The Influence of High-elaborative, Emotion-rich Reminiscing on Children’s
Development of Autobiographical Memory and Emotion Knowledge
Penny Wareham
Submitted in fulfillment of the requirements of the degree of PhD at the
School of Psychology, University of New South Wales
August 2007 ABSTRACT
High elaborative parent-child reminiscing plays a significant role in preschoolers’
development of autobiographical memory, and, given the emotional salience of many
past events, may also contribute to the development of emotion knowledge and other
socio-cognitive skills. Additionally, whilst research has traditionally focused on
reminiscing style, emotional content may also be important for child outcomes. In Study
1, a naturalistic paradigm was employed to examine associations of parents’ reminiscing
style and emotion references with children’s emotion knowledge. Twenty-five parent- child dyads each discussed four emotionally salient past events. It was found that high elaborative parents more often discussed emotions causes than did other parents; in turn, a high elaborative style and discussion of emotion causes were each uniquely associated with children’s emotion knowledge.
In Study 2 an experimental paradigm was used to examine the impact of emotion- oriented reminiscing on 88 children’s memory for a staged, emotion-rich event. Two days after participating in the event, children reminisced with an experimenter in one of four ways. Emotion-cause, emotion-expression, and no-emotion reminiscing were all high
elaborative but differed in emotion content. Minimal reminiscing was low elaborative.
Children who participated in emotion-cause reminiscing and, to a lesser extent, emotion-
expression reminiscing, recalled significantly more emotional and non-emotional
information about the event than did children who participated in no-emotion or minimal
reminiscing. Study 3 aimed to extend the findings of Studies 1 and 2 by training mothers to
reminisce using a high elaborative style and emotion content. 80 dyads initially
participated; 44 completed all stages. After training, mothers and children in the reminiscing condition each used a more elaborative style and discussed emotions more than did their counterparts in a powerful ‘child directed play’ control condition. These differences were sustained across six months, at which time children in the reminiscing
condition also showed better emotion cause knowledge than did children in the control
condition. Taken together, these findings suggest that children’s autobiographical
memory and emotion knowledge may each develop via shared reminiscing interactions in
the preschool years. In each case, the role of high-elaborative, emotion-rich reminiscing
is highlighted.
CHAPTER 1: COGNITIVE AND SOCIO-EMOTIONAL DEVELOPMENT IN YOUNG CHILDHOOD...... 1 THE ROLE OF PARENT-CHILD DISCUSSION IN CHILDREN’S DEVELOPMENT ...... 2 The Sociocultural Theoretical Framework ...... 4 THE DEVELOPMENT OF AUTOBIOGRAPHICAL MEMORY ...... 7 The Definition of Autobiographical Memory...... 8 The Emergence of Autobiographical Memory: Theoretical Perspectives ...... 9 The Importance of Autobiographical Memory...... 15 Autobiographical Memory and Autobiographical Recall...... 17 Parents’ Reminiscing Style...... 18 Individual elements of a high elaborative style...... 20 Close-ended questions: Are they elaborative? ...... 21 The Impact of Parent Reminiscing Style on Children’s Autobiographical Memory24 Elaborative Structure, Autonomy Support, and Autobiographical Memory ...... 30 The Mechanisms of the Elaborative Style...... 32 Memorial Benefit of Elaborative Reminiscing ...... 34 Child Influences on Reminiscing Style...... 36 Children’s Contributions to Reminiscing Style and Recall ...... 36 The Impact of Child Gender on Reminiscing Style ...... 37 Associations between Attachment and Reminiscing Style ...... 38 The Impact of Child Temperament on Reminiscing Style...... 38 The Impact of Culture on Reminiscing Style...... 39 Parents’ Narrative Content...... 40 Discussion of Emotional Experiences in the Past ...... 42 The Impact of Child Gender on Emotion Talk ...... 44 Future Research...... 45 THE DEVELOPMENT OF EMOTION KNOWLEDGE ...... 45 The Importance of Emotion Knowledge for Preschoolers...... 46 Findings: Parent-child Discussion of Emotions and Emotion Knowledge...... 47 The Impact of Discussing Emotion Causes above and beyond other Emotion Talk48 Associations between Emotion Knowledge, Language, and Theory of Mind...... 49 Influences on Emotion Knowledge...... 51 The Impact of Child Gender on Emotion Knowledge ...... 51 The Impact of Culture on Emotion Knowledge...... 52 INTEGRATING RESEARCH ON PAST EVENT DISCUSSIONS, AUTOBIOGRAPHICAL MEMORY, AND EMOTION KNOWLEDGE: THE PRESENT RESEARCH...... 53 Associations between Parent-child Reminiscing and Emotion Knowledge...... 54 The Impact of Causal and Non-causal Emotion Reminiscing on Recall ...... 55 Facilitating Recall and Emotion Knowledge via Reminiscing Training...... 56 CHAPTER 2: ASSOCIATIONS BETWEEN AN ELABORATIVE STYLE, EMOTION REFERENCES, AND EMOTION KNOWLEDGE ...... 58 METHOD...... 63 Participants ...... 63 Measures and Procedure ...... 63 Memory Conversations ...... 64 Children’s Emotion Knowledge...... 64 Children’s Language ...... 65 Coding ...... 66
RESULTS...... 67 Preliminary Analyses and Sample Characteristics...... 67 Correlation Analyses of Reminiscing Style and Content, and Children’s Language and Emotion Knowledge ...... 69 Associations between Parents and Children’s Memory Conversation Variables ....70 Associations between Parents’ Style and Content Memory Conversation Variables ...... 70 DISCUSSION...... 72 CHAPTER 3: EMOTION ORIENTED REMINISCING AND CHILDREN’S RECALL FOR A NOVEL EVENT...... 78 METHOD...... 83 Participants ...... 83 Materials ...... 84 Zoo Event Props...... 84 Procedure...... 85 Zoo Event ...... 85 Reminiscing Conversation ...... 85 Memory Interview...... 89 Children’s Emotion Knowledge...... 89 Children’s Language ...... 90 Coding ...... 91 RESULTS...... 92 Regression and Preliminary ANOVA Analyses...... 92 Effects of Emotional Reminiscing on Recall ...... 94 Statistical Strategy...... 94 Correct Total Recall ...... 95 Correct Free Recall...... 97 Additional Recall during Prompted Recall and Direct Questioning...... 100 Summary ...... 100 DISCUSSION...... 101 CHAPTER 4: TRAINING MOTHERS IN HIGH ELABORATIVE, EMOTION-RICH REMINISCING: FACILITATING AUTOBIOGRAPHICAL MEMORY AND EMOTION KNOWLEDGE...... 109 METHOD...... 114 Participants ...... 114 Measures and Procedure ...... 115 Assessments ...... 116 Memory Conversations ...... 116 Children’s Emotion Knowledge...... 116 Children’s Language ...... 117 Training...... 117 Training Video ...... 118 Training Booklet ...... 118 Assessment Coding...... 119 RESULTS...... 120 Mothers’ Reminiscing Style and Content, Children’s Reminiscing Style and Content, and Children’s Emotion Knowledge ...... 121 Mothers’ Shared Reminiscing Style and Content ...... 122
Children’s Shared Reminiscing Style and Content...... 125 Children’s Independent Reminiscing Contributions...... 126 Children’s Emotion Knowledge...... 127 Comparison of Results using the LOFC Method and the Final Sub-sample Only ...... 128 Supplementary Analysis of the Association between Mothers’ Reminiscing and Children’s Independent Recall...... 128 DISCUSSION...... 130 CHAPTER 5: AN INTEGRATION OF FINDINGS, AND THEIR IMPLICATIONS FOR CHILDREN’S COGNITIVE AND SOCIO-EMOTIONAL DEVELOPMENT .. 139 FINDINGS OF THE CURRENT RESEARCH...... 140 Children’s Internalisation of Reminiscing Style and Content...... 142 The Influence of Reminiscing on Recall for a Single, Discussed Event...... 145 Associations between Parents’ Reminiscing Style and Content...... 148 The Importance of Emotion Causes for Autobiographical Memory and Emotion Knowledge...... 150 THEORETICAL IMPLICATIONS FOR THE SOCIOCULTURAL THEORY AND THE DEVELOPMENT OF EMOTION KNOWLEDGE ...... 154 Implications for the Sociocultural Theory ...... 154 Implications for the Development of Emotion Knowledge...... 156 APPLIED IMPLICATIONS OF THE PRESENT RESEARCH ...... 159 METHODOLOGICAL ISSUES...... 162 The Coding of Close-Ended Questions ...... 162 The Coding of High Elaborative Style: Proportions or Totals? ...... 163 The Assessment of Emotion Knowledge...... 164 SUMMARY AND CONCLUSION...... 167 REFERENCES...... 169 APPENDIX A: DENHAM’S EMOTION KNOWLEDGE TASK VIGNETTES ...... 193 STEREOTYPICAL VIGNETTES ...... 193 NON-STEREOTYPICAL VIGNETTES ...... 194 APPENDIX B: MEMORY QUESTIONS USED DURING THE DIRECT QUESTIONING PHASE OF STUDY 2...... 196 APPENDIX C: INTERVENTION TRAINING SCHEDULES USED IN STUDY 3 .... 197 REMINISCING TRAINING ...... 197 Reminiscing Training Session 1 Guidelines (approximately 1 hour) ...... 197 Reminiscing Training Follow-up Session Guidelines (approximately 1/2 hour)...... 198 CONTROL TRAINING...... 200 Control Training Session Guidelines (approximately 1 hour)...... 200 Control Training Follow-up Session Guidelines (approximately 1/2 hour)...... 201 APPENDIX D: INTERVENTION TRAINING BOOKLETS PROVIDED TO PARENTS IN STUDY 3 ...... 203 REMINISCING TRAINING BOOKLET...... 203 CONTROL TRAINING BOOKLET...... 207
1
CHAPTER 1: COGNITIVE AND SOCIO-EMOTIONAL DEVELOPMENT IN YOUNG
CHILDHOOD
“Remembering past events is a universally familiar experience. It is also a uniquely
human one. As far as we know, members of no other species possess quite the same
ability to experience again now, in a different situation and perhaps a different form,
happenings from the past” (Tulving, 1983, p.1)
“If much of our self-concept is defined though our autobiographical life story, then
our emotional reactions to these experiences, as they were occurring and in the
present, provide the glue that connects our past to our present and makes these
experiences meaningful” (Fivush, Berlin, Sales, Mennuti-Washburn, & Cassidy,
2003, p.179).
The preschool years, from ages 3 to 6, are a critical time for children’s socio-cognitive development. Memory skills, including those that underpin autobiographical memory, develop across this period (Nelson & Fivush, 2004), such that 4 and 5 year old children are generally able to remember and report experiences in their lives that occurred at least one year earlier (Peterson, 2002; Salmon & Pipe, 2000). Understanding of minds (or theory of mind) and emotion knowledge also grow: children understand the impact of desires on behaviour at age 3 but do not understand the impact of beliefs on behaviour and emotion until age 4 and 7 respectively, and, although children across the preschool years understand simple emotions and recognise the expressions of these emotions well, their understanding of mixed emotions, of complex social and self-conscious emotions such as guilt, and of the nature of emotion causes, develops rapidly between ages 3 and 5 (Denham, 1998; Hughes & Dunn, 2002; Pons, 2
Harris, & de Rosnay, 2004; Wellman, Cross, & Watson, 2001). Finally, a multitude of other skills also develop across this period, including language skills and emergent literacy skills necessary for future reading and writing (Whitehurst & Lonigan, 1998), knowledge of numbers necessary for future arithmetic (Gelman & Gallistel, 1978), and a child’s sense of self as being continuous throughout time (Povinelli, 1995).
The early and extended development of these skills is important for self, social, and other outcomes. Given that emotion knowledge is an important component of emotion regulation, for example, it comes as no surprise that children’s ability to regulate their experience and expression of emotions increases throughout the preschool period, together with the growth of their emotion knowledge (Brophy & Dunn, 2002; Denham, 1998).
Emotion regulation at preschool is an important tool for maintaining self-control, and predicts children’s social functioning at school up to six years later (Eisenberg, Smith, Sadovsky, &
Spinrad, 2004). By investigating the crucial factors that influence development across the preschool years, it is the aim of this thesis to gain insight into the ways that aspects of this development might best be facilitated. Research will focus primarily on children’s development of autobiographical memory and emotion knowledge.
The Role of Parent-Child Discussion in Children’s Development
What are the important influences on young children’s lives that facilitate the development of cognitive and socio-emotional skills? A strong body of research highlights the influence on cognitive development of the ways in which parents, particularly mothers, talk about everyday experiences with their young children. For example, parent-child discussion about children’s past and present experiences helps children represent their experiences as organised and coherent narrative stories, and therefore significantly influences how they come to both remember and independently report each experience (e.g. Nelson & Fivush, 2004;
Fivush, 1991). Furthermore, research shows that parent-child discussion in a variety of 3 contexts enhances cognitive skills other than memory, such as emergent literacy development
(e.g. Beals, 2001; Reese, 1995). Reese (1995) asked 24 mother-child dyads to participate in shared book reading and shared reminiscing when children were 40-, 46-, and 58-months-old.
