I find impossible to understand, Learning by heart doesn’t work for me in the long run because I need emotion to remember something. A number of learners reported using social strategies, such as studying in pairs or groups and practising with native speakers. On the other hand, although a number of learners were able to identify their emotions which accompanied them in learning English, no one really admitted to trying to manage them in any way. When it comes to the correspondence between LLS and personality traits, the analysis of data obtained in the series of semi-structured interviews lent support to the findings described in the above sections of the present chapter. First of all, some learners who were highly open to experience and attributed a particular role to Openness reported using a wide range of LLS. Their willingness to explore new strategies could be, perhaps, best described with the words of one of the participants: It’s difficult to learn a language if you’re narrow-minded, unwilling to try new things or you just think that you already know everything. On the other hand, a student whose level of Openness was diagnosed as low acknowledged: I know what helps me and I don’t look for other ways to study because what I do seems to be working. Also, a number of the metacognitive and cognitive strategies reported by the interviewed highly open learners required their users to dedicate a considerable amount of time in the long run. For instance, this was probably the case of watching English films or reading books in English. One student literally confessed: I’m really addicted to watching films in English and another one commented: If it wasn’t for English, I wouldn’t ‘t be able to do half the things I do. I mean, mainly entertainment. Another observed tendency which supported the conclusions from statistical analysis of the interaction between LLS and personality traits consisted in highly extraverted students’ preference for social strategies. Declaring the willingness to learn through interaction with others, they appreciated the benefits of both learning together (e.g. revising for tests, checking each other’s answers) and participating in conversations in English

238 with foreigners. Some of them established long-term friendships with foreigners and specifically referred to them as pen-pals. Others referred to the satisfaction they experienced when engaging in real-life conversations rather than classroom activities imitating genu- ine communication. Another reported benefit of social learning consisted in receiving immediate feedback, both on one’s overall command of English and with reference to its certain specific aspects required for successful communication. Finally, a number of learners reflected on the advantages of being a conscientious language learner. However, while a number of the interviewed learners did refer the trait and its underlying facets, such as being systematic or persistent, no clear correspondence could be established between its level and participants’ preference for any category of LLS.

5.2.2. The relationships between language learners’ LLS use and attainment

A number of positive relationships were observed between the investigated students’ level of English and their reports of strategy use. To start with, it was found that the frequency of general use of language learning strategies (GULLS) in groups characterised by relatively higher level of competence in English exceeded the frequency in groups of lower-level learners as the mean values of GULLS measured for each level of proficiency demonstrated a clear rising pattern from 2.70 for A1+ learners to 2.91 for B2 learners. Moreover, one-way ANOVA indicated the existence of statistically significant differences between learners characterised by different CEFR levels (p < 0.05). As the variances in the groups were homogeneous, post-hoc tests were run in order to account for differences in the frequency of strategy use between learners at specific levels of competence. According to the Tukey’s test results, statistically significant differences existed between A1+ and B2 students, and A2 and B2 students. The difference in the frequency of applying LLS between B1 students and B2 students was not statistically significant. The differences in general strategy use among the investigated students across proficiency levels are shown in Figure 20.

239 Strategy use (GULLS)

2,95

2,9 2.91 2,85

2,8 2.85

2,75 2.79 2,7

2,65 2.7 2,6

2,55 A1+ A2 B1 B2

Fig. 20. General frequency of strategy use across proficiency levels.

The use of specific categories of strategies was also subject to variation across levels of proficiency, in the cases of the normally-distributed cognitive and metacognitive scales, and the remaining strategy scales, whose distributions were significantly different from normal. On the one hand, the results of the Kruskal-Wallis test failed to confirm the existence of statistically-significant differences in strategy use across participants’’ level of proficiency in English with regard to the frequency of memory, compensation, affective and social strategies (p > 0.05). However, participants’ attainment operationalised as their CEFR level was found to be relevant for their use of cognitive and metacognitive strategies. Since the variances across the compared groups were not homogeneous in either of the two considered strategy categories, as indicated by the results of Levene’s test (p < 0.05), robust tests for the equality of means were conducted instead of post-hoc tests. According to the results of the Brown-Forsythe F test, the frequency of participants’ use of both cognitive and metacognitive strategies differed significantly for learners whose level of attainment was labelled with different CEFR levels. The mean values for both scales strategies, calculated across levels of competence, are presented in Table 40.

Table 40. Use of cognitive and metacognitive strategies across levels of proficiency.

Scale A1+ A2 B1 B2

Cognitive strategies 2.6087 2.8277 2.9645 3.0311

Metacognitive strategies 2.7398 2.8017 2.9130 3.0253

240

In order to account for pairwise comparisons, a Kruskal-Wallis test was conducted, which revealed the following findings: the distribution of the reported mean frequency of applying cognitive strategies differed significantly between the following groups of participants: A1-B1, A1-B2, and A2-B2 (p < 0.05), and overall, participants whose level of attainment was relatively higher used cognitive strategies more frequently than lower-level participants; the distribution of the reported mean frequency of applying metacognitive strategies was significantly different between the investigated A2 and B2 learners as B2 learners employed metacognitive strategies significantly more frequently than A2 learners (p < 0.05). In terms of the interviewed students’ strategic preference for certain strategy categories, it was observed that the learners whose level of attainment was diagnosed as B1 and B2 (thus relatively higher-level learners) tended to report applying cognitive, metacognitive, and social strategies relatively frequently. A number of them attempted to choose entertainment, such as films or books, in English rather than in their mother tongue. In particular, one person described the meaning of English in her life as the medium of entertainment which makes it possible for me to read books or newspapers and expressed satisfaction with being able to use it in real-life communication. Another participant provided the example of studying when participating in online forums and ex- changing ideas, which, as he claimed, was accompanied by paying attention to the words and phrases used by other uses of English in order to improve his English vocabulary connected with IT. Higher-level (B1, B2) learners also reported reflecting on their strategy use and appeared to display a greater awareness of the need to analyse the efficiency of their learning. This became evident in statements such as I analyse what is efficient in learning English; I make sure I use English because if I don’t, I tend to forget it; I have these second thoughts, like, for example, I could have said it in a better way, or I should have used a different form.

241 5.2.3. Students’ personality traits and attainment

With reference to personality traits, one-way analysis of variance and its nonparametric equivalent, the Kruskal-Wallis H test, did not reveal any significant differences in the level of either of the ‘big’ five traits across proficiency levels (p values exceeded 0.05). However, this finding does not diminish the importance of personality in the use of LLS. Accepting Hornowska’s (2003) definition of personality, according to which, personality can be understood as a set of permanent psychic characteristics and internal mechanisms which regulate human behaviour, it is logical to assume that learners’ paths to proficiency differ as they are likely to reflect the personality differences between individuals. The findings from the qualitative part of the study, which support this claim, are presented below for each of the ‘big’ five personality traits except for Agreeableness, which was not mentioned in students’ verbal reports. To start with, learners who were diagnosed as introverts (Introversion levels in stens 1-3) typically reported their preference for pair work and working in small teams over working together with all the other groupmates. They appreciated these types of interaction before participating in whole-group discussions or answering course instructor’s questions in front of the other participants of their classes. Some interviewees specifically referred to the alleged easiness of their more extroverted groupmates with which the latter established new friendships and entered social interactions, admitting that they actually envied them and speculating that they could do better as language learners if they were equipped with that characteristic. A highly neurotic person pointed out that they were likely to feel pessimistic about their performance and irritated “when everything goes wrong” - a possible manifestation of trait or state anxiety which may be accompanied by hostile behavior, anger, withdrawal from interaction, and, if the feelings persist, even continual shame, and self-consciousness (Zawadzki et al. 2010). The learner feared that their anxiety could, in fact, worsen their performance in class, limit the degree of comprehension in receptive tasks, increase their unwillingness to communicate54, and, finally reduce the amount of output in English. Some comments made by other participants also confirmed the impact of anxiety on the

54 Willingness to communicate can be understood as the as a function of “proximal and distal variables comprising psychological, linguistic, educational and communicative dimensions of language” (Mystkow- ska-Wiertelak and Pawlak 2014).

242 learning process included the following: There’s a battle inside me. I think I know the answer, but I remain quiet because I fear making a mistake; I get really frustrated whenever a test’s coming and I know I understand nothing; I fear I don’t have the guts to be spontaneous; I feel that my unwillingness to make mistakes really limits me; Pessimism and being anxious can really impede my learning sometimes. The theme of Openness to Experience echoed through a vast majority of interviews as nearly all participants referred to openness as a desirable quality of a language learner. The only participant who did not specifically refer to Openness, did, actually, mention some its underlying traits, specifically curiosity and imagination. This suggests that the participants of the study themselves perceived openness as a kind of gift, and a facilitator in language learning which is likely to have a favourable effect on their level of proficiency in English. Those who were highly open to experience were also more agile as strategy users and, consequently, language learners. Moreover, the beneficial impact of being open was also evident in learners’ reports of the usefulness of English in learning content subjects and vice versa as the following claims appeared: Delving into the merits we also gather extra knowledge of English; What really helps me is making the study of English suit my interests; I simply know I have to know English to study cognitive science. Moreover, some participants even pointed out that their decision which learning strategy to use originated from their intellectual interests or preferences. Students who tried to find examples of learners who were highly open to experience paid attention to the fact that some individuals were more likely to be open than others because of their family backgrounds. For instance, an example of a friend was introduced whose father happened to be a businessman, frequently relocated from one country to another. Both the businessman and his family were in a way forced to encounter a variety of cultures, which, according to the participant of the study, contributed to a greater openness to experience of the mentioned friend, and equipped the friend with a quality of a successful language learner. Some students pointed to the phenomenon of “learnt” openness, which was either their own case or the case of a good language learner. To use the very words of one participant, What really helps you master a language is the kind of openness, willingness to communicate, and curiosity of the world. I think it’s something you can also learn in a way. An anecdote was also told about a boy who was the son of a businessman moving from one country to another every few years. The participant knew that the boy was a

243 successful language learner of some languages and concluded that in a way, he learnt to be more open. Other participants also recalled their own learning experience and concluded that they gradually became more open which helped them communicate more freely and more effectively. Moreover, the learners reported that as they became more open, communicating in English became more pleasurable. Other participants of the study noticed that their curiosity of the world motivated them to learn English, which they perceived as the global language. These reflections were compliable with the descriptions of individuals diagnosed as highly open to experience, such as their willingness to experience a lot in life and having a rich spiritual life (experiencing deeply rather than briefly) (McCrae 1993: 577). Describing the manifestations of their own openness to experience, participants of the study referred to the willingness “to be in the natural environment where English is spoken”, their experiences as international exchange students, or vocational experience. The only person whose level of Openness to experience was rather low expressed her belief that she was “thinking too logically to succeed in learning English” - here, logical reasoning seems to be juxtaposed with imagination, which constitutes another underlying facet of Openness to experience. On the other hand, positive emotions were expressed by a few learners who viewed learning English as an opportunity to learn something new in general. This group included individuals whose levels of Openness to experience were relatively high. An interesting observation was made by another learner who was highly open to experience. The person noticed that they had not explicitly learnt anything for a long time, but rather broadened their vocabulary “by chance”, for example, through reading articles or watching films in English. Only if contextual information was insufficient to decode the meaning of the encountered new word or phrase did they refer to a dictionary in order to check the meaning. The person mentioned above as well as other relatively more advanced students stressed that they were motivated to boost their level of competence in English in order to reach their aims, and that learning the language was principally a matter of their own choice, a source of satisfaction, or even a planned phase in their development as a person. According to another valuable comment made by a person who was highly open to experience, it was virtually impossible for them to learn anything that was enforced as a duty rather than the learner’s own choice. Improving the command of English was therefore treated as a necessary prerequisite of expanding students’

244 professional knowledge (cognitive scientists) or joining multinational companies whose official language was English. Students also reflected that they learnt English in order to make friends abroad, remain in touch with the friends they have already made, become more mobile, be able to joint various entertainment events worldwide, and as it was said in one of the comments, have “contact with the world”. Conscientiousness was repeatedly mentioned by some interviewees as a desired trait for foreign language learners. Regarding its underlying facets, or orderliness was believed to particularly supportive for the efficiency of the learning process while laziness was considered as an obstacle in raising one’s competence. What is important, the comments that learners made referring to Conscientiousness addressed learning vocabulary or grammar rules rather than the so-called naturalistic acquisition of language. In particular, some interviewees quoted difficulties in learning, remembering, or recalling new English words, or experiencing difficulties trying to comprehend the English grammar as a system. Another difficulty reported by the students was self-discipline problems, which could be interpreted as lack of motivation to learn. These problems were mainly situational, most frequently encountered before progress tests while making decisions to attend English classes. On the other hand, highly conscientious students were worried about making mistakes, which affected their willingness to communicate inside and outside the classroom. One person even reflected: I really envy those who don’t experience the language barrier preventing them from using English. Another one blamed her perfectionism for the feeling of frustration: I just want to be perfect and I get so frustrated when things don’t work out. To summarise the investigations of the interplay between learners’ personality traits and their level of attainment, it could be concluded that while no significant differences in the level of personality traits were found across participants’ level of language proficiency in the quantitative part of the study, nearly all personality traits were mentioned by the learners investigated in the semi-structured interviews. Moreover, some of them, including Openness to Experience, Extraversion, and Conscientiousness, were believed to be facilitative, whereas others, such as Introversion, were viewed as impedi- ments in developing linguistic competence. In the specific case of Neuroticism, participants did not refer to the cardinal trait as such, but pointed to one of its underlying facets, consisting in the tendency to experience negative emotions. Among those, anxiety was

245 specifically identified as a major obstacle in participating in classroom activities and hence, limiting the opportunities to make progress. The results discussed in sections 5.2.2 and 5.2.3 indicate that statistically significant relationships existed between learners’ level of proficiency, and the frequency characterising their application of LLS in general, as well as cognitive and metacognitive strategies in particular. Given the lack of statistically significant relationships between participants’ attainment and the levels of their personality traits, it could be hypothesised that language learning strategies are a latent variable that moderates the interaction between learners’ level of personality traits and their level of proficiency in English as a foreign language. At the same time, taking into account the insights from the qualitative study, it could be speculated that certain combinations of personality traits could be more likely to enhance more frequent strategy use, at least with regard to some categories of LLS, and through that, prove to be more beneficial for developing their competence in English. Section 5.2.4 is dedicated to the exploration of that possibility, as it seeks to describe the profiles of strategy users and characterise them in terms of both the levels of each the ‘big’ five personality traits and the frequency of applying specific categories of LLS.

5.2.4. Student profiles

In order to provide the answer to the final research question, clustering techniques described in section 4.4.4.7 were employed. By applying the two-step clustering technique, initial pre-clusters of learners were distinguished, which were afterwards used to create a categorical variable, cluster membership, in the hierarchical analysis. The obtained results were then compared with the results of the k-means non-hierarchical clustering procedure in order to triangulate both groups of findings. Along with the GULLS results entered as ordinal values, 1 (low use), 2 (medium use), and 3 (frequent use), the input variables in cluster analysis included the levels of four out of five personality traits, which proved to be significant predictors of the frequency of strategy use in the analysis of regression, described in subsection 5.2.2.3 of the present Chapter: Openness to Experience, Extra- version, Conscientiousness, and Neuroticism.

246 Results of two-step cluster analysis

Based on calculating the log-likelihood distance measure, the use of the algorithm described in section 4.4.4.7 (Chapter Four) resulted in the emergence of three clusters of learners, strictly corresponding to their SILL Profile results, interpreted according to the guidelines by Oxford (1990: 300). Although the clusters of frequent and rare strategy users (3 and 1) were smaller than the cluster of average strategy users, the model fit was fair, as shown in Fig. 21. Table 41 contains preliminary information concerning cluster distribution, including the number of learners classified within each cluster, the fraction of the combined population, and the percentage of the overall population of the investigated learners.

Table 41. Cluster distribution % of N Combined % of Total 1 93 14.0% 12.4% Cluster 2 502 75.4% 67.0% 3 71 10.7% 9.5% Combined 666 100.0% 88.9% Excluded 83 11.1% Cases Total 749 100.0%

Fig. 21. Two-step cluster model summary.

247 Members of cluster one were rare strategy users, whose GULLS scores amounted to 2.4 or less, characterised by the lowest level of Openness to Experience, Extraversion, and Conscientiousness, and the highest level of Neuroticism in comparison to the other groups. They constituted approximately 14% of the investigated population. Members of cluster two were average strategy users, whose mean GULLS values ranged from 2.5 to 3.4, characterised by the lowest level of Neuroticism, and the values of the three remaining investigated traits, Openness to Experience, Extraversion, and Conscientiousness, lower than those for cluster one members, but higher than those typical of cluster three members. Finally, members of cluster three were frequent strategy users whose GULLS scores amounted to 3.5 and above, but at the same time also highly open to experience, relatively more extrovert, and slightly more conscientious than members of the remaining clusters. They included nearly 11% of the investigated population of learners. Their level of Neuroticism was slightly higher than the mean value for calculated for the cluster of average strategy users, but lower than the value calculated for rare strategy users. Overall, the quality of the clusters obtained through the application of the two-step cluster algorithm could be described as fair, with the coefficient of silhouette measure of cohesion and separation amounting to 0.455. Table 42 contains the values of the centroids for the four considered personality traits across GULLS categories.

Table 42. Distribution of personality traits across clusters of learners Cluster Openness to Ex- Extraversion Conscientiousness Neuroticism perience Mean Std. Mean Std. Mean Std. Mean Std. Devia- Deviation Deviation Deviation tion 1 24.63 6.06 26.51 7.88 28.44 6.93 24.48 9.27 2 28.39 6.21 28.06 6.90 29.61 7.63 23.32 8.73 3 31.61 7.33 30.80 7.49 30.52 8.63 24.01 9.27 Com- 28.20 6.55 7.18 29.54 7.65 23.56 8.87 bined 28.13

The above distributions were tested for significance by means of a procedure which reflected the one used in the analysis of the correspondence between participants’ frequency of strategy use and the psychological interpretation of the level of their

55 A coefficient of 0.5 or above indicates a good cluster quality, while a coefficient below 0.3 indicates a poor one.

248 personality traits described in subsection 5.2.2.1. of the present chapter. In three out of four cases of the investigated personality traits, including Extraversion, Conscientious- ness, and Neuroticism, it was possible to conduct one-way ANOVA as the distributions of the levels of these traits did not significantly differ from normal. The results of one- way ANOVA confirmed statistically significant differences in the levels of Extraversion across the isolated clusters of learners (p < 0.017). More specifically, statistically significant differences were found between clusters 1 and 3, and 2 and 3, as indicated by the Tukey’s post-hoc test. In other words, frequent strategy users were significantly more extrovert than rare or average strategy users. In the case of Openness to Experience, the Kruskal-Wallis test was run due to the fact that the distribution of the trait level was significantly different from normal. The results of the test revealed that statistically differences in the level of the trait existed across all clusters (p < 0.05). The most frequent strategy users were also the most open to experience, while the least frequent strategy users were also the least open to experience.

Results of k-means cluster analysis

The final stage of quantitative analysis involved conducting non-hierarchical clustering with the use of the k-means procedure. It involved determining a fixed number of clusters, which was set as three, basing on the previous findings from two-step cluster analysis. Respondents were clustered according to the level of their personality traits as well as the frequencies of using each of the six strategy scales. The maximum number of iterations was set as 10, and cluster membership was saved as a separate variable in order to expand the range of characteristics for each respondent. Due to discrepancies in the range of values between personality scales and strategy scales (the former ranging from 0 to 40 as sums, the latter ranging from 0 to 5 as mean values), data was standardised and new, standardised variables were created. The distances between clusters were obtained through iteration and classification and computed using simple Euclidean distance. Even- tually, three clusters were isolated which included 212, 255, and 198 students. 84 students were missed out in this analysis since their personality traits or frequency of using either of the strategy scale did not centre around the centroid values in the three clusters. The ratio of the biggest cluster size to the smallest cluster size amounted to 1.29, which

249 signalled a good model fit (ratio value less than 2.0). Table 43 contains detailed characteristics of the created clusters, including the values of the F statistic, which describes the degree of relative variation between clusters considering each of the input variable. The Z_scores represent standardised values while the values from the final column indicate the significance of the F statistic.

Table 43. ANOVA for final clusters

Cluster Error Scale F Sig. Mean Mean Square df Square df Zscore: Memory strategies 137.32 2.00 0.59 662.00 232.39 0.000 Zscore: Cognitive strategies 167.85 2.00 0.49 662.00 343.83 0.000 Zscore: Compensation strat- 103.04 2.00 0.72 662.00 143.89 0.000 egies Zscore: Metacognitive strat- 168.07 2.00 0.51 662.00 328.90 0.000 egies Zscore: Affective strategies 103.60 2.00 0.66 662.00 157.42 0.000 Zscore: Sociocultural strate- 141.03 2.00 0.57 662.00 248.57 0.000 gies Zscore: Neuroticism 1.44 2.00 0.30 662.00 4.76 0.009 Zscore: Openness to Experi- 44.22 2.00 0.87 662.00 50.69 0.000 ence Zscore: Conscientiousness 34.81 2.00 0.89 662.00 38.96 0.000 Zscore: Extraversion 43.60 2.00 0.89 662.00 48.92 0.000 Zscore: Agreeableness 18.24 2.00 0.93 662.00 19.55 0.000

The descriptive statistics for each cluster are presented in Table 44. Clusters are marked as Q1, Q2, and Q3. The enlisted categories involve all six strategy scales, the aggregated strategy frequency of use (GULLS) and all the levels of all of the ‘big’ five personality traits. For each category, the mean value is shown across the isolated clusters along with the value of the standard deviation (SD). For clarity purposes, lowest values of the means per category are marked red, medium values are marked yellow, and highest values are marked green. The superscript letters serve as significance markers of differences across clusters. Different letters for any pair of clusters mean that significant differences exist between the two compared clusters in the specific case of a given variable56.

56 In order to account for the significance of the differences across clusters, one-way ANOVA was planned for scales whose distributions did not significantly vary from the normal distribution, including the scales of cognitive and metacognitive strategies, GULLS, Conscientiousness, Extroversion, and Neuroticism (see Chapter 3, section 5.1), and its non-parametric exquivalent, the Kruskal-Wallis test, was run for the remaining scales, whose distributions were significantly different from normal. Since the results of Levene’s test indicated that variances across clusters to be compared in the analysis of variance were not homogeneous

250 Table 44. Centroids for final clusters

Q1 (N= 212) Q2 (N= 255) Q3 (N=198) Variable Mean SD Mean SD Mean SD

Memory strategies 2.09a 0.47 2.54 b 0.44 3.04 c 0.42

Cognitive strategies 2.32 a 0.38 3.03 b 0.41 3.41 c 0.49

Compensation strategies 2.75 a 0.55 3.36 b 0.51 3.63 c 0.57

Metacognitive strategies 2.24 a 0.49 2.97 b 0.50 3.55 c 0.58

Affective strategies 2.19 a 0.39 2.54 b 0.42 2.93 c 0.47

Social strategies 2.41 a 0.56 3.10 b 0.53 3.64 c 0.60

GULLS_mean 2.39 a 0.26 2.98 b 0.20 3.41 c 0.30

Neuroticism 23.83 a 9.07 24.57 ab 8.56 22.03 ac 8.82

Openness to Experience 25.31 a 5.97 28.11 b 5.90 31.36 c 6.44

Conscientiousness 28.64 a 6.70 27.38 a 7.35 33.26 b 7.69

Extraversion 27.18 a 7.13 25.89 a 6.41 31.96 b 6.57

Agreeableness 28.42 a 6.01 26.98 a 6.37 30.61 b 5.97

As can be seen, the participants belonging to cluster one were rare strategy users whose level of Openness to Experience tended to be significantly lower than the level of the trait which characterised the members of the other clusters. They employed all categories of each and every category of LLS significantly less frequently than the members of the other clusters. At the same time, their level of Neuroticism was, as a rule, significantly lower than the level which characterised the members of cluster Q2. The levels of Conscientiousness, Agreeableness and Extraversion, centred around values which were significantly lower than those characterising Q3. The members of Q3 applied all categories of LLS significantly more often than members of the two remaining clusters. They were also significantly more open, extrovert, conscientious, and agreeable, but least neurotic in comparison to their counterparts from the other clusters. Overall, a consistent pattern was observed in the use of all categories of LLS and the levels of Openness to Experience and Neuroticism. With respect to the former trait, a positive relationship was observed between its level and the frequency of use of LLS in general and all their

in all the analysed cases, the Kruskal-Wallis test was applied in order to account for the significance in all the analysed comparisons. Significance values have been adjusted by the Bonferroni correction for multiple tests.

251 categories. With regard to the latter, a negative relationship appeared evident. The levels of the remaining traits: Conscientiousness, Extraversion, and Agreeableness demonstrate a clear pattern along with the frequency of strategy use with respect to the statistically significant findings, and the most frequent strategy users were also significantly more conscientious, extrovert, and agreeable than members of the two remaining clusters. The findings from the semi-structured interviews seem to be consistent with those described above. Firstly, relatively more advanced learners (B1, B2) did all refer to either the impact of Openness to Experience or its underlying traits. Secondly, a number of them insisted that they preferred learning in groups because of their extraverted personality or believed that they could easily engage into conversations with foreigners because social- ising with other speakers of English did not pose a difficulty for them. On the other hand, people who were not extraverted still did manage to prevent learning English from becoming a solitary experience. Some of them claimed that it was easier for them to apply instant messengers rather than use English in face to face conversations. Others indicated their preference for practising with their friends or classmates of their own choice rather over interacting with strangers. In one specific case, an introverted person explained that while she tended to opt for learning on her own, but did not really mind, and in fact, enjoyed, getting a chance to prove herself when travelling to the US and talking to Amer- icans. Finally, whereas a number of learners reflected on the importance of being conscientious in learning English, their self-perception in this respect was extremely critical. Some of them literally called themselves lazy and wished they had been more determined in their efforts to master the target language. Notwithstanding, it needs to be emphasised that according to the results of the cluster analysis, while a specific combination of traits can, indeed, result in a more frequent (and, possibly, more efficient) use of LLS, excep- tions to the rule are not rare, as illustrated by the number of cases beyond clusters.

5.3. Discussion

The organisation of the present section reflects that of the previous one in the sense that each of the following parts is dedicated to a specific research question. Subsection 5.3.1 thus involves a discussion of the findings concerning the relationships between the levels of the investigated students’ personality traits and the frequency of applying LLS as such

252 and particular strategy categories. While the impact of each of the ‘big’ five personality traits is discussed separately, some overlap is unavoidable in the specific case of social strategies, for which it is particularly challenging to distinguish between the interplay of underlying factors of Extraversion and Openness to Experience.