Regression analyses showed significant relationships between the way in which mothers talked to their children at these earlier time-points and children’s literacy skills at 70-months, particularly for print concepts, vocabulary, and story comprehension skills, even once children’s participation in the conversations had been accounted for. Furthermore, although reminiscing conversations do not involve exposure to print, relationships between mother- child discussion and children’s literacy were particularly strong.
Consistent with this research focused on cognitive development, a large body of research shows that parent-child discussion about emotions in a variety of contexts, including family talk, picture book reading, and mothers’ explanations of their own emotions, predicts children’s concurrent and future emotion knowledge (e.g. Denham, Mitchell-Copland,
Strandberg, Auerbach, & Blair, 1997; Denham, Zoller, & Couchod, 1994; Dunn, Brown, &
Beardsall, 1991; Garner, 1997). Moreover, parents’ references to mental states such as desires and beliefs predict children’s theory of mind (e.g. Meins, 1997; Meins & Fernyhough, 1999;
Peterson & Slaughter, 2003; Ruffman, Slade, & Crowe, 2002), as do both mothers’ contributions to reminiscing conversations, and the role that language plays in allowing children to compare two different representations of a situation (Reese & Cleveland, in press;
Lohmann & Tomasello, 2003a).
Together, cognitive and socio-emotional research highlights that everyday conversations between parents and their young children may play a critical role in enhancing development across domains. Of course, acknowledgement of the biological bases to cognition is also necessary for a full explanation of development (Blasi, 1996). Nonetheless, is clear that development must be considered within the social and cultural context in which it 4 occurs (Nelson, 1996; Rogoff & Gardner, 1984). This may particularly be the case throughout the preschool years: it is between the ages of 2 and 6, Nelson argues, that “biology hands over development to the social world” (1996, p.325).
The Sociocultural Theoretical Framework
Much of the current research investigating the way in which parent-child discussion facilitates development is underpinned by the sociocultural theory (also known as the social interaction theory) (e.g. Fivush, 1991; Nelson, 1993a, 1993b, 1996; Reese, 2002a, 2002b;
Reese & Fivush, 1993). The sociocultural theory is in turn influenced strongly by Vygotsky’s seminal notion of cognitive development (1962, 1978), and by the more recent views of
Rogoff (1984, 1990); a neo-Vygotskian psychologist. Although Vygotsky’s ideas have most often been applied to the understanding of language and memory skill development, they are also relevant to other representational development including that of emotion and of mind.
Vygotsky postulated that all individual mental representations, including memories, thoughts, knowledge, and ideas, have sociocultural origins. The social or cultural context in which a child is raised determines the specific representational skills and information that will be most valuable for that child. Furthermore, a child’s culture provides him or her with important
‘psychological tools’ that mediate his or her cognitive development, and the most important of these tools is language.
In what way does language mediate a child’s cognitive development? Vygotsky
(1978) reasons that language is not only a tool with which children can reflect upon and represent ideas, memories, and knowledge in their own minds, but it also allows them to participate in ‘culturally organised activities’; namely, in social exchanges with others. In exchanges with a more knowledgeable partner, such as a parent or a teacher, the child is guided and supported in a way that allows him or her to extend the memory, conversational, linguistic, numerical, or other skills being used, and to contribute to the exchange more than 5 he or she would be capable of independently. For example, children learn to problem solve with partners who are able to support or ‘scaffold’ their progress by adjusting the difficulty of problem solving tasks as needed, and by providing experience in the joint solution of the problem (Wertsch, Minick & Arnes, 1984; Wood, Bruner, & Ross, 1976). Scaffolding contrasts with explanation and with demonstration, which do not actively involve the child
(Rogoff, 1984, 1990; Wersch & Stone, 1979).
By regularly participating in social exchanges with more knowledgeable others, children come to internalise the skills that are being scaffolded, practiced and extended
(Rogoff, 1990). Nonetheless, not all social interaction facilitates this internalization to the same degree. Vygotsky (1978) terms the difference between a child’s current level of development and their potential level of development, if facilitated by a capable adult, the
‘zone of proximal development’. It is therefore participation in social exchanges (or
‘culturally organised activities’), with more capable others, within the zone of proximal development, which best facilitates young children’s internalization of cognitive skills.
Parents’ scaffolding of children’s skills within the zone of proximal development has been shown to contribute to development of skills as varied as learning to talk and learning to weave (Greenfield, 1984). Scaffolding outside the zone of proximal development is likely not to extend children’s developing skills to the same degree. Rogoff (1990) likens the development of these skills to an apprenticeship between a child and a skilful partner. It is not the mere presence of a partner that matters, but the nature of the interaction between partners.
Drawing on Vygotsky’s work, sociocultural theorists such as Nelson (1993a, 1993b,
1996), Fivush (1991; Reese & Fivush, 1993), and Reese (2002a, 2002b) propose that the quality of parent-child discourse significantly shapes children’s representational development throughout the preschool years. With respect to autobiographical memory, for example,
Nelson (1996, p.172) describes development as involving a “dialectical process of 6 collaborative construction of remembered events in linguistic form, rather than simply acquiring models from parent to child”. Together with the parent, the child is also an active participant in his or her own development. Whilst focused heavily on autobiographical memory development, particularly in the context of parent-child reminiscing, the sociocultural theory has also been applied to the development of other representational skills such as language, emergent literacy, self understanding, and theory of mind (e.g. Nelson,
1996; Peterson & Slaughter, 2003; Reese, 1995; Reese & Cleveland, in press). In each case language-based interactions between parents and their children have been identified as having a critical influence on children’s development, with individual differences in interaction quality determining the extent to which development will occur.
Despite the strong influence of Vygotsky’s work on the current social sociocultural theory, there remains one significant difference between the two. Newcombe and Reese
(2004) note that a classic Vygotskian model predicts that parents will provide less guidance and support as children become more proficient in a given skill. Consistent with this prediction, Rogoff (1990) suggests that expert partners may allow or even require children to take a greater role in a social exchange as those children become more competent at the skill they are being guided in. The sociocultural theory, however, suggests that whilst parents sensitive to their children’s development may encourage those children to contribute more to an interaction as they become capable of doing so, this does not mean the parents’ own contributions will decrease. There is some evidence to support this view. When examining the use of internal state terms and other ‘evaluatives’ during reminiscing, for example,
Newcombe and Reese (2004) found that parents tend to provide even more guidance and support as their children become more proficient at using these terms, perhaps in response to their child’s greater involvement in the interaction. The same can also be said for children’s more general development of autobiographical recall. As parents encourage their children in 7 conversations about their past and present experiences, they facilitate their children’s internalisation of how and what to retain in order to become skilled rememberers. As their children come to contribute more to these conversations, parents are given greater encouragement and opportunity to guide and support their children even further. This pattern reflects a spiral model of parent-child interaction contributing to children’s cognitive development (Newcombe & Reese, 2004).
The aim of this thesis is to investigate children’s development of autobiographical memory and emotion knowledge during reminiscing, within the sociocultural framework.
Below, the development of autobiographical memory and the development of emotion knowledge are reviewed in turn. A large body of research suggests that the way in which parents reminisce with their children, in terms of both style and content, impacts autobiographical memory development. Furthermore, recent research suggests that the style and content of parents’ reminiscing may also impact children’s developing understanding of emotion. Following a review of this autobiographical memory and emotion knowledge literature, gaps in our current knowledge are highlighted and the current research is proposed.
The Development of Autobiographical Memory
Over the past two decades, emphasis has been placed on ecologically valid cognitive research. In the case of memory research, this emphasis has resulted in a reduction of research on deliberate or strategic memory tests, some (but not all) of which test memory for meaningless stimuli, such as unrelated word lists, over the course of one experimental session only; and an expansion of research focusing on naturally occurring memory for real-world events measured over weeks, months, or years (Bjorklund, 2004; Siegler, 2004). Nonetheless, different aspects of memory should not be considered independent, and research on one aspect can therefore inform research on other aspects: all require encoding and retrieval skills, for example, and, in the case of explicit memory, the ability to search memory for specific 8 information. Recent work by Rudek and Haden (2005) demonstrates that many predictors of children’s autobiographical memory also predict their strategic memory for word lists, thus supporting the notion that different memory skills may develop together.
The following section defines autobiographical memory within the context of other aspects of memory and clarifies its importance for both children and adults, before considering the intertwined relationship between autobiographical memory and autobiographical recall. The different styles of autobiographical recall that parents use with their children, and the content of this recall, are each examined with regard to children’s developing memory skills. Unless otherwise stated, a sociocultural framework is assumed.
The Definition of Autobiographical Memory
To understand autobiographical memory fully, it is helpful to define it within the broader context of long-term memory. There is no one clear definition of long-term memory
(Reese, 2002a). Many researchers believe that long-term memory consists of different and discrete systems, such that autobiographical memory is a sub-system of memory separate to non-autobiographical memory (e.g. Squire, 1987, 1995; Nelson, 1996; Nelson & Fivush,
2004; see Squire, 1987, 1995, for a description of his declarative/non-declarative memory framework, and Tulving, 1972, for his distinction between episodic and semantic memory).
Other researchers argue instead for one memory system, defined by different memory processes (e.g. Howe & Courage, 1997; Roediger, 1990). In any case, researchers agree that long-term memory may involve multiple skills (e.g. Nelson & Fivush, 2004; see Schneider &
Bjorklund, 1998, for review).
Autobiographical memory is, broadly speaking, ones explicit memory for a personally experienced event or events that occurred in the past (Nelson & Fivush, 2004). Unlike other forms of explicit memory there a sense of the self, remembering one specific time (Courage &
Howe, 2004; Nelson & Fivush, 2004). Autobiographical memory is therefore primarily 9 episodic, in that it is temporally specific, but it may also integrate aspects of semantic memory; that is, our store of factual knowledge (Reese, 2002a; Tessler & Nelson 1994;
Tulving, 1972). The episodic nature of autobiographical memory serves to distinguish it from generic event memory, which, being a source of general knowledge, is best described as semantic (Nelson, 1996; Tulving, 1972; Schneider & Bjorklund, 1998). In generic event memory, children and adults alike use generalised schemata to provide a ‘script’ for the standard elements of common experiences such as ‘going to a restaurant’ (Bartlett, 1932;
Schank & Abelson, 1977). Unique elements of the experience are not recalled, however, and there is therefore no sense of the self remembering one specific time (Nelson, 1996).
Nelson (1996) describes autobiographical memory as “…enduring, chronologically sequenced memories for significant events in ones own life” (p.162). The memories that each individual recalls provide information not only about the event but also about one’s preferences, interests, abilities, and values (Cockroft & Reese, in press). It is because autobiographical memories are self-involved, personal, and significant that they are interesting to recall and to discuss with others.
The Emergence of Autobiographical Memory: Theoretical Perspectives
The following section explains the emergence of autobiographical memory according to both the sociocultural theory and an alternative theory: the ‘cognitive self’ theory.
Interestingly, it is only recently that the emergence of autobiographical memory has been thought of and studied in developmental terms. Previously, research was primarily concerned not with the explanation of how and why young children develop autobiographical memory, but with the explanation of infantile/childhood amnesia; that is, why adults are unable to recall events experienced before they reach the preschool years (Nelson, 1993b).
It should also be noted that both the sociocultural theory and the cognitive self theory assume that neurological development has occurred in the basic memory structures 10 responsible for long-term memory (Conway & Pleydell-Pearce, 2000). Indeed, research shows that these basic memory structures are functional before the first year of age (see C.A.
Nelson, 1995, for review). For example, in the deferred imitation paradigm, infants as young as 9-months-old demonstrate explicit long-term retention for a unique, staged set of actions modelled to them by an experimenter by nonverbally reproducing this sequence with as little as one exposure, and without prior opportunity to practice, after a period of delay (Bauer &
Wewerka, 1995; Carver & Bauer, 1999; Hayne & Herbert, 2004; Melzoff, 1985, 1995).
Whilst these findings relate more specifically to presence of explicit memory in infants, they necessarily also indicate the presence of functioning long-term memory structures (Howe &
Courage, 1997).
Major tenets of the sociocultural theory. According to the sociocultural theory, autobiographical memory emerges gradually and continuously when children are approximately 3 or 4 years of age (Nelson, 1993b, 1996; Nelson & Fivush, 2004; Reese,
2002a). Prior to this, explicit memory is limited to semantic memory (general knowledge) and to nonverbal, non-autobiographical event memory, as evidenced by the deferred imitation paradigm (Nelson, 1996). Many contributors and precursors to the emergence of autobiographical memory are proposed, including not only neurological development of memory structures but also language and narrative skills, parent-child event talk, an understanding of self, and, most recently, an understanding of mind (e.g. Kleinknect & Beike,
2004; Nelson, 1996; Reese, 2002a, 2002b; Welch-Ross, 1997; see Nelson & Fivush, 2004, for review). In the case of the latter, for example, an understanding of mind (or theory of mind) may be necessary for children to truly understand the nature of autobiographical memory; that is, that autobiographical memories are subjective mental representations and constitute an important source of knowledge (Kleinknect & Beike, 2004; Reese, 2002b; Welch-Ross,
1997). No one contributor or precursor is considered responsible for the emergence of 11 autobiographical memory on its own; rather, multiple influences may be important (Harley &
Reese, 1999; Nelson & Fivush, 2004; see Reese, 2002b, for a review of single and multiple influence theorists).