5.3.1. Personality and LLS use

The present subsection is dedicated to the relationships between the levels of personality traits and the reported frequency of using LLS in general and across categories. Relevant findings are discussed for each of the ‘big’ five personality traits, including Openness to Experience, Extraversion, Conscientiousness, Neuroticism, and Agreeableness, in the order that reflects the strength of their relationship. This is preceded with a preliminary section aimed to provide commentary on the investigated students’ strategic preferences.

Commentary on the investigated students’ strategic preferences

The investigated students demonstrated some consistent patterns with respect to the reported frequency of applying LLS. According to the results of both Pilot Study One and the quantitative part of the main study, compensation strategies were the most frequently employed scale by the analysed group of students. Similar findings were obtained by Yang (1993), who investigated a group of 500 Taiwanese students. At the same time, the results obtained by Magogwe and Oliver (2007) in a study of Batswana students were considerably different, which signals the importance of learners’ cultural background and the mainstream ELT methodologies in their countries. Specifically, in Poland there has been a shift from accuracy to fluency in terms of the (upper) secondary57 school-leaving exam criteria and hence, more learners might be more interested in conveying meaning even if they do not know a specific word or phrase or lack appropriate structures. The

57 The Polish educational system is currently being reorganised. At the time when the main study was conducted, everyone was required to attend a 6-year primary school and a 3-year lower secondary school. Afterwards, they could choose between an upper secondary school or a modern technical school if they wanted to go to college or university. However, in 2019 lower secondary schools ceased to exist while primary schools were extended to eight years. Also, upper secondary schools became secondary schools.

253 least frequently used strategies were affective strategies, which, again, might be due to cultural reasons. In an earlier study conducted on another group of Polish students, the author of the present dissertation also found that learners used affective strategies the least frequently, pointing to their lack of familiarity with strategies involving reward rather than punishment (Przybył 2016). In the main study, apart from compensation strategies, students reported using social strategies relatively more often than any of the categories of strategies belonging to the cognitive domain, that is memory, cognitive and metacognitive strategies. This finding corresponds to the previous one, and can be explained in a similar manner, i.e., through referring to the general tendencies in ELT methodology in Poland, emphasising the role of communication in learning English as a foreign language.

The relationships between learners’ personality traits and LLS

Referring to the relationships between the level of learners’ personality traits and their strategic choices, it needs to be emphasised in the first place that a number of findings signal the particular importance of Openness to Experience as a trait whose level may constitute a critical factor responsible for triggering strategy use on the part of the learner. As indicated by Dollinger (1996: 25), the domain is responsible for organising people’s views on a vast number of activities, as well as goals and preferences. More specifically, as stated by Costa and McCrae (1995: 31), one of the underlying facets of Openness refers to ideas, among which a high degree of intellectual capacity, valuing intellectual matters, and having a wide range of interests is extremely likely to facilitate strategy use. In view of the results of the main study, but also the results of Pilot Study One, Openness to Experience seems to trigger the use of the widest range of language learning strategies. Firstly, the comparisons of strategy use across groups of individuals whose levels of Openness to Experience were classified as low, medium, and high, according to the psychological interpretation (Zawadzki et al. 2010) indicated statistically significant differences for every single strategy category and for the measure of overall strategy use, that is GULLS. Highly open participants outperformed other participants in the use of memory, cognitive, metacognitive, compensation, affective, and social strategies, and the overall use of LLS measured as GULLS. In three specific cases of cognitive strategies, social strategies, and GULLS, the differences were also statistically significant between

254 the groups of individuals whose Openness levels were diagnosed as medium, and those whose levels of the trait were diagnosed as low, consistently indicating the positive relationship between the level of the trait and the frequency of applying LLS. To sum up, the level of the trait is likely to determine both the frequency and scope of learners’ strategy use. Secondly, the level of Openness to Experience accounted for a major proportion of strategy variance, as indicated by the results of the stepwise regression, and moderately correlated with the mean frequency of strategy use in general, and the mean frequency of metacognitive strategy use. These findings can particularly easily be linked to a whole group of underlying facets of Openness including creativity, imagination, and intellectual curiosity (Zawadzki et al. 2010). The link between these characteristics and the use of language learning strategies seems quite clear: a high frequency of using language learning strategies in general (GULLS) corresponds to learners’ active imagination (as manifested in the use of memory strategies, such as linking the sound of a word with its image in order to make oneself remember the word, or compensation strategies, such as creating new words), their attentiveness to inner feelings (as manifested in the application of affective strategies in general), preference for variety, consisting in using a number of different types of language learning strategies, strategies from each of the six categories, or three dimensions, and intellectual curiosity, consisting in trying out new strategies, and searching for strategies which fit a particular language task better; a high frequency of metacognitive strategy use involves making intensified efforts directed at maximising strategy use in general by, for example, using English more frequently, broadening one’s knowledge about strategies or looking for opportunities to communicate and read in English; it is an obvious conclusion that the use of these strategies can be more efficient if the person who employs them shows intellectual curiosity, exhibits preference for variety (e.g. through developing both, productive and receptive skills in English, or resorting to different modalities, e.g. reading in English or listening to or interacting with other speakers of English; also, the investigated students who were relatively more open to experience employed such strategies as imitating native speakers of English, watching films or programmes in English, reading in English for pleasure, looking for patterns in English, using English in as

255 many ways as possible, looking for opportunities to read in English, asking questions in English, and even learn about the culture of English-speaking countries. Thirdly, comparing the above results of the study to earlier results of studies investigating the relationships between LLS and learners’ levels of Openness to Experience poses a real challenge due to the fact that little attention has been dedicated to the problem. The impact of Openness to Experience on foreign language development is still only vaguely acknowledged in reviews of studies dedicated to the role of learners’ personality, while its impact on the application of LLS is hardly ever discussed. For example, Dörnyei and Ryan (2015: 25) described Openness to Experience as a trait whose level could directly affect the learning process as such, but at the same time supported Kappe and Van Der Flier’s (2012: 615) view, according to which Conscientiousness, rather than Open- ness to Experience, should be deemed as the strongest predictor of academic performance. A number of studies referred to in the final section of Chapter One can be used as evidence of the growing interest in the impact of the trait on certain specific aspects of foreign language development, including Willingness to Communicate, and various measures of performance (Ożańska-Ponikwia and Dewaele 2012; Pourfeiz 2015; Piechurska-Kuciel 2018). However, with the exception of a limited number of studies whose findings were not frequently referred to due to methodological inaccuracies, the links between the level of Openness and the use of LLS were virtually unreported. The frequency of applying social strategies demonstrated significant weak correlations with the levels of two personality traits, i.e., Openness to Experience, and Extra- version, which suggests a dual nature of these strategies. It is obvious that social trategies require interaction with other speakers of English; however, these speakers might be either native speakers or other learners/speakers of English, and the opportunity to use these strategies might arise either in naturalistic settings, or in the classroom. Perhaps these different settings correspond to different personality traits. Getting involved in classroom interaction or in interaction with one’s friends or acquaintances involves more characteristics typically associated with Extraversion, which, according to the Five-Factor Model (Zawadzki et al. 2010) involves such facets as being energetic, displaying positive emotions, assertiveness, sociability, or seeking stimulation in the company of others relate to interacting with other in general rather than interacting with strangers. Hence, those of the investigated students who were relatively more extravert made efforts to find interlocutors to hold conversations in English or asked questions in English. Entering

256 conversations with native speakers or other speakers of English might be a sign of not merely one’s sociability or willingness to thrive in the company of others, but also preference for variety, which is a component of Openness to Experience. On the other hand, one could link extroverts’ more frequent use of social strategies to the results of Griffiths’ (1991) study, which indicated that the mode of learning preferred by extroverts tended to be more discovery-driven, as opposed to the more reception-driven mode of learning preferred by introverts, and Oz’s (2014) study, according to which extraverted students demonstrated a greater desire to talk to strangers in a foreign language. The participants whose level of Extraversion was interpreted as high in terms of the standards provided by Zawadzki et al. (2010) were also significantly more frequent users of social strategies than the remaining participants, and significantly more frequent users of metacognitive strategies and LLS as such, measured as GULLS, than the participants whose level of Extraversion was psychologically interpreted as low. With regard to extroverts’ use of metacognitive strategies, the findings support those from Fayyaz’s study (2011), which revealed a positive relationship between learners’ level of metacognitive awareness with regard to listening skills and their level of the trait. It could thus be assumed that being an extrovert facilitates the use of social learning strategies, and cooperative strategies in general, which consequently strengthens learners’ overall strategy repertoire. These conclusions are supported by the existence of statistically significant correlation between learners’ level of Extraversion and the frequency of using social language learning strategies. Moreover, according to the results discussed in section 5.2.1.4 of the present chapter, Extraversion proved to be one of the significant predictors of strategy use in the stepwise regression model. Finally, the results of cluster analysis also position Extraversion as a trait whose high level, according to the psychological interpretation, corresponds to significantly more frequent use all categories of LLS. All these findings can be related to the results of earlier studies by Konttinen (1970), according to which extraverts are more likely to develop foreign language skills in general and the findings of Busch (1982) and Strong (1983), according to which extravert children are able to intake a foreign language than introvert children. They also correspond to insights from the studies conducted by Dewaele and Furnham (2000), according to which a high level of Extraversion facilitates oral production and the study by Dewaele (2004), according to which extraverts tend to use more colloquial expressions when having a conversation in a foreign language. Also, a high level of Extraversion was related to the use of

257 some affective strategies, such as using emoticons when communicating on the Internet or rewarding one’s own good performance when studying English, which confirms Grif- fith’s (1991) conclusion according to which extraverts show a stronger preference for rewarding their performance when learning than introverts. The beliefs about the usefulness of being extravert in learning English as a foreign language were expressed in the interviews, in which learners who considered themselves introverts expressed envy about the easiness with which extraverts interacted with others and wished they could employ more social strategies and collaborative strategies. It seems that in terms of the perceived usefulness of possessing a high level of the trait, not much has changed since the study conducted by Naiman et al. (1978), according to which 31% of good language learners directly linked the level of Extraversion to the acquisition of oral skills in a foreign language. At the same time, it is worth emphasising that the beliefs about the positive impact of Extraversion on foreign language achievement were empirically confirmed by Oya and et al. (2004), who demonstrated that teachers assessing learners’ performance form better global impression of extravert students, and Ockey (2009), who found that better scores in oral tasks correlated with students’ level of assertiveness (one of the underlying facets of Extraversion). At the same time, it needs to be remembered that a high level of Extraversion can manifest itself in seeking stimulation in the presence of others, and therefore may impede the use of cognitive strategies if they are not supported by other strategies, which provide learners with opportunities for interaction (outside the classroom) or group work, pair work, or whole-class discussion (in the classroom). Consequently, introverts may outperform extraverts in certain aspects of developing a foreign language, such as vocabulary acquisition, as indicated by Carell et al. (1996) and a high level of Extraversion can negatively correspond to learners’ overall language aptitude, as contested by Ockey (2009). On the other hand, extraverts may demonstrate greater lexical complexity when performing in speaking task, as found by Van Daele et al. (2006). Traditionally recognised as the trait of good learners, Conscientiousness was expected to facilitate frequent use of LLS as such as well as foster the use of strategy categories whose application particularly rests on learners’ laboriousness. The results of the main study partly confirmed these expectations. More specifically, a high level of Con- scientiousness was found to correspond to significantly more frequent use of memory and metacognitive LLS, positively correlate with the use of LLS in general, and serve as a

258 valid predictor of the variance in the frequency of strategy use. Referred to by Biedroń (2011) as the will to achieve, Conscientiousness does certainly play an important part in planning and organising one’s learning. It is also frequently viewed by students as a desirable quality of language learner, as voiced by the participants of the qualitative study. Curiously, a number of interviewees expressed their concerns about not being conscientious enough and wished they had dedicated more time to learning English before they had commenced studying at university. This finding certainly supports the results of Pour- feiz’s (2015) study, according to which the level of Conscientiousness can serve as a valid predictor of the affective-evaluative component of the Attitudes towards Foreign Lan- guage Learning Inventory developed by Vandewaetere and Desmet (2009). The impact on the use of LLS and foreign language development is far from unequivocally positive or negative. In the of interviews conducted for the purpose of the study, some of the investigated learners who viewed themselves as conscientious expressed their fears of making mistakes when they communicated in English in the classroom. This could be attributed to highly conscientious people’s inclinations towards perfectionism, which might not be facilitative in the specific case of foreign language development due to its affective implications. Moreover, the beliefs about the beneficial impact of a high level of Consci- entiousness on academic performance expressed by Dörnyei and Ryan (2015: 31) seem a bit far-fetched, as investigations of the impact of personality on attainment operationalised by Khodaddy and Zabihi (2011) as learners’ GPA in English, failed to confirm a positive relationship between the two variables. In addition, the results of studies investigating the links between the level of Conscientiousness and language learning aptitude are inconclusive. On the one hand, the outcomes of the study conducted by Fayyaz (2011) implied that the level of the trait positively correlated with the level of language aptitude, but on the other hand, no significant differences in the level of Conscientiousness were detected between the group of gifted and non-gifted learners by Biedroń (2011). Simi- larly, although some significant relationships were found between the level of the trait and the frequency of strategy use, these were limited to two specific categories rather than indicated a clear tendency. The strength of the relationship between Conscientiousness and strategy use was considerably lower than the strength of the relationship between strategy use and participants’ level of Openness to Experience and parallel observations could be drawn regarding the impact of the two traits in cluster analysis. At the same time, the cluster analysis in the main study indicated that Conscientiousness need not be viewed

259 as a ‘rival’ trait of Openness. In fact learners who were more frequent strategy users were also characterised by significantly higher values of both trait levels. However, the impact Openness was both stronger and more consistent across all clusters. One more time, it could be argued that foreign language learning differs remarkably from learning other subjects at school or at university and so do factors which determine learners’ attainment in these two fields, including language learning strategies. According to the results of the main study, the level of an individual’s Neuroticism influences the participants’ use of strategies in general, as shown in the stepwise regression model, but the impact is unsubstantial in comparison with the other personality traits, as indicated by the lowest standardised b value. Also, no statistically significant differences exist between frequent, average, and infrequent strategy users in terms of their Neu- roticism levels, as indicated by the Kruskal-Wallis test for independent samples (p > 0.05). At the same time, it needs to be remembered that individuals who are characterised by a high level of Neuroticism may suffer from maladjustment, have low stress tolerance, suffer from various types of anxiety, behave in a hostile way, or even develop depression (Magnavita 2004: 82), which may limit their learning ability. The results of the main study partly comply with these indications as some of the correlations between the frequency of use of affective strategies and the level of Neuroticism, although weak in terms of the strength of the relationship, confirm the nature of the relationship. For instance, learners who are relatively more neurotic tend to use English less frequently whenever they risk making a mistake (ϱ = -0.23, p < 0.01), and they more frequently notice that they become tense while learning and/or using English (ϱ = 0.23, p < 0.01). A negative reference to the impact of anxiety on performing as a language learner also appeared in one of the interviews. A student diagnosed with a high level of Neuroticism (sten 7) com- plained about being anxious when using English in the classroom, afraid of making a mistake, and noticed that anxiety impeded their comprehension and seriously limited their communicative ability. Perhaps one way to arrive at more specific results would be to investigate the relationships between the frequency of strategy use and the underlying facets of Neuroticism, that is anxiety, emotionality, or moodiness and linking them to self-perception (cf. Thomson 2016). As far as the impact of the level of Agreeableness on strategy application is concerned, it appears that the trait does not considerably affect the frequency with which language learners use any strategies. Firstly, no significant differences in strategy use

260 were found between learners with high, moderate, and low level of the trait. Secondly, its level did not significantly correlate with any of the SILL categories, and neither did it prove to be a significant predictor of the frequency of strategy application in the stepwise regression model. Although some correlations with the level of Agreeableness were observed for particular strategies, no consistent pattern in terms of the impact of the trait could be traced. In fact, it was a rather curious finding that the level of the trait weakly correlated with learners’ reported frequency of watching films and programmes or reading in English, but at the same time, one that could not easily explicable by merely considering the underlying facets of Agreeableness. When taken into account in the k-means analysis, the level of Agreeableness appeared to follow the pattern observed for Extraver- sion and Conscientiousness, that is frequent strategy users were more agreeable than rare strategy users. However, given the results of the analyses of correlations and stepwise regression, the impact of the trait on strategy use is considerably weaker than the impact of the other four ‘big’ pillars of personality.

5.3.2. The relationships between learners’ use of LLS and language attainment

In terms of the relationship between participants’ level of proficiency and their general frequency of applying LLS, the results of the quantitative part of the main study lend support to the outcomes of the studies conducted by Griffiths (2003), and Lai (2009), or Gerami and Baighlou (2011), according to which relatively more proficient learners employed language learning strategies significantly more frequently than their relatively less proficient counterparts. One particular similarity between the results of the former study and the results of the main study consists in the fact that statistically significant differences in the frequency of strategy application across levels of proficiency were found between the most and the least proficient language learners rather than between any two levels (specifically B2 and A1+, and B2 and A2 in the main study). It could be inferred that the development of the strategy repertoire is a complex process and perhaps it takes a certain amount of time for the progress in strategy use to become evident. Category-specific, statistically significant differences in the frequency of LLS application were identified between participants of main study with regard to the strategies from the cognitive domain, including both cognitive (between levels A1 and B1, A1 and

261 B2, and A2 and B2) and metacognitive (between levels B2 A2) strategies. This finding corresponds to a number of outcomes of earlier studies. To start with, Oxford and Ehrman (1995) traced a positive relationship between learners’ reported use of cognitive strategies and their end-of-term proficiency ratings. In another study, Gregersen et al. (2001) identified a consistent tendency of significantly more frequent application of cognitive, metacognitive, and social strategies among relatively more proficient learners. A similar outcome was reported by Lai (2009) who found that both cognitive and metacognitive strategies were significantly more often used by relatively more proficient language learners. It is worth emphasising that, similarly to the main study, Lai’s (2009) study involved a group of university students learning English as a foreign language. Several later studies, for instance research conducted by Gerami and Baighlou (2011), Salahshour et al. (2013), Kunasaraphan (2015), or Habók and Magyar (2018) confirmed the relatively more proficient students’ preference for metacognitive strategies over other categories. Although the correlational nature of the relationship does not allow researchers to simply conclude that frequent use of LLS is beneficial for language learning attainment, as asserted by Halbach (2000) and Pawlak (2008ab), it could be concluded that the frequent use of strategies, especially within the cognitive domain, typically constitutes the characteristic of high, rather than low-level learners. This claim is supported by the results of one-way ANOVA and its non-parametric equivalent, the Kruskal-Wallis test, according to which the dependent variables including the frequency of LLS use in general (GULLS), and the frequencies of students’ use of cognitive and metacognitive strategies are significantly higher in groups of relatively more advanced students. While certain studies, such as those conducted by Phillips (1991) or Magogwe and Oliver (2007) provided evidence of some inconsistencies regarding the positive relationship between learners’ attainment and their overall use of LLS, no studies exist, to the best knowledge of the author of the present dissertation, which would undermine the consistent, positive relationship between language learners’ level of proficiency and their application of LLS belonging to the cognitive realm. Finally, the results of the main study which link the frequency of strategy use to learners’ level of proficiency can be related to the milestone study by Griffiths (2003) which aimed to pinpoint the strategies used by particularly successful language learners. In the study Griffiths (2003) identified a number of both cognitive and metacognitive strategies in the ‘plus’ category of strategies whose application differentiated successful

262 learners. Specifically, the metacognitive strategies which Griffiths incorporated in the ‘plus’ category included efforts oriented towards maximising the opportunities of either communicating in English or reading in English, as well as strategies allowing learners to utilise available resource, and strategies aimed at managing one’s own feelings and emotions (Items 35, 36, 39 and 40 in Oxford’s (1990) original SILL ver. 7.0 questionnaire). With reference to the “PLUS” cognitive and metacognitive strategies identified by Griffiths (2003), it was observed that lower-level learners used them less frequently than higher level learners. Mean frequencies regarding looking for interlocutors amounted to 2.21 for A1+ students, 2.29 for A2 students, 2.48 for B1 students and 2.63 for B2, and similarly, the values were 2.11, 2.16, 2.49 and 2.72 respectively for looking for reading opportunities and 1.81, 2.13, 2.41 and 2.80 for extensive reading. No clear-cut pattern was, however, observed for trying to relax when tense, whose mean values amounted to 2.26, 2.82, 2.75 and 2.69 for the four CEFR levels respectively. The discrepancy is an example of the irregularities referred to by Oxford (1996) with reference to the patterns in the frequency of using affective strategies. It is unclear whether learners really need them after they have achieved a certain level of proficiency or if their greater use is conditioned by language learning anxiety (in other words, students who do not experience any type of anxiety while learning English need not use these strategies).

5.3.3. The relationships between learners’ personality and attainment

Regarding the relationship between the levels of participants’ personality traits and their level of proficiency in English, the main study is based on the assumption that investigating the link between personality and language learners’ level of proficiency is not (and ought not to be) limited to tracing the relationships between learners’ personality and various measures of proficiency. Instead, it also involves a much broader range of individual factors, such as learners’ temperament, learning styles, or language learning strategies. Although the results of the quantitative part of the main study did not reveal statistically significant relationships between the levels of learners’ personality trait and the levels of their proficiency in English, the impact of learners’ personal characteristics on the process of foreign language development was referred to by the interviewed students. The learners themselves clearly recognised the particular role of Openness to Experience

263 and Conscientiousness. In this way, the present investigation corroborated the results of earlier studies assigning a particular role the levels of these traits in educational achievement, such as those conducted by Chamorro-Premuzik (2006) and Bratko et al. (2006). At the same time, since the relationship failed to be directly explained by any statistical procedure, it could be stipulated that some other factors must mediate between the two constructs, as inferred by Boekaerts (1996) and De Raad and Schouwenburg (1996). The result of the main study therefore also support to the assertion made by Ledzińska and Czerniawska (2011: 158) that a high level of Conscientiousness is likely to increase the degree to which learners fulfil their duties as participants of different courses, and that a high level of Openness to Experience can positively affect students’ willingness to engage in educational challenges. The process of learning a foreign language cannot be explained by treating personality traits as sole predictors of learners’ behaviour, performance, or even strategy use. The final construct constitutes one of the pillars of learner autonomy, which, according to Macaro (2008: 60), is shaped by both psycholinguistic and sociolinguistic factors, and manifests itself in the learner’s ability to make strategic decisions. As far as the sociolinguistic factors are concerned, Schmenk (2005: 107) expressed a belief that learner autonomy was not a homogeneous construct, but, instead, subject to “the heterogeneity of local environments arising from interaction and blending of local and global practices” and advocated for a compromise between globally recognised and locally experienced needs of autonomous learners. With regard to psycholinguistic factors, the influences of personality on learner’s strategic choices must not be neglected. The findings of the quantitative part of the analysis suggest the existence of statistically significant relationships between the use of language learning strategies and learners’ personality traits. Given the simultaneous lack of significant, direct relationship between learners’ personality traits and their level of competence, and the significant relationship between learners’ level of competence and their use of language learning strategies, it is possible to assume that language learning strategies, perhaps among other variables, mediate the way that learners’ personality affects their level of language competence.

264 5.3.4. Clusters of LLS users

Aiming to identify clusters of language learners formed on the basis of their personality traits and reported frequencies of LLS application, the final part of the main study brought results that could partly be interpreted as a confirmation of findings in the fields of educational psychology and the psychology of the language learner. However, at the same time, it also expanded the perception of the foreign language learner by taking into account all their ‘big’ five personality dimensions and their complete strategy repertoire. The members of the three clusters differed significantly from each other in virtually all aspects which were analysed. In particular, the members of cluster three applied all categories of LLS significantly more frequently than members of the two other clusters and were significantly more open to experience, extravert, conscientious, and agreeable than members of the two other clusters, and less neurotic than members of cluster two. Also, no significant differences in the level of Neuroticism existed between members of cluster one and cluster three. These findings imply that, firstly, the analysed personality traits can work in a complimentary rather than substitutive manner. In other words, the fact that a given language learner is a frequent strategy user results from a combination of being relatively open to experience and conscientious rather than being either open or conscientious. A typical frequent strategy user is also relatively more extravert and more agreeable than a typical average or rare strategy user. Relating the assumptions presented above to former studies poses a challenge as, traditionally, the profiles of foreign language learners did not use to be constructed on the basis of the levels of the learners’ personality traits and rarely, if ever, on the basis of their reported use of LLS. Historically, some theoretical concepts of learner-specific variables in foreign language development did not explicitly include either of these characteristics and in this respect, the study might be regarded as innovative. However, beliefs concerning the impact of some underlying facets of the ‘big’ five personality traits have been present in publications on foreign language learning for at least several decades. In particular, the results of the main study lend partial support to Jankowski’s (1973) views, according to which imagination (a component of Openness to Experience) and ability to concentrate (a component of Conscientiousness) are both likely to facilitate learning a foreign language and people’s imaginative capacity can be linked to their efficiency at remembering details along with the ability to plan the learning process (metacognitive

265 strategy). Moreover, they could prevent learners from experiencing fatigue or boredom and losing concentration. More recently, profiles of foreign language learners have been created on the basis of their belonging to a particular generation and degree to which they reflect the characteristics of good language learners (Bozavli 2016), their level of linguistic aptitude (Michońska-Stadnik 2017), their level of proficiency during the period of studies (Kucharczyk 2017) or expertise in terms of new technology (Jedynak 2017). While neither of these profiles suffices to describe the foreign language learner at a 21st- century university, each of them, along with the profile created on the basis of learners’ personality traits and reported use of LLS, adds up to the holistic picture of the foreign language learner.

5.3.5. Limitations of the study

While the methodology applied in the main study has contributed to providing answers to all the research questions raised in Chapter Four of the present dissertation, some improvements could be considered in future investigations of the relationships between learners’ personality traits and their reported use of LLS on the basis of the encountered difficulties or experienced limitations. Generally, it could be asserted that an attempt to account for individual factors influencing university EFL students’ strategic preferences remains a challenge since efforts to generalise findings over entire university populations require large-scale studies. These, in turn, especially in the case of questionnaire-based studies, involve the risk of participants’ being unable to report their feelings, behaviours or thoughts through adequate responses as signalled by Atieno (2009). The statement seems adequate when addressing the field of individual variation in foreign language development with respect to reliance on questionnaires in general and in the specific case of applying the adaptation of SILL ver. 7.0 (Oxford 1990), being one of the two quantitative instruments applied in the main study. Some criticisms concerning the allegedly insufficient theoretical background or potential psychometric weaknesses of the latter (cf. Dörnyei 2003; Tseng et al. 2006; Rose 2015) have been referred to in section 4.4.3.1 of Chapter 4. It could, perhaps, be added that even after introducing a number of amendments to the questionnaire, it remains an instrument created about three decades ago and its substantial benefit consisting in the opportunity of comparing study results with a

266 major amount of other SILL-based research, it may gradually cease to be a sufficient representation of the universe of content connected with contemporary language learning strategies. Although Oxford (2017) advocated for a major shift in strategy research, from the investigations of general tendencies to studies dedicated to contextualised use of LLS with respect to specific skills or tasks, it would be a premature conclusion to discontinue large-scale quantitative research. Similarly, as emphasised by Rose (2012), abandoning studies into LLS for the sake of investigating learners’ self-regulation appears to be pre- cipitous if not reckless. Instead, while continuing the investigations aimed to account for language learners’ self-regulation, attempts could be made to improve the degree to which strategy questionnaires meet the goodness criteria for psychometric tests. In particular, attempts should be made to improve the reliability of the affective strategy scale since strategy researchers themselves have recognised the need to investigate learners’ affective states in greater depth (cf. Oxford 2015; Pawlak and Oxford 2018), and reported relatively low reliability of the affective scale of the SILL ver. 7.0, especially in comparison to other strategy scales (cf. Hsiao and Oxford 2002, Castillo and Córdova 2014, Przybył 2016, 2017). The values of the reliability coefficient, Cronbach’s alpha, were also rather low when calculated for the data sets in the main study even though the scale had been subject to modification on the basis of the conclusions from Pilot Study One. Efforts should therefore be continued to provide a more adequate representation for the affective scale of LLS. However, perhaps instead of addressing a panel of experts again, students’ reflections could be shared in interviews or journals so as to allow for a more concatenative approach as understood by Ellis (1994, 2008). Although the main study involved a qualitative component for triangulation purposes, the distribution of the levels of personality traits in the interviewed group of participants was not even. Conducted in a group of volunteers, the interviews were, indeed, characterised by a certain degree of intimate familiarity between the interviewer and the participants of the study, some of whom were former students of the interviewer. There- fore, on the one hand, the students participating in the series of interviews were less likely to experience inhibitions or encounter difficulties in reflecting on their language learning strategies or personal characteristics, but on the other hand, they may have been interesting in being perceived in a positive, rather than realistic, manner. While the design of the interview schedule was aimed to counteract the practice of enhancing a biased self on the

267 part of participants, it cannot be ruled out that some reports of learning practices were exaggerated or biased.

5.4. Conclusion

A number of conclusions can be drawn from the results of the main study, referring to the investigated students’ strategic preferences, the links between LLS and learners’ personality traits, the relationship between both of these variables with the level of proficiency, and, finally, the outcomes of the cluster analyses. To start with, the investigated students’ overall preference for social and compensation strategies indicates that they tend to rely on collaboration and are not easily discouraged from learning or using English even when they have to cope with scarce resources as foreign language users. This is reflected in the most frequent use of two strategies: When I do not understand, I ask the speaker to slow down, repeat, or clarify what was said and When I cannot think of the correct expression to say or write, I find a different way to express the idea; for example I use a synonym or describe the idea. Since successful strategy training clearly depends on adopting a learner-oriented, interactive, and communicative approach by EFL teachers, as suggested by Droździał-Szelest (2008: 414), it could be argued that the investigated university students would be willing to embrace further strategic instruction. This assumption is supported by the finding according to which participants’ strategy repertoire already indicates preference for communication-based and interactive learning. It could be also inferred that the affective dimension of foreign language learning is neglected by university students, as approximately 40% of the participants proved to be rare users of these strategies. In particular, the students tended not to share their feelings about learning English as a foreign language. Moreover, the role of reflecting on one’s feelings about learning English as a foreign language seems underappreciated by the investigated students, with regard to both spoken and written communication. These findings seem to partly contradict a former one, consisting in the considerably greater popularity of social strategies. However, the very use of social strategies does not require students to handle emotions or reflect on their beliefs about foreign language learning or relevant attitudes, while the use of affective strategies certainly does. The difficulties experienced by language learners in the use and orchestration of affective strategies might

268 therefore boil down to the lack of recognition of the role of affect and self-reflection in learning a foreign language rather than their unwillingness to share their feelings as such. In other words, it appears that the act of sharing them becomes considerably more challenging whenever one’s own feelings about their language self are at stake. The scarcity of frameworks which conceptualise affect might be one of the reasons to blame, as both EFL teachers and researchers are not typically educated as experts in the area, and thus, may have to rely on intuition or reach out for available schemes. These, on the one hand, could shed some light, provide a theoretical basis, or even be embraced by some foreign language learners and teachers, but on the other hand, they are likely to be subjective, if not biased. While emotional well-being has been empirically proven to correlate with language proficiency in robust studies based on large samples of language learners (cf. Ehrman and Oxford 1995; Piechurska-Kuciel 2008), it is hard to disagree with the view expressed by Kębłowska (2012: 157) that contemporary linguists have often neglected the area, especially in comparison to other fields of learner-focused investigations. In- deed, a number of handbooks of second language acquisition or foreign language learning do not include any consideration of the role of affect in foreign language development, or merely provide an explanation of Krashen’s (1981a) concept of affective filter (cf. Doughty and Long 2003; De Bot et al. (2005); Van Pattern and Williams (2007), Kaplan 2010). Also, whereas adopting a certain framework for investigating emotions is inevitable, the current status quo, as summarised by Panksepp (2008) rejects both behaviour- only and cognition-only-based visions of emotions. This view is at least partly shared by Dörnyei (2009), according to whom emotions are inseparable from cognitive appraisals as the latter have not only a cognitive, but also an evaluative and interpretative function. To make matters even more complex, a recent theory of emotions views them as a developmental phenomenon, constructed by individuals on the basis of their idiosyncratic experiences, and subject to both physiological and environmental influences (Barrett 2017). Reflecting on the links between the investigated students’ personality traits and their strategy repertoire, the results of the present dissertation partly support the view expressed by Gerrig and Zimbardo (2009), according to which the Five Factor Model of personality allows predicting academic achievement, though not always in a straightfor- ward manner. A number of direct references to the links between students’ personality traits and their level of language proficiency in English were found in learners’ reflections on the facilitators of developing their competence in English as a foreign language in the

269 interviews conducted as part of the main study. However, these verbal reports do not directly correspond to the results of the quantitative part of the main study given the lack of statistically significant correlations between the investigated students’ level of proficiency in English and the levels of their personality traits. At the same time, when treating LLS as a latent variable, mediating between language learners’ personal characteristics and their level of proficiency, a number of parallels can be drawn to the findings of earlier studies linking personality traits to educational attainment (Ledzińska and Czerniawska 2011; Dörnyei and Ryan 2015). Also, a number of findings from the main study can be referred to functional implications58 of personality traits described by Strelau and Zawadzki (2008: 810). Firstly, the positive impact of high levels of Conscientiousness on educational achievement and performance, although not to be denied, needs to be reconsidered in the specific case of foreign language learning at university settings, particularly with respect to strategy use. First of all, the strength of the relationship between the levels of Consci- entiousness and strategy use is exceeded by that between strategy use and Openness to Experience, which constitutes an extension of the observations by Chamorro-Premuzik (2006)59. Secondly, high levels of Openness to Experience are becoming more and more widely recognised as a desirable trait for students, particularly at university level (Gray and Watson 2002), and by the students themselves as facilitative in foreign language development, as can be inferred from the interviews conducted as part of the main study. While no statistically significant correlation was found between the level of Openness and the investigated students’ level of proficiency in English in the main study, the use of language learning strategies was significantly influenced by the level of the trait, and statistically significant differences existed with respect to the level of Openness between frequent, medium, and rare strategy users with respect to all the investigated strategy categories and overall strategy use. This highlights the importance of Openness in tertiary education and, for a following time, supports the distinction between the predictors of academic performance and achievement in tertiary education in general and regarding foreign languages. Along with the results of previous studies linking the level of

58 A functional interpretation of the level of an individual’s personality traits consists in accounting for the role of personality traits in the process of the individual’s adaptation to environmental requirements (Zawadzki et al. 2010: 68) 59 According to Chamorro-Premuzik (2006), the role of Openness to Experience in education increases with each educational stage and becomes equal to that of Conscientiousness at university level.

270 Openness and or its underlying traits to variables which directly condition success in foreign language development, such as language aptitude (Tyszkowa 1990) or willingness to communicate (Piechurska-Kuciel 2018), the results of the main study rank Openness as the most vital trait for foreign language development. While the impact of Conscientiousness was proved to be the best predictor of academic achievement, particularly operationalised as grades in a course, it was found to be superseded by the impact of Openness in the case of academic performance in specific tasks. Also, much as foreign language learning remains an easily distinguished area of education, it certainly shares some features with linguistic education in general at the same time. Perhaps the role of Openness in developing competence in a foreign language can be understood more holistically after considering the results of the study conducted by Schwartz et al. (2013). According to its results, Openness to Experience is the trait whose level can best be predicted on the basis of language use, involving such characteristics in the case of highly open individuals as markers of aesthetic interests, misspelt words or contractions in the specific case of online written contributions. Finally, cluster analysis revealed that in terms of personality traits, the most frequent strategy users investigated in the main study tended to be both highly conscientious and highly open to experience. At the same time, they were also emotionally stable rather than neurotic and extravert rather than introvert, which leads to the conclusion that the potentially welcome traits can be analysed in clusters and their nature, does not competitive, but complementary in the specific case of foreign language learners. This finding supports Dörnyei and Ryan’s (2015: 28) assertion, according to which “successful language learners can combine their personality features to best effect by utilizing their specific strengths and compensating for their possible weaknesses in adjustment to the particular learning environment”, but at the same time provides positive evidence for Costa and McCrae’s (1998) view of the trait-dependence of educational, learning and instruction processes. Specifically, learners with high levels of Conscientiousness and Openness to Experience can be portrayed as good students thanks to their diligence and aspiration level to excel as opposed to individuals whose levels of the traits are low, characterised as reluctant scholars.

271 Conclusions, directions for future research and pedagogical implications

Increasingly adopted in the area of foreign language development, the assumptions of humanistic education include the recognition of the particular role of psychological knowledge and its beneficial effects on the learning process. According to Oxford (2011, 2017), person knowledge constitutes one of the pillars of strategic knowledge and reflection is a vital component in learners’ strategic self-regulation. Both the rationale for conducting the main study and its results remain fully consistent with these claims. Seeking to embrace the opportunities provided by the employed introspective techniques, the author of the present dissertation strongly believes that elaborating on learners’ application of LLS and linking it to their personality traits can contribute to the personalisation and, more broadly, individualisation, of the educational process. Moreover, it might also facilitate the process of teachers’ preparation for classes and make classes more effective. Finally, it could contribute to the development of learner’s autonomy provided that the learners can be skilfully guided by their teachers (cf. Ellis and Sinclair 1989; Esch 1997; Michońska-Stadnik 2008). A considerable number of results obtained in the main study confirmed that LLS use can be at least partly accounted for from the perspective of learners’ personality traits. To begin with, a number of statistically significant relationships were observed as regards the frequency of strategy use by learners characterised by high, medium, and low levels of particular personality traits. Among them, Openness to Experience turned out to be the trait whose level significantly influenced the frequency the frequency of each of the six strategy categories (memory, cognitive, compensation, metacognitive, affective, and social) and the overall frequency of learners’ use of LLS. The levels of Extraversion were found to differ significantly across group of rare and frequent strategy users in general, but also with reference to the categories of metacognitive and social strategies. The final trait whose levels proved to be meaningful for the frequency of strategy use was Consci- entiousness, whose level mattered particularly with regard to learners’ use of memory and metacognitive strategies. The importance of Openness to Experience for strategy use was confirmed in the analysis of correlations. The level of the trait was found to moderately correlate with the overall frequency of LLS use as well as the frequency of learners’ use

272 of metacognitive strategies and both relationships were statistically significant. Moreo- ver, it was also the level of Openness which was proved to account for the highest proportion of strategy variance in comparison with the levels of the other ‘big’ four personality traits. At the same time, except for Agreeableness, the other traits were also significant constituents in the stepwise regression model which ultimately explained 16% of variance in strategy use. In terms of linking the investigated variables to learners’ proficiency, the obtained results differed considerably. In the case of LLS use, the results of ANOVA and its non- parametric equivalent, the Kruskal-Wallis test, indicated that the frequency of LLS use was significantly different across groups of learners distinguished on the basis of their attainment, operationalised as CEFR levels. According to the observed tendency, frequent strategy use was characteristic of more proficient learners. This remained true not merely for the overall reported use of LLS, but also for the categories of cognitive and metacognitive strategies. On the other hand, personality traits were not found to have a direct impact on learners’ attainment. However, their relevance for language development was emphasised by the language learners themselves in virtually all the conducted interviews. Their participants recalled the benefits of being open and extravert on multiple occasions and generally perceived Conscientiousness as a highly beneficial characteristic. Although the source of the investigated students’ preference for social and compensation strategies remains debatable, their wide use does indicate their relative popularity with a considerable proportion of students. It could also suggest that learners become familiar with them before they start their education at university. One clear implication for foreign language teachers at universities consists in the noticing the opportunity to make use of different types of both compensation and communication strategies in foreign language classrooms at universities. It appears that since the students themselves at least try to cope with their L2 shortcomings, no need exists to be afraid of insisting on communicating in the target language instead of resorting to students’ mother tongue. Moreover, it could be presumed that interacting with students outside the classroom, for instance in emails or online course forums, could largely rely on communication strategies which favour the use of L2, such as circumlocution or encouraging clarification requests, rather than those minimising the use of L2, such as language switch or avoidance.

273 The infrequent use of affective LLS by the investigated university students could, first of all, be deemed as worrying as affect exerts a powerful influence on virtually all key indicators of success in foreign language learning and teaching and it is believed that recognising one’s affectivity is vital for such affective personality factors as self-confidence, self-esteem, and attitudes (Gabryś-Barker 2010: 44). Without acknowledging that the success of learning additional languages is emotionally driven just as success in any other area which one attempts to master (Schumann 1975), and without reflecting on one’s emotional state and well-being, a language learner may never succeed in exercising control over self-referential judgements, experience low situational self-esteem and anxiety and struggle with their own self-efficacy beliefs (cf. Oxford 2002). Given the importance of affectivity, which, according to Gabryś-Barker (2010: 45) “functions as a stimulus for action and type of approach taken as well as a monitor and controller of cognitive processing, grounded in an individual learning situation”, investigations of affect are by all means expected and welcome. Indeed, affect was listed by Grabe (2002: 12) among the variables which could be the source of language problems and it has become an area of growing scientific interest in the 21st century. Recently, Pawlak and Oxford (2018) emphasised the need for further studies investigating the role of affect in foreign language development, particularly referring to the strategy domain. It seems justified to expect that these studies will also include the development of new instruments dedicated to the investigation of the use of affective strategies, whose measurement proved to be particularly prone to error and challenging in terms of meeting the criteria of sufficient reliability and validity. In fact, the availability of tools designed to measure people’s emotional states is growing with reference to both online self-reporting instruments (cf. Betella and Verschure 2016) and those developed by neuroscientists (cf. Dieckmann and Unfried 2014). Perhaps some of them could be used in revisiting the scope of affective learning strategies. A number of implications arise with regard to the outcomes of the main study concerning the relationships between language learners’ personality traits and their application of LLS. First of all, it needs to be acknowledged that this area of investigation constitutes another field which distinguishes foreign language development from learning a content subject. On the one hand, it is commonly agreed that the level of Conscientious- ness can boost one’s academic performance and the trait is believed to be main the predictor of overall achievement across all levels of education (cf. Kappe and Van Der Flier

274 2012; Dörnyei and Ryan 2015). On the other hand, the results of the main study fail to grant role of utmost importance to the level of the trait in either its relationship to the level of attainment in English or its relationship with the frequency with LLS use. Also, the strength of correlational relationships between Conscientiousness and both LLS use and attainment is relatively lower than in the case of Openness of Experience. Moreover, the impact of being open to experience on the application of metacognitive strategies is the most evident as shown by the results of the main study. Since the level of Openness was shown to account for the largest proportion of the variance in strategy application in comparison to other personality traits and its high level was found to positively affect both the frequency of strategy application in general and the versatility of the strategy repertoire, the potential of Openness for facilitating strategy use certainly needs to be acknowledged. While it has been established that the levels of the ‘big’ five traits are, to a large extent, biologically determined by such factors as genes, sex, aging, and brain patholo- gies, as summarised by Costa and McCrae (2017: 28), it has also been proved that the level of Openness can be influenced by cognitive training. The results of longitudinal research conducted by Jackson et al. (2012) indicated that the level of Openness increased as a result of cognitive training implemented as a 16-week programme in inductive reasoning accompanied by weekly practice of doing crosswords and solving Sudoku puzzles. Although no similar results have been obtained in a group of young adults so far, a similar development cannot be ruled out and a study which would examine the effects of cognitive training on foreign language learners’ level of Openness to Experience would certainly benefit all the participants of the language learning process. It appears that elements of cognitive training could well be introduced to the EFL classroom. Firstly, it would be technically and economically feasible to accommodate elements of cognitive training even within a single term of a foreign language course, which generally lasts for at least 15 weeks at Polish universities. Some elements of such training could, indeed, involve crosswords puzzles which are quite commonly included as vocabulary practice activities in EFL books and supplementary materials for teachers, for either individual or pair work. Preference for the latter option seems reasonable as it enhances the use of both social and compensation strategies, which are generally relatively more willingly used by university students than any other strategy categories according to the results of the main study. Typically, a crossword done in pairs requires a student to explain a word to their partner

275 who knows how many letters the word consists of and ask for clarification if they cannot guess the word explained by their partner. Of course, cognitive training could go far beyond the practice of doing crosswords and embrace the already referred-to strategies of inferencing, in the sense that Bialystok (1978: 78) defined them, that is understanding them as strategies which enable language learners to arrive at certain new linguistic information. More generally, it is highly probable that foreign language learners at universities would benefit from the application of guided discovery in the sense that Marton (1988: 123) framed it, that is involving testing hypotheses about the target language and rule formation taking place in learners’ minds rather their passive absorption. Turning the language learning process into a series of discoveries assisted by the teacher seems tempting, but it is certainly not a new idea. On the contrary, it can easily be related to two widely acknowledged theoretical models. The first one, proposed by Vygotsky (1986) assumes involves building the necessary scaffolding in the development of learners’ zone of proximal development. The second one can be associated with the cognitive approach to language learning, recognising the benefits of integrating the components of communication, that is internal, behavioural, and external representations, by actualising the potential of information, as suggested by Dakowska (2013). The implications concerning the relationships between the levels of the remaining investigated personality traits and the frequency with which learners used LLS are by no means as clear-cut as those concerning Openness to Experience. Moreover, to the best knowledge of the author of the present dissertation, no studies exist which would prove evidence that the levels of the other ‘big’ five personality traits can be modified. On the other hand, the finding that highly conscientious and extraverted students tend to be significantly more frequent strategy users in general and with reference to specific categories cannot be simply ignored. The fact that high levels of either Extraversion or Conscien- tiousness can positively affect the frequency of applying memory, metacognitive, and social strategies, and thus boost learners’ overall strategy use, could be useful for both foreign language learners and their teachers. Whereas an individual cannot really influence the level of their personality traits, with the exception of Openness described above, the theoretical foundations of LLS have always emphasised the volitional aspect of strategy use and the extendibility of one’s strategy repertoire (Cohen 1998, 2014; Oxford 1990, 2011, 2017). Becoming aware of one’s personality traits can therefore serve as a starting point for reflecting on how they affect the process of developing a foreign language in

276 one’s particular case. Specifically, individuals whose levels of Conscientiousness or Ex- traversion are low might be interested in reviewing a list of LLS, such as the one provided in the SILL ver. 7.0 by Oxford (1990), in order to compensate for the potential deficiencies in strategy use induced by their personality characteristics by resorting to alternative strategies, which they might not have been aware of before. Although a certain degree of expertise in the field of human personality appears inevitable here, the use of specialist psychological tests which require the assistance of a professional psychologist (C category according to the Psychological Test Laboratory of the Polish Psychological Associ- ation) is not indispensable. First of all, one cannot easily dismiss the importance of observation in the specific case of foreign language teachers. Experienced teachers can easily notice at least some of their learners’ personal characteristics, including their cardinal traits, which, by definition, are reflected in a person’s daily behaviour and their passions or priorities. To use Allport’s (1961) nomenclature, they might identify their learners’ cardinal traits, and, possibly, their central traits as well. Moreover, both the learners themselves and their teachers can freely use other tests, not restricted to psychologists and it once encouraged, students may be willing to reflect on their results. One example is TIPI, a ten-item personality test, which only consists of ten questions, as well as a key, and is available both in English and in Polish (Sorokowska et al. 2014). Alter- natively still, teachers may simply stick to one of a number of personality tests widely available in English coursebooks and carefully analyse the results of these tests provided students agree to share them. In either case, the necessary condition, which consists in the inclusion of personality as one of the topics referred to in the curriculum, is met in a number of courses. Having that in mind, teachers may, and perhaps, should, adjust the strategy instruction and not only teach strategies, but also familiarise learners with other pillars of person knowledge (Oxford 2011). In such a way, they could also recognise the holistic character of both the language learner and LLS (Oxford 1990, 2011, 2017). It appears that emphasising the conscious character of strategies could benefit the use of both the strategies which are employed naturally, as part of learners’ preferences induced by their personality characteristics, and those which could only be chosen by learners after providing explicit instruction by the teacher. The results of the main study linking the frequency of applying LLS to the investigated students’ level of proficiency (RQ 2) provide some further information on the nature of the relationship between the two variables. On the one hand, the observed trend

277 seems clear, and it can be summarised that according to the results of the main study, relatively more proficient learners are typically frequent strategy users while their less proficient counterparts use LLS relatively less frequently. On the other hand, while the trend is consistent for all proficiency levels, statistically significant differences only concern levels which are not adjacent (e.g. A1+ - B2 and A2 - B2 for GULLS or A1-B1, A2- B2, or A1-B2 for cognitive strategies). It could thus be argued that the differences in the frequency of strategy use evolve gradually rather than rapidly while the level of language proficiency increases. It could therefore be assumed that since learners’ progress is of evolutionary rather than revolutionary, it could be induced by a gradual increase in the efficiency of app applying the strategy repertoire by language learners. This assumption is supported by the outcomes of the comparative analysis of the results of the main study and the results of the study conducted by Griffiths (2003) in order to identify the ‘plus’ strategies characterising good language learners. According to the analysis, the use of both cognitive and metacognitive strategies appears to be a crucial marker of good (understood as more proficient) language learners. It is neither easy nor obvious to suggest how to translate the above findings into the everyday practice of learning and teaching English as a foreign language. However, it seems clear that students whose levels of some key personality traits, including Open- ness to Experience and Conscientiousness, are low according to psychological interpretation, are particularly in need of assistance from their teachers. This assistance could start with providing a scaffolding for the process of developing their level of proficiency in English, involving both stimuli that would encourage students to challenge themselves intellectually, and a framework which would compensate for the deficiencies in the level of Conscientiousness. The former could then include a set of activities addressing various categories of students’ motivation, especially as its intrinsic component is known to distinguish high achievers from underachievers (cf. Vefali 2008). Also, given the fact that the perception of motivation has shifted from a constant ID to a dynamic variable as a result of studies conducted in the classroom context (cf. Pawlak 2012; Waninge et al. 2014), EFL teachers can, indeed, feel empowered in their attempts to boost students’ interest in the lesson and actively stimulate the level of learners’ involvement in classroom activities. Popular activities intended to diversify the types of interaction in the classroom, and make the learning process more stimulating, involve language games, such as board games, card games, quizzes, crosswords and puzzles, dominoes, bingo or role plays, but

278 even activities which may seem to constitute plain and standard elements of the language classroom can boost learners’ intellectual activity and facilitate guided the already mentioned guided discovery provided that they do not become monotonous, and are well- planned ahead. Along with the authors of resource books for English teachers, for instance Seymour and Popova (2005) or O’Dell and Head (2003), the author of the present dissertation strongly believes that a good variety of such classroom activities as simple categorisation tasks, activities involving information gap, spotting differences, completing grids, or even guessing words by mimicking or defining them, can all positively affect learners’ interest and stimulate their motivation to face intellectual challenges. Since learners differ considerably with respect to the level of their personality traits and other IDs, it becomes evident that not only are teachers’ positive attitude and conviction of vital importance, but also the tasks as such need to vary enough so as to cater for the needs of as many learners as possible. The group of students who are likely to require considerably more assistance than their counterparts as low strategy users are students characterised by low levels of Con- scientiousness. It seems that in their specific case the focus of assistance which could be provided by foreign language teachers should, in the first place, include guidance on organising and controlling the language development process. In fact, it might resemble the support suggested for students with psychological or cognitive deficiencies during a series of trainings conducted by the AMU Office for Students with Disabilities in the two last academic years. It could be argued that the set of principles described below could benefit not only students with specific intellectual deficiencies, but also all participants of the language classroom. They could also specifically prove very helpful for students with Asperger’s syndrome, students anxious about failure, students with excessively high levels of perfectionism, those suffering from concentration loss, dyslexia, bipolar disorder, or schizophrenia. The principles include: providing a clear introduction of the structure of the language course; making demands transparent and predicable; looking for ways of reducing students’ stress levels; carefully approaching the issues connected with group integration; guaranteeing access to information through planning resources; ensuring good communication.