Whilst the sociocultural theory allows for multiple contributors to the emergence of autobiographical memory, the importance of both language and parent-child discussion are central themes (Nelson & Fivush, 2004). Based on Vygotsky’s (1978) notion of language as a
‘tool’ for development, sociocultural theorists suggest that language is vital for the development of a coherent and structured underlying memory representation, and not just for the communication of memory. It is with language that an event is both encoded and represented in memory: language, according to this view, partially constitutes representation
(e.g. Harley & Reese, 1999; Nelson & Fivush, 2004; Tessler & Nelson, 1994). Furthermore, according to sociocultural theory, parent-child discussion about everyday events is the critical means by which both autobiographical memory and many of the contributors to autobiographical memory are shaped (Nelson, 1996; Nelson & Fivush, 2004; Wang, Hutt,
Kulkofsky, McDermott, & Wei, 2006). By regularly participating in discussions with more knowledgeable others, children come to internalise the memory, self, and other representational skills that are being scaffolded, practiced and extended. The quality of this discussion determines the extent to which internalization occurs (Rogoff, 1990; Vygotsky,
1978).
Major challenges to the sociocultural theory. As part of their ‘cognitive self’ theory, and in contrast to the sociocultural framework, Howe and Courage suggest that although children’s receptive and expressive language is necessary for the appropriate expression of memory to others in a socially determined narrative form, and may therefore limit the degree to which very young children can demonstrate their autobiographical recall, language is not necessary for the emergence of autobiographical memory itself (e.g. Courage & Howe, 2004; 12
Howe & Courage, 1997). Instead, they argue, children’s development of a ‘cognitive self’ is the key development necessary for the onset of autobiographical memory.
Children’s understanding of themselves as unique, ‘thinking and knowing’ individuals, termed ‘the cognitive self’, develops gradually across early childhood (Courage
& Howe, 2004; Howe & Courage, 1997; Howe, Courage, & Edison, 2003). The cognitive self first emerges between 18 and 24 months of age, when children begin to recognise themselves in a mirror and respond to marks placed on their face (Howe & Courage, 1993). Without a sense of self there is no “me”; no personal frame of reference to which personal memories can be referenced and stored as autobiographical memory (e.g. Courage & Howe, 2004; Howe et al., 2003). Long-term event memory before this age is possible, as evidenced by the elicited and deferred imitation paradigms, but not as a personal memory for a specific experience
(Courage & Howe, 2004).
Povinelli and colleagues agree with Howe and Courage’s position that a cognitive self is necessary for the emergence of autobiographical memory, but state that a full cognitive self requires not just self-recognition, as measured by the mirror test, but also the additional understanding that the self remains consistent throughout time (e.g. Povinelli, 1995; Povinelli,
Landau, & Perilloux, 1996). It is only when children become aware that they are the same person now who experienced an event a day or year ago that they are able to form a specific, personal memory of that event (Povinelli et al., 1996; Reese, 2002a). Povinelli et al. (1996) tested children’s understanding of the self-in-time by surreptitiously placing a sticker upon each child’s head. When shown the video after a short delay, children aged 42 to 48 months of age demonstrated an awareness of the self in time by reaching for their head. Although younger children are able to recognise themselves in the mirror test, they did not demonstrate an understanding of the cognitive self-in-time, and thus, according to Povinelli and colleagues
(1996), may not be capable of relating events in the past to their present self. 13
Current research. There are two bodies of research that can provide insight into whether the sociocultural or the cognitive self view is best supported. In the first, the importance of language for memory is addressed. In the second, the relative contributions of the cognitive self and other contributors to autobiographical memory are compared.
If it is the case that language is necessary only for the verbal expression of memory, as suggested by the ‘cognitive self’ view, then the underlying representational memory structure should develop irrespective of language ability. If however, language is necessary for the underlying representation of an event, as suggested by the sociocultural view, then language should predict the amount which children are able to both verbally and nonverbally recall.
Two studies, conducted by Simcock and Hayne (2002, 2003), compare these possibilities. In each study, children aged 2- to 4-years-old participated in a unique event - ‘the magic shrinking machine’ - which involved turning the machine on, take a large toy from a bag, place the toy in the machine, pull a lever, and collect a matching smaller toy from another part of the machine. Productive vocabulary was measured concurrently, and both verbal recall and nonverbal re-enactment were measured after a delay.
In the first study (2002), after a delay of either six months or one year, it was found that no child used any word to describe the event at test that had not been part of his or her productive vocabulary at encoding. Thus, language is a critical aspect of the underlying event representation. In the second study (2003), a similar pattern of results emerged. After a delay of just 24 hours it was found that children with more advanced language skills not only showed better verbal recall, they also showed better nonverbal recall. That nonverbal recall was enhanced by language, despite language being unnecessary for the expression of this recall, suggests perhaps that language contributes to the underlying representation of the event. Importantly, as verbal recall lagged behind both nonverbal recall and language development in both studies, despite the different delay periods, Simcock and Hayne conclude 14 that language acquisition is a necessary, but not sufficient, condition for the emergence of autobiographical memory (2002, 2003). This finding is consistent with the sociocultural emphasis on language.
To test the sociocultural view against the ‘cognitive self’ view more directly, Reese
(Harley & Reese, 1999; Reese, 2002a) used a longitudinal design in which 58 children, initially aged 19 months, and their mothers completed assessments every six months. Tests included children’s language development and self recognition, and mothers were also asked to reminisce with their children about everyday events. Children were categorised as being early or late developers of a cognitive self, based on their performance on the self-recognition mirror test at 19 months. Children’s placeholder contributions during reminiscing at 19 months and mothers’ memory elaborations during reminiscing at 25 months were each found to uniquely predict children’s shared memory elaborations at 32 months. Early developers of a cognitive self had greater shared recall skills than did late developers of a cognitive self, but only if they also had low initial language skills. These findings suggest that the development of the cognitive self is indeed associated with autobiographical memory, but only in the context of language. As children became older the roles of language and conversation are highlighted (Reese, 2002a).
In sum, research tends to support the sociocultural theory. Factors such as language skill, narrative skill, the understanding of self, and an understanding of mind may each be associated with the emergence of autobiographical memory. In particular, research shows that whilst there does not seem to be any single cause responsible for the emergence of autobiographical memory, the influences of self, superseded by language and reminiscing, are important (Harley & Reese, 1999; Nelson & Fivush, 2004; Reese, 2002b). Moreover, each of these factors has also been associated with children’s active participation in parent- or adult- scaffolded narrative interactions (e.g. Bird & Reese, 2006; Nelson, 1996; Nelson & Fivush, 15
2004; Wang et al., 2006). Biologically determined cognitive development is therefore necessary, yet parent-child discussion is the critical means by which autobiographical memory is shaped (Nelson & Fivush, 2004).
The Importance of Autobiographical Memory
Research and theory in a wide variety of psychological domains, including clinical, developmental, cognitive, and social, suggests that the primary functions of autobiographical memory can be broadly summarised as being self, directive, or social in nature (Bluck, 2003;
Pillemer, 1992; Wang & Conway, 2004). As prescribed by the literature, these functions have discrete labels and will be described separately here. It is important to note, however, that there is strong overlap between the functions. Individual memories are often thought serve multiple functions (Bluck, 2003).
Many sociocultural researchers highlight the critical role of autobiographical memory in developing a sense of self (e.g. Wang, 2004; Welch-Ross, Fasig, & Farar, 1999).
Autobiographical memories are by definition personally meaningful and self-relevant, and therefore reflect one’s preferences, interests, abilities, and values (Cockroft & Reese, in press). Although the specific emphases of the various theoretical positions differ greatly (e.g. see Fivush, 1991; Howe & Courage, 1997), discrete autobiographical memories are thought to combine to produce a coherent autobiography, also termed an ‘autobiographical life story’ or
‘self history’, from which children can draw a sense of who they are (Cockroft & Reese, in press; Fivush et al., 2003). This relationship also is almost certainly bi-directional; such that a stronger sense of self helps tailor autobiographical memories to a greater degree (Cockroft &
Reese, in press; Eder, 1989; Hudson, 1986; see p.15 of this thesis for the contribution of self to the emergence of autobiographical memory). The notion that a child’s concept of self is developed, in part, from his or her personal memories is supported by findings of cultural, 16 gender, and individual differences in the experiences recalled by different children (Nelson &
Fivush, 2004; Peterson & McCabe, 1994; Wang & Conway, 2004).
Autobiographical memory may also have important directive functions (Bluck, 2003).
Using transcripts of adult memories as illustration, Pillemer (2003) demonstrated that autobiographical memories provide models to inform and guide behaviour during current activities. By understanding the causes and consequences of past experiences held in memory, we are exposed to a wealth of extra information with which to solve various problems as they arise throughout life (Pillemer, 2003; Webster & Cappeliez, 1993). Successful past deeds can be drawn upon as a source of information and inspiration, whereas unsuccessful past deeds can be learnt from and avoided. When autobiographical memories are impoverished or incomplete, individuals may be deprived of a rich source of past learning to guide current cognitive and behavioural responses.
Finally, autobiographical memory, and, more specifically, autobiographical recall, may serve fundamental social functions across childhood and adulthood. Reminiscing conversations do not simply occur for the sake of it, nor to provide a conversational partner with information about run-of-the-mill past events, but to share information about one’s preferences, interests, abilities, and values within the context of salient, personally meaningful past events (Cockroft & Reese, in press; Miller, 1994). Consistent with this view, both younger and older adults report sharing their important or personally relevant past events with others (Pasupathi, 2003). Alea and Bluck (2003) suggest that reminiscing about past experiences held in autobiographical memory allows the development and maintenance of intimacy in relationships, the illustration of a point of advice, the elicitation of empathy, and the provision of sympathy. The level of detail and amount of emotion characterising spoken recall are indicative of the extent to which the social functions of reminiscing will be served
(Alea & Bluck, 2003; see also Nelson, 1996). 17
Considering these self, directive, and social functions, the development of rich and complete autobiographical memories may imply healthy child and adult development in diverse yet important ways
Autobiographical Memory and Autobiographical Recall
According to the sociocultural framework, learning to reminisce with others is both a memory skill and a conversation skill (Fivush & Fromhoff, 1988). To become skilled at reminiscing, children must learn to communicate their experiences via conversations with others by structuring them as though they were narrative stories (Fivush & Fromhoff, 1988;
Reese & Fivush, 1993), and by adhering to topics their culture deems important (Reese &
Fivush, 1993). A good narrative not only organises event details in a coherent structure, it also emphasises high points of the event, the child’s affective responses, intentions of participants, causal relations between actions, and salient occurrences (Nelson, 1993a). By learning to structure their spoken recall in narrative form, children also learn to structure their underlying episodic memory, or representation, of the experience in an organised and coherent way
(Harley & Reese, 1999; Nelson, 1993a; Reese & Fivush, 1993).
Naturalistic and experimental research highlights the role that parents play in encouraging their children to participate in discussions about their past and ongoing experiences (Nelson & Fivush, 2004; Fivush, 1991). Parents heavily scaffold early conversations with their children by providing narrative structure and content, as well as task guidance, monitoring, and feedback (Dodici, Draper, & C.A. Peterson, 2003; Reese & Fivush,
1993). Importantly, this scaffolding is continuously revised in order to respond to children’s increasing memory skill (Rogoff, 1990). By facilitating parent-child discussion in this way, parents are able to build on the often-minimal pieces of information that their children contribute to the discussion, and to guide their children in how best to represent their experiences in an organised and coherent manner (Nelson, 1996; Nelson & Fivush, 2004; 18
Fivush, 1991). As children become more skilled in remembering, at around 65 months of age
(Cleveland & Reese, 2005), their ability to represent and discuss their experiences begins to emerge independently of parental support and guidance (Fivush, 1991, 1993).
Parents’ Reminiscing Style
A broad body of research has investigated the quality of parent-child event scaffolding. In these studies, which are described in detail below (pp. 30-37), parents are typically asked to choose and discuss recent, one-time, everyday past events with their child that do not already include an internal storyline (e.g. trips to the movies). The parent’s and the child’s contributions to these discussions are each coded. This research shows that although all parents scaffold their children’s developing autobiographical memories to a certain extent, there are differences in both the degree to which they do, and the nature of this scaffolding.
These differences are reflected by parents’ reminiscing style (Peterson & McCabe, 1994;
Reese, 2002b; Reese & Fivush, 1993). By the time children are in preschool, the reminiscing style that their parent uses, relative to other parents, has been found to remain quite stable across time (Harley & Reese, 1999; Tessler & Nelson, 1994). It should be noted that most research to date has examined the narrative style of mothers only. In Reese and Fivush’s
(1993) study of reminiscing style within 24 two-parent families, however, it was found that styles were not associated with parent gender. Findings relating to mothers may therefore also be generalised to fathers.