279 It is crucial to teach strategic skills to those individuals who are infrequent strategy users, as their low levels of Openness to Experience or Conscientiousness may result in a limited strategy repertoire and dependence on readily provided tactics rather than be a factor contributing to building learner’s autonomy. More generally, it could be argued that EFL teachers’ cannot confine themselves to merely planning and implementing the teaching process. What is an indispensable part of this process is the need analysis of the participants as well as reflection dedicated to the stimulation of their development as whole persons. This is especially necessary for individuals who are relatively less open to experience and display low intellectual curiosity, as possessing these traits can result in applying a limited strategy repertoire and, therefore, in serious constraint of learner’s autonomy. Finally, teachers working with young adults may consider explaining the mechanism of self-regulation to students as, given the characteristics of young adults, the latter are able to reflect on their competence as language learners, and their autonomy. Such actions could improve the recognition of the affective dimension in language learning, which seems to be a deficit rather than a strength, given the lowest use of the affective scale of language learning strategies in comparison to other strategy scales. Developing the skill of using affective strategies selectively can result in both diminishing the feeling of defeat in case of low or unsatisfactory educational achievements and in providing positive reinforcement in case of achieving educational success. As the frequency of affective strategy use was not significantly different across levels, it can be concluded that the low use of these strategies is a general problem rather than a problem reflecting differences in learners’ level of competence. The use of these strategies may depend on the very need to employ them rather than specific task demands, and as Gregersen et al. (2001) suggest, the relationship between using affective strategies and learners’ level of competence is not linear. In fact less proficient learners might use them more often as they might more often need them. Given the relatively low reliability of the affective scale, perhaps the whole concept of affective strategies should be reconsidered, the scale should be reinves- tigated, or, as Macaro (2001, 2006) postulates, affective strategies should not be considered as a separate scale of strategies, but rather as an aspect (or dimension) of learners’ efficiency of using either cognitive or metacognitive strategies. A number of characteristics considered to be individual variables in second language learning are obvious to the language teacher ‘at first sight’. They include the learners’ gender and their level of competence (either simply as the entry information at the

280 beginning of the academic year or, after some time, recognised basing on a learner’s performance), but also features such as students’ place of origin, employment status, and personality. Moreover, any foreign language teacher is not deprived of possibilities of investigating earners’ personality or their background, as mentioned earlier. Not only does it help personalise lesson contents, but also provides the teacher with specific knowledge that can be used to make predictions concerning learners’ strategy use. For the language teacher in order to help their learners become efficient strategy users, it is inevitable to analyse their performance in the classroom, track their progress, but also discuss the use of strategies by the learners, as some strategies may remain unknown to the teacher. As learners’ personality directly influences their performance in the classroom and underlies their strategic choices, the language teacher can benefit from expanding their knowledge about the language learner as a whole person. From a more practical point of view, knowing one’s students well may facilitate the choice of proper classroom procedures or types of interaction. Even if the teacher experiences a temporary failure, reflecting on learners’ characteristics could prove beneficial by making reflection more efficient and, thus, counteracting burnout. From the psychological point of view, the more extensive knowledge EFL teachers have about their students, the more likely they are to make use of their mentalising ability with respect to each and every other classroom participant. Some suggestions for future research could be made on the foundations of the results of the main study and after reviewing its rationale. They could be grouped under the three major categories. To begin with, researchers could aim to boost the fulfilment of the goodness criteria for psychometric tests applied in the present dissertation. Alter- native versions of both personality and strategy inventories could thus be considered. In- stead of NEO-FFI (Costa and McCrae 1992), one of its alternatives, such as NEO-PI-R (Costa and McCrae 2008). Its use might not merely result in an improved reliability and validity, but also enable researchers to establish the correspondence between the underlying facets of the ‘big’ five personality traits and learners’ strategic choices. As far as the SILL ver. 7.0 is considered, it could be subject to further adaptations. In particular, the reliability of its affective scale turned out to be far from expected and perhaps a study aiming to identify learners’ affective strategies and, later on, incorporate them into the amended questionnaire, would be particularly beneficial. This could take place through a

281 series of interviews in order to map the universe of contents to be incorporated in the questionnaire. Secondly, research in Polish settings could address the reasons behind learners’ neglect of affective strategies and, perhaps, provide EFL teachers with suggestions how to increase students’ interest in the emotions which they experience in language learning. While studies in investigating the role of affect are by all means expected (Kębłowska 2012, Oxford 2017), they could be impeded by lack of appropriate tools. Perhaps relying on recent suggestions of introducing positive psychology to investigations of foreign language investigations (MacIntyre and Mercer 2014) could be worth considering in this respect. Finally, it appears that a need exists to conduct research into the correspondence between LLS and self-regulation. Studies investigating the nature of the relationship between them could contribute to a more holistic treatment of language learners, but also prevent wasting a great deal of findings from past strategy research (Rose 2012). Moreo- ver, it seems reasonable to expect that reconciling self-regulation and strategy research could enhance a more pedagogical orientation of strategy research and thus meet contemporary expectations (Pawlak and Oxford 2018).

282 References

Abbs, Brian and Ingrid Freebairn. 1977. Starting strategies. London: Longman. Abbs, Brian and Ingrid Freebairn. 1979. Building strategies. London: Longman. Abbs, Brian and Ingrid Freebairn. 1980. Developing strategies. London: Longman. Abbs, Brian and Ingrid Freebairn. 1982. Opening strategies. London: Longman. Abbs, Brian and Ingrid Freebairn. 1982. Studying strategies. London: Longman. Abraham, Roberta G. and Roberta J. Vann. 1987. “Strategies of two language learners: A case study”, in: Anita Wenden and Joan Rubin (eds.), Learner strategies in language learning. Englewood Cliffs, New Jersey: Prentice Hall International, 85-102. Aharony, Noa. 2006. “The use of deep and surface learning strategies among students learning English as a foreign language in an Internet environment”, British Jour- nal of Educational Psychology 76, 4: 851-866. Allan, Dave. Oxford Placement Test. Oxford: Oxford University Press. Allport, Gordon W. 1961. Pattern and growth in personality. Oxford: Holt, Rinehart and Winston. Allport, Gordon W. and Henry S. Odbert. 1936. “Trait-names: A psycho-lexical study", Psychological Monographs 47, 1: i-171. Alpert, Richard and Ralph Norman Haber. 1960. “Anxiety in academic achievement situations”, The Journal of Abnormal and Social Psychology 61, 2: 207-215. Altan, Mustafa Zülküf. 2003. “Language Learning Strategies and Foreign Language Achievement”, Education and Science 28, 129: 25-31. Anastasi, Anna and Sussan Urbina. 1997. Psychology testing. Upper Saddle River: Prentice Hall. Anderson, John. 1982. “Acquisition of cognitive skill”, Psychological Review 89, 4: 369-406. Anderson, John. 2000. Learning and memory. An integrated approach. New York: John Wiley and Sons. Anthis, Kristine and Joseph C. La Voie. 2006. “Readiness to change: A longitudinal study of changes in adult identity”, Journal of Research in Personality 40, 2: 209-219.

283 Anugkakul, Gunniga and Suwaree Yordchim. 2014. “Language learning strategies of Chinese students at Suan Sunandha Rajabhat University in Thailand”, Interna- tional Journal of Social, Behavioral, Educational, Economic, Business and In- dustrial Engineering, 2511-2514. Ardasheva, Yuliya and Thomas R. Tretter. 2013. “Strategy inventory for language learning–ELL student form: testing for factorial validity”, The Modern Lan- guage Journal 97, 2: 474-489. Arnett, Jeffrey Jensen. 2000. “Emerging adulthood: A theory of development from the late teens through the twenties”, American Psychologist 55, 5: 469-480. Aronson, Elliot, Timothy D. Wilson, Robin M. Akert. 2014. Social psychology. Upper Saddle River, New Jersey: Pearson Education Inc. Ashton, Michael C. 2017. Individual differences and personality. London: Academic Press. Ashton, Michael C., Kibeom Lee and Sampo V. Paunonen. 2002.“What is the central feature of extraversion? Social attention versus reward sensitivity”, Journal of Personality and Social Psychology 83, 1: 245-252. Atieno, Ochieng Pamela. 2009. “An analysis of the strengths and limitation of qualitative and quantitative research paradigms”, Problems of Education in the 21st Century 13, 1: 13-38. Ayotte, Brian J., Guy G. Potter, Heather T. Williams, David C. Steffens and Hayden B. Bosworth. 2009. “The moderating role of personality factors in the relationship between depression and neuropsychological functioning among older adults”, International Journal of Geriatric Psychiatry 24, 9: 1010-1019. Bachman, Lyle F. 2002. “Some reflections on task-based language performance assessment”, Language Testing 19, 4: 453-476. Baddeley, Alan. 1997. Human memory: Theory and practice. Hove: Psychology Press. Bailey, Kenneth D. 1994.Typologies and taxonomies: an introduction to classification techniques. London: Sage Publications. Balzarotti, Stefania, O. P. John and J. J. Gross. 2010.“Italian adaptation of the emotion regulation questionnaire”, European Journal of Psychological Assessment 26: 61-67. Bandura, Albert. 1986. Social foundations of thought and action: A social cognitive theory. Englewood Cliffs: Prentice-Hall.

284 Bandura, Albert and Dale H. Schunk. 1981.“Cultivating competence, self-efficacy, and intrinsic interest through proximal self-motivation”, Journal of Personality and Social Psychology 41, 3: 586-598. Bandura, Albert and Walters, Richard H. 1977. Social learning theory. Englewood Cliffs: Prentice Hall. Barone, David F., Michel Hersen and Vincent B. Van Hasselt (eds.). 1998. Advanced personality. New York: Springer. Barrett, Martyn. 2017. “Early lexical development”, in: Paul Fletcher and Brian MacWhinney (eds.). The handbook of child language. Oxford: Blackwell, 361- 392. Bates, Timothy C. and Alexandra Shieles. 2003. “Crystallized intelligence as a product of speed and drive for experience: the relationship of inspection time and openness to g and Gc”, Intelligence 31, 3: 275-287. Baumeister, Roy and Jessica L. Alquist. 2009. “Self-regulation as a limited resource: Strength model of control and depletion”, in: Joseph P. Forgas, Roy F. Baumeis- ter and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psy- chology Press, 21-34. Bedyńska, Sylwia, and Aneta Brzezicka-Rotkiewicz (eds.). 2007. Statystyczny drogo- wskaz: praktyczny poradnik analizy danych w naukach społecznych na przykła- dach z psychologii. [Statistical guide: A practical guide of data analysis in social sciences based on psychological examples]. Warszawa: Wydawnictwo Szkoły Wyższej Psychologii Społecznej ”Academica”. Bègue, Laurent, Jean-Léon Beauvois, Didier Courbet, Dominique Oberlé, Johan Lepage and Aaron A. Duke. 2015. “Personality predicts obedience in a Milgram paradigm”, Journal of Personality 83, 3: 299-306. Belâev, Boris Vasil’evič. 1969. Zarys psychologii nauczania języków obcych [An outline of foreign language teaching psychology]. (Translated by Lucjan Mirecki.) Warszawa: Państwowe Zakłady Wydawnictw Szkolnych. Benson, Phil. 2011. Teaching and researching autonomy. (2nd edition.) Harlow: Long- man. Benson, Phil and Winnie Lor. 1999.“Conceptions of language and language learning”, System 27, 4: 459-472.

285 Betella, Alberto and Paul FMJ Verschure. 2016. “The affective slider: A digital self-assessment scale for the measurement of human emotions", PloS One 11, 2: e0148037. Beukeboom, Camiel J., Martin Tanis and Ivar E. Vermeulen. 2013. “The language of extraversion: Extraverted people talk more abstractly, introverts are more con- crete”, Journal of Language and Social Psychology 32, 2: 191-201. Bhagat, Vidya and Ramyashilpa D. Nayak. 2014. “Neuroticism and academic performance of medical students”, International Journal of Humanities and Social Sci- ence Invention 3, 1: 51-55. Bialystok, Ellen. 1978. “A theoretical model of second language learning”, Language Learning 28, 1: 69-83. Bialystok, Ellen. 1981. “The role of conscious strategies in second language proficiency”, The Modern Language Journal 65, 1: 24-35. Bialystok, Ellen. 1983. “Some factors in the selection and implementation of communication strategies”, in Claus Færch and Gabriele Casper (eds.), Strategies in interlanguage communication. London: Longman, 100-118. Bialystok, Ellen and Maria Fröhlich. 1977. “Second language learning and teaching in classroom settings: The learning study, year one” (https://files.eric.ed.gov/fulltext/ED180228.pdf) (date of access: 14 Feb. 2018). Bialystok, Ellen and Maria Fröhlich. 1978. “Variables of classroom achievement in second language learning”, The Modern Language Journal 62, 7: 327-336. Biedroń, Adriana. 2011. “Personality factors as predictors of foreign language aptitude”, Studies in Second Language Learning and Teaching 1, 4: 467-489. Biedroń, Adriana and Mirosław Pawlak. 2016. “The interface between research on individual difference variables and teaching practice: The case of cognitive factors and personality”, Studies in Second Language Learning and Teaching 6, 3: 3 95- 422. Bitter, James R. 2011. Contributions to Adlerian psychology. Bloomington: Xlibris. Blanton, Hart and Deborah L. Hall. 2009. “Punishing difference and rewarding diversity: A deviance regulation analysis of social structure”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 273-288.

286 Boekaerts, Monique. 1996. “Personality and the psychology of learning”, European Journal of Personality 10, 5: 377-404. Boekaerts, Monique. 1997. “Self-regulated learning: A new concept embraced by researchers, policy makers, educators, teachers, and students”, Learning and In- struction 7, 2: 161-186. Bouchard, Thomas Jr. and John C. Loehlin. 2001. “Genes, evolution, and personality”, Behavior Genetics 31, 3: 243-273. Bozavli, Ebubekir. 2016. “Understanding of Foreign Language Learning of Generation Y”. Journal of Education and Practice 26, 7: 69-76. Bratko, Denis, Tomas Chamorro-Premuzic and Zrnka Saks. 2006. “Personality and school performance: Incremental validity of self-and peer-ratings over intelligence”, Personality and Individual Differences 41, 1: 131-142. Braun, Virginia and Victoria Clarke. 2006. “Using thematic analysis in psychology”, Qualitative Research in Psychology 3, 2: 77-101. Bredemeier, Keith, Howard Berenbaum, Steven B. Most, and Daniel J. Simons. 2011. “Links between neuroticism, emotional distress, and disengaging attention: Evi- dence from a single-target RSVP task”, Cognition & Emotion 25, 8: 1510-1519. Brown, Ann L. 1982. Learning, remembering, and understanding. Technical Report No. 244. (https://pdfs.seman- ticscholar.org/194a/406e0d10e8c91aa357b6bb0394b9ce34d7fe.pdf) (date of access 02 Jul. 2018) Brown, James D. 1988. Understanding research in second language learning. Cam- bridge: Cambridge University Press. Brzezińska, Anna I., Karolina Appelt and Beata Ziółkowska. 2016. Psychologia rozwoju człowieka. [Psychology of human development]. Gdańsk: Gdańskie Wy- dawnictwo Psychologiczne. Brzezińska, Anna (ed). 2005. Psychologiczne portrety człowieka. Praktyczna psychologia rozwojowa. [Human psychological portraits. Practical developmental psychology]. Gdańsk: Gdańskie Wydawnictwo Psychologiczne. Brzeziński, Jerzy. 2004. Metodologia badań pedagogicznych. [Methodology of pedago- gic research]. Warszawa: PWN. Brzeziński, Jerzy and Tomasz Maruszewski. 1978. „Metoda sędziów kompetentnych i jej zastosowanie w badaniach pedagogicznych” [”The expert judge panel

287 method and its application in pedagogical research”], Kwartalnik Pedagogiczny 23, 1: 569-587. Bugental, James FT. 1964. “The third force in psychology”, Journal of Humanistic Psy- chology 4,1: 19-26. Burnette, Jeni L., Ernest H. O’Boyle, Eric M. VanEpps, Jeffrey M. Pollack and Eli J. Finkel. 2013. “Mind-sets matter: A meta-analytic review of implicit theories and self-regulation”, Psychological Bulletin 139, 3: 655-701. Busch, Deborah. 1982. “Introversion-extraversion and the EFL proficiency of Japanese students”, Language Learning 32, 1: 109-132. Cambridge Dictionary. 2018. Cambridge: Cambridge University Press (https://dictionary.cambridge.org) (date of access: 19 Feb. 2018). Canagarajah, Suresh A. 1999. Resisting linguistic imperialism in English teaching. Ox- ford: Oxford University Press. Canale, Michael and Merrill Swain. 1980. “Theoretical bases of communicative approaches to second language teaching and testing”, Applied Linguistics 1, 1: 1- 47. Carrell, Patricia L., Moneta S. Prince and Gusti G. Astika. 1996. “Personality types and language learning in an EFL context”, Language Learning 46, 1: 75-99. Carrier, Carol, Verna Higson, Victor Klimoski and Eric Peterson. 1984. “The effects of facilitative and debilitative achievement anxiety on notetaking”, The Journal of Educational Research 77, 3: 133-138. Carroll, John B. 1962. “The prediction of success in intensive foreign language training”, in: Nathaniel L. Gage (ed.), Handbook of research on teaching. Chicago: Rand McNally, 87-136. Carroll, John B. 1965. “The contributions of psychological theory and educational research to the teaching of foreign languages”, Modern Language Journal 49,5: 273-281. Carroll, John B. and Stanley Sapon. 2002. Modern language aptitude test MLAT: Man- ual 2002 edition. Bethesda: Second Language Testing. Carson, Joan G. and Ana Longhini. 2002.“Focusing on learning styles and strategies: A diary study in an immersion setting”, Language Learning 52, 2: 401-438. Carton, Aaron S. 1966. The "method of inference" in foreign language study. New York: City University New York, Research Foundation.

288 Carton, Aaron S. 1971. “Inferencing: A process in using and learning language”, in: Paul Pimsleur and Terence Quinn (eds.), The psychology of second language learning. Cambridge: Cambridge University Press, 45-58. Carver, Charles S. 2001. “Affect and the functional bases of behavior: On the dimensional structure of affective experience”, Personality and Social Psychology Re- view 5, 4: 345-356. Carver, Charles S. and Michael F. Scheier. 1999. “Themes and issues in the self-regulation of behavior”, in: Robert S. Wyer, Jr and Thomas K. Srull (eds.), Advances in social cognition. Vol 12. Mahwah: Lawrence Erlbaum Associates, Inc., 1- 105. Carver, Charles S. and Michael F. Scheier. 2009. “Action, affect, multitasking, and layers of control”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 109-126. Carver, Charles S. and Teri L. White. 1994. “Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: the BIS/BAS scales”, Journal of Personality and Social Psychology 67, 2: 319-333. Carver, David. 1984. “Plans, learner strategies and self direction in language learning”, System 12, 2: 123-131. Castilho, Simone M. Psyche, matter and crop circles: An approach by analytical psychology. Hauppauge: Nova Science Publishers. Cattell, Raymond B. 1943. “The description of personality: Basic traits resolved into clusters”, The Journal of Abnormal and Social Psychology 38, 4: 476-506. Cattell, Raymond B. 1956. “Validation and intensification of the sixteen personality factor questionnaire”, Journal of Clinical Psychology 12, 3: 205-214. Cattell, Raymond B. 1973. Personality and mood by questionnaire. San Fransisco: Jossey-Bass. Chamorro-Premuzic, Tomas. 2006. “Creativity versus conscientiousness: which is a better predictor of student performance?”, Applied Cognitive Psychology 20, 4: 521-531. Chamorro-Premuzik, Tomas. 2011. Personality and individual differences. Chichester: BPS Blackwell.

289 Chamorro-Premuzic, Tomas and Adrian Furnham. 2003. “Personality predicts academic performance: Evidence from two longitudinal university samples”, Journal of Research in Personality 37, 4: 319–338. Chamot, Anna Uhl. 1987. A study of learning strategies in foreign language instruction. First year report. Washington, DC: Office of International Education. Chamot, Anna Uhl and Michael J. O’Malley. 1986. A cognitive academic language learning approach: An ESL content-based curriculum. Washington, DC: Na- tional Clearinghouse for Bilingual Education. Chamot, Anna Uhl and Pamela Beard El-Dinary. 1999. “Children’s learning strategies in language immersion classrooms”, The Modern Language Journal 83, 3: 319- 338. Chandra, Charu and Armen Tumanyan. 2005. “Supply chain system taxonomy: A framework and methodology”, Human Systems Management 24, 4: 245-258. Chastain, Kenneth. 1975. “Affective and ability factors in second-language acquisition”, Language Learning 25, 1: 153-161. Chaudron, Craig. 1988. Second language classrooms: Research on teaching and learning. Cambridge: Cambridge University Press. Chularut, Pasana and Teresa K. DeBacker. 2004. “The influence of concept mapping on achievement, self-regulation, and self-efficacy in students of English as a second language”, Contemporary Educational Psychology 29, 3: 248-263. Coffield, Frank, David Moseley, Elaine Hall and Kathryn Ecclestone. 2004. Learning styles and pedagogy in post-16 learning: A systematic and critical review. Lon- don: Learning and Skills Research Centre. Coffield, Frank, Richard Steer, Ann Hodgson, Ken Spours, Sheila Edward and Ian Fin- lay. 2005. “A new learning and skills landscape? The central role of the Learn- ing and Skills Council”, Journal of Education Policy 20, 5: 631-656. Cohen, Andrew D. 1995. Second language learning and use strategies: Clarifying the issues. Minneapolis: Minnesota University, Centre for Advanced Research on Language Acquisition. Cohen, Andrew D. 1998. Strategies in learning and using a second language. London and New York: Longman.

290 Cohen, Andrew D. 2003. “The learner’s side of foreign language learning: Where do styles, strategies, and tasks meet?”, International Review of Applied Linguistics in Language Teaching 41, 4: 279-291. Cohen, Andrew D. 2005. “Strategies for learning and performing L2 speech acts”, In- tercultural P1ragmatics 2, 3: 275-301. Cohen, Andrew D. 2014. Strategies in learning and using a second language. (2nd edition.) London: Routledge. Cohen, Andrew D. and Ernesto Macaro. 2007. Learner strategies. Oxford: Oxford Uni- versity Press. Cohen, Andrew D. and Zoltán Dörnyei. 2002. “Focus on the language learner: Motiva- tion, styles, and strategies”, in: Norbert Schmitt (ed.), An introduction to applied linguistics. London: Arnold, 170-190. Cohen, Henri and Claire Lefebvre (eds.). 2005. Handbook of categorization in cognitive science. St. Louis: Elsevier. Cohen, Louis, Lawrence Manion and Keith Morrison. 2005. Research methods in education. London and New York: Routledge Falmer. Coker, Crystal and Florin Mihai. 2017. “Personality traits and second language acquisition: The influence of the Enneagram on adult ESOL students”, TESOL Journal 8, 2: 432-449. Cook, Vivian. 2008. Second language learning and language teaching. London: Hod- der Education. Corr, Philip J. (ed.). 2008. The reinforcement sensitivity theory of personality. Cam- bridge: Cambridge University Press. Cortina, Jose M. 1993. “What is coefficient alpha? An examination of theory and applications”, Journal of Applied Psychology 78: 98–104. Costa, Paul T. Jr. and Robert R. McCrae. 1985. The NEO Personality Inventory manual. Odessa, FL: Psychological Assessment Resources. Costa, Paul T. Jr. and Robert R. McCrae. 1992. Revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) manual. Odessa, FL: Psychological Assessment Resources. Costa, Paul T. Jr. and Robert R. McCrae. 1994. “Set like plaster? Evidence for the stability of adult personality”, in: Heatherton, Todd F. and Joel Lee Weinberger,

291 Can personality change? Washington: American Psychological Association: 21- 40. Costa, Paul T. Jr. and Robert R. McCrae. 1995. “Domains and facets: Hierarchical personality assessment using the Revised NEO Personality Inventory”, Journal of Personality Assessment 64, 1: 21-50. Costa Jr, Paul T. and Robert R. McCrae. 1998. “Six approaches to the explication of facet‐level traits: examples from conscientiousness”, European Journal of Per- sonality 12: 2, 117-134. Costa, Paul T. Jr and Robert R. McCrae. 2017. “The NEO inventories as instruments of psychological theory”, in: Thomas A. Widiger, (ed.), The Oxford handbook of the five factor model of personality. Oxford: Oxford University Press, 11-38. Costa, Paul T. Jr., Robert R. McCrae and David A. Dye. 1991. “Facet scales for Agreea- bleness and Conscientiousness: A revision of the NEO Personality Inventory”, Personality and individual Differences 12, 9: 887-898. Cotterall, Sara. 1995. “Developing a course strategy for learner autonomy”, ELT Jour- nal 49, 3: 219-227. Cramer, Duncan. 1991. Personality and psychotherapy: Theory, practice, and research. Milton Keynes: Open University Press. Crocker, Jennifer and Riia K. Luhtanen. 2003. “Level of self-esteem and contingencies of self-worth: Unique effects on academic, social, and financial problems in college students” Personality and Social Psychology Bulletin 29, 6: 701-712. Cronbach, Lee J. 1956. “Assessment of individual differences”, Annual Review of Psy- chology 71, 1: 173-196. Cronbach, Lee J. 1970. Essentials of psychological testing. New York: Harper & Row. Cronbach, Lee J. and Paul E. Meehl. 1955. “Construct validity in psychological tests”, Psychological Bulletin 52, 4: 281-302. Csizér, Kata and Judit Kormos. 2009. “Learning experiences, selves and motivated learning behaviour: A comparative analysis of structural models for Hungarian secondary and university learners of English”, in: Zoltán Dörnyei and Ema Ush- ioda (eds.), Motivation, language identity and the L2 self. Bristol: Multilingual Matters, 98-119. Dakowska, Maria. 2005. Teaching English as a foreign language: a guide for profes- sionals. Warszawa: PWN.

292 Dakowska, Maria. 2013. “Foreign language didactics encounters cognitive science”, in: Mirosław Pawlak and Krystyna Droździał-Szelest (eds.). Psycholinguistic and sociolinguistic perspectives on second language learning and teaching. Berlin: Springer, 3-25. Daniels, David and Virginia Price. 2009. The essential enneagram. The definite personality test and self-discovery guide. New York: Harper One. Dansereau, Donald F. 1985. “Learning strategy research”, in: Judith W. Segal, Susan F. Chipman and Robert Glaser (eds.), Thinking and learning skills: Relating instruction to research (Vol. l). Hillsdale: Erlbaum: 209-239. De Boot, Kees, Wander Lowie and Marjolijn Verspoor. 2005. Second language acquisition: An advanced resource book. London: Routledge. De Bruin, Anique B. H and Tamara van Gog. 2012. “Improving self-monitoring and self-regulation: From cognitive psychology to the classroom”, Learning and In- struction 22, 2: 245-252. De Haan, Lydia, Esther Kuipers, Yvanca Kuerten, Margriet van Laar, Berend Olivier and Joris Cornelis Verster. 2011.“The RT-18: a new screening tool to assess young adult risk-taking behaviour”, International Journal of General Medicine 4: 575-584. De Raad, Boele and Henri C. Schouwenburg. 1996. “Personality in learning and education: a review”, European Journal of Personality 10, 5: 303-336. Del Ángel Castillo, Martha C. and Katherina E. Gallardo Córdova. 2014. “Language learning strategies and academic success: a Mexican perspective”, Universitas Psychologica 13, 2: 703-713. Delamont, Sara, Paul Atkinson and Lesley Pugsley. 2010. “The concept smacks of magic: Fighting familiarity today”, Teaching and Teacher Education 26, 1: 3-10. Derlega, Valerian, Barbara A. Winstead and Warren H. Jones. 1991. Personality: Con- temporary theory and research. Chicago: Nelson-Hall. Dewaele, Jean-Marc. 2004. “The acquisition of sociolinguistic competence in French as a foreign language: an overview”, Journal of French Language Studies 14, 3: 301-319. Dewaele, Jean‐Marc and Adrian Furnham. 1999. “Extraversion: The unloved variable in applied linguistic research”, Language Learning 49, 3: 509-544.