Reminiscing styles have been conceptualised in several different ways over the past two decades, but with distinct similarities between conceptualisations. Parents who construct full and elaborative stories about ongoing or past experiences, and invite children to join them in the construction of these stories, have in past research been variously referred to as having an elaborative (e.g. Fivush & Fromhoff, 1988; Reese & Fivush, 1993; Thompson, Laible, &
Ontai, 2003), reminiscing (e.g. Engel, 1986, in Fivush & Fromhoff, 1988), topic extending 19
(e.g. McCabe & Peterson, 1991), or narrative (e.g. Tessler & Nelson, 1994) style (Fivush,
1991; Reese & Fivush, 1993). In contrast, parents who attempt to draw particular pieces of information from their children by using repeated questions or providing scant information, and who frequently switch topic, have variously been referred to as being pragmatic (e.g.
Thompson et al., 2003), repetitive (e.g. Fivush & Fromhoff, 1988; Reese & Fivush, 1993), practical (e.g. Engel, 1986, in Fivush & Fromhoff, 1988), or paradigmatic (e.g. Tessler &
Nelson, 1994) in style (Fivush, 1991; Laible & Thompson, 2000).
There are two ways in which these early notions of narrative style have been refined in more recent research. Firstly, it has become more widely acknowledged that narrative styles are not dichotomous, but rather, are on a continuum. All parents show aspects of each style; it is simply that the extent to which these aspects are shown varies (Fivush, 1991; Fivush,
Haden, & Reese, 2006). Secondly, in acknowledgement of this continuum, researchers have recently referred to one single dimension of narrative style: elaboration. Narrative styles therefore vary from highly elaborative to very low in elaboration (e.g. Haden, Ornstein,
Eckerman, & Didow, 2001; Harley & Reese, 1999; Hudson, 1993; Nelson, 1996; Nelson &
Fivush, 2004; Reese, 2002a, 2002b). Although the degree to which parents are elaborative, relative to other parents, remains stable across time, there is a tendency for parents to become more elaborative as children become more competent and skilled participants in reminiscing conversations (Newcombe & Reese, 2004).
What are the elements of a high elaborative style or a low elaborative style? Arising from a strong body of research findings, there is general agreement in the literature that high elaborative parents consistently elicit longer, more embellished discussions of events with their children by frequently asking open-ended questions, encouraging talk about aspects of the events in which the child is interested, providing detailed descriptions of the event, and positively evaluating child responses. Low elaborative parents, on the other hand, elicit 20 shorter and less complete discussions of events by repeating their own questions and statements as though in search of a particular response that the child has not yet provided (e.g.
Boland, Haden, & Ornstein, 2003; Reese, 2002b). Thus, it is not the case that low elaborative parents provide no elaboration or contextual support to their children, as was indicated by the dichotomous approach previously employed to conceptualise narrative style; rather, they provide less elaboration and contextual support than do high elaborative parents.
Individual elements of a high elaborative style
The individual elements of a high elaborative style are rarely considered in isolation.
Fivush et al. (2006) teased apart these elements. They stated, “whereas it is clear that elaborative reminiscing broadly defined plays an important role in multiple aspects of developmental outcome, it is critical to understand that an elaborative reminiscing style is actually a conglomeration of many elements” (p.28).
Ornstein, Haden, and Hedrick (2004) suggest that encoding of an event as it unfolds may be especially facilitated when attention is drawn to particular event components via an open-ended question and subsequent response. The same is likely to be true of open-ended questioning during reminiscing. Open-ended questions encourage children to put an experience into their own words, which may, firstly, aid in the encoding and therefore retrieval of the particular event, and, secondly, according to sociocultural theory, may over time aid in the internalisation of generalised retrieval skills (Fivush et al., 2006). In this sense, coding a high elaborative style as one broad dimension (comprising, for example, of open- ended questioning and information statements) may be problematic, as it is possible for a parent who asks few or no questions to still be coded as high elaborative through the exclusive use of many information statements. Through the exclusive use of information statements, the parent may model to the child a clear and coherent narrative story. Without asking questions of the child, however, the parent does not give their child the opportunity to 21 contribute to the construction of the narrative, or to view the narrative as their own (Fivush et al., 2006). Fivush et al. (2006) argue “this form of elaboration would not be particularly effective as it would neither engage the child nor help the child to learn through participation.
Modelling is not enough; the child must be actively engaged in order to internalise the requisite skills” (p. 30).
In turn, when even high elaborative questioning is not accompanied by the provision of event information, the effect may be closer to that of a low than a high elaborative style
(Fivush et al., 2006). Event information is necessary for the scaffolding of an event representation, for two reasons. Firstly, as the parent is a partner in the collaborative reconstruction of the event, it is clear that the information he or she contributes to this process with have direct impact on the collaborative outcome. This is particularly the case in the very early preschool years, when children provide minimal event information themselves.
Secondly, and just as importantly, however, children may use event information provided by their parent as a memory cue to aid in their own provision of event information (e.g. Reese &
Brown, 2000). Repeated questioning without the provision of event information may appear interrogatory, and lessen the degree to which a child feels inclined to contribute. Therefore, questioning and information statements are complementary, such that each is necessary but not sufficient for the enhancement of autobiographical memory development.
Close-ended questions: Are they elaborative?
Despite the similarity between various researchers’ accounts of elaboration, there is not universal agreement. Specifically, it is unclear to what degree close-ended questions should be considered part of a high or low elaborative style. These questions do not typically include tag questions, in which a parent makes an information statement he or she believes to be true and then asks the child to agree (e.g. “We went to the park, didn’t we?”), but do include yes/no questions in which the child must either confirm or deny information – which 22 may or may not be true - provided by the parent (e.g. “Did we go to the park?”). Although less common, alternative choice questions (e.g. “Did we go to the park or the school?”) and
‘fill in the blank’ questions (e.g. “And then we went to the ….?”) may also be also included.
Many sociocultural researchers do not distinguish between open- and close-ended questions, or consider both wh-questions and yes/no questions to be ‘question elaborations’
(e.g. Fivush et al., 2003; Leichtman, Pillemer, Wang, Koreishi, & Han, 2000; Reese &
Brown, 2000; Reese & Fivush, 1993; Wang & Fivush, 2005). Other researchers either include only open-ended wh-questions in their definition and coding of a highly elaborative style (e.g.
Boland et al., 2003; Harley & Reese, 1999), or explicitly rate close-ended questions as being low in elaboration (e.g. Laible, 2004a, 2004b).
In light of viewing elaboration as a single continuum, it seems pertinent when considering close-ended questions to ask not ‘are such questions elaborative?’ but ‘to what extent are such questions more or less elaborative than other elements of narrative style?’ For preschool children at least, open-ended questions are considered to be highly elaborative elements of style as they invite the child to actively search memory and contribute new information to the conversation (Fivush et al., 2006). In contrast, close-ended questions limit the degree to which preschool children may contribute. This is not to say that they do not elicit any event information from the child – on the contrary, the child is still being invited to respond – however, this response is necessarily less detailed and linguistically complex than is an open-ended response of the same nature. Interestingly, some of the researchers who consider close-ended yes/no questions to be part of a more elaborative narrative style also consider ‘fill in the blank’ questions to be low in elaboration (e.g. Fivush et al., 2003; Reese
& Fivush, 1993), despite ‘fill in the blank’ questions, like yes/no questions, requiring only a single word in response. 23
Information statements are also an important part of a more elaborative style, according to most conceptualizations of elaboration, as they contribute to the joint construction of a coherent event representation by providing rich contextual detail. Following this reasoning, Fivush et al. (2006) state that close-ended yes/no questions can also be elaborative, to some degree, if they too include additional information or focus on a new aspect of the event. An information statement, however, provides concrete information believed by the mother to be true (e.g. “You saw lots of giraffes!”), whereas in a close-ended question, the parent may attempt to engage the child by providing information he or she knows is false (e.g. “Did you see any monsters?”). Furthermore, although information statements do not directly prompt the child to respond with new information, the nature of conversation is such that simply making a statement and pausing invites the conversational partner to respond appropriately. For children old enough, an information statement may therefore allow them the opportunity to respond in more detail than might a close-ended question of the same nature, as a close-ended response has not been specified.
Fivush et al. (2006) suggest whilst close-ended questions certainly provide and elicit less information than do other elements of a high elaborative style, they should be considered high elaborative when used with young children. Research shows that parents adjust the way they engage in discussions about past experiences as children get older (e.g. Hudson, 1990).
When a toddler is just beginning to contribute to event discussions, his or her parent may ask close-ended yes/no questions in an attempt to elicit this contribution. As children enter the preschool years, there is a shift toward open-ended wh-questions (Farrant & Reese, 2000;
Haden et al., 2006). This shift reflects sensitivity to children’s developing abilities to engage in reminiscing, with parents “upping the ante” by asking children to contribute more as they become capable of recalling information in more detail (Fivush et al., 2006, p.1578). One possible drawback of this view, however, is that if the definition and structural elements of a 24 high elaborative style changes dependent on children’s ability to contribute to a conversation
(and parents’ sensitivity to this ability), then it is never possible to independently verify the impact that this style has on development. It may be more meaningful to argue that the degree to which the structural elements of a high elaborative style are appropriate and useful for children, and therefore engaged in by sensitive parents, may differ according to children’s age and reminiscing ability.
In sum, close-ended questions may be somewhat elaborative in nature, but are objectively less elaborative than are other structural elements of narrative style such as open- ended questions and information statements. Despite being objectively lower in elaboration than other elements of narrative style, however, they may be used by parents of very young children as a developmentally sensitive way to entice the child into a conversation to which he or she may be less capable of providing an open-ended answer. By the preschool years, open- ended questions are more developmentally appropriate for engaging children in the process of reminiscing.
Notwithstanding the different approaches that researchers have taken when conceptualising and coding close-ended questions, strong and consistent individual differences in the degree to which mothers use various other elements of narrative talk, including open-ended questions, information statements, evaluation, and repetition, mean that coding parents’ degree of elaboration still retains a good degree of consistency both within and between studies.
The Impact of Parent Reminiscing Style on Children’s Autobiographical Memory
Many studies have shown that as high elaborative parents teach their toddler and preschool-aged children to report both past and present events in a more detailed and complete manner than do low elaborative parents, their children come to independently organise their reports of past experiences in a more elaborative manner than do children of 25 low elaborative parents. This is the case irrespective of whether close-ended questions have been coded as being high or low elaborative in style, whether a full high elaborative style or only open-ended questions are considered, and importantly, is also the case both for experiences that have previously been discussed and for experiences that have not been discussed (see Boland et al., 2003; Farrant & Reese, 2000; Hudson, 1993; Leichtman et al.,
2000). Moreover, although children’s early self-awareness, language, memory, and attachment security also correlate with their subsequent ability to reminisce about personally experienced events, maternal reminiscing style has been found to uniquely predict children’s reminiscing (Reese, 2002b).
The majority of studies investigating individual differences in elaborative style have been conducted using concurrent correlations. In an early study by Fivush and Fromhoff
(1988), for example, 10 mothers and their 30- to 35-month-old children reminisced about everyday past events. Five mothers were identified as being high elaborative, and children of these mothers recalled significantly more event information during their shared past event discussions than did children of low elaborative mothers. In a similar study, Hudson (1993) found that 27-month-old children were more responsive and recalled more information about enjoyable past experiences during discussions with their mother if their mother was high elaborative. It is plausible, based on these results alone, that children of high elaborative parents are not better at recalling autobiographical information than are children of low elaborative parents, but are simply given an unfair advantage during recall (see Nelson, 1996).
High elaborative parents are by definition likely to ask more questions of their children and to provide their children with more memory cues in the form of event details than are low elaborative parents, thus giving children greater opportunity to recall information about the experience. Importantly, however, Hudson (1993) also measured children’s independent autobiographical recall skills with an experimenter and found that children of high elaborative 26 mothers also provided more information about four enjoyable, previously undiscussed experiences, despite the experimenter providing the same prompts for all children.
Taking Fivush and Fromhoff’s, Hudson’s, and a wealth of similar early concurrent correlational findings into account (e.g. Fivush & Fromhoff, 1988; Hudson, 1990, 1993;
Leichtman et al., 2000; Reese & Brown, 2000; Reese & Fivush, 1993; Tessler & Nelson,
1994), the association between parents reminiscing style and children’s autobiographical recall is clear. Not as clear, however, is the direction of this association. A drawback of much of the reminiscing style and autobiographical memory literature is a reliance on concurrent correlations, which cannot be used to establish causality (Dunn, 2006; Fivush, 1991). It is tempting to conclude from the findings of such studies that maternal elaboration contributes to children’s autobiographical memory development. Harris, however, argues that
“demonstrating the process of transmission from parent to child is not easy” (2006, p.158).
Specifically, findings do not rule out the possibility that a sensitive parent will adjust his or her narrative style in order to match their child’s memory competence (Harris, 2006).
Children who are better at recalling their personally experienced past events, as demonstrated during shared reminiscing with parents and independent reminiscing with experimenters, may be viewed by their parents as being capable of responding to multiple questions and additional event details. Parents of children who recall less of their personally experienced past events, on the other hand, may ask only close-ended questions, repeatedly, so as to help the child find the ‘right’ answer but not exceed the child’s ability to respond. In order to address these concerns and determine the direction of effect, longitudinal and experimental studies are examined.
Longitudinal findings. Given that correlations generally allow for multiple interpretations (that is, maternal elaboration causes children’s autobiographical recall; children’s autobiographical recall causes maternal elaboration; there is a bi-directional 27 relationship; or both are caused by an extraneous variable), studies capable of providing causal information are required to determine which of these interpretations is most appropriate. Using a longitudinal design, Reese, Haden, and Fivush (1993) examined mother- child reminiscing when children were 40, 46, 58, and 70 months of age. Importantly, whilst all mothers became more elaborative over time, consistent individual differences in reminiscing style were apparent. Cross-lagged correlations showed that these individual differences were related to children’s concurrent and subsequent contributions to their shared memory conversations, with more elaborative mothers having children who are able to contribute more memory information themselves. As maternal style at early ages predicts children’s subsequent memory, but the reverse does not hold true, this finding provides evidence to suggest a direction of effect from mother to child. By the time children were 58- and 70-months-old, bidirectional influences between children’s memory contributions and mothers’ reminiscing style were observed.
Harley and Reese (1999; Reese, 2002a) also conducted a longitudinal study examining the relationship between maternal reminiscing and children’s autobiographical memory; however, their study differed from Reese et al.’s (1993) earlier study in two ways. Firstly, children were aged just 19 months at the first time-point, and 32 months at the final time- point. Secondly, children’s independent recall was measured together with their shared recall.
Whilst maternal elaboration predicted children’s contributions to shared recall, independent recall was unaffected by maternal reminiscing style. Harley & Reese (1999) conclude that their sample may be too young: independent recall for events previously undiscussed is not typically developed until late in the preschool period.
Experimental findings. A small number of experimental and intervention studies also provide direct causal evidence that elaborative discussion of an event leads children to develop a stronger and, perhaps, more accurate event representation. These studies do not rule 28 out the possibility of a bi-directional relationship between maternal elaboration and children’s autobiographical recall, as was found by Reese et al. (1993), but do not support the notion that the direction of effect goes only from child to parent.
In McGuigan and Salmon’s (2004) experimental study, 3- and 5-year-old children participated in a staged, ‘visit to the pretend zoo’ event. In this event, children visited five animals and completed various activities at each animal. The overarching aim of the visit was to find the lost baby elephant. The event itself - that is, the costumes, props, and activities - remained the same for all children; however the timing of talk was manipulated. Children were assigned to high elaborative pre-, during-, or post- event talk with ‘the zookeeper’, in which objects, actions, and goals were labelled and described three days before, during, or three days after the event respectively and empty talk was used at other times; or to an empty talk control condition, in which minimal, non-informative language was used at all times.
Children who participated in high elaborative talk with the zookeeper during and especially after the event later correctly recalled more of the event than did children who participated in high elaborative talk before the event, or who participated in minimal talk at all times.
Furthermore, items that children rather than first discussed as part of the high-elaborative pre-, during- or post-event talk were better recalled than were items that ‘the zookeeper’ initiated.
Thus, both the timing of talk and the relative contributions of adult and child are important for recall.
Conroy and Salmon (2006) extended these findings to also examine low-elaborative rather than empty talk. Five- and six-year-old children experienced a staged ‘visiting the pirate’ event (similar to the ‘visit to the pretend zoo’ event), containing two logically and two arbitrarily connected scenes of five components each. Children were allocated to a high- elaborative condition, low-elaborative condition, or no discussion control condition. For three consecutive days after the event, children in the high- and low-elaborative conditions 29 discussed five logical and five arbitrary components with the experimenter, leaving five logical and five arbitrary components undiscussed. Although high-elaborative reminiscing better facilitated recall for discussed components of the event than did low-elaborative reminiscing, the degree of connection between scenes was important for the recall of non- discussed components. For arbitrarily connected scenes, memory for non-discussed components was impaired relative to the no-discussion control condition, with the degree of impairment being equal after high- and low-elaborative discussion. For logically connected scenes, memory for non-discussed components was not impaired after discussion of either style. This is the first experimental study to compare high elaborative reminiscing with both low elaborative reminiscing and a no-reminiscing control. Although the findings of
McGuigan and Salmon’s (2004) and Conroy and Salmon’s (2006) studies refer to experimenter- rather than parent-guided memory for one event only, and cannot examine bidirectional influences, they are important to the current discussion in that they provide evidence to suggest a direction of effect from adult to child (see also McGuigan & Salmon,
2006; Ornstein & Haden, 2006).
Three intervention studies have also been used to establish the causal influence of adult narrative style, either during or after an event, on children’s autobiographical memory development. Each produces similar findings. In the earliest of these studies, Peterson et al.
(1999) successfully encouraged mothers to ask their preschool children more open-ended and context-eliciting questions during reminiscing: children of trained mothers subsequently produced autobiographical memories containing more context-eliciting descriptions than did children of control mothers. Using a similar design to Peterson et al. (1999), yet focusing on during-event talk, Boland et al. (2003) found that mothers encouraged to use high elaborative talk whilst participating in a shared ‘pretend camping’ activity had children who were subsequently able to independently recall more of the camping activity after both one day and 30 three weeks than were children of untrained mothers. Finally, Reese and Newcombe (2007) trained mothers to ask open-ended questions and confirm their toddler-aged children’s utterances during reminiscing. One year later, children of mothers who were trained provided richer shared memories than did children of mothers who were not, and, for those with high
(but not low) self-awareness, also showed better independent recall.
In McGuigan and Salmon’s (2004) and Conroy and Salmon’s (2006) experimental studies, together with Boland et al.’s (2003) intervention study, memory was investigated only for the event that had been discussed (that is, the zoo event, the pirate event, and the camping event respectively). By also obtaining a post-training measure of children’s recall of undiscussed events to an impartial experimenter, Peterson et al. (1999) and Reese and
Newcombe (2007) provide important causal evidence that long-term elaborative style impacts children’s underlying development of autobiographical memory for events in general, as previously suggested by Hudson (1993). This evidence is particularly strong in the case of
Reese and Newcombe’s (2007) study, as the accuracy of children’s independent memories was also verified by their mothers. Across both levels of self-recognition and condition allocation, an average of 86% of children’s independent memory utterances were rated by parents as maybe or definitely correct. Thus, it is not simply the case that children who participate in highly elaborative event talk recall more about those events in isolation. Further experimental and intervention studies are required to support the preliminary causal results that have been found to date, and to further elucidate the potential bi-directional relationship between a high elaborative style and children’s autobiographical recall.
Elaborative Structure, Autonomy Support, and Autobiographical Memory
Engel (1986, as cited in Cleveland & Reese, 2005) first suggested that ‘reminiscers’ have the goal of engaging children in reminiscing for its own sake, whereas ‘practical rememberers’ have the goal of demonstrating their children’s memory performance. Many 31 researchers now suggest that high elaborative parents reminisce with the goal of engaging their child, and low elaborative parents reminisce with the goal of eliciting a particular memory response (see Fivush et al., 2006, for review). Nonetheless, the nature of these goals during reminiscing is rarely included in elaborative coding schemes. Moreover, the success of elaborative training studies that do not address reminiscing goals suggests that the goal with which one reminisces may not be essential for enhancing children’s memory (Cleveland &
Reese, 2005).
To address the importance of reminiscing goals more directly, Cleveland and Reese
(2005) separately analysed ‘elaborative structure’ and ‘autonomy support’ in mothers.
Elaborative structure refers to the narrative guidance that mothers give their children in the form of additional information and open-ended elaborative questions. Autonomy support refers to mothers’ willingness to follow in on the child’s perspective. Mothers are therefore considered to be autonomy supportive if they are child-oriented in their reminiscing, and support the child’s interests rather than their own. They are considered to be low in autonomy support if they are controlling and support their own interests (Grolnick, Fodi, & Bridges,
1984).
Importantly, Cleveland and Reese (2005) found that elaborative structure and autonomy support each emerged as unique and independent components of reminiscing style during the preschool years. When children were 40 months old, those whose mothers provided either high elaborative structure or high autonomy support provided more information during shared recall than did those whose mothers provided low elaborative structure or low autonomy support, and those whose mothers provided both high structure and high autonomy support recalled the most of all. When children were 65 months old, however, recall was no longer affected by autonomy support. Therefore, whilst a high elaborative style is important past the age of 6, the extent to which parents support children’s contributions to 32 reminiscing is important primarily at younger ages. This may be because children at younger ages are still highly dependent on the guidance of a more capable adult. As they reach an age at which they are capable of independent recall, they may support their own interests to a greater degree, and the autonomy support provided by parents becomes less necessary.
The Mechanisms of the Elaborative Style
Influenced by Vygotsky (1978)’s notion of cognitive development, social interaction theorists Fivush, Pipe, Murachver, and Reese suggest, “development [of autobiographical memory] occurs as adults lure children into interactions just beyond what they can achieve on their own” (1997, p.35, emphasis added). By eliciting children’s participation in discussions about their experiences, parents gain the opportunity to support and guide their recall.
Furthermore, according to Fivush and Fromhoff (1988), by encouraging young children to participate in discussions about their experiences, parents inform children that reminiscing about the past is a valuable social tool. Four- to 6-year-old children’s memory contributions in turn predict the degree to which mothers will use a high elaborative style (Reese et al., 1993), such that both conversational partners are influenced and accommodated by the other (Fivush et al., 2006). The developmental mechanism of change from this perspective is the language interaction participated in by parent and child (Nelson, 1996). The nature of this language interaction - that is, the style used by the parent and the child’s subsequent responses - determine the degree to which change occurs and memory skills are advanced.
It is clear from previous research that the style with which mothers talk about past and present events affects the degree to which their children later recall both discussed and undiscussed events. It is less clear, however, whether children of high elaborative parents develop richer underlying representations of these events, or simply come to report them in more detail (Harley & Reese, 1999). McGuigan and Salmon (2006) address this question in an experimental study comparing the influence of high elaborative and empty discussion on 33 both verbal and nonverbal recall. They found that high elaborative talk during an event enhanced 5- and 6-year-old’s verbal recall and nonverbal re-enactment of the event, irrespective of whether verbal recall or nonverbal re-enactment was tested first. Given that nonverbal recall, or re-enactment, was also enhanced, it seems likely that high elaborative compared to empty talk – at least during the event - helped children to encode a richer underlying representation of the event. Furthermore, Craik and Tulving (1979)’s seminal model of elaborative processing suggests that semantic material that is elaboratively processed in a rich, extended manner is represented in memory more richly than is semantic material less elaboratively processed, as elaborative processing strengthens the links between associated memory propositions. With regards to autobiographical memory, Harley and Reese
(1999, p.1340) state that “We do know that children of high elaborative parents are being exposed to and are engaging in richer and more complex rehearsals of past events [than are children of less elaborative parents] with each conversation … it seems likely that elaborative rehearsals would serve to strengthen and elaborate the original memory trace in much the same way as other forms of elaborative processing enhance recall”.
In support of this view, Fivush and Fromhoff (1988) propose that high elaborative parents do not simply provide more information to their children than do low elaborative parents; they provide a different kind of information. Specifically, they give a fuller, more embellished, child sensitive account of the experience being discussed. Children who participate with their parents in high elaborative discussions about their experiences are likely to more strongly view reminiscing as important and enjoyable than are children who participate in low elaborative discussions. Children of high elaborative parents will therefore be more easily encouraged to participate than will children of low elaborative parents.
Moreover, it is through embellishment that elaborative parents achieve two important aims.
Firstly, they distinguish the experience being discussed from other similar experiences. 34
Secondly, they inform children what specifically it is about the experience that is distinctive, memorable, and interesting to talk about. To distinguish between high and low elaborative parents, Fivush and Fromhoff (1988) give the example of a pig. The child of a high elaborative parent doesn’t just see a pig, but a big fat pig rolling in the mud. The child is provided with both a distinctive memory cue, enriched such that it will be more accessible in the future (Ornstein et al., 2004), and a reason for why the event is interesting to remember in the first place.
Memorial Benefit of Elaborative Reminiscing
Research has shown that high elaborative parent-child discussion, both during and after an event, is associated with children’s organised reports of their experiences, yet only a very small number of studies have systematically examined whether high elaborative discussion of an event is more beneficial during or after an event has occurred. These studies have produced mixed results.
In McGuigan and Salmon’s (2004) experimental study described above (p.25), it was found that reminiscing more greatly affected 3- and 5-year-olds’ recall than did during-event or pre-event discussion. This was particularly the case for 5-year-old children, who were later able to recall to an experimenter twice as many items of information about the trip, especially self-generated items, if they participated in reminiscing rather than pre- or during-event discussion. McGuigan and Salmon suggest that high elaborative reminiscing, especially where children are active participants, is particularly beneficial for recall as it allows children two spaced exposures to the event (the event itself, and the post-event discussion). This suggestion is consistent with the spacing effect identified in learning studies, in which optimal encoding and recall occurs when material is presented during spaced rather than massed presentations (Bahrick, 2000; Dempster, 1996). Whilst Price, Connolly, and Gordon (2006) found only mixed evidence for a spacing effect in the recall of a staged event repeated four 35 times, Bellezza and Young (1989) suggest that the spacing effect is most effective when the encoding context of subsequent exposures is somewhat different to that of the original exposure, such that the subsequent exposure both reinstates and expands the memory representation. Presumably, the different modalities of the event itself and the post-event talk in McGuigan and Salmon’s (2004) study achieve this aim, allowing children the opportunity to reflect upon and evaluate their experience.