293 Dewaele, Jean-Marc and Adrian Furnham. 2000. “Personality and speech production: a pilot study of second language learners”, Personality and Individual Differences 28, 2: 355-365. DeYoung, Colin G. 2010. “Personality neuroscience and the biology of traits”, Social and Personality Psychology Compass 4, 12: 1165-1180. DeYoung, Colin G., Lena C. Quilty, and Jordan B. Peterson. 2007. "Between facets and domains: 10 aspects of the Big Five." Journal of Personality and Social Psy- chology 93, 5: 880-896. Dieckmann, Anja and Matthias Unfried. 2014. “Writ large on your face: Observing emotions using automatic facial analysis”, GfK Marketing Intelligence Review 6, 1: 52-58. Dollinger, Stephen J., Frederick T. L. Leong, and Shawna K. Ulicni. 1996. “On traits and values: With special reference to openness to experience”, Journal of Re- search in Personality 30, 1: 23-41. Doughty, Catherine and Michael H. Long (eds.). 2003. The handbook of second language acquisition. Oxford: Blackwell Publishing Dörnyei, Zoltán. 2001. Motivational strategies in the language classroom. Cambridge: Cambridge University Press. Dörnyei, Zoltán. 2003a. “Attitudes, orientations, and motivations in language learning: Advances in theory, research, and applications”, Language Learning 53, 1: 3-32. Dörnyei, Zoltan. 2003b. Questionnaires in second language research: Construction, administration, and processing. Mahwah, NJ: Lawrence Erlbaum Associates Pub- lishers. Dörnyei, Zoltán. 2005. The psychology of the language learner. London: Lawrence Erl- baum Associates. Dörnyei, Zoltán. 2009. “Individual differences: Interplay of learner characteristics and learning environment”, Language Learning 59, 1: 230-248. Dörnyei, Zoltán and Ema Ushioda. 2009. Motivation, language identity and the L2 self. Bristol: Multilingual Matters. Dörnyei, Zoltán and István Otto. 1998. “Motivation in action: A process model of L2 motivation”, Working Papers in Applied Linguistics Vol. 4. London: Thames Valley University, 43-69.

294 Dörnyei, Zoltán and Peter Skehan. 2003. “Individual differences in second language learning”, in: Catherine Doughty and Michael H. Long (eds.). The handbook of second language acquisition. Oxford: Blackwell Publishing: 589-630. Dörnyei, Zoltán and Stephen Ryan. 2015. The psychology of the language learner revisited. New York: Routledge. Drake, Kim, Jay Belsky and R. M. Pasco Fearon. 2014. “From early attachment to engagement with learning in school: The role of self-regulation and persistence”, Developmental Psychology 50, 5: 1350-1361. Drapela, Victor J. 1995. A review of personality theories. Springfield: Charles C Thomas Publisher. Dreyer, Carisma and Rebecca L. Oxford. 1996. “Learning strategies and other predictors of ESL proficiency among Afrikaans speakers in South Africa”, in: Rebecca L. Oxford (ed.) Language learning strategies around the world: Cross-cultural perspectives. Manoa: Second Language Teaching and Curriculum Center, Uni- versity of Hawai: 61-74. Droździał-Szelest, Krystyna. 1993. “Strategies of second language learners: some research findings and their pedagogical implications”, Glottodidactica 21: 43-52. Droździał-Szelest, Krystyna. 1997. Language learning strategies in the process of acquiring a foreign language. Poznań: Motivex. Droździał-Szelest, Krystyna. 2004. “Strategie uczenia się języka obcego: badania a rzeczywistość edukacyjna” [Language learning strategies: Research and educational reality], in: Mirosław Pawlak (ed.), Autonomia w nauce języka obcego [Autonomy in foreign language learning]. Poznań-Kalisz: Wydział Pedago- giczno-Artystyczny UAM w Poznaniu, 31-43. Duckworth, Angela L. and Martin E.P. Seligman. 2005. ”Self-discipline outdoes IQ in predicting academic performance of adolescents”, Psychological Science 16, 12: 939-944. Duckworth, Angela L. and Martin E.P. Seligman. 2006.“Self-discipline gives girls the edge: Gender in self-discipline, grades, and achievement test scores”, Journal of Educational Psychology 98, 1: 198-208. Duckworth, Angela Lee and Stephanie M. Carlson. 2013, “Self-regulation and school success", in: Bryan W. Sokol, Frederick M. E. Grouzet and Ulrich Müller (eds.),

295 Self-regulation and autonomy: Social and developmental dimensions of human conduct. Cambridge: Cambridge University Press, 209-230. Duff, Angus, Elizabeth Boyle, Karen Dunleavy and John Ferguson. 2004. “The relationship between personality, approach to learning and academic performance”, Personality and Individual Differences 36, 8: 1907-1920. Dunkel, Harold B. 1947. “The effect of personality on language achievement”, Journal of Educational Psychology 38, 3: 177-182. Dunn, Rita, Kenneth Dunn and Gary E. Price. 1975. Learning styles inventory. Law- rence: Price Systems. Edgcumbe, Rose. 2000. Anna Freud: A view of development, disturbance and therapeutic techniques. London: Routledge. Ehrman, Madeline E. and Rebecca Oxford. 1988. “Effects of sex differences, career choice, and psychological type on adult language learning strategies”, The Mod- ern Language Journal 73, 1: 1-13. Ehrman, Madeline E. and Rebecca Oxford. 1990.“Adult language learning styles and strategies in an intensive training setting”, The Modern Language Journal 74, 3: 311-327. Ehrman, Madeline E. and Rebecca L. Oxford. 1995. “Cognition plus: Correlates of language learning success”, The Modern Language Journal, 79, 1: 67-89. Eisinga, Rob, Manfred Te Grotenhuis and Ben Pelzer. 2013.“The reliability of a two- item scale: Pearson, Cronbach, or Spearman-Brown?”, International Journal of Public Health 58, 4: 637-642. Elgort, Irina and Paul Warren. 2014. “L2 vocabulary learning from reading: Explicit and tacit lexical knowledge and the role of learner and item variables”, Lan- guage Learning 64, 2: 365-414. Ellis, Albert, Mike Abrams, and Lidia Abrams. 2009. Personality theories: Critical perspectives. Thousand Oaks: SAGE Publications, Inc. Ellis, Gail and Barbara Sinclair. 1989. Learning to learn English. A course in learner training. Cambridge: Cambridge University Press. Ellis, Rod. 1994. The study of second language acquisition. Oxford: Oxford University Press. Ellis, Rod. 2003. Task-based language learning and teaching. Oxford: Oxford Univer- sity Press.

296 Ellis, Rod. 2008. The study of second language acquisition. (2nd edition.) Oxford: Ox- ford University Press. Engler, Barbara. 1979. Personality theories: An introduction. Boston, Houghton Miffin. Entwistle, Noel (ed.). 1990. Handbook of educational ideas and practices. London: Routledge. Eppler, Martin J. and Jeanne Mengis. 2011. Drawing Distinction: The visualization of classification in qualitative research. (http://citeseerx.ist.psu.edu/viewdoc/down- load?doi=10.1.1.471.6838&rep=rep1&type=pdf) (date of access 01 Jul. 2018) Epstein, Robert. 1991. “Skinner, creativity, and the problem of Spontaneous behavior”, Psychological Science 2,6: 362-370. Erikson, Erik H. 1958. “The nature of clinical evidence”, Daedalus 87,4: 65-87. Erikson, Erik H. 1968. Identity: Youth and crisis. London: Faber & Faber. Esch, Emily. 1997. “Learner training for autonomous language learning”, in: Phil Ben- son and Peter Voller (eds.). Autonomy and independence in language learning. London: Routledge, 164-176. Ewen, Robert B. 2003. An introduction to theories of personality. (6th edition.) London: Psychology Press. Eysenck, Hans J. 1959. Manual of the Maudsley personality inventory. London: Univer- sity of London Press. Eysenck, Hans J. 1960. “Dimensions of personality, psychiatric syndromes, and mathe- matical models”, The Journal of Mental Science 106, 443: 581-589. Eysenck, Hans J. 1964. “The measurement of personality: A new inventory”, Journal of the Indian Academy of Applied Psychology 1, 1: 1-11. Eysenck, Hans J. 1967a. “On the unitary nature of extraversion”, Acta Psychologica 26, 4: 383-390. Eysenck, Hans J. 1967b. The biological basis of personality. London: Charles C. Thomas Publisher. Eysenck, Hans J. 1982. Personality, genetics, and behaviour: Selected papers. New York: Praeger. Eysenck, Hans J. 1990. “Genetic and environmental contributions to individual differences: The three major dimensions of personality”, Journal of Personality 58, 1: 245-261.

297 Eysenck Hans J. and Sibil B.G. Eysenck 1975. Manual of the Eysenck Personality Questionnaire. London: Hodder and Stoughton. Eysenck, Sybil B., G. Easting and Paul R. Pearson. 1984.“Age norms for impulsiveness, venturesomeness and empathy in children”, Personality and Individual Differ- ences 5, 3: 315-321. Eysenck, Sybil B., T. Russell and Hans J. Eysenck. 1970. “Extraversion, intelligence, and ability to draw a person”, Perceptual and Motor Skills, 30, 3: 925-926. Fayyaz, Waseem. 2011. “Personality traits and the metacognitive listening skills of English as a foreign language in Pakistan”, Journal of Behavioural Sciences 21, 2: 59-76. Ferguson, George A. and Yoshio Takane. 2002. Statistical analysis in psychology and education. New York: McGraw-Hill. Fishbach, Ayelet. 2009. “The dynamics of self-regulation”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation.New York: Psychology Press, 163-182. Fiske, Donald W. 1949. “Consistency of the factorial structures of personality ratings from different sources”, The Journal of Abnormal and Social Psychology 44, 3: 329-344. Fitzsimons, Gráinne M., Justin Friesen, Edward Orehek and Arie W. Kruglanski. 2009. “Progress-induced goal shifting as a self-regulatory strategy”, in: Joseph P. For- gas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 183-200. Fleeson, William and Jutta Heckhausen. 1997.“More or less “me” in past, present, and future: Perceived lifetime personality during adulthood”, Psychology and Aging 12, 1:125. Forgas, Joseph P., Roy Baumeister and Dianne M. Tice (eds.). 2009. Psychology of self- regulation. New York: Psychology Press. Förster, Jens and NiraLiberman. 2009. “Goal gradients: Challenges to a basic principle of motivation”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 147-162. Fowler, William S. and Norman Coe. 1977. Test your English. Nairobi: Nelson. Freud, Anna. 1992. The ego and the mechanisms of defence. London: Karnac Books.

298 Freud, Sigmund. 1923. Das Ich und das Es. [The Ego and the Id]. Vienna: W. W. Nor- ton & Company. Freud, Sigmund and Josef Breuer. 2004. Studies in hysteria. London: Penguin Books. Friese, Malte, Michaela Wänke and Wilhelm Hofmann. 2009. “Unscrambling self-regulatory behaviour determination: The interplay of impulse strength, reflective processes, and control resources”, in: Joseph P. Forgas, Roy F. Baumeister and Di- anne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 53-72. Frijda, Nico H. 1986. The emotions. Cambridge: Cambridge University Press. Furnham, Adrian F. 1997. “Knowing and faking one’s five-factor personality score”, Journal of Personality Assessment 69, 1: 229-243. Gagné, Ellen D., Carol Walker Yekovich and Frank R. Yekovich. 1993. The cognitive psychology of school learning. New York: HarperCollins College Publishers. Gao, Xuesong. 2006. “Understanding changes in Chinese students’ uses of learning strategies in China and Britain: A socio-cultural re-interpretation”, System 34, 1: 55-67. Gao, Xuesong. 2007. “Has language learning strategy research come to an end? A response to Tseng et al. (2006)”, Applied Linguistics 28, 4: 615-620. Gardner, Robert C. 2010. Motivation and second language acquisition: The socio-educational model. Vol. 10. New York: Peter Lang. Gardner, Robert C. 2013. “Motivation”, in: Peter Robinson (ed.), The Routledge Ency- clopedia of Second Language Acquisition. London: Routledge, 443-447. Gardner, Robert C., Paul F. Tremblay and Anne Masgoret. 1997. “Towards a full model of second language learning: An empirical investigation”, The Modern Lan- guage Journal 81, 3: 344-362. Gardner, Robert. C., Smythe, Padric. C., Clément, Richard. and Gliksman, Louis. 1976. “Social and psychological factors in second language acquisition” Canadian Modern Language Review 32, 3: 198-213. Gabryś-Barker, Danuta. 2010. “Emotion versus cognition, or what psycho and neurolin- guistics tell us about affectivity in second language acquisition”, in: Janusz Ar- abski and Adam Wojtaszek (eds.). Neurolinguistic and psycholinguistic perspectives on SLA. Bristol: Multilingual Matters: 44-63.

299 Gass, Susan M. and Evangeline Varonis. 1986.“Sex differences in NNS/NNS interactions”, in: Richard R. Day (ed.). Talking to learn: Conversation in second language acquisition. Cambridge (MA): Newbury House Publishers, 327-351. Gass, Susan M. and Larry Selinker. 2008. Second language acquisition: An introduc- tory course topics in applied psycholinguistics. Vol 1. New York: Routledge. Gerami, Mohammad H. and Shiva M. Baighlou. 2011. “Language learning strategies used by successful and unsuccessful Iranian EFL students”, Procedia-Social and Behavioral Sciences 29: 1567-1576. Gerrig, Richard J. and Philip G. Zimbardo. 2009. Psychology and life. Boston: Pearson. Główka, Danuta. 2011. “Mix? Yes, but how? Mixed methods research illustrated”, in: Mirosław Pawlak (ed.), Extending the boundaries of research on second language learning and teaching. Berlin Heidelberg: Springer, 289-300. Goldberg, Lewis R. 1981. “Unconfounding situational attributions from uncertain, neutral, and ambiguous ones: A psychometric analysis of descriptions of oneself and various types of others”, Journal of Personality and Social Psychology 41, 3: 517-552. Goldberg, Lewis R. 1990. “An alternative description of personality: The big-five factor structure”, Journal of Personality and Social Psychology 59, 6: 1216-1229. Goldberg, Lewis R., John A. Johnson, Herbert W. Eber, Robert Hogan, Michael C. Ashton, C. Robert Cloninger and Harrison G. Gough. 2006. “The International Personality Item Pool and the future of public-domain personality measures”, Journal of Research in Personality 40, 1: 84-96. Góralski, Andrzej. 1976. Metody opisu i wnioskowania statystycznego w psychologii (wydanie drugie) [Methods of statisticaldescription and inference in psychology] (2nd edition). Warszawa: Państwowe Wydawnictwo Naukowe. Gosling, Samuel D, Peter J. Rentfrow and William B. Swann Jr. 2003. “A very brief measure of the five personality domains”. Journal of Research in Personality 37: 504-528. Grabe, William. 2002. “Dilemmas for the development of second language reading abilities”, in: Jack C. Richards and Willy A. Renandya (eds.). Methodology in language teaching: An anthology of current practice. Cambridge: Cambridge Uni- versity Press: 276-286.

300 Grainger, Peter R. 1997. “Language-learning strategies for learners of Japanese: Investi- gating ethnicity”, Foreign Language Annals 30, 3: 378-385. Grainger, Peter R. 2012. “The impact of cultural background on the choice of language learning strategies in the JFL context”, System 40, 4: 483-493. Gray, Elizabeth K. and David Watson. 2002. “General and specific traits of personality and their relation to sleep and academic performance”, Journal of Personality 70, 2: 177-206. Gray, Jeffrey A. 1970. “The psychophysiological basis of introversion-extraversion”, Behaviour Research and Therapy 8, 3: 249-266. Gray, Jeffrey A. 1981. “A critique of Eysenck’s theory of personality”, in: Eysenck, Hans, J. A model for personality. Berlin Heidelberg: Springer: 246-276. Gray, Jeffrey A. 1982. “On mapping anxiety” Behavioral and Brain Sciences 5, 3: 506- 534. Gray, Peter. 1999. Psychology. (3rd edition). New York: Worth Publishers. Graziano, William G. and Meara M. Habashi. 2015. The Oxford Handbook of Prosocial Behavior. Oxford: Oxford University Press. Graziano, William G. and Renée M. Tobin. 2017. “Extroversion and the Five Factor Model”, in: Thomas Widiger, The Oxford handbook of the five factor model of personality. Oxford: Oxford University Press, 105-132. Graziano, William G. and Meara M. Habashi. 2017. “Agreeableness and the Five Factor Model”, in: Widiger, Thomas. (ed.), The Oxford handbook of the five factor model of personality: 105-132. Oxford: Oxford University Press. Green, John M., and Rebecca Oxford. 1995. “A closer look at learning strategies, L2 proficiency, and gender”, TESOL Quarterly 29, 2: 261-297. Greenberg, Jay R. and Stephen Mitchell. 1983. Object relations in psychoanalytic theory. Cambridge (US): Harvard University Press. Gregersen, Tammy and Peter MacIntyre. 2014. Capitalizing on language learners’ individuality: From premise to practice. Bristol: Multilingual Matters. Gregersen, Tammy, Vera R. Martínez, Pino P. Rojas and Alvarado, L. Espinoza. 2001. “Can foreign language learning strategies turn into crutches? A pilot study on the use of strategies by successful and unsuccessful language learners”, Revista Signos 34, 49-50: 101-111.

301 Grenfell, Michael and Harris Vee. 1999. Modern languages and learning strategies: In theory and practice. London: Routledge. Griffiths, Carol. 2003. “Patterns of language learning strategy use”, System 31, 3: 367- 383. Griffiths, Carol. 2004. Language-learning strategies: Theory and research. St Helens: Centre for Research in International Education. Griffiths, Carol. 2012.“Learning styles: Traversing the quagmire”, in: Sarah Mercer, Stephen Ryan and Marion Williams (eds.). Psychology for language learning: Insights from research, theory and practice. London: Palgrave Macmillan, 151- 168. Griffiths, Carol. 2013. The strategy factor in successful language learning. Bristol: Multilingual Matters. Griffiths, Carol. 2015. “What have we learnt from ‘good language learners’?”, ELT Journal 69, 4: 425-433. Griffiths, Carol and Rebecca L. Oxford. 2014. “The twenty-first century landscape of language learning strategies: Introduction to this special issue”, System 43: 1-10. Griffiths, Carol (ed.). 2008. Lessons from good language learners. Cambridge: Cam- bridge University Press. Griffiths, Roger. 1991. “Personality and second-language learning: Theory, research and practice”, in: Eugenius Sadtono (ed.), Language acquisition and the second/foreign language. Singapore: Southeast Asian Ministers of Education Or- ganization Regional Language Centre, 103-135. Gross, James J. (ed). 2014. Handbook of emotion regulation. New York: The Guildford Press. Gross, Richard. 2009. Themes, issues, and debates in psychology. (3rd edition). Milton Park: Hodder Education. Groth-Marnat, Gary. 2003. Handbook of psychological assessment. (4th edition.) New Jersey: John Wiley and Sons. Grusec, Joan E. 1992. “Social learning theory and developmental psychology: The lega- cies of Robert Sears and Albert Bandura”, Developmental Psychology 28, 5: 776-786. Gu, Yongqi. 2012. “Learning strategies: Prototypical core and dimensions of variation”, Studies in Self-Access Learning Journal 3, 4: 330-356.

302 Guilford, Joy P. 1950. Fundamental statistics in psychology and education. York: McGraw-Hill Book Company. Guilford, Joy P. 1959. Personality. New York: McGraw-Hill Book Company. Guion, Robert M. 1980. “On trinitarian doctrines of validity”, Professional Psychology: 385-398. Guiora, Alexander Z., Robert C. L. Brannon and Cecelia Y. Dull. “Empathy and second language learning”, Language Learning 22, 1: 111-130. Guiraud, Pierre. 1959. Les caractères statistiques du vocabulaire : essai de méthodolo- gie. [Statistical characteristics of vocabulary: A methodological essay]. Paris : Presses Universitaires de France. Habók, Anita and Andrea Magyar. 2018. “The effect of language learning strategies on proficiency, attitudes and school achievement”, Frontiers in Psychology 8, 1-8. Hajimohammadi, Reza and Jayakaran Mukundan. 2011. “Impact of self-correction on extrovert and introvert students in EFL writing progress”, English Language Teaching 4, 2: 161-168. Hakan, Karatas, Balyer Aydin and Alci Bulent. 2015.“An investigation of undergraduates’ language learning strategies”,Procedia-Social and Behavioral Sciences 197: 1348-1354. Hakimi, Soraya, Elaheh Hejazi and Masoud Gholamali Lavasani. 2011. “The relationships between personality traits and students’ academic achievement”, Procedia- Social and Behavioral Sciences 29: 836-845. Halbach, Ana. 2000. “Finding out about students’ learning strategies by looking at their diaries: A case study”, System 28, 1: 85-96. Han, ZhaoHong (ed.). 2008. Understanding second language process. Clevendon: Mul- tilingual Matters. Harmer, Jeremy. 2001. The practice of English language teaching. Harlow: Longman. Harris, Vee. 2001. Helping learners learn: exploring strategy instruction in language classrooms across Europe. Strasbourg: Council of Europe Publishing. Haynes, Stephen N., David Richard and Edward S. Kubany. 1995.“Content validity in psychological assessment: A functional approach to concepts and methods”, Psychological assessment 7, 3: 238. Hefferon, Kate and Ilona Boniwell. 2011. Positive psychology: Theory, research, and applications. Maidenhead: McGraw-Hill International.

303 Heyde, Adelaide W. 1977. “The relationship between self-esteem and the oral production of a second language”, in: H. Douglas Brown, Carlos A. Yorio and Ruth H. Crymes (eds.), On TESOL ‘77: Teaching and Learning English as a Second Language: Trends in Research and Practice. Washington: Teachers of English to Speakers of Other Languages, 226-240. Higgins, E. Tory. 1996. “Ideas, oughts, and regulatory focus: Affect and motivation from distinct pains and pleasures”, in: Peter M. Gollwitzer and John A. Bargh (eds.). The psychology of action: Linking cognition and motivation to behavior. New York: Guilford Press, 91-114. Higgins, E. Tory. 1997.“Beyond pleasure and pain”, American Psychologist 52, 12: 1280-1300. Hirsh, Jacob B. and Michael Inzlicht. 2008. “The devil you know: Neuroticism predicts neural response to uncertainty”, Psychological Science 19, 10: 962-967. Hjelle, Larry A. and Daniel. J. Ziegler. 1976. Personality theory: Basic assumptions, research, and theory. New York: McGraw-Hill Humanities, Social Sciences & World Languages. Hoare, Carol H. 2002. Erikson on development in adulthood: New insights from the unpublished papers. Oxford: Oxford University Press. Hoff, Joseph G. and R. Michael Paige. 2008.“A strategies-based approach to culture and language learning in education abroad programming”, Frontiers: The Interdisci- plinary Journal of Study Abroad 17: 89-106. Hornowska, Elżbieta. 2007. Testy psychologiczne. Tom 6. Teoria i praktyka. [Psycholo- gical tests. Volume 6. Theory and practice]. (4th edition.) Warszawa: Wydawnictwo Naukowe Scholar. Horwitz, Elaine K. 1985. “Using student beliefs about language learning and teaching in the foreign language methods course”, Foreign Language Annals 18, 4: 333- 340. Horwitz, Elaine K., Michael B. Horwitz and Joann Cope. 1986. “Foreign language classroom anxiety”, The Modern Language Journal 70, 2: 125-132. Hsiao, Tsung–Yuan and Rebecca L. Oxford. 2002. “Comparing theories of language learning strategies: A confirmatory factor analysis”, The Modern Language Journal 86, 3: 368-383. Huffman, Karen. 2008. Psychology in action. Hoboken: John Wiley & Sons.

304 Hurd, Stella. 2006. “Towards a better understanding of the dynamic role of the distance language learner: Learner perceptions of personality, motivation, roles, and approaches”, Distance Education 27, 3: 303-329. Hymes, Dell. 1972. “On communicative competence”, Sociolinguistics 269293: 269- 293. Jackson, Carrie N. 2018. “Second language structural priming: A critical review and directions for future research”, Second Language Research 34, 4: 539-552. Jackson, Joshua J. and Brent W. Roberts. 2017. Conscientiousness. in: Widiger, Thomas. (ed.), The Oxford handbook of the five factor model of personality. Ox- ford: Oxford University Press: 133-150. Jackson, Joshua J., Dustin Wood, Tim Bogg, Kate E. Walton, Peter D. Harms and Brent W. Roberts. 2010. “What do conscientious people do? Development and validation of the Behavioral Indicators of Conscientiousness (BIC)”, Journal of Re- search in Personality 44, 4: 501-511. Jackson, Joshua J., Patrick L. Hill, Brennan R. Payne, Brent W. Roberts, and Elizabeth A. L. Stine-Morrow. 2012. “Can an old dog learn (and want to experience) new tricks? Cognitive training increases openness to experience in older adults”, Psy- chology and Aging 27, 2: 286-292. Jackson, Joshua J., Tim Bogg, Kate E. Walton, Dustin Wood, Peter D. Harms, Jennifer Lodi-Smith, Grant W. Edmonds and Brent W. Roberts. 2009. “Not all conscientiousness scales change alike: A multimethod, multisample study of age differences in the facets of conscientiousness”, Journal of Personality and Social Psy- chology 96, 2: 446-459. Jakobovits, Leon A. 1970. Foreign language learning: A psycholinguistic analysis of the issues. Michigan: Newbury House. Jankowski, Bogusław A. 1973. Nauka języka obcego. [Learning a foreign language]. (1st edition.) Warszawa: Wiedza Powszechna. Jedynak, Małgorzata (ed.). 2017. Specyficzne potrzeby studentów szkół wyższych a nauczanie języków obcych: tom 1. Nowe doświadczenia i wyzwania [Specific needs of university students and foreign language education: Volume 1. New experiences and challenges]. Wrocław: Studium Praktycznej Nauki Języków Obcych Uniwersytetu Wrocławskiego.