In an experimental study conducted by Ornstein and Haden (2006), however, high elaborative reminiscing did not show a consistent advantage over high elaborative discussion during a staged event. Fifty-four children aged 4 years participated with an experimenter in one of four conditions characterising both during- and post-event talk: elaborative/elaborative, elaborative/empty, empty/elaborative, and empty/empty. During reminiscing, children who participated in high elaborative reminiscing recalled significantly more information about the event than did children who participated in empty reminiscing, irrespective of their during- event talk style. During a separate memory interview three weeks later, however, children who participated in high elaborative talk during the event recalled significantly more information than did children who participated in empty talk, yet reminiscing style had no significant effect on recall. It is unclear why Ornstein and Haden’s (2006) findings differ from
McGuigan and Salmon’s (2004): both use Western samples of similar ages, and the staged events that children participated in each study were both structured and interactive. Further research is required to clarify these findings.
Whilst no significant interactions between during event and reminiscing style were found by Ornstein and Haden, perhaps due to the relatively small number of participants in each of the four cells, the authors highlight a non-significant finding whereby children who experienced high elaborative talk both during the event and during reminiscing seemed to 36 recall more than did children in any other condition. Although non-significant, this finding suggests that dual or multiple exposures to the event is the strongest form of reinstatement.
Child Influences on Reminiscing Style
Influenced by Vygotsky (1978), sociocultural theorists propose that autobiographical memory development occurs within social exchanges between parent and child (e.g. Nelson,
1993a, 1993b; Reese & Fivush, 1993). Whilst coding schemes often focus on parents’ questions and comments to children, the interactive nature of this conversation is important.
Firstly, the high elaborative reminiscing style may contribute to a more interactive exchange between parent and child. Research shows that mothers who use structural elements of the high elaborative style, such as wh-questions and information statements, may also be more likely to confirm and evaluate children’s responses in the conversation (Bauer & Burch, 2004;
Reese et al., 1993). Although confirmations do not in themselves contribute to children’s developing memory skills, they do engage children in the language interaction. Secondly, it has been found that individual differences in reminiscing style are moderated by child-related influences. These influences include children’s own contributions to memory conversations, gender, and temperament, together with parent-child attachment.
Children’s Contributions to Reminiscing Style and Recall
Although research conducted by sociocultural theorists typically focuses on the impact of parent characteristics, such as narrative style, on children’s development, a key premise of this theory is that children’s contributions to their interactions with others are also vitally important. Farrant and Reese (2000) found that when children were 2 years of age, their interest in participating in memory conversations predicted the number of open-ended questions that mothers would ask of them during shared reminiscing. Furthermore, both maternal reminiscing style and children’s interest in participating in memory conversations contributed to young children’s shared memory elaborations, which in turn later contributed 37 to their independent memory at 40 months. By the time children are 58-months-old, their contributions to reminiscing conversations predict maternal reminiscing style at 70 months
(Reese et al., 1993), and at 3 and 5 years, children are more likely to recall aspects of an event that they themselves have contributed to an earlier reminiscing conversation than aspects that an adult partner has contributed (McGuigan & Salmon, 2004; see Ornstein et al., 2004, for similar findings relating to talk during an event in which children who more frequently respond to their mother’s open-ended questions have better subsequent memory of the event).
The latter finding by McGuigan and Salmon is consistent with previous research of a
‘generation effect’: children tend to better recall information that they themselves generate than information that they have read or heard (e.g. McNamara & Healy, 1995; Slamecka &
Graf, 1978). Combined, findings show that children contribute to both their parents’ reminiscing style and their own subsequent recall (Farrant & Reese, 2000).
The Impact of Child Gender on Reminiscing Style
Consistent with the notion of gender based socialisation differences in childhood, several studies have found that both mothers and fathers talk more overall with preschool aged girls than with preschool aged boys, and are more highly elaborative and evaluative with girls than with boys (e.g. Fivush et al., 2003; Reese & Fivush, 1993; Reese, Haden & Fivush,
1996). For instance, Reese and Fivush (1993) found that both mothers and fathers talked to their 40-month-old children in a more elaborative manner if their child was a daughter rather than a son. Similarly, Fivush et al. (2003) found that mothers reminisced about emotion- oriented events with their 42- to 48-month-old daughters in a more highly elaborative style, and with more evaluative comments, than with sons. In contrast to these findings, Wang and
Fivush (2005) found few gender differences in either American or Chinese dyads when asked to reminisce about positive and stressful events. Nonetheless, no study conducted using a
Western sample has found parents to reminisce more elaboratively with sons than with 38 daughters. Thus, if gender differences exist, young girls appear to be given greater scaffolding and support with which to develop autobiographical memory than young boys. Importantly,
Fivush et al. (2006) emphasise that despite possible differences in the way that parents reminisce with daughters and with sons, there is no evidence to suggest gender differences in children’s willingness or ability to reminisce.
Associations between Attachment and Reminiscing Style
Recent research indicates that a child’s secure attachment to their mother predicts their autobiographical memory development. Securely attached children participate to a greater extent in reminiscing conversations (Laible, 2004a). Furthermore, mothers of children who are securely attached are more highly elaborative when talking about the children’s past experiences than are mothers of insecurely attached children, and the memory elaborations that they use tend more often to follow the child’s interests (Laible, 2004a). These findings are consistent with the spiral model of parent-child interaction proposed by sociocultural theorists (e.g. Newcombe & Reese, 2004). Importantly, Reese and Farrant (2003) found long- term associations between maternal reminiscing style and children’s recall only in dyads with securely attached children.
The Impact of Child Temperament on Reminiscing Style
Preliminary findings indicate that a child’s temperament also impacts the degree to which their mother will elaborate with them during reminiscing. Measures of temperament vary, but generally include dimensions such as activity level, emotionality, and sociability
(Lewis, 1999). Laible (2004a) found that mothers were more highly elaborative and discussed negative emotions more with children who were, according to parent ratings, high in negative reactivity. Mothers appear to be sensitive to their child’s greater need for assistance in regulating negative emotions, and adjust their reminiscing accordingly. Furthermore, mothers of children who showed greater effortful control (that is, the ability to inhibit a prepotent 39 response) and less extraversion were more highly elaborative than were mothers of children with less ability to consciously regulate and control their behaviour (Laible, 2004a). This finding indicates that, consistent with the spiral model of parent-child interaction, mothers are encouraged to elaborate more with children whose temperaments permit them to attend and focus. According to Lewis (1999), children who are highly impulsive and active or who are easily distracted may elicit a shorter, more direct, less elaborative reminiscing style from their mothers in order to control the child’s behaviour and keep them on task.
The Impact of Culture on Reminiscing Style
Cultural differences in patterns of parent-child discourse clearly illustrate the importance of elaborative mother-child reminiscing on children’s autobiographical memory development (Peterson & McCabe, 1994; see Leichtman, Wang, & Pillemer, 2003, for review). For instance, it has been found that European American mothers talk about past events to their 3 year old children up to three times as often as do Korean and Chinese mothers, and are more highly elaborative than are Korean and Chinese mothers (Han,
Leichtman, & Wang, 1998). Furthermore, memory conversations between European
American dyads may be more interactive: Wang, Leichtman and Davies (2000) present research that European mothers and children each elaborate on one other’s memory responses more than do Chinese mothers and children. Chinese mothers more frequently direct the discussion and the child’s responses by posing and repeating factual questions, and focusing on behavioural standards.
Importantly, these differences in reminiscing correspond to differences in children’s autobiographical memory development. European American children’s autobiographical memory develops at around 39 months of age, whereas Asian children’s autobiographical memory does not typically develop until 56 months of age (Nelson, 1996). Moreover,
European American children provide longer, more specific, and more highly elaborative 40 memories than do Chinese children (Han et al., 1998; Wang, 2004). In cultures with a strong oral tradition, such as New Zealand Maori, autobiographical memory is developed at an even earlier age than in Pakeha (European-origin) New Zealanders (MacDonald, Uesiliana, &
Hayne, 2000).
It is possible that the social functions served by autobiographical memory are different across cultures, and that these differences contribute to the cultural differences in maternal reminiscing and children’s autobiographical memory development. Consistent with the norms of Eastern society, in which collectivity and adherence to social norms are valued above individuality, Wang (2004) found that Chinese children not only recalled less of their past events than did European American children, but also focused on social interactions and daily routines, particularly with regards to their own social role, more greatly than did European
American children. European American children focused more strongly on their own individual preferences and experiences (also see Leichtman et al., 2003; Wang et al., 2000).
Consistent with these findings, Wang (2006) found an important influence of culture on children’s autobiographical memory, mediated by maternal style and child self-concept.
Parents’ Narrative Content
Further demonstrating the powerful influence of parent-child discussion on children’s autobiographical recall, a number of studies have shown that narrative content included by mothers in early conversations with their children tends to be the same content that their children use more predominantly in their own narratives (Fivush, 1993; Sales, Fivush, &
Peterson, 2003). This is true for both how children come to talk about the past, and what they come to talk about.
When considering how parents and their children recall the past, it has been found that the narrative devices - that is, the generalised type of content information - used by parents to organise their shared narratives tend to be the same devices used by their children to organise 41 both shared and independent narratives. These narrative devices include the provision of orienting information (Newcombe & Reese, 2004; Peterson et al., 1999; Peterson & McCabe,
1994), complex temporal information (Fivush, 1991) and evaluative information such as intensifiers (e.g. “It was very good”) and emotions (Kuebli, Butler, & Fivush, 1995;
Newcombe & Reese, 2004; Sales et al., 2003). For example, in a small longitudinal study of
10 mother-child dyads, Peterson and McCabe (1994) found that mothers who most frequently prompted their 2-year-old children for orienting information during reminiscing, such as when and where things had occurred, had children who later provided more orienting information during reminiscing to an experimenter than did other children. Following this, Peterson et al.
(1999) successfully trained mothers to ask more orienting and context eliciting questions. As expected, children’s orienting and context-setting descriptions also increased (see p.29). In the latter study dyads were both older than those in the former study, at 43 months of age, and economically disadvantaged; thus, results can be generalised across ages and social standing.
Orienting, temporal, and evaluative information all serve to place an experience in context.
This context that is important not only for distinguishing the memory of a particular experience from memories of similar experiences, but also for communicating ones’ experiences to others in a detailed manner that is easily understood.
Parents’ discussion input is also important for children to learn what information to include in their autobiographical accounts of events (Fivush, 1991). In a small but influential study, Tessler and Nelson (1994) invited 10 mothers and their children aged three-and-a-half to the museum. Five mothers were instructed to discuss the event with their child as they normally would, and five were instructed to respond minimally to their child. Interviews conducted one week later revealed that children were significantly better at recalling objects that had been jointly discussed than they were at recalling other objects (Tessler & Nelson,
1994). Haden et al. (2001) later obtained similar findings using three staged play activities 42 including a pretend camping trip, pretend bird watching, and the pretend opening of an ice- cream shop, suggesting that children glean the importance of particular aspects of an event from their mothers before they commit them to memory for subsequent recall.
Notwithstanding the role that a parent’s discussion input plays in his or her child’s subsequent recall, it is important to consider this role in light of the social interaction between parent and child. Haden et al. (2001) and Tessler and Nelson (1994) each found that children better recall features of activities that are discussed by both mother and child than are features of activities that are discussed by the mother or the child only. In fact, consistent with
McGuigan and Salmon’s (2004) findings, in which children better recalled items of an event that they themselves had contributed to the discussion about the event than that the experimenter had contributed, children in Tessler and Nelson’s study were entirely unable to recall objects that the mother had discussed but the child had not. It therefore seems that preschool aged children do not simply insert their parent’s narrative discussion content into their own narratives. By preschool, Nelson (1996) argues, adults help build on the sometimes skeletal event sequences that the child brings to the conversation, but children also play an active role in both encoding and recall (see Fivush, 1991, 1993, for similar arguments). Thus, although parents may guide reminiscing, the process is collaborative.
Discussion of Emotional Experiences in the Past
The inclusion of emotional content in past event narratives has received particular attention in recent years. As was earlier stated, autobiographical memory is concerned primarily with episodes that have personal meaning. Following this, Fivush et al. (2003, p.179) argue that “If much of our self-concept is defined though our autobiographical life story, then our emotional reactions to these experiences, as they were occurring and in the present, provide the glue that connects our past to our present and makes these experiences meaningful”. In other words, it is the emotional aspect of events that we recall that makes the 43 events personally meaningful and therefore worthy of remembering (Fivush, 1993). As illustration of the importance of emotions in autobiographical memory, Pasupathi (2003) argues that nearly all highly emotional events that are experienced are later disclosed to others.
There are a number of potential socialization benefits of engaging in emotionally rich reminiscing. Through emotion-laden reminiscing, children are provided with a linguistic representation of their emotional experiences and, within a secure attachment relationship, may therefore develop the ability to understand and convey a range of experiences and emotional responses to others in an organized and coherent form (O’Kearney & Dadds, 2005;
Wareham & Salmon, 2006). These abilities, in turn, are likely to provide at least some of the skills necessary to engage with others in conversation and to report any difficulties or concerns that might be experienced (Wareham & Salmon, 2006).
Research shows that mothers who refer frequently to emotions and their causes during past and present event discussion typically have children who later refer more frequently to emotions and their causes than do their peers (e.g. Sales & Fivush, 2005; Welch-Ross, Fasig,
& Farar, 1999). In the context of past-event discussion, for example, Kuebli et al. (1995) found that mothers who talked more about emotions with their 40-month-old children had children who used more emotion words at 58 months of age. By 70 months of age, there was a strong correlation between mothers and children’s concurrent emotion references. Children who talked more about emotions at 40 months of age, however, were unable to elicit more emotion talk from their mothers than were other children. Although Kuebli et al.’s results were correlational, the fact that mothers’ emotional references predicted children’s subsequent emotional references, and not vice versa, suggests that emotional scaffolding may have a causal influence on young children’s recall of the emotional aspects of their past experiences. 44
The Impact of Child Gender on Emotion Talk
Interestingly, women show superior overall recall of experiences associated with emotion than do men, and, when the content of an event is controlled, also recall more emotional information specifically (Adams, Kuebli, Boyle, & Fivush, 1995; Bloise &
Johnson, 2007). Furthermore, throughout adulthood, women are generally considered to be more responsible than men for maintaining positive interactions (Fivush, 1993). Women express emotion more often, and report experiencing and valuing emotions at a greater level than do men (Fivush et al., 2003; Kuebli et al., 1995). It is widely argued that these gender differences observed in adulthood stem at least partially from young girls being socialised to attend to, display, and discuss emotions to a greater extent than young boys in childhood
(Hertsgaard & Mathews, 1993; Kuebli et al., 1995).
There is evidence to suggest gender differences in both the degree to which parents talk about emotions with their young children, and the way in which they do so. By observing families in their homes, Dunn et al. (1987) found that mothers encouraged more communication about feelings by girls than by boys at as early as 18 months of age. By 24 months of age, girls made more references to feelings than did boys (see also Cervantes &
Callanan, 1998). Furthermore, Fivush, Brotman, Buckner, and Goodman (2000) found that both mothers and fathers used more emotional utterances when talking about sad events with daughters than they did with sons, and Fivush and colleagues (Fivush, Brotman, Buckner, &
Goodman, 2000; Fivush et al., 2003) found that mothers and fathers are more likely to place emotional experiences in an interpersonal context with four-year-old girls than with four-year- old boys, particularly for negative experiences involving anger or fear. Two suggestions emerge from these findings. First, mothers and fathers each adjust the emotion content of these conversations depending on their child’s gender. Second, young girls may be forming a richer and more complex knowledge of emotions than might young boys, and might also be 45 more capable than boys of applying this knowledge to the range of emotions that others display, as well as to their own emotional experiences.
Future Research
To date, research conducted within the sociocultural framework has focused primarily on associations between parents’ and children’s reminiscing style and content, particularly during reminiscing, and the influence these have on autobiographical memory development.
To a degree, language and literacy development, and, more recently, theory of mind development, have also been examined (e.g. Reese, 1995; Reese & Cleveland, in press). It is the aim of this thesis to extend this literature and examine ways in which these interactions between parents and their children might facilitate the internalisation not just of memory, but also of other socio-cognitive skill. That emotions characterise autobiographical memories, and that parents who discuss emotions frequently have children who do the same, suggests that reminiscing conversations may provide children with the opportunity to internalise emotion- oriented information. The impact of both style and content elements of reminiscing will therefore be examined with regards to children’s emotion knowledge.
The Development of Emotion Knowledge
As autobiographical memory and language develop across the preschool years, so too does children’s emotion knowledge (Bennet, Bendersky, & Lewis, 2005). Given that young children’s own emotions and others’ emotions are both central experiences in their young lives, developing an understanding of these emotions is critical (Denham, 1998). Children’s emotion knowledge is characterised by an awareness and understanding of their own’ and others’ emotional states, expressions, causes, and outcomes, and, although cognitive skills are involved, socio-emotional skills are more dominant (Denham, 1998).
Research shows that children learn from ‘socialising agents’, primarily their parents, about the appropriate expression of emotions, the nature of emotional expressions and 46 situations, means of coping with emotions, and potential reactions to others’ positive and negative emotions (Denham, 1998). According to the emotion coaching hypothesis, children’s growth in emotion exploration and knowledge is directly facilitated not only by parents’ direct modelling of emotion expressions and reaction to children’s own emotion expressions, but also by their ongoing discussion and explanation of emotions (Denham, 1998; Denham et al.,
1994). The emotion coaching hypothesis is thus broadly consistent with the central premise of sociocultural theory which, following Vygotsky (1978), states that children’s cognitive and representational development occurs during parent-child discussions about the past and present. Following sociocultural theory, furthermore, both parents’ and children’s roles in these interactions are central to development (Fivush et al., 2006; Symons, 2004).
The Importance of Emotion Knowledge for Preschoolers
The growth of emotion knowledge in young children serves two vital functions.
Firstly, emotion knowledge is essential to the development of emotion regulation, the process of monitoring and modifying both the experience and the expression of emotions (Kopp,
1989; Pasupathi, 2003). Adaptive emotion regulation involves the ability to experience and communicate their emotions in a way that allows important goals such as safety, maintenance of positive social interactions, and perceived competence, to be met (Bridges, Denham, &
Ganiban, 2004). As evidence that emotion knowledge is important for children’s ability to adaptively regulate their own emotions over time, Izard et al. (2001) found that in low socio- economic samples, poor emotion knowledge at age 5 is related to internalising behaviour at age 9. Although rarely as obvious to parents and teachers as externalising disorders, internalising disorders are not only damaging in their own right, but can also lead to serious behaviour problems (Izard et al., 2001).
Secondly, emotion knowledge is a vital tool for children to not just understand and regulate their own emotions, but also to predict others’ emotions and subsequent emotion 47 related behaviour (Denham et al., 1994; Kuebli et al., 1995; Thompson et al., 2003). By understanding others’ emotional actions, children are better able to adjust their responses accordingly. Research suggests that preschoolers’ emotion knowledge correlates positively with their concurrent empathic behaviour, prosocial behaviour, and peer status; both in middle and upper class families (Denham, 1986; Denham et al., 2003; Hughes & Dunn, 1998), and in working class families (Izard et al., 2001). In a recent study by Denham et al. (2003), for example, 143 children’s emotion knowledge, emotion regulation, and social competence were measured at age 3 to 4 and at age 5 to 6. To measure emotion knowledge, Denham’s Emotion
Knowledge Task was used: children were asked to identify emotional expressions on a puppet’s face, and then identify the emotion of the puppet in acted out scenarios by putting the corresponding face on the puppet (Denham, 1986). To measure emotion regulation, children’s expression of emotion and reactions to others’ emotions were observed during free play in their classrooms over a number of trials. Finally, to measure social competence, teacher ratings of competence and peers’ ratings of likeability were used. Whilst language was not measured, children’s emotion knowledge at ages 3 and 4 positively predicted social competence at ages 5 and 6, and venting of emotions at ages 3 and 4 was a negative predictor of both concurrent and subsequent social competence.
Findings: Parent-child Discussion of Emotions and Emotion Knowledge
A sizable body of literature supports the notion that language-based interactions between parents and their children are critical to the development of emotion knowledge. In a laboratory study conducted by Denham and colleagues (1994), for example, Denham’s
Emotion Knowledge Task was used to measure emotion knowledge when children were 42 and 57 months of age. At the first of these time-points, Denham et al. also directed mothers of the preschoolers to act as though they were very sad, and then as though they were very angry. The conversations that occurred between mother and child during this activity were 48 coded, and a regression analysis with current and future emotion knowledge was conducted.
Independent of age and language ability, mothers’ explanations of their sad and angry displays to their children were significantly associated with their children’s current and future emotion knowledge.
Similarly, Dunn, Brown, and Beardsall (1991) found when observing families in their homes that there is a wide variation in the frequency with which 36-month-old children talk about feelings, and corresponding differences in their mothers’ talk about feeling states. By age 6, children who grew up in families that engage in frequent feeling-state talk were better at making judgments about the emotions of unfamiliar adults than were children not as frequently exposed to feeling-state talk. Dunn et al. (1991) suggest that by talking about feeling states, children are able to distance themselves from the emotional experience and thus better reflect upon them. These studies indicate that parent-child and family discussion each significantly predict children’s emotion knowledge.
The Impact of Discussing Emotion Causes above and beyond other Emotion Talk
Importantly, research suggests that the way in which parents discuss emotions with their children might have a greater association with emotional knowledge than might the overall frequency of emotion references uttered. By asking low-income mothers to view a wordless picture book with their preschool aged children, for example, Garner, Carlson Jones,
Gaddy, and Rennie (1997) examined the association between different types of maternal emotion references and children’s understanding of emotion. Emotion references were coded as unelaborated comments, explanations about the causes and consequences of emotions, or empathy related statements. Over and above the effect of children’s own emotion-related comments, mothers’ empathy related statements were found to predict children’s emotion situation knowledge (the ability to use situational and facial cues to determine how a story character felt), and mothers’ explanations of the causes and consequences of emotions were 49 uniquely related to children’s emotional role-taking ability (the ability to use facial cues incongruent with situational cues to determine how a story character felt). Unelaborated emotion comments were not related to any aspect of emotion knowledge. These results support the notion that parents’ emotion references might be differentially related to preschoolers’ emotion knowledge. Specifically, emotion talk that was rich and extended was more closely related than was other emotion language.
A small number of other studies, examining children’s talk with both their mothers and with older siblings, also support the finding that discussion and explanation of the causes of emotions is a stronger predictor of emotion knowledge than is the sheer frequency of emotion talk, or general talkativeness (Brown & Dunn, 1996; Martin & Green, 2005).
Emotion explanations link emotion states to causal information, providing more easily interpretable and socially relevant information (Cervantes & Callanan, 1998). The association between causal discourse and emotion knowledge is stable across the preschool years, and distinguishes those children who, at age 6, can explain ambivalent feelings from those who cannot (Brown & Dunn, 1996).
Associations between Emotion Knowledge, Language, and Theory of Mind
Two key aspects of the understanding of oneself in the social world are the understanding of emotions and of minds (de Rosnay & Hughes, 2006). Understanding of minds, or theory of mind, includes the ability to understand another person’s actions in terms of desires, thoughts, emotions, and beliefs. Like emotion knowledge, it develops rapidly during the preschool years (Denham, 1998; Harris, de Rosnay, & Pons, 2005; Wellman et al.,
2001). During the preschool years, significant associations are observed between emotion knowledge and theory of mind (Hughes & Dunn, 1998). During middle childhood, furthermore, emotion knowledge and theory of mind skills combine as children come to understand the role that beliefs and desires play in determining emotions, and the difference 50 between inwardly felt and outwardly expressed emotions (Pons et al., 2004; Pons & Harris,
2005).
Although emotion knowledge and theory of mind are significantly associated with one another, the nature of the relationship between them is debated (Izard, 2004). Research suggests that both are important for optimal social functioning, and both are influenced by children’s language skill and their social interactions with others (e.g., Astington & Jenkins,
1999; Bosacki & Moore, 2004; Cutting & Dunn, 1999; Lohmann & Tomasello, 2003). For example, just as the discussion of emotions predicts children’s concurrent and future emotion knowledge, mental state references in a variety of contexts, particular kinds of syntax which enable the report of false belief (e.g., embedded propositions such as “John thinks it’s a cat”), and conversations in which children encounter different subjective viewpoints may each predict children’s theory of mind (Astington & Jenkins, 1999; Harris, 1999; Harris et al.,
2005; Lohmann & Tomasello, 2003; Meins, 1997; Ruffman et al., 2002; Symons, Peterson,
Slaughter, Roche, & Doyle, 2005). Furthermore, preliminary research demonstrates a significant contribution of mothers’ high elaborative style to both emotion knowledge and theory of mind (Laible, 2004a, 2004b; Reese & Cleveland, in press; Lohmann & Tomasello,
2003a).
de Rosnay and Hughes (2006) use the umbrella term ‘socio-cognitive understanding’ to refer to children’s emotion knowledge, theory of mind, and other skills related to psychological understanding. Consistent with Vygotsky’s notion of language as a tool for development, Carpendale, Lewis, Susswein, and Lunn (in press) propose that language is not only externally related to socio-cognitive understanding, for example, by supporting executive functions such as working memory and by allowing parent-child interactions in which development is facilitated; it is also internally related to socio-cognitive understanding in that it partially constitutes such understanding (see also Bosacki & Moore, 2004). External and 51 internal functions of language are interlinked: by merely engaging in the process of reminiscing and other language-based interactions, children may acquire important information about the nature of mental representation, associated with the understanding of emotion and of mind (Thompson, 2006). Variants of this view are proposed by several researchers. For example, in considering the possible reasons for the strong association between children’s language skill and their emotion knowledge, Pons and colleagues suggest that language as an “instrument of representation” facilitates children’s ability to understand
“objects of representation” such as emotions. Thereafter, greater understanding enables more effective communication of emotions and a “virtuous cycle” is established (Pons, Lawson,
Harris, & de Rosnay, 2003, p.352).