305 Jensen, Arthur R. 1958. ”The Maudsley personality inventory”, Acta Psychologica, 14, 1: 314-325. Johnson, David W. and Ardyth A. Norem-Hebeisen. 1979. “A measure of cooperative, competitive, and individualistic attitudes”, The Journal of Social Psychology 109, 2: 253-261. Jones, Craig H. and John R. Slate. 1992. Technical manual for the study habits inventory. [Unpublished ms, Arkansas State University]. Jung, Carl G. 1921. Psychology of the unconscious: A study of the transformations and symbolisms of the libido; a contribution to the history of the evolution of thought. Moffat: Yard and Company. Jurkovič, Violeta. 2010. “Effect of explicit language learning strategy instruction on language-test and self-assessment scores”, English Language Teaching 3, 1: 16- 27. Kamper, Gerrit D., Eric B. Mahlobo and Eleanor Lemmer. 2003. “The relationship between standardised test performance and language learning strategies in English Second Language: A case study”, Journal for Language Teaching 37, 2: 164- 178. Kao, Po-Chi and Philip Craigie. 2014. “Effects of English usage on Facebook and personality traits on achievement of students learning English as a foreign language”, Social Behavior and Personality: An International Journal 42, 1: 17-24. Kaplan, Robert B. (ed.). 2010. The Oxford handbook of applied linguistics. (2nd edition). Oxford: Oxford University Press. Kappe, Rutger, and Henk Van Der Flier. 2012. “Predicting academic success in higher Education: What’s more important than being smart?”, European Journal of Psychology of Education 27, 4: 605-619. Kayaoğlu, Naci M. 2013. “Impact of extroversion and introversion on language-learning behaviors”, Social Behavior and Personality: An International Journal 41, 5: 819-825. Kębłowska, Magdalena. 2012. “The place of affect in second language acquisition”, in: Mirosław Pawlak (ed.). New perspectives on individual differences in language learning and teaching. Berlin: Springer, 157-167. Kelly, George A. 1963. A theory of personality: The psychology of personal constructs. New York: W.W. Norton and Company.

306 Kelly, George A. 1991. The psychology of personal constructs. Vol. 3. London: Routledge. Khamkhien, Attapol. 2010. “Factors affecting language learning strategy reported usage by Thai and Vietnamese EFL learners”, Electronic Journal of Foreign Language Teaching 7, 1: 66-85. Khodaddy, Ebrahim and Reza Zabihi. 2011. “School performance, cultural, social and personality factors and their relationships with majoring in foreign and first languages”, English Language Teaching 4, 3: 63-73. Klassen, Robert M., Lindsey L. Krawchuk and Sukaina Rajani. 2008.“Academic procrastination of undergraduates: Low self-efficacy to self-regulate predicts higher levels of procrastination”, Contemporary Educational Psychology 33, 4: 915- 931. Kline, Paul. 2000. Psychological Testing. (2nd edition.) London: Routledge. Kolb, David. 1984. Experiential learning: Experience as the source of learning and development. Englewood Cliffs: Prentice Hall. Komarraju, Meera, Steven J. Karau, Ronald R. Schmeck and Alen Avdic. 2011. “The Big Five personality traits, learning styles, and academic achievement”, Person- ality and Individual Differences 51, 4: 472-477. Konttinen, Raimo. 1970. Opiskelijoiden englannin kielen taitojen ja niiden oppimisen yhteydet verbaaliseen lahjakkuuteen ja persoonallisuuden piirteisiin. [The relationships between the students’ English language skills and their learning are verbal talent and personality traits]. Jyväskylä: Kasvatustieteiden tutkimuslaitos Jyväskylän yliopisto. Körner, Annett, Zofia Czajkowska, Cornelia Albani, Martin Drapeau, Michael Geyer and Elmar Braehler. 2015.“Efficient and valid assessment of personality traits: Population norms of a brief version of the NEO Five-Factor Inventory (NEO- FFI)”, Archives of Psychiatry and Psychotherapy 1: 21-32. Kostons, Danny, Tamara Van Gog and Fred Paas. 2012. “Training self-assessment and task-selection skills: A cognitive approach to improving self-regulated learning”, Learning and Instruction 22, 2: 121-132. Krahé, Barbara, Jana Becker and Jana Zöllter. 2008.“Contextual cues as a source of response bias in personality questionnaires: The case of the NEO-FFI”, European Journal of Personality 22, 8: 655-673.

307 Krashen, Stephen D. 1981a. “Formal and informal linguistic environments in language TESOL Quarterly 10, 2: 157-168. Krashen, Stephen D. 1981b. Second language acquisition and second language learning. Oxford: Pergamon Press. Krashen, Stephen D. and Herbert W. Seliger. 1976. “The role of formal and informal environments in second language learning: A pilot study”, Linguistics 14, 172: 15-22. Kucharczyk, Radosław. 2017. “The new profile of the students in Romance Philology and their specific needs. A few words about the training strategy”, in: Małgor- zata Jedynak (ed.) Specyficzne potrzeby studentów szkół wyższych a nauczanie języków obcych: tom 1. Nowe doświadczenia i wyzwania [Specific needs of university students and foreign language education: Volume 1. New experiences and challenges]. Wrocław: Studium Praktycznej Nauki Języków Obcych Uni- wersytetu Wrocławskiego, 43-56. Kuder, Frederic G. and Marion W. Richardson. 1937. “The theory of the estimation of test reliability." Psychometrika 2, 3: 151-160. Kunasaraphan, Kanokrat. 2015. “English learning strategy and proficiency level of the first year students”, Procedia-Social and Behavioral Sciences 197: 1853-1858. Lahey, Benjamin B. 2009. “Public health significance of neuroticism”, American Psy- chologist 64, 4: 241-256. Lai, Ying-Chun. 2009. “Language learning strategy use and English proficiency of university freshmen in Taiwan”, TESOL Quarterly 43, 2: 255-280. Lalonde, Richard N. and Richard C. Gardner. 1984. “Investigating a causal model of second language acquisition: Where does personality fit?”, Canadian Journal of Behavioural Science 16, 3: 224-237. Landau, Sabine, and Brian Everitt. 2004. A handbook of statistical analyses using SPSS. Vol. 1. Boca Raton: Chapman & Hall/CRC. Larsen, Randy, David Buss and Andreas Wiesmaijer. 2013. Personality psychology: Domains of knowledge about human nature. New York: McGraw-Hill Educa- tion. Larson-Hall, Jenifer. 2010. A guide to doing statistics in second language research using SPSS. New York and London: Routledge.

308 Lauriola, Marco, Jordan A. Litman, Patrick Mussel, Riccardo De Santis, Howard M. Crowson, and Robert R. Hoffman. 2015. “Epistemic curiosity and self-regulation”, Personality and Individual Differences 83: 202-207. Laursen, Brett, Lea Pulkkinen and Ryan Adams. 2002. “The antecedents and correlates of Agreeableness in adulthood”, Developmental Psychology 38, 4: 591-603. Lawshe, Charles H. 1975. “A quantitative approach to content validity”, Personnel psychology 28, 4: 563-575. Leaver, Betty Lou, Madeline Ehrman and Boris Shekhtman. 2005. Achieving success in second language acquisition. Cambridge: Cambridge University Press. Ledzińska, Maria and Ewa Czerniawska. 2011. Psychologia nauczania. Ujęcie poznaw- cze. [The psychology of teaching. A cognitive approach.] Warszawa: Wydaw- nictwo Naukowe PWN. Lee, Juyeon, and Michael Heinz. 2016. “English language learning strategies reported by advanced language learners”, Journal of International Education Research 12, 2: 67-76. Lee, Rang-Kyoung and Rebecca L. Oxford. 2008. “Understanding EFL learners’ strategy use and strategy awareness”, Asian EFL Journal 10, 1: 7-32. Leino, Anna-Liisa. 1972. “English school achievements and some student characteristics I. On the relationship of personality and intelligence variables to English school achievements”, Research Bulletin 33. Helsinki: Institute of Education, University of Helsinki. Lesiak-Bielawska, Elżbieta D. 2013. Różnice indywidualne w procesie akwizycji języka. [Individual differences in the process of language acquisition]. Warszawa: Ofi- cyna Wydawnicza Łośgraf. Liang, Ting. 2009. “Language learning strategies - the theoretical framework and some suggestions for learner training practice”, English Language Teaching 2, 4: 199- 206. Litman, Jordan A., and Patrick Mussel. 2013.“Validity of the interest-and deprivation- type epistemic curiosity model in Germany”, Journal of Individual Differences 34, 2: 59–68. Liu, Heidi Han-Ting and Young-Sun Lee. 2015.“Measuring self-regulation in second language learning: A Rasch analysis”, SAGE Open 5, 3: 1-12.

309 LoCastro,Virginia. 1994.“Learning strategies and learning environments”, TESOL Quarterly 28, 2: 409-414. Loehlin, John C. and Nicholas G. Martin. “Personality types: A twin study”, Personality and Individual Differences 122: 99-103. Loevinger, Jane. 1957. “Objective tests as instruments of psychological theory”, Psy- chological Reports 3,7: 635-694. Longman Dictionary of Contemporary English. 2018. London: Pearson. (https://www.ldoceonline.com) (date of access: 19 Feb. 2018). Lowie, Wander, Marijn van Dijk, Huiping Chan and Marjolijn Verspoor. 2017. “Find- ing the key to successful L2 learning in groups and individuals”, Studies in Sec- ond Language Learning and Teaching 7, 1: 127-148. Lucas, Richard E. and M. Brent Donnellan. 2011. “Personality development across the life span: longitudinal analyses with a national sample from Germany”, Journal of Personality and Social Psychology 101, 4: 847-867. Lüdtke, Oliver, Brent W. Roberts, Ulrich Trautwein and Gabriel Nagy. 2011. “A random walk down university avenue: life paths, life events, and personality trait change at the transition to university life”, Journal of Personality and Social Psychology 101, 3: 620-637. Lukaszewski, Aaron W. and James R. Roney. 2011. “The origins of extraversion: Joint effects of facultative calibration and genetic polymorphism” Personality and So- cial Psychology Bulletin 37, 3: 409-421. Ma, Rui and Rebecca L. Oxford. 2014.“A diary study focusing on listening and speaking: The evolving interaction of learning styles and learning strategies in a motivated, advanced ESL learner”, System 43: 101-113. Macaro, Ernesto. 2001. Learning strategies in foreign and second language classrooms. London and New York: Continuum. Macaro, Ernesto. 2006. “Strategies for language learning and for language use: Revis- ing the theoretical framework”, The Modern Language Journal 90, 3: 320-337 Macaro, Ernesto. 2008. “The shifting dimensions of language learner autonomy”, in: Terry Lamb and Hayo Reinders (eds.) Learner and teacher autonomy: Concepts, realities and responses. Amsterdam: John Benjamins Publishing Company, 47- 62.

310 MacIntyre, Peter D., Susan C. Baker, Clément, Richard and Sarah Conrod. 2001. “Will- ingness to communicate, social support, and language-learning orientations of immersion students”, Studies in Second Language Acquisition, 23, 3: 369-388. MacIntyre, Peter D. and Robert C. Gardner. 1991. “Language anxiety: Its relationship to other anxieties and to processing in native and second languages”, Language Learning 41, 4: 513-534. MacIntyre, Peter D. and Sarah Mercer. 2014. “Introducing positive psychology to SLA”, Studies in Second Language Learning and Teaching 4,2: 153-173. MacIntyre, Peter D., Zoltán Dörnyei, Richard Clément and Kimberly A. Noels. 1998. “Conceptualizing willingness to communicate in a L2: A situational model of L2 confidence and affiliation”, The Modern Language Journal 82, 4: 545-562. MacIntyre, Peter D, Tammy Gregersen and Sarah Mercer (eds.). 2016. Positive psychology in SLA. Bristol: Channel View Publications. Macmillan Dictionary. 2018. Basingstoke: Macmillan. (https://www.macmillandiction- ary.com/) (date of access: 19 Feb. 2018). Magnavita, Jeffrey J. 2004. Handbook of personality disorders. Hoboken: John Wiley & Sons. Magnusson, David and Bertil Törestad. 1993. “A holistic view of personality: A model revisited”, Annual Review of Psychology 44,1: 427-452. Magogwe, Joel Mokuedi and Rhonda§ Oliver. 2007.“The relationship between language learning strategies, proficiency, age and self-efficacy beliefs: A study of language learners in Botswana”, System 35, 3: 338-352. Marsden, Emma and Rowena Kasprowicz. 2017. “Foreign language educators’ exposure to research: Reported experiences, exposure via Citations, and a proposal for action”, The Modern Language Journal 101, 4: 613-642. Marton, Ference and Roger Säljö. 1976. “On qualitative differences in learning: I - Out- come and process”, British Journal of Educational Psychology 46, 1: 4-11. Marton, Waldemar. 1988. Methods in English language learning. Frameworks and Op- tions. London: Prentice Hall. Maslow, Abraham H. 1999. Toward a psychology of being. New York : J. Wiley & Sons.

311 Matthews, Gerald and Kirby Gilliland. 1999. “The personality theories of HJ Eysenck and JA Gray: A comparative review”, Personality and Individual Differences 26, 4: 583-626. McAdams, Dan P. and Jennifer L. Pals. 2006. “A new Big Five: fundamental principles for an integrative science of personality”, American Psychologist 61, 3: 204-217. McCrae, Robert R. 1993. “Moderated analyses of longitudinal personality stability", Journal of Personality and Social Psychology 65, 3: 577-585. McCrae, Robert R. 2010. “The place of the FFM in personality psychology”, Psycho- logical Inquiry 21, 1: 57-64. McCrae, Robert R., and Paul T. Costa Jr. 1985. “Updating Norman’s “adequacy taxonomy”: Intelligence and personality dimensions in natural language and in questionnaires”, Journal of Personality and Social Psychology 49, 3: 710-721. McCrae, Robert R. and Paul T. Costa Jr. 1996. “Toward a new generation of personality theories: Theoretical contexts for the five-factor model”, in: Jerry S. Wiggins (ed.), The five-factor model of personality: Theoretical perspectives. New York: Guiford Press, 51-87. McCrae, Robert R. and Paul T. Costa Jr. 1997. “Personality trait structure as a human universal”, American Psychologist 52, 5: 509-516. McCrae, Robert R. and Paul T. Costa Jr. 2007. “Brief versions of the NEO-PI-3”, Jour- nal of Individual Differences 28, 3: 116. McCrae, Robert R. and Paul T. Costa Jr. 1988. “Age, personality, and the spontaneous self-concept”, Journal of Gerontology 43, 6: 177-185. McCrae, Robert R. and Paul T. Costa Jr. 2003. Personality in adulthood: A five-factor theory perspective. New York: Guilford Press. McCrae, Robert R., Paul T. Costa, Margarida Pedroso de Lima, António Simões, Fritz Ostendorf, Alois Angleitner, Iris Marušić, Denis Bratko, Gian V. Caprara, Clau- dio Barbaranelli, Joon-Ho Chae, Ralph L. Piedmont. 1999. “Age differences in personality across the adult life span: parallels in five cultures”, Developmental Psychology 35, 2: 466-477. McCrae, Robert. R. and Paul T. Costa Jr. 2008. “Empirical and theoretical status of the five-factor model of personality traits”, The SAGE Handbook of Personality Theory and Assessment 1, 273-294.

312 McCroskey, James C. and Virginia P. Richmond. 1987. “Willingness to communicate”, in: John A. Daly and James C. McCroskey (eds.). Personality and interpersonal communication. Newbury Park: SAGE Publications, 129-156. McGroarty, Mary. 1987. Patterns of University Foreign Language Learning: Elemen- tary Spanish and Japanese. Los Angeles: Center for Language Education and Research, University of California. McLean, Kate C. and Monisha Pasupathi. 2006. “Collaborative narration of the past and extraversion”, Journal of Research in Personality 40, 6: 1219-1231. McNemar, Quinn. 1962. Psychological Statistics. New York and London: John Wiley and Sons. Meisenberg, Gerhard. 2015. “Do we have valid country-level measures of personality?”, Mankind Quarterly 55, 4: 360-382. Messick, Samuel. 1989.“Meaning and values in test validation: The science and ethics of assessment”, Educational Researcher 18, 2: 5-11. Messick, Samuel. 1994. Validity of psychological assessment: Validation of inferences from persons’ responses and performances as scientific inquiry into score meaning. Princeton: Educational Testing Service. Messick, Samuel. 1998. Consequences of test interpretation and use: The fusion of validity and values in psychological assessment. Princeton: Educational Testing Service. Michońska-Stadnik, Anna. 2008. “Identyfikacja oraz trening strategii dla rozwoju autonomii” [Strategy identification and training for autonomy development], in:Pawlak, Mirosław (ed.), Autonomia w nauce języka obcego – co osiągnęliśmy i dokąd zmierzamy [Autonomy in foreign language learning – what we have achieved and where we are heading]. Poznań-Kalisz-Konin: Wydział Pedago- giczno-Artystyczny UAM and Państwowa Wyższa Szkoła Zawodowa w Koni- nie: 393-404. Michońska-Stadnik, Anna. 2017. “Students with language aptitude - challenges for tertiary language teachers”, in: Małgorzata Jedynak (ed.) Specyficzne potrzeby stu- dentów szkół wyższych a nauczanie języków obcych: tom 1. Nowe doświadczenia i wyzwania [Specific needs of university students and foreign language education: Volume 1. New experiences and challenges]. Wrocław: Studium Praktycz- nej Nauki Języków Obcych Uniwersytetu Wrocławskiego, 7-14.

313 Mischel, Harriet N. and Walter Mischel (eds.). 1973. Readings in personality. New York: Holt, Rinehart and Winston. Mischel, Walter, Yuichi Shoda and Ozlem Ayduk. 2008. Introduction to personality: Toward an integrative science of the person. Hoboken: Wiley. Moreno, Audrey and Kevin Bartlett. 2016. “Personality profiles, learning styles, and the Japanese student: An exploratory study”, Kwansei Gakuin University Humani- ties Review 21: 73-80. Morgan, Brianand Matthew Clarke. 2011.“Identity in second language teaching and learning”, in: Eli Hinkel (ed.), Handbook of research in second language learning and teaching. New York: Routledge, 817-836. Muck, Peter M., Benedikt Hell and Samuel D. Gosling. 2007. “Construct validation of a short five-factor model instrument”, European Journal of Psychological Assess- ment 23, 3: 166-175. Murphy, Kevin R. and Charles O. Davidshofer. 2005. Psychological testing: principles and applications. Upper Saddle River: Pearson Education. Murray, Henry A. and Clyde E. Kluckhohn, Clyde E. 1953. Personality in nature, society, and culture. New York: Alfred A. Knopf. Myers, Isabel B. and Katherine Briggs. 1976. The Myers-Briggs Type Indicator. Palo- Alto: Consulting Psychologists Press. Myers, Isabel B. and Mary H. McCaulley. 1985. Manual: A guide to the development and use of the Myers-Briggs Type Indicator. Palo Alto: Consulting Psychologists Press. Mystkowska-Wiertelak, Anna and Mirosław Pawlak. 2014. “Fluctuations in learners’ willingness to communicate during communicative task performance: Condi- tions and tendencies”, Research in Language 12, 3: 245-260. Naiman, Neil, Maria Fröhlich, M., Hans H. Stern and Angie Todesco. 1978. The good language learner. Toronto: Modern Language Center, Ontario Institute for Stud- ies in Education. Neeman, Jennifer and Susan Harter. 1986. Self-perception profile for college students. Denver: University of Denver. Nenkov, Gergana Y., J. Jeffrey Inman and John Hulland. 2008. “Considering the Fu- ture: The Conceptualization and Measurement of Elaboration on Potential Out- comes”, Journal of Consumer Research 35, 1: 126–141.

314 Nettle, Daniel. 2005. “An evolutionary approach to the extraversion continuum", Evolu- tion and Human Behavior 26, 4: 363-373. Newman, Barbara M. and Philip R. Newman. 2017. Development through life: A psychosocial approach. Boston: Cengage Learning. Noftle, Erik E. and Richard W. Robins. 2007. “Personality predictors of academic outcomes: Big five correlates of GPA and SAT scores”, Journal of Personality and Social Psychology 93, 1: 116-130. Norman, Warren T. 1963. “Toward an adequate taxonomy of personality attributes: Replicated factor structure in peer nomination personality ratings”, The Journal of Abnormal and Social Psychology 66, 6: 574-583. Nowakowska, Maria. 1975. Psychologia ilościowa z elementami naukometrii. [Quanti- tative psychology and elements of scientometrics]. Warszawa: Państwowe Wy- dawnictwo Naukowe. Nunan, David. 1997. “Does learner strategy training make a difference?”, Lenguas Modernas 24: 123-142. Nunan, David. 2004. Task-based language teaching. Cambridge: Cambridge University Press. Nye, Christopher D., Mathias Allemand, Samuel D. Gosling, Jeff Potter, and Brent W. Roberts. 2016. “Personality trait differences between young and middle‐aged adults: Measurement artifacts or actual trends?”, Journal of Personality 84, 4: 473-492. Nyikos, Martha. 1987. The effect of color and imagery as mnemonic strategies on learning and retention of lexical items in German. [Unpublished Ph.D. Purdue Univer- sity] Nyikos, Martha. 2008. “Gender and good language learners”, in: Griffiths, Carol (ed.). Les- sons from good language learners. Cambridge: Cambridge University Press, 73-82. Nyikos, Martha and Rebecca L. Oxford. 1993. “A factor analytic study of language learning strategy use: Interpretations from information-processing theory and social psychology”, The Modern Language Journal 77, 1: 11-22. O’Dell, Felicity and Katie Head. 2003. Games for vocabulary practice. Cambridge: Cambridge University Press. O’Malley, Michael J. and Anna U. Chamot. 1990. Learning strategies in second language acquisition. Cambridge: Cambridge University Press.

315 O’Malley, Michael J., Rocco P. Russo, Anna U. Chamot, Gloria Stewener-Manzanares and Lisa Kupper. 1985. “Learning strategies used by high school students learning English as a second language”, Language Learning 35, 1: 21-46. Ockey, Gary J. 2009. “The effects of group members’ personalities on a test taker’s L2 group oral discussion test scores”, Language Testing 26, 2: 161-186. Ockey, Gary J. 2011. “Self-consciousness and assertiveness as explanatory variables of L2 oral ability: A latent variable approach”, Language Learning 61, 3: 968-989. Oettingen, Gebriele and Peter M. Gollwitzer. 2009. “Making goal pursuit effective: Ex- pentancy-dependent goal setting and planned goal striving”, in: Joseph P. For- gas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 127-146. Olivares‐Cuhat, Gabriela. 2002. “Learning strategies and achievement in the Spanish writing classroom: A case study”, Foreign Language Annals 35, 5: 561-570. Oller, John W. 1976a. “A program for language testing research”, in: H. Douglas Brown (ed.), Papers in second language acquisition: Proceedings of the sixth annual conference on applied linguistics at the University of Michigan. Lan- guage Learning Special issue 4, 141-165. Oller, John W. 1976b. “Evidence for a general language proficiency factor: An expec- tancy grammar”, Die Neueren Sprachen 75, 2: 165-174. Oller, John W., Alan J. Hudson and Phyllis Fei Liu. 1977. “Attitudes and attained proficiency in ESL: A sociolinguistic study of native speakers of Chinese in the United States”, Language Learning 27, 1: 1-23. Onwuegbuzie, Anthony J., Phillip Bailey and Christine E. Daley. 1997. Foreign Lan- guage Anxiety among College Students. (Paper presented at the Annual Meeting of the Mid-South Educational Research Association, Memphis, 12 November 1997). Onwuegbuzie, Anthony J., Phillip Bailey and Christine E. Daley. 2000. “Cognitive, affective, personality, and demographic predictors of foreign-language achievement”, The Journal of Educational Research 94, 1: 3-15. Oskarsson, Mats. 1975. The relationship between foreign language proficiency and various psychological variables. (Paper presented at the 4th International Congress of Applied Linguistics, Stuttgart, Germany, 25-30 August 1975).

316 Oxford Dictionary of English. 2018. Oxford: Oxford University Press. (https://en.ox- forddictionaries.com) (date of access: 19 Feb. 2018). Oxford, Rebecca L. 1986. Development and psychometric testing of the Strategy Inven- tory for Language Learning (SILL). Alexandria: U.S. Army Research Institute for the Behavioural and Social Sciences. Oxford, Rebecca L. 1989a. Strategy inventory for language learning. Alexandria, VA: Oxford Associates. Oxford, Rebecca L. 1989b. The role of styles and strategies in second language learning. (https://files.eric.ed.gov/fulltext/ED317087.pdf) (date of access: 21 Aug. 2018). Oxford, Rebecca L. 1989c. “Use of language learning strategies: A synthesis of studies with implications for strategy training”, System 17, 2: 235-247. Oxford, Rebecca L. 1990. Language learning strategies: What every teacher should know. Boston: Heinle and Heinle Publishers. Oxford, Rebecca L. 1993. “Instructional implications of gender differences in second/foreign language (L2) learning styles and strategies”, Applied Language Learning 4: 65-94. Oxford, Rebecca L. 1996. “Employing a questionnaire to assess the use of language learning strategies”, Applied Language Learning 7, 1: 28-47. Oxford, Rebecca L. 1999. “Relationships between second language learning strategies and language proficiency in the context of learner autonomy and self-regulation”, Revista Canaria de Estudios Ingleses 38, 1: 108-126. Oxford, Rebecca L. 2002. “Language learning strategies in a nutshell: Update and ESL suggestions”, in: Jack C. Richards and Willy A. Renandya (eds.). Methodology in language teaching: An anthology of current practice. Cambridge: Cambridge University Press, 124-132. Oxford, Rebecca L. 2003. Language learning styles and strategies: An overview. (http://www.videa.ca/wp-content/uploads/2015/08/LANGUAGE-LEARNING- STYLES-AND-STRATEGIES1.pdf) (date of access: 21 Aug. 2018). Oxford, Rebecca L. 2008. “Hero with a thousand faces: Learner autonomy, learning strategies and learning tactics in independent language learning”, in: Stella Hurd and Tim Lewis (eds.), Language learning strategies in independent settings. Bristol, Multilingual Matters: 41-63.

317 Oxford, Rebecca L. 2011. Teaching and researching language learning strategies. Har- low: Longman. Oxford, Rebecca L. 2013. English teachers helping students become strategic and successful: Ten guidelines. (http://appi.pt/act/congsemin/27cong.htm ) (date of access 30 June 2014). Oxford, Rebecca. 2015. “Emotion as the amplifier and the primary motive: Some theories of emotion with relevance to language learning”, Studies in Second Lan- guage Learning and Teaching 5, 3: 371-393. Oxford, Rebecca L. 2017. Teaching and researching language learning strategies. Self- regulation in context. (2nd edition). London: Routledge. Oxford, Rebecca L. and Martha Nyikos. 1989. “Variables affecting choice of language learning strategies by university students”, The Modern Language Journal 73, 3: 291-300. Oxford, Rebecca L. and Andrew D. Cohen. 1992. “Language learning strategies: Cru- cial issues of concept and classification”, Applied Language Learning 3: 1-35. Oxford, Rebecca L. and Carmen M. Amerstorfer (eds.). 2018. Language learning strategies and individual learner characteristics: Situating strategy use in diverse contexts. London: Bloomsbury Publishing. Oxford, Rebecca L. and Christina Gkonou. 2018.“Interwoven: Culture, language, and learning strategies”, Studies in Second Language Learning and Teaching 8, 2: 403-426. Oxford, Rebecca L. and David Crookall. 1989. “Research on language learning strategies: Methods, findings, and instructional issues”, The Modern Language Jour- nal 73, 4: 404-419. Oxford, Rebecca L. and Judith A. Burry-Stock. 1995.“Assessing the use of language learning strategies worldwide with the ESL/EFL version of the Strategy Inven- tory for Language Learning (SILL)”, System 23, 1: 1-23. Oxford, Rebecca L. and Madeline E. Ehrman. 1995. “Adults’ language learning strategies in an intensive foreign language program in the United States”, System 23, 3: 359-386. Oxford, Rebecca L. and Martha Nyikos. 1989. “Variables affecting choice of language learning strategies by university students”, The Modern Language Journal 73, 3: 291-300.

318 Oxford, Rebecca L., Carol Griffiths, Ana Longhini, Andrew D. Cohen, Ernesto Macaro and Vee Harris. 2014a. “Experts’ personal metaphors and similes about language learning strategies”, System 43: 11-29. Oxford, Rebecca L., Joan Rubin, Anna Uhl Chamot, Karen Schramm, Roberta Lavine, Pamela Gunning, and Carisma Ne. 2014b. “The learning strategy prism: Per- spectives of learning strategy experts”, System 43, 1-4: 30-49. Oxford, Rebecca L., Yunkyoung Cho, Santoi Leung and Kim Hae-Jin. 2004. “Effect of the presence and difficulty of task on strategy use: An exploratory study”, Inter- national Review of Applied Linguistics in Language Teaching 42, 1: 1-48. Oxford, Rebecca, Martha Nyikos and Madeline Ehrman. 1988.“Vive la difference? Re- flections on sex differences in use of language learning strategies”, Foreign Lan- guage Annals 21, 4: 321-329. Oxford, Rebecca L. (ed.). 1996. Language learning strategies around the world: Cross- cultural perspectives. Manoa: Second Language Teaching & Curriculum Center. Oya, Taeko, Emmanuel Manalo and Jackie Greenwood. 2004. “The influence of personality and anxiety on the oral performance of Japanese speakers of English”, Applied Cognitive Psychology 18, 7: 841-855. Oyserman, Daphna and Leah James. 2009. “Possible selves: From content to process”, in: Keith D.Markman, William MP Klein and Julie A. Suhr, Handbook of imagination and mental simulation. New York: Psychology Press, 373-394. Oz, Huseyin. 2014. “Big Five personality traits and willingness to communicate among foreign language learners in Turkey”, Social Behavior and Personality: An In- ternational Journal 42, 9: 1473-1482. Ożańska-Ponikwia, Katarzyna and Jean-Marc Dewaele. 2012. “Personality and L2 use: The advantage of being open-minded and self-confident in an immigration context”, EUROSLA Yearbook 12, 1: 112-134. Paas, Fred GWC and Jeroen JG Van Merriënboer. 1993. “The efficiency of instructional conditions: An approach to combine mental effort and performance measures”, Human Factors 35, 4: 737-743. Panksepp, Jaak. 2008. “The power of the word may reside in the power of affect”, Inte- grative Psychological and Behavioral Science 42, 1: 47-55. Park, Gi-Pyo. 2011. “The validation process of the SILL: A confirmatory factor analysis”, English Language Teaching 4, 4: 21-27.

319 Parel, Rolande. 2004. “The impact of lexical inferencing strategies on second language reading proficiency”, Reading and Writing 17, 6: 847-873. Pawlak, Mirosław. 2010. “Designing and piloting a tool for the measurement of the use of pronunciation learning strategies”, Research in Language 8: 1-14. Pawlak, Mirosław. 2011. “Research into language learning strategies: Taking stock and looking ahead”, in: Janusz Arabski and Adam Wojtaszek (eds.), Individual learner differences in SLA. Bristol: Multilingual Matters, 17-37. Pawlak, Mirosław. 2012. “The dynamic nature of motivation in language learning: A classroom perspective”, Studies in Second Language Learning and Teaching 2, 2: 249-278. Pawlak, Mirosław. 2018. “Investigating the use of speaking strategies in the performance of two communicative tasks: The importance of communicative goal”, Studies in Second Language Learning and Teaching 8, 2: 269-291. Pawlak, Mirosław and Anna Mystkowska-Wiertelak. 2015. “Investigating the dynamic nature of L2 willingness to communicate”, System 50: 1-9. Pawlak, Mirosław and Rebecca Oxford. 2018. “Conclusion: The future of research into language learning strategies”, Studies in Second Language Learning and Teach- ing 8, 2: 525-535. Pawlak, Mirosław and Zuzanna Kiermasz. 2018. “The use of language learning strategies in a second and third language: The case of foreign language majors”, Stud- ies in Second Language Learning and Teaching 8, 2: 427-443. Pawlak, Mirosław (ed.). 2004. Autonomia w nauce języka obcego [Autonomy in foreign language learning]. Poznań-Kalisz: Wydział Pedagogiczno-Artystyczny UAM w Poznaniu Pawlak, Mirosław (ed.). 2008a. Autonomia w nauce języka obcego – co osiągnęliśmy i dokąd zmierzamy? [Autonomy in foreign language learning – what have we ac- complished and where are we headinfg?]. Poznań-Kalisz-Konin: Wydział Peda- gogiczno-Artystyczny UAM w Poznaniu and Państwowa Wyższa Szkoła Zawo- dowa w Koninie. Pawlak, Mirosław (ed.). 2008b. Investigating English language learning and teaching. Kalisz: Faculty of Fine Arts and Pedagogy in Kalisz. Pawlak, Mirosław (ed.). 2011a. Autonomia w nauce języka obcego – uczeń a nauczyciel [Autonomy in foreign language learning – the language learner and the language

320 teacher]. Poznań-Kalisz-Konin: Wydział Pedagogiczno-Artystyczny UAM w Poznaniu and Państwowa Wyższa Szkoła Zawodowa w Koninie. Pawlak, Mirosław (ed.). 2011b. Extending the Boundaries of Research on Second Lan- guage Learning and Teaching. Berlin Heidelberg: Springer. Pawlak, Mirosław (ed.). 2012. New perspectives on individual differences in language learning and teaching. Heidelberg: Springer. Peabody, Dean and Boele De Raad. 2002. “The substantive nature of psycholexical personality factors: a comparison across languages”, Journal of Personality and So- cial Psychology 83, 4: 983-997. Peacock, Matthew, and Belinda Ho. 2003. “Student language learning strategies across eight disciplines”, International Journal of Applied Linguistics 13, 2: 179-200. Perugini, Marco and Marcello Gallucci. 1997. “A hierarchical faceted model of the Big Five”, European Journal of Personality 11, 4: 279-301. Phillips, June K. 1991. “Upgrading the target language proficiency levels of foreign language teachers”, ERIC Digest (https://files.eric.ed.gov/fulltext/ED328082.pdf) (date of access 05May 2019). Pica, Teresa, Ruth Kanagy and Joseph Falodun. 1993. “Choosing and using communication tasks for second language instruction”, in: Graham Crookes and Susan M. Gass (eds.). Tasks and language learning: integrating theory and practice. Clevedon: Multilingual Matters, 9-31. Piechurska-Kuciel, Ewa. 2008. Language anxiety in secondary grammar school students. Opole: Wydawnictwo Uniwersytetu Opolskiego. Piechurska-Kuciel, Ewa. 2018. “Openness to experience as a predictor of L2 WTC”, System 72, 190-200. Pieschl, Stephanie, Elmar Stahl, Tom Murray and Rainer Bromme. 2012.“Is adaptation to task complexity really beneficial for performance?”, Learning and Instruction 22, 4: 281-289. Pineda, Jorge E. 2010. “Identifying language learning strategies: An exploratory study”, Gist Education and Language Research Journal 4, 1: 94-106. Pintrich, Paul R. 2000. “The role of goal orientation in self-regulated learning”, in: Monique Boekaerts, Paul R. Pintrich and Moshe Zeidner (eds.). Handbook of self-regulation. Burlington: Elsevier Inc., 451-502.

321 Pintrich, Paul R. and Elisabeth Van De Groot. 1990. “Motivational and self-regulated learning components of classroom academic performance”, Journal of Educa- tional Psychology 82, 1: 33-40. Pintrich, Paul R. and Teresa Garcia. 1991a. “Student goal orientation and self-regulation in the college classroom”, in: Paul R. Pintrich, Don Brown and Claire E. Wein- stein (eds.)., Student motivation, cognition, and learning: Essays in honor of Wilbert J. McKeachie. Hillsdale: Lawrence Erlbaum Associates, 371-402. Pintrich, Paul R., David A. F. Smith, Teresa Garcia and Wilbert J. McKeachie. 1991b. A manual for the use of the Motivated Strategies for Learning Questionnaire (MSLQ). Washington: National Center for Research to Improve Postsecondary Teaching and Learning, Ann Arbor. Plimsleur, Paul, Ludwig Mosberg and Andrew L. Morrison. 1962. “Student factors in foreign language learning”, The Modern Language Journal 46, 4: 160-170. Politzer, Robert L. 1983. “An exploratory study of self-reported language learning behaviors and their relation to achievement”, Studies in Second Language Acquisi- tion 6, 1: 54-68. Politzer, Robert L. and Mary McGroarty. 1985. “An exploratory study of learning behaviors and their relationship to gains in linguistic and communicative competence”, TESOL Quarterly 19, 1: 103-123. Poropat, Arthur E. 2009. “A meta-analysis of the five-factor model of personality and academic performance”, Psychological Bulletin 135, 2: 322-338. Pourfeiz, Jafar. 2015. “Exploring the relationship between global personality traits and attitudes toward foreign language learning”, Procedia-Social and Behavioral Sciences 186: 467-473. Power, Robert A. and Michael Pluess. 2015. “Heritability estimates of the Big Five personality traits based on common genetic variants”, Translational Psychiatry 5, 7: 1-4. Pressley, Michael and Joel R. Levin. 1977. “Developmental differences in subjects’ as- sociative-learning strategies and performance: Assessing a hypothesis”, Journal of Experimental Child Psychology 24, 3: 431-439. Przybył, Jakub. 2016. “Cultural variation in the use of language learning strategies: A comparative study”, Konińskie Studia Językowe 4, 4: 439-461.

322 Przybył, Jakub. 2017. “The determinants of strategy use by students learning English at Poznan universities”, Neofilolog 48, 1: 89-103. Purpura, James E. 1997. “An analysis of the relationships between test takers’ cognitive and metacognitive strategy use and second language test performance”, Lan- guage Learning 47, 2: 289-325. Purpura, James E. 1998. “Investigating the effects of strategy use and second language test performance with high-and low-ability test takers: A structural equation modelling approach”, Language Testing 15, 3: 333-379. Quilty, Lena C., Colin G. DeYoung, Jonathan M. Oakman and R. Michael Bagby. 2014. “Extraversion and behavioral activation: integrating the components of approach”, Journal of Personality Assessment 96, 1: 87-94. Ramírez-Esparza, Nairán, Kathryn Harris, John Hellermann, Clemence Richard, Patri- cia K. Kuhl and Steve Reder. 2012. “Socio-interactive practices and personality in adult learners of English with little formal education”, Language Learning 62, 2: 541-570. Ranta, Leila. 2008. “Aptitude and good language learners”, in: Carol Griffiths (ed.). 2008. Lessons from good language learners. Cambridge: Cambridge University Press, 142-155. Reinders, Hayo. 2011. “Towards a classroom pedagogy for learner autonomy: A framework of independent language learning skills”, Australian Journal of Teacher Education (Online) 35, 5: 40-55. Renau, Vanessa Ruiz, Ursula Oberst, Samuel D. Gosling, Jordi Rusiņol and Andrés Chamarro Lusar. 2013. “Translation and validation of the Ten-Item-Personality Inventory into Spanish and Catalan”, Aloma: Revista de Psicologia, Cičncies de l’educació i de l’esport 31, 2: 85-97. Rennie, David L. 2007. “Methodical hermeneutics and humanistic psychology”, The Humanistic Psychologist 35,1: 1-14. Revelle, William. 1997. “Extraversion and impulsivity: The lost dimension”, in: Ny- borg, Helmuth. The scientific study of human nature: Tribute to Hans J. Eysenck at eighty. Oxford: Elsevier Science: 189-212. Revelle, William and Joshua Wilt. 2008. “Personality is more than reinforcement sensitivity”, European Journal of Personality, 22, 5: 407-409.

323 Richards, Jack C. 2002. “Theories of teaching in language teaching”, in: J.C. Richards and W.A. Renadya (eds.). Methodology in language teaching. An anthology of current practice: 19-25, Cambridge: Cambridge University Press. Richards, Jack C. 2015. Key issues in language teaching. Cambridge: Cambridge Uni- versity Press. Richards, Keith. 2009. “Trends in qualitative research in language teaching since 2000”, Language Teaching 42, 2: 147-180. Riding, Richard and Indra Cheema. 1991. “Cognitive styles—an overview and integration”, Educational Psychology 11, 3-4: 193-215. Riemann, Rainer, Alois Angleitner and Jan Strelau. 1997. “Genetic and environmental influences on personality: A study of twins reared together using the self-and peer report NEO-FFI scales”, Journal of Personality 65, 3: 449-475. Rigney, Joseph W. 1978. “Learning strategies: A theoretical perspective”, in: Harold F. O’Neil (ed.). Learning strategies. New York: Academic Press, Inc.: 165-205. Ritchie, Stuart J., Michelle Luciano, Narelle K. Hansell, Margaret J. Wright and Timo- thy C. Bates. 2013. “The relationship of reading ability to creativity: Positive, not negative associations”, Learning and Individual Differences 26: 171-176. Roberts, Brent W. and Daniel Mroczek. 2008. “Personality trait change in adulthood”, Current Directions in Psychological Science 17, 1: 31-35. Roberts, Brent W. and Wendy F. DelVecchio. 2000. “The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies”, Psychological Bulletin 126, 1: 3-25. Roberts, Brent W., Oleksandr S. Chernyshenko, Stephen Stark and Lewis R. Goldberg. 2005. “The structure of conscientiousness: An empirical investigation based on seven major personality questionnaires”, Personnel Psychology 58, 1: 103-139. Robins, Richard W., R. Chris Fraley, Brent W. Roberts and Kali H. Trzesniewski. 2001. “A longitudinal study of personality change in young adulthood”, Journal of Personality 69, 4: 617-640. Robinson, Peter (ed.). 2002. Individual differences and instructed language learning: Language learning and language teaching. Vol. 2. Amsterdam: John Benjamins. Rogers, Carl R. 1959. “A theory of therapy, personality and interpersonal relationships as developed in the client-centered framework”, in: Sigmund Koch (ed.), Psy- chology: A Study of a Science. Vol. 3. New York: McGraw-Hill, 184-256.

324 Rogers, Carl R. 1969. Freedom to learn. A view of what education might become. Co- lumbus: Charles E. Merrill. Rogers, Carl R. 1981. “The foundations of the person-centered approach”, Dialectics and Humanism 8, 1: 5-16. Rogers, Carl R. 2011. On becoming a person: A therapist’s view of psychotherapy. Bos- ton: Houghton Mifflin Company. Rohwer Jr, William D. and Joan P. Bean. 1973. “Sentence effects and noun-pair learning: A developmental interaction during adolescence”, Journal of Experimental Child Psychology 15, 3: 521-33. Romero, Eq. 1995. “From Darwin to Skinner - learning psychology and operant conditioning historical genesis”, Psicothema 7, 3: 543-556. Rose, Heath. 2012. “Reconceptualizing strategic learning inthe face of self-regulation: throwing language learning strategies out with the bathwater”, Applied Linguis- tics 33, 1: 92–98. Rose, Heath. 2015. “Researching language learner strategies”, in: Brian Paltridge and Aek Phakiti (eds.). 2015. Research methods in applied linguistics: A practical resource. London: Bloomsbury Publishing: 421-438. Rose, Heath and Lesley Harbon. 2013. “Self-regulation in second language learning: An investigation of the kanji learning task”, Foreign Language Annals 46, 1: 96- 107. Rose, Heath, Jessica G. Briggs, Jill A. Boggs, Lia Sergio and Natalia Ivanova-Slavian- skaia. 2018. “A systematic review of language learner strategy research in the face of self-regulation”, System 72: 151-163. Rothbart, Mary K., Stephan A. Ahadi, Karen L. Hershey and Phillip Fisher. 2001. “In- vestigations of temperament at three to seven years: The Children’s Behavior Questionnaire”, Child Development 72, 5: 1394-1408. Rubin, Joan. 1975. “What the “good language learner” can teach us”, TESOL Quarterly, March, 1: 41-51. Rubin, Joan. 1981. “Study of cognitive processes in second language learning”, Applied Linguistics 2, 2: 117-131. Rubin, Joan. 1987. “Learner strategies: Theoretical assumptions, research history and typology”, in: Anita Wenden and Joan Rubin (eds.). Learner strategies in language learning. New York: Prentice Hall International, 15-30.

325 Rubin, Joan. 2008. “Reflections”, in: Carol Griffiths (ed.). 2008. Lessons from good language learners. Cambridge: Cambridge University Press: 10-15. Ryan, Richard M. and Edward L. Deci. 2001.“On happiness and human potentials: A review of research on hedonic and eudaimonic well-being”, Annual Review of Psychology 52, 1: 141-166. Ryan, Stephen. 2005. Motivational factors questionnaire. Nottingham: School of Eng- lish Studies, University of Nottingham. Rychlak, Joseph F. 1968. A philosophy of science for personality theory. Boston: Houghton, Mifflin and Company. Salahshour, Farzad, Mahnaz Sharifi and Neda Salahshour. 2013. “The relationship between language learning strategy use, language proficiency level and learner gender”, Procedia-Social and Behavioral Sciences 70: 634-643. Säljö, Roger. 1979. “Learning about learning” Higher Education 8, 4: 443-451. Sampson, Richard. 2012. “The language-learning self, self-enhancement activities, and self-perceptual change”, Language Teaching Research, 16, 3: 317-335. Sansone, Carol. 2009. “What’s interest got to do with it? Potential trade-offs in the self- regulation of motivation”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 35- 52. Sansone, Carol and Judith M. Harackiewicz (eds.). 2000.Intrinsic and extrinsic motivation: The search for optimal motivation and performance. New York: Adademic Press. Schacter, Daniel, Daniel Gilbert, Daniel Wegner. 2012. Psychology: European edition. London: Palgrave Macmillan. Schaller, Mark and Damian R. Murray. 2008. “Pathogens, personality, and culture: Dis- ease prevalence predicts worldwide variability in sociosexuality, extraversion, and openness to experience”, Journal of Personality and Social Psychology 95, 1: 212-221. Scheier, Michael F. and Charles S. Carver. 1992. “Effects of optimism on psychological and physical well-being: Theoretical overview and empirical update”, Cognitive Therapy and Research 16,2: 201-228. Schmenk, Barbara. 2005. “Globalizing learner autonomy”, TESOL Quarterly 39, 1: 107-118.

326 Schmidt, Richard. 1994. “Deconstructing consciousness in search of useful definitions for applied linguistics”, AILA Review 11: 11-16. Schmidt, Richard. 1995. Attention and awareness in foreign language learning. Manoa: University of Hawai’i Second Language Teaching and Curriculum Center. Schmidt, Richard and Yuichi Watanabe. 2001. “Motivation, strategy use, and pedagogical preferences in foreign language learning”, Motivation and Second Language Acquisition 23: 313-359. Schultz, Duane P. 1981. Theories of personality. Belmont: Wadsworth. Schultz, Duane P. and Sydney Ellen Schultz. 2005. Theories of personality. Boston: Cengage Learning. Schultze, Ulrike, and Michel Avital. 2011. “Designing interviews to generate rich data for information systems research”, Information and organisation 21, 1: 1-16. Schumann, John H. 1975. “Affective factors and the problem of age in second language acquisition”, Language learning 25, 2: 209-235. Schunk, Dale H. and Barry J. Zimmerman. 1994. Self-regulation of learning and performance: Issues and educational applications. Hillsdale: Lawrence Erlbaum Asso- ciates, Inc. Schwartz, H. Andrew, Johannes C. Eichstaedt, Margaret L. Kern, Lukasz Dziurzynski, Stephanie M. Ramones, Megha Agrawal, Achal Shah, Michal Kosinski, David Stillwell, Martin E. P. Seligman and Lyle H. Ungar. 2013. “Personality, gender, and age in the language of social media: The open-vocabulary approach”, PloS One 8, 9: e73791. Sedikides, Constantine. 2009. “On self-protection and self-enhancement regulation: The role of self-improvement and social norms”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 73-92. Seker, Meral. 2016. “The use of self-regulation strategies by foreign language learners and its role in language achievement”, Language Teaching Research 20, 5: 600- 618. Selinker, Larry.1972. “Interlanguage”, International Review of Applied Linguistics in Language Teaching 10, 1-4: 209–231. Selinker, Larry, Merrill Swain, and Guy Dumas. 1975. “The interlanguage hypothesis extended to children”, Language Learning 25, 1: 139-152.

327 Selvam, Puveneswary, Ernie Syahida Binti, Mohd Kamal,Vincent Nathan G. Swamina- than and Shasthrika Baskaran. 2016. Effects of debilitative and facilitative anxiety on speaking in second language among Malaysian ESL learners. (http://webcache.googleusercon- tent.com/search?q=cache:http://www.umcced.edu.my/psh2016/web_pa- per/2.3.1.pdf&gws_rd=cr&dcr=0&ei=s1tkWuCxFc_TsAeyhbvgDw) (date of access: 21 Jan. 2017). Seymour, David and Maria Popova. 2005. 700 classroom activities. Oxford: Macmillan. Sharp, Daryl. 1998. “Jungian psychology unplugged: My life as an elephant”. Studies in Jungian Psychology by Jungian Analysts, Toronto: Inner City Books. Sharp, Emily Schoenhofen, Chandra A. Reynolds, Nancy L. Pedersen and Margaret Gatz. 2010. “Cognitive engagement and cognitive aging: Is openness protec- tive?”, Psychology and Aging 25, 1: 60-73. Singleton, David and Justyna Leśniewska. 2012. “Age and SLA: Research highways and bye-ways”, in: Mirosław Pawlak (ed.). New perspectives on individual differences in language learning and teaching. Berlin: Springer: 97-113. Sitzman, Traci and Katherine Ely. 2010. “Sometimes you need a reminder: The effects of prompting self-regulation on regulatory processes, learning, and attrition” Journal of Applied Psychology 95, 1: 132-144. Skehan, Peter. 1989. Individual differences in language learning. London: Edward Ar- nold. Skehan, Peter. 1991.“Individual differences in second language learning”, Studies in Second Language Acquisition 13, 2: 275-298. Skehan, Peter. 1996. “A framework for the implementation of task-based instruction”, Applied Linguistics 17, 1: 38-62. Skehan, Peter. 2002.“Theorising and updating aptitude”, in: Peter Robinson (ed.). Indi- vidual differences and instructed language learning: Language learning and language teaching. Vol. 2. Amsterdam: John Benjamins, 69-94. Skinner, Burrhus F. 1953. Science and human behavior. New York: Simon and Schus- ter. Skinner, Burrhus. F. 1974. About behaviorism. New York: Random House, Inc.

328 Smith, M. Brewster. 1950. “The phenomenological approach in personality theory: Some critical remarks”, The Journal of Abnormal and Social Psychology 45, 3: 516-522. Sneed, Carl D., Robert R. McCrae and David C. Funder. 1998. “Lay conceptions of the five-factor model and its indicators”, Personality and Social Psychology Bulletin 24, 2: 115-126. Snyder, Charles R., and Shane J. Lopez (eds.). 2002. The handbook of positive psychology. Oxford: Oxford University Press. Sorokowska, Agnieszka, Aleksandra Słowińska, Anita Zbieg and Piotr Sorokowski. 2014. Polska adaptacja testu Ten Item Personality Inventory (TIPI)–TIPI-PL– wersja standardowa i internetowa. [Polish adaptation of the Ten Item Personal- ity Inventory (TIPI): TIPI-PL: Standard and internet version]. Wrocław: Wroc- lab. Soto, Christopher J. and Oliver P. John. 2014. “Traits in transition: The structure of par- ent-reported personality traits from early childhood to early adulthood”, Journal of Personality 82, 3: 182-199. Spielberger, Charles D. 1962. The effects of manifest anxiety on the academic achievement of college students. New York: Mental Hygiene. Staats, Walter W. 1971. Child learning, intelligence, and personality. New York: Har- per & Row. Staats, Walter W. 1996. Behavior and personality: Psychological behaviorism. New York: Springer. Stelmack, Robert M. 1990. “Biological bases of extraversion psychophysiological evidence”, Journal of Personality 58, 1: 293-311. Stern, Hans H. 1975. “What can we learn from the good language learner?”, Canadian Modern Language Review 31, 4: 304-318. Stern, Hans H. 1992. Issues and options in language teaching. Oxford: Oxford Univer- sity Press. Sternberg, Robert J. and Karin Sternberg. 2012. Cognitive psychology. (6th edition.) Wadsworth: Cengage Learning. Stevick, Earl W. 1989. Success with foreign languages: Seven who achieved it and what worked for them. Hertfordshire: Prentice Hall. Strelau, Jan. 1983. Temperament - personality - activity. London: Academic Press.

329 Strelau, Jan and Bogdan Zawadzki. 2008. “Temperament from a psychometric perspective: Theory and measurement”, in: Gregory J. Boyle, Gerald Matthews and Donald H. Saklofske (eds.). The SAGE handbook of personality theory and assessment: Personality measurement and testing. (Vol. 2). London: Sage Publica- tions Ltd., 352-73. Strelau, Jan (ed.). 2008. Psychologia. Podręcznik akademicki. [Psychology. An academic course book]. Gdańsk: Gdańskie Wydawnictwo Psychologiczne. Strong, Michael. 1983. “Social styles and the second language acquisition of Spanish- speaking kindergartners”, TESOL Quarterly 17, 2: 241-258. Strunk, Kamden K. and Misty R. Steele. 2011. “Relative contributions of self-efficacy, self-regulation, and self-handicapping in predicting student procrastination”, Psychological Reports 109, 3: 983-989. Stuart-Hamilton, Ian. 2007. Dictionary of psychological testing, assessment and treatment. London and Philadelphia: Jessica Kingsley Publishers. Studenska, Anna. 2011a. “Educational level, gender and foreign language learning self- regulation difficulty”, Procedia-Social and Behavioral Sciences, 29: 1349-1358. Studenska, Anna. 2011b. “Personality and parenting styles as predictor of self-regulation in foreign language learning”, in Janusz Arabski and Adam Wojtaszek (eds.), Individual differences in second language acquisition. Bristol: Multilin- gual Matters, 74-94. Sutin, Angelina R. 2017. “Openness”, in: Thomas Widiger (ed.), The Oxford handbook of the five factor model of personality. Oxford: Oxford University Press, 83-104. Szyszka, Magdalena. 2015. “Good English pronunciation users and their pronunciation learning strategies”, Research in Language 13, 1: 93-106. Tackett, Jennifer L. and Benjamin B. Lahey. 2017. “Neuroticism”, in: Thomas Widiger (ed.), The Oxford handbook of the five factor model of personality. Oxford: Ox- ford University Press, 39-56. Taguchi, Tatsuya, Michael Magidand Mostafa Papi. 2009. ”The L2 motivational self systemamong Japanese, Chinese and Iranian learners of English: A compara- tivestudy”, in: Zoltán Dörnyei and Ema Ushioda (eds.), Motivation, language identityand the L2 self. Bristol: Multilingual Matters, 66-97. Takeuchi, Osamu. 1993. “Language learning strategies and their relationship to achievement in English as a foreign language”, Language Laboratory 30, 17-34.

330 Takeuchi, Osamu, Carol Griffiths and Do Coyle. 2007. “Applying strategies to contexts: the role of individual, situational, and group differences”, in: Andrew D. Cohen and Ernesto Macaro (eds.). Language learner strategies. Oxford: Oxford Uni- versity Press, 69-92. Tangney, June P., Roy Baumeister and Boone, Angie L. 2004.” High self-control predicts good adjustment, less pathology, better grades, and interpersonal success”, Journal of Personality 72, 2: 271–324. Tarone, Elaine. 1981. “Some thoughts on the notion of communication strategy”, TESOL Quarterly 15, 3: 285-295. Tarone, Elaine. 1983. “On the variability of interlanguage systems”, Applied Linguistics 4, 2: 142-164. Taylor, Alan, John R. Stevens and J. William Asher. 2006. “The effects of Explicit Reading Strategy Training on L2 reading comprehension: A meta-analysis”, in: Lourdes Ortega and John Michael Norris (eds.). Synthesizing research on language learning and teaching. Amsterdam: John Benjamins Publishing Co, 213- 244. Taylor, Linda L., John C. Catford, Alexander Z. Guiora and Harlan L. Lane. 1971. “Psychological variables and ability to pronounce a second language”, Language and Speech 14, 2: 146-157. Tellegen, Auke. 1985. “Structures of mood and personality and their relevance to assessing anxiety, with an emphasis on self-report”, in: Tuma, A. Hussain and Jack D. Maser, (eds.), Anxiety and the anxiety disorders. Hillsdale: Lawrence Erl- baum Associates, 681-706. Tellegen, Auke, David T. Lykken, Thomas J. Bouchard, Kimerly J. Wilcox, Nancy L. Segal and Stephen Rich. 1988. “Personality similarity in twins reared apart and together”, Journal of Personality and Social Psychology 54, 6: 1031-1039. Teng, Lin S. and Lawrence J. Zhang. 2016. “A questionnaire-based validation of multidimensional models of self-regulated learning strategies”, The Modern Lan- guage Journal 100: 674-701. Terracciano, Antonio, Robert R. McCrae and Paul T. Costa. 2010. “Intra-individual change in personality stability and age”, Journal of Research in Personality 44, 1: 31-37.

331 Thiede, Keith W., Mary Anderson and David Therriault. 2003.“Accuracy of metacognitive monitoring affects learning of texts”, Journal of Educational Psychology 95, 1: 66-73.. Thorndike, Edward. 1911. “A study in memorizing various materials by the reconstruc- tion method”, Psychological Bulletin 8, 7: 254-256. Thorndike, Edward. 1931. “On the learning of rules in the study of a foreign language”, The German Quarterly 4, 3: 89-95. Thurstone, Louis L. 1934. The theory of multiple factors. Oxford: Edwards Bros. Tice, Dianne M. 2009. “How emotions affect self-regulation”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation.New York: Psychology Press, 201-216. Tragant, Elsa and Mia Victori. 2006. “Reported strategy use and age”, Age and the Rate of Foreign Language Learning 19: 208-236. Tragant, Elsa and Mia Victori. 2012. “Language learning strategies, course grades, and age in EFL secondary school learners”, Language Awareness 21, 3: 293-308. Tragant, Elsa, Marilyn S. Thompson and Mia Victori. 2013. “Understanding foreign language learning strategies: A validation study”, System 41, 1: 95-108. Trapmann, Sabrina, Benedikt Hell, Jan-Oliver W. Hirn and Heinz Schuler. 2007. “Meta-analysis of the relationship between the Big Five and academic success at university”, Zeitschrift für Psychologie [Journal of Psychology] 215, 2: 132-151. Trice, Ashton D. 1985. “An academic locus of control scale for college students”, Per- ceptual and Motor Skills 61, 3: 1043-1046. Tseng, Wen-Ta, Zoltán Dörnyei and Norbert Schmitt. 2006.“A new approach to assessing strategic learning: The case of self-regulation in vocabulary acquisition”, Applied Linguistics 27, 1: 78-102. Tupes, Ernest C. and Raymond C. Christal. 1958. Stability of personality trait rating factors obtained under diverse conditions. USAF Wright Air Development Cen- ter. Tupes, Ernest C. and Raymond E. Christal. 1961. Recurrent personality factors based on trait ratings. No. ASD-TR-61-97. United States Defense Technical Infor- mation Center. Tyszkowa, Maria. 1990. Zdolności, osobowość i działalność uczniów. [Learners’ abilities, personality and activity], Warszawa: Wydawnictwo Naukowe PWN.

332 Uhrig, Karl. 2015. “Learning styles and strategies for language use in the context of academic reading tasks”, System 50: 21-31. Usuki, Miyuki. 2000. Promoting learner autonomy: Learning from the Japanese language learners’ perspectives, Washington: ERIC Clearinghouse on Language and Linguistics. Uztosun, Mehmet Sercan. 2014.“The impact of language learning experience on language learner strategy use in Turkish EFL context”, International Journal on New Trends in Education and Their Implications 5, 1: 157-168. Van Daele, Siska, Alex Housen, Michel Pierrard and Luc De Bruyn. 2006. “The effect of extraversion on oral L2 proficiency”, EUROSLA Yearbook 6, 1: 213-236. Vandergrift, Larry, Christine Goh, Catherine J. Mareschal and Marzieh H. Tafaghodtari. 2006. “The metacognitive awareness listening questionnaire: Development and validation”, Language Learning 56, 3: 431-462. Vandewaetere, Mieke and Piet Desmet. 2009. “Introducing psychometrical validation of questionnaires in CALL research: The case of measuring attitude towards CALL”, Computer Assisted Language Learning 22, 4: 349-380. Varasteh, Hanieh, Afsaneh Ghanizadeh and Omid Akbari. 2016. “The role of task value, effort-regulation, and ambiguity tolerance in predicting EFL learners’ test anxiety, learning strategies, and language achievement”, Psychological Studies 61, 1: 2-12. Vefali, Gülşen M. 2008. “Do we know our language learners?”, in: Mirosław Pawlak (ed.). Investigating English language learning and teaching. Poznań-Kalisz: Faculty of Pedagogy and Fine Arts in Kalisz, Adam Mickiewicz University in Poznań, 293-306. Véliz, Mauricio C. 2012. “Language learning strategies (LLSs) and L2 motivation associated with L2 pronunciation”, Literatura y Lingüística 25: 193-220. Verhoeven, Ludo and Anne Vermeer. 2002. “Communicative competence and personality dimensions in first and second language learners”, Applied Psycholinguistics 23, 3: 361-374. Vohs, Kathleen D. and Brandon J. Schmeichel. 2003. “Self-regulation and extended now: Controlling the self alters the subjective experience of time”, Journal of Personality and Social Psychology 85, 2: 217-230.

333 Vohs, Kathleen D., Jannine D. Lasaleta and Bob Fennis. 2009. “Self-regulation in the interpersonal sphere”, in: Joseph P. Forgas, Roy F. Baumeister and Dianne M. Tice (eds.). Psychology of self-regulation. New York: Psychology Press, 289- 302. Vygotsky, Lev.S. 1978. Mind in society: The development of higher psychological processes. Cambridge, MA: MIT Press. Vygotsky, Lev.S. 1986. Thought and language. Cambridge, MA: MIT Press. Wakabayashi, Akio. 2014.“A sixth personality domain that is independent of the Big Five domains: The psychometric properties of the HEXACO Personality Inven- tory in a Japanese sample”, Japanese Psychological Research 56, 3: 211-223. Wakamoto, Natsumi. 2000. “Language learning strategy and personality variables: Fo- cusing on extroversion and introversion”, IRAL 38, 1: 71-81. Waninge, Freerkien, Zoltan Dörnyei and Kees De Bot. 2014. “Motivational dynamics in language learning: Change, stability, and context”, The Modern Language Jour- nal 98, 3: 704-723. Warren, Ricks, Robert McLellarn and Albert Ellis. 1987. “Albert Ellis’ personal responses to the Survey of Rational-Emotive Therapists”, Journal of Rational Emotive Therapy 5 ,2: 92-107. Waterman, Alan S. 2013. “The humanistic psychology-humanistic psychology divide”, American Psychologist 68, 3: 124-133. Waterman, Alan S. 2014. “Further reflections on the humanistic psychology-humanistic psychology divide”. American Psychologist 69, 1: 92-94. Watson, David and Lee A. Clark. 1997. “Extraversion and its positive emotional core”, in: Hogan, Robert, John A. Johnson and Stephen R. Briggs (eds.), Handbook of personality psychology. San Diego: Academic Pres, 767-793. Wechsler, David. 1955. Manual for the Wechsler Adult Intelligence Scale. New York: Psychological Corporation. Weinstein, Claire E. and Brenda T. Rogers. 1985. “Comprehension monitoring as a learning strategy”, in: Charles Y’dewalle (ed.), Cognition, information processing, and motivation. New York: Elsevier, 619-629. Weinstein, Claire E., Jenefer Husman and Douglas R. Dierking. 2000. “Self-regulation interventions with a focus on language learning strategies”, in: Monique

334 Boekaerts, Paul R. Pintrich and Moshe Zeidner (eds.). Handbook of self-regulation. Burlington: Elsevier Inc, 22: 727-747. Wenden, Anita L. 1986. “Helping language learners think about learning”, ELT Journal 40, 1: 3-12. Wenden, Anita L. 1991. Learner strategies for learner autonomy: Planning and implementing learner training for language learners. New York: Prentice Hall. Wenden, Anita and Joan Rubin (eds.). 1987. Learner strategies in language learning. New York: Prentice Hall International. Wesche, Marjorie B. 1977. Learning behaviors of successful adult students on intensive language training. Paper presented at Los Angeles Second Language Acquisition Forum, (Los Angeles, 11-13 February). White, Cynthia. 2008a. “Beliefs about language learning”, in: Carol Griffiths (ed.). 2008. Lessons from good language learners. Cambridge: Cambridge University Press: 121-130. White, Cynthia. 2008b. “Language learning strategies in independent language learning: An overview", in: Stella Hurd and Tim Lewis Language learning strategies in independent settings. Bristol: Multilingual Matters, 3-24. Whiteside, Stephen P. and Donald R. Lynam. 2001. “The five factor model and impulsivity: Using a structural model of personality to understand impulsivity”, Per- sonality and Individual Differences 30, 4: 669-689. Widiger, Thomas A. 2009. “In defence of borderline personality disorder”, Personality and Mental Health 3, 2: 120-123. Widiger, Thomas A. (ed.). 2017. The Oxford handbook of the five factor model of personality. Oxford: Oxford University Press. Wiggins, Jerry S. and Paul D. Trapnell. 1997. “Personality structure: The return of the big five”, in: Robert Hogan, John A. Johnson and, Stephen R. Briggs (eds.), Handbook of personality psychology. San Diego: Academic Press, 737-765. Wilczyńska, Weronika. 1999. Uczyć się czy być nauczanym? O autonomii w przyswaja- niu języka obcego [To learn or to be taught? About autonomy in foreign language acquisition]. Warszawa: PWN. Williams, Marion, Sarah Mercer and Stephen Ryan. 2015. Exploring psychology in language learning and teaching. Oxford: Oxford University Press.

335 Wilt, Joshua and William Revelle. 2015. “Affect, behavior, cognition, and desire in the Big Five: An analysis of item content and structure”, European Journal of Per- sonality 29, 4: 478-497. Wilt, Joshua and William Revelle. 2017. Extraversion. in: Thomas A. Widiger (ed.), The Oxford handbook of the five factor model of personality. Oxford: Oxford University Press, 11-38. Winne, Philip H. 2011. “A cognitive and metacognitive analysis of self-regulated learning”, in: Barry J. Zimmerman and Dale H. Schunk (eds.). Handbook of self-regulation of learning and performance. London: Routledge, 29-46. Winne, Philip H. 2018. “Theorizing and researching levels of processing in self-regulated learning”, British Journal of Educational Psychology 88: 9–20. Winne, Philip H. and Nancy E. Perry. 2000. “Measuring self-regulated learning”, in: Monique Boekaerts, Paul R. Pintrich and Moshe Zeidner (eds.). Handbook of self-regulation. Burlington: Elsevier Inc., 531-566. Wittenborn, J. Richard, R. P. Larsen and R. L. Mogil. 1945. “An empirical evaluation of study habits for college courses in French and Spanish”, Journal of Educational Psychology 36, 8: 449-474. Wojtowicz, Marcin. 2006. Test zdolności językowych. [Foreign language aptitude test]. Warszawa: Pracownia Testów Psychologicznych Polskiego Towarzystwa Psy- chologicznego. Wolters, Christopher A. 1999. “The relation between high school students’ motivational regulation and their use of learning strategies, effort, and classroom performance”, Learning and Individual Differences 11, 3: 281-299. Wong, Lillian L. C and David Nunan. 2011. “The learning styles and strategies of effective language learners”, System 39, 2: 144-163. Woodrow, Lindy. 2005. “The challenge of measuring language learning strategies”, Foreign Language Annals 38, 1: 90-98. Woodrow, Lindy. 2006a. “Anxiety and speaking English as a second language”, RELC Journal 37, 3: 308-328. Woodrow, Lindy. 2006b. “A model of adaptive language learning”, The Modern Lan- guage Journal 90, 3: 297-319. Wundt, Wilhelm. 1893. Die Grundzüge der physiologischen Psychologie. [The basics of physiological psychology]. Leipzig: Verlag von Wilhelm Engelmann.

336 Yang, Nae-Dong. 1993. Understanding Chinese students’ language beliefs and learning strategy use. (https://eric.ed.gov/?id=ED371589) (date of access: 15 May 2018). Yang, Nae-Dong. 1999. “The relationship between EFL learners’ beliefs and learning strategy use”, System 27, 4: 515-535. Yang, Yu, Stephen J. Read, Thomas F. Denson, T. F., Yiyuan Xu, Jin Zhang and Wil- liam C. Pedersen. 2014. “The key ingredients of personality traits situations, behaviors, and explanations”, Personality and Social Psychology Bulletin 40, 1: 79–91. Yılmaz, Cevdet. 2010. “The relationship between language learning strategies, gender, proficiency and self-efficacy beliefs: a study of ELT learners in Turkey”, Proce- dia-Social and Behavioral Sciences 2, 2: 682-687. Zawadzki, Bogdan. 1970. Wstęp do teorii osobowości. [Introduction to the theory of personality]. Warszawa: Państwowe Wydawanictwo Naukowe. Zawadzki, Bogdan, Jan Strelau, Piotr Szczepaniak and Magdalena Śliwińska. 1998. In- wentarz osobowości NEO-FFI Costy i McCrae. Adaptacja polska. Podręcznik. [The NEO-FFI personality inventory by Costa and McCrae. Polish adaptation. Manual]. Warszawa: Pracownia Testów Psychologicznych Polskiego Towarzy- stwa Psychologicznego. Zawadzki, Bogdan, Jan Strelau, Piotr Szczepaniak and Magdalena Śliwińska. 2010. In- wentarz osobowości NEO-FFI Paula T. Costy Jr i Roberta R. McCrae: adaptacja polska: podręcznik. [NEO-FFI personality inventory: Polish adaptation: Ma- nual]. Warszawa: Pracownia Testów Psychologicznych Polskiego Towarzystwa Psychologicznego. Ziegler, Nick. 2015.“The predictive value of the self-regulating capacity in vocabulary learning scale”, Applied Linguistics 36,5: 641-647. Zimmerman, Barry J. 2002. “Becoming a Self-Regulated Learner: An Overview”, The- ory into Practice 41, 2: 64-70. Zimmerman, Barry J. and Anastasia Kitsantas. 2005. “Homework practices and academic achievement: The mediating role of self-efficacy and perceived responsibility belief”, Contemporary Educational Psychology, 30: 397-417. Zimmerman, Barry J. and Anastasia Kitsantas. 2007. “Reliability and validity of self- efficacy for learning form (SELF) scores of college students”, Zeitschrift für Psychologie [Journal of Psychology] 215, 3: 157-163.

337 Zimmerman, Barry J. and Anastasia Kitsantas. 2014. “Comparing students’ self-discipline and self-regulation measures and their prediction of academic achievement”, Contemporary Educational Psychology 39: 145–155. Zimmerman, Barry J. and Manuel Martinez Pons. 1986. “Development of a structured interview for assessing student use of self-regulated learning strategies”, Ameri- can Educational Research Journal 23, 4: 614-628. Zuroff, David C. 1986. “Was Gordon Allport a trait theorist? “, Journal of Personality and Social Psychology 51, 5: 993.

338 Appendices

A. NEO-FFI validity calculations. Bivariate position-scale correlations.

NEO- 32 0.68 p < 0.0041 Scale FFI r signifi- 37 0.71 p < 0.0041 item value cance 42 0.58 p < 0.0041 3 0.38 p < 0.0041 47 0.44 p < 0.0041 8 0.16 p < 0.0041 52 0.58 p < 0.0041 13 0.61 p < 0.0041 57 0.36 p < 0.0041 18 0.35 p < 0.0041 23 0.55 p < 0.0041 Openness 28 0.36 p < 0.0041 NEO- to 33 0.34 p < 0.0041 Scale FFI r Experience 38 0.20 p < 0.0041 item value significance 43 0.67 p < 0.0041 4 0.45 p < 0.0041 48 0.58 p < 0.0041 9 0.41 p < 0.0041 53 0.46 p < 0.0041 14 0.57 p < 0.0041 58 0.53 p < 0.0041 19 0.42 p < 0.0041 24 0.52 p < 0.0041 Agreeable- 29 0.29 p < 0.0041 NEO- ness 34 0.34 p < 0.0041 Scale FFI r signifi- 39 0.67 p < 0.0041 item value cance 44 0.43 p < 0.0041 5 0.48 p < 0.0041 49 0.46 p < 0.0041 10 0.61 p < 0.0041 54 0.45 p < 0.0041 15 0.65 p < 0.0041 59 0.56 p < 0.0041 20 0.55 p < 0.0041 25 0.64 p < 0.0041 Conscien- 30 0.58 p < 0.0041 NEO- tiousness 35 0.62 p < 0.0041 Scale FFI r 40 0.45 p < 0.0041 item value significance 45 0.49 p < 0.0041 1 0.63 p < 0.0041 50 0.68 p < 0.0041 6 0.48 p < 0.0041 55 0.64 p < 0.0041 11 0.67 p < 0.0041 60 0.46 p < 0.0041 16 0.70 p < 0.0041 21 0.57 p < 0.0041 26 0.43 p < 0.0041 Neuroticism NEO- 31 0.58 p < 0.0041 Scale FFI r signifi- 36 0.56 p < 0.0041 item value cance 41 0.58 p < 0.0041 Extraversion 2 0.66 p < 0.0041 46 0.56 p < 0.0041 7 0.42 p < 0.0041 51 0.61 p < 0.0041 12 0.26 p < 0.0041 56 0.51 p < 0.0041 17 0.62 p < 0.0041 22 0.62 p < 0.0041 27 0.50 p < 0.0041

339 B. SILL ver. 7.0 validity calculations. Bivariate position-scale correlations.

SILL r signifi- Subscale item value cance 1 0.54 p < 0.0055 2 0.514 p < 0.0055 NEO- 3 0.572 p < 0.0055 Subscale FFI r 4 0.617 p < 0.0055 Memory item value significance 5 0.436 p < 0.0055 strategies 1 0.679 p < 0.0055 6 0.327 p < 0.0055 2 0.591 p < 0.0055 7 0.496 p < 0.0055 3 0.659 p < 0.0055 8 0.377 p < 0.0055 4 0.656 p < 0.0055 Metacogni- 9 0.534 p < 0.0055 5 0.599 p < 0.0055 tive strategies 6 0.701 p < 0.0055 NEO- 7 0.666 p < 0.0055 Subscale FFI r signifi- 8 0.644 p < 0.0055 item value cance 9 0.646 p < 0.0055 1 0.282 p < 0.0036 2 0.681 p < 0.0036 NEO- 3 0.605 p < 0.0036 Subscale FFI r 4 0.610 p < 0.0036 item value significance 5 0.650 p < 0.0036 1 0.539 p < 0.0055 6 0.592 p < 0.0036 2 0.312 p < 0.0055 Cognitive 7 0.637 p < 0.0036 3 0.523 p < 0.0055 strategies 8 0.604 p < 0.0036 4 0.468 p < 0.0055 Affective 9 0.239 p < 0.0036 5 0.279 p < 0.0055 strategies 10 0.428 p < 0.0036 6 0.530 p < 0.0055 11 0.564 p < 0.0036 7 0.515 p < 0.0055 12 0.440 p < 0.0036 8 0.370 p < 0.0055 13 0.508 p < 0.0036 9 0.361 p < 0.0055 14 0.556 p < 0.0036

NEO- NEO- Subscale FFI r signifi- Subscale FFI r signifi- item value cance item value cance 1 0.420 p < 0.0041 1 0.585 p < 0.0062 2 0.691 p < 0.0041 Social 2 0.588 p < 0.0062 3 0.596 p < 0.0041 3 0.582 p < 0.0062 strategies Compensa- 4 0.729 p < 0.0041 4 0.539 p < 0.0062 tion strategies 5 0.708 p < 0.0041 5 0.602 p < 0.0062 6 0.574 p < 0.0041 6 0.552 p < 0.0062 7 0.258 p < 0.0062 8 0.116 p < 0.0062 Critical coefficient values for LLS Memory strategies rc = 0.101

Cognitive strategies rc =0.106 Scale Compensation strategies rc =0.1 Metacognitive strategies rc = 0.101 Affective strategies rc = 0.101 Social strategies rc = 0.096

340 C. Interview schedule (translated into English) Interview schedule: The relationships between LLS and the learner’s personality

1. Introduction. a. How important is English for you? b. Do you think you can actually learn a language?

2. Strategy repertoire. a. How often and when do you learn English? b. To what degree do you plan your learning? c. Do you happen to memorise English words, phrases or grammar rules? If so, when? d. How do you learn English vocabulary? e. Can you use the vocabulary you have learnt efficiently? If so, how? f. Do you reflect on the efficiency of your language learning strategies? g. How do you try to understand English grammar? Are you interested in increasing the range of various structures that you can apply? h. Are you particularly interested in improving one of the four basic skills (reading, writing, listening, speaking)? If so, do you do anything to foster the improvement? i. Can you identify and reflect on the emotions which you experience when learning English? j. Do you make any efforts to optimise these emotions? k. To what degree do you consider your learning an interpersonal experience? l. Do you use the Internet to learn English? Do you (also) use it to communicate in English? m. Do you think that communicating in English allows you to improve your command of the language?

3. Personality. a. What type of people i sit easy to learn English for? b. Do you happen any of the qualities you have just mentioned? c. Do you use some learning strategies than others? d. Which of your characteristics make it difficult for you to learn English?

341 e. Do you learn English because it is a duty or because it has been your choice?

4. Conclusion a. Do you think that good intrapersonal knowledge allows the learner to learn Engish better? b. Does increasing knowledge of the language learner allow the English course instructor to teach better? c. Acknowledgements.

342 D. Validity tables 1: NEO-FFI.

NEO- 17 0.62 p < 0.0041 Subscale FFI r signifi- 22 0.62 p < 0.0041 item value cance 27 0.50 p < 0.0041 p < 32 0.68 p < 0.0041 3 0.38 0.0041 37 0.71 p < 0.0041 p < 42 0.58 p < 0.0041 8 0.16 0.0041 47 0.44 p < 0.0041 p < 52 0.58 p < 0.0041 13 0.61 0.0041 57 0.36 p < 0.0041 p < 18 0.35 0.0041 p < 23 0.55 0.0041 p < NEO- Openness 28 0.36 0.0041 Subscale FFI r to p < item value significance Experience 33 0.34 0.0041 4 0.45 p < 0.0041 p < 9 0.41 p < 0.0041 38 0.20 0.0041 14 0.57 p < 0.0041 p < 19 0.42 p < 0.0041 43 0.67 0.0041 24 0.52 p < 0.0041 p < Agreeable- 29 0.29 p < 0.0041 48 0.58 0.0041 ness 34 0.34 p < 0.0041 p < 39 0.67 p < 0.0041 53 0.46 0.0041 44 0.43 p < 0.0041 p < 49 0.46 p < 0.0041 58 0.53 0.0041 54 0.45 p < 0.0041 59 0.56 p < 0.0041

NEO- Subscale FFI r signifi- NEO- item value cance Subscale FFI r 5 0.48 p < 0.0041 item value significance 10 0.61 p < 0.0041 1 0.63 p < 0.0041 15 0.65 p < 0.0041 6 0.48 p < 0.0041 20 0.55 p < 0.0041 11 0.67 p < 0.0041 25 0.64 p < 0.0041 16 0.70 p < 0.0041 Conscien- 30 0.58 p < 0.0041 21 0.57 p < 0.0041 26 0.43 p < 0.0041 tiousness 35 0.62 p < 0.0041 Neuroticism 40 0.45 p < 0.0041 31 0.58 p < 0.0041 45 0.49 p < 0.0041 36 0.56 p < 0.0041 50 0.68 p < 0.0041 41 0.58 p < 0.0041 55 0.64 p < 0.0041 46 0.56 p < 0.0041 60 0.46 p < 0.0041 51 0.61 p < 0.0041 56 0.51 p < 0.0041

NEO- Subscale FFI r signifi- item value cance Extraversion 2 0.66 p < 0.0041 7 0.42 p < 0.0041 12 0.26 p < 0.0041

343