Notwithstanding the similarities between emotion knowledge and theory of mind, and the strong association of each to language, Dunn and her colleagues (Dunn, 2000; Cutting &
Dunn, 1999) propose that the precursors and sequelae of each differ. For example, whereas early emotion knowledge is associated with peer popularity, early mind-reading skills are related to sensitivity to criticism and to more sophisticated role play with peers. Thus, socio- cognitive understanding is best conceptualised as consisting of inter-related yet distinct skills
(de Rosnay & Hughes, 2006).
Influences on Emotion Knowledge
Systematic factors that contribute to children’s development of emotion knowledge have not been heavily researched. There is, however, preliminary evidence to suggest that both gender and culture contribute to this development.
The Impact of Child Gender on Emotion Knowledge
If girls participate in richer, more elaborative discussions of emotions than do boys, as has been suggested by past research (e.g. Dunn et al., 1987; Fivush et al., 2003; Reese &
Fivush, 1993), then they may also display superior emotion knowledge. Experimental 52 evidence for gender differences in emotion knowledge is mixed. Differences have not consistently been found (e.g. Wang, 2003a), but, when found, they invariably demonstrate that preschool girls display better-developed emotion knowledge than do preschool boys. For example, Hughes and Dunn (2002) demonstrated that 4- to 7-year-old girls show greater differentiation between sadness and anger than do age-matched boys. Likewise, Kuebli et al.
(1995) found that although there was no gender difference in emotion references at 40 months, 70-month-old girls not only talked about emotion more frequently, but also referred to a greater variety of emotions than did their male counterparts. Finally, Martin and Green
(2005) found that mothers total emotion talk and use of causal explanations when storytelling both predicted 41-month-old boys’ emotion knowledge, but that no maternal variables predicted girls’ emotion knowledge (see Cutting & Dunn, 1999, and Bosacki & Moore, 2004, for similar findings that preschoolers’ receptive vocabulary was associated with emotion knowledge only in boys). Martin and Green suggest that girls may have received more information about emotions at younger ages than have boys, and therefore developed superior emotion knowledge whereby emotion talk at 41 months of age is no longer as strong a predictor of this knowledge as it may once have been.
The Impact of Culture on Emotion Knowledge
Only a small amount of research has examined cultural differences in the acquisition of emotion knowledge. In research by Wang and colleagues, it has been found that 4-, 6-, and
8-year-old European American children make more emotion references during reminiscing than do age-matched Chinese children (Wang, 2004), and that European American children have higher emotion knowledge than do Chinese children at age 2.5-3.5 years (Wang et al.,
2006). These cultural differences may be due to differences in emotion socialisation practices.
Parents in European and European American cultures tend to focus on helping children express and understand emotions, whereas Chinese parents use interactions with their children 53 to emphasise psychological discipline and behavioural standards (Wang & Fivush, 2005;
Wang et al, 2006). Thus, European and European American children are exposed to greater emotion coaching and scaffolding than are Chinese children, and may go on to develop a more sophisticated understanding of emotions.
Cultural differences have also been found between British and indigenous South
American children’s development of emotion knowledge. Tenenbaum, Visscher, Pons, and
Harris (2004) found that 4- to 7-year old and 8- to 11-year-old Quechua children, from a rural
Peruvian agricultural village, were less accurate than on items of the Test of Emotional
Comprehension (a test taking into account many different aspects of emotion understanding) than were British children of the same age. Nonetheless, a similar pattern of development was evident. For example, children were better at emotion recognition than they were at understanding of emotion regulation, and older children were more accurate at identifying emotions connected to individual desires and to moral misdemeanour than were younger children. Tenenbaum et al. suggest that the Peruvian children’s slower development of emotion knowledge may relate not only to a relative lack of formal education, but also in part to care arrangements amongst the Quechua in which young children are primarily cared for by
– and thus interact more frequently with - siblings and not parents. It is thus clear from the research of both Wang and colleagues and Tenenbaum and colleagues that, like autobiographical memory, emotion knowledge develops within a culturally determined social framework.
Integrating Research on Past Event Discussions, Autobiographical Memory, and Emotion
Knowledge: The Present Research
This thesis aimed to understand better the ways in which high elaborative and emotion-rich discussion of everyday experiences contribute to children’s development of autobiographical memory and emotion knowledge. In Study 1, associations between parent- 54 child reminiscing, children’s shared and independent recall, and children’s emotion knowledge were examined. In Study 2, reminiscing was manipulated to include causal emotion talk, non-causal emotion talk, or no emotion content, and the causal impact of reminiscing on children’s memory for a staged event was examined. In Study 3, mothers were trained to use a high elaborative, emotion-oriented reminiscing style, and the effectiveness of training in facilitating children’s shared and independent recall and emotion knowledge was evaluated.
Associations between Parent-child Reminiscing and Emotion Knowledge
To date, little research has examined the possible impact that reminiscing might have on children’s development of emotion knowledge, despite both a plethora of findings suggesting an association between emotion knowledge and language-based interactions in other contexts, and the important role of emotions in characterising an event and imbuing it with meaning. Given the emotional salience of the events that we recall, and the association between parents’ and children’s emotion references, reminiscing may also be a fruitful opportunity for children’s development of emotion knowledge (Denham & Burton, 2003).
If it is the case that emotion knowledge develops within a reminiscing context, some forms of reminiscing may be more effective than others. Just as the explanation of emotion causes during ongoing events and activities has been found to better predict emotion knowledge than has other emotion references (Brown & Dunn, 1996; Garner et al., 1997;
Martin & Green, 2005), so too might the explanation of emotion causes during past event discussions be more important than other emotion references for children’s development of emotion knowledge. Furthermore, as for the development of autobiographical memory, the extent to which reminiscing is high elaborative might also determine the extent to which the development of emotion knowledge is facilitated. 55
To investigate these possibilities, Study 1 served the dual purpose of replicating a broad body of research on the autobiographical memory of preschool-aged children, using an
Australian sample, and extending this research into the socio-emotional domain. Associations between parents’ elaborative style and emotion references during discussion of an emotionally salient past event, children’s shared and independent recall, and children’s emotion knowledge were examined. The findings of this study provided a theoretical basis from which to design an intervention enhancing the development of children’s autobiographical memory and emotion knowledge (Study 3).
The Impact of Causal and Non-causal Emotion Reminiscing on Recall
Whilst observational and experimental research has focused on the elaborative style, the impact of different emotion references during reminiscing has not previously been examined with respect to children’s recall for an event. By reminiscing about the emotions that characterise an event, it may be that the salience and memorability of the overall event representation is strengthened. This may particularly be the case for emotion cause explanations, which explicitly link emotion states to other event information (Cervantes &
Callanan, 1998). Much research shows that a strong degree of connectedness between the components of an event, for example, by virtue of logical rather than arbitrary connections, produces better recall (Bauer & Mandler, 1989; Conroy & Salmon, 2006).
Study 2 aimed to examine in more detail the association between emotion-oriented reminiscing and children’s recall, and, by using an experimental paradigm, to establish causal relationships between these variables. Children aged 3- to 6-years each participated in a staged ‘visit to the pretend zoo’ event (adapted from McGuigan & Salmon, 2004). During the event they met five pretend animals, each experiencing a particular emotion. High elaborative reminiscing that focused on emotions, and, in particular, emotion causes, was expected to benefit children’s subsequent recall of both emotional and non-emotional event information 56 more greatly than was either high elaborative talk unrelated to emotion, or low elaborative
‘minimal’ talk. The contributions of pre-existing emotion knowledge to recall were also examined.
Facilitating Recall and Emotion Knowledge via Reminiscing Training
It has often been stated in developmental literature that more effort is needed to translate the findings from basic research into programs aimed at facilitating development
(e.g. Izard et al., 2001). If, in Studies 1 and 2, high elaborative, emotion-oriented reminiscing is shown to facilitate the development of both emotion knowledge and autobiographical memory, then it may be the case that children’s cognitive and socio-emotional development can be enhanced by training parents to reminisce in this way. Intervention studies by Boland et al. (2003), Peterson et al. (1999), and Reese & Newcombe (2007) each show that parents can be effectively trained to use a high elaborative style both during ongoing events and during reminiscing. Nonetheless, each of these previous studies has used a waitlist control. In the context of reminiscing training, a waitlist control cannot account for the additional time that a parent may spend interacting with either the experimenter or with their own child. The possibility that children of trained parents demonstrated superior autobiographical recall due to differences in training method, and not because of the style of discussion per se, cannot definitively be ruled out. Furthermore, no study to date has examined whether parents can also be trained to discuss emotions when using a high elaborative style, and no study has examined the effect of training on children’s emotion knowledge development.
In Study 3, an intervention was implemented in which mothers were trained reminisce using a high elaborative style and discussing the emotions and emotion causes that characterise each event. A powerful control was used, in which mothers were trained to play by following their child’s lead. Although the control provided no instruction with regards to the high elaborative style or emotion references, it did encourage mothers in the two 57 conditions to attend to their children sensitively for an equal length of time. Both immediately and six months after training, children’s shared and independent recall were assessed together with their emotion knowledge. If successful, the intervention should represent a potential avenue by which children’s autobiographical memory and emotion knowledge can be enhanced. 58
CHAPTER 2: ASSOCIATIONS BETWEEN AN ELABORATIVE STYLE, EMOTION
REFERENCES, AND EMOTION KNOWLEDGE
In Study 1, associations between parents’ high elaborative style and emotion references during reminiscing, children’s high elaborative utterances and emotion references during shared and independent recall, and children’s emotion knowledge were investigated.
As discussed in Chapter 1, a large and robust body of work based upon the sociocultural theory highlights the profound influence of reminiscing style on children’s autobiographical memory development (see Fivush et al., 2006; Nelson & Fivush, 2004, for reviews). Parents high in elaboration use detailed descriptions and open-ended questions to construct rich and embellished stories about past experiences with their children (e.g. Fivush & Fromhoff, 1988;
Harley & Reese, 1999; Reese & Fivush, 1993). Children of these parents not only organise their shared recall in a more elaborative manner than do children of low elaborative parents, but, by age 5-6 years, have internalized the high elaborative style and use it to organise their independent recall (e.g. Cleveland & Reese, 2005; Hudson, 1993; Leichtman et al., 2000;
Reese & Newcombe, 2007). Notwithstanding these important findings, however, less attention has been dedicated to the associations between reminiscing and other aspects of children’s socio-cognitive functioning, such as emotion knowledge (Denham et al., 2003;
Fivush et al., 2003; see Fivush et al., 2006; Wareham & Salmon, 2006, for reviews). The present study addresses this gap.
The suggestion that emotion knowledge might develop within reminiscing conversations integrates elements of both the sociocultural theory, which theorises that children come to internalise memory and other skills via social interaction with more knowledgeable others; and the emotion-coaching hypothesis, which suggests more 59 specifically that children learn about emotions through discussions with their parents and other ‘socialising agents’ (e.g. Denham, 1998; Denham & Auerbach, 1995; Nelson, 1996).
Furthermore, just as sociocultural research supports the role of shared reminiscing in the development of autobiographical memory, a large body of research based upon the emotion coaching hypothesis shows that mothers’ discussion of emotion in ongoing contexts, such as family interactions, shared book reading, and shared storytelling, plays an important role in children’s emotion knowledge development (e.g. Denham et al., 1994; Dunn, et al., 1991).
Children’s language ability and participation in these interactions may also be critical to development (de Rosnay, Pons, Harris, & Morrell, 2004; Fivush et al., 2006; Pons et al.,
2003; Symonds, 2004).
There are several reasons to suggest that children may internalize important information about emotions, and thus increase their development of emotion knowledge, as they engage in emotion-rich reminiscing with their parents (Denham & Burton, 2003; Welch-
Ross et al., 1999). First, autobiographical memory, by definition, is focused primarily on episodes that have personal meaning and relevance (Fivush et al., 2003). A number of researchers suggest that it is the emotions characterizing these memories that provide this meaning (Alea & Bluck, 2003; Fivush, 1993; Pasupathi, 2003; Fivush et al., 1993). Whilst other contexts can also be emotion-rich, reminiscing may allow parents and their children the opportunity to reflect upon emotions and their causes at a time when children are no longer in a state of emotional arousal, which might otherwise serve to distract both parent and child
(Bretherton, Fritz, Zahn-Waxler, & Ridgeway, 1986; Denham & Burton, 2003; Fivush, 1993,
Fivush et al., 2000; Thompson et al., 2003). Second, as the event has passed, reminiscing allows parents and their children to discuss the consequences of their emotions (Fivush et al.,
2000). Third, reminiscing might make emotion talk more memorable than might emotion talk at the time of an event. Experimental work has shown that, at least under some 60 circumstances, children remember event information better if they reminisced about the event after it had occurred than if they discussed the event as it was occurring (McGuigan &
Salmon, 2004; but see Ornstein & Haden, 2006).
If children come to internalize important information about emotions during reminiscing about significant personal experiences with their parents, then the style that a parent uses during reminiscing should impact the degree to which this occurs. It is possible that parents who are highly elaborative during reminiscing might also discuss these emotions in a more detailed and extended manner than might low elaborative parents, by explaining their causes and consequences in a rich and complete manner. Research shows that discussion of the causes of emotion in everyday contexts is a stronger predictor of emotion knowledge than is overall emotion talk or general talkativeness (e.g. Brown & Dunn, 1996; Garner et al.,
1997; Martin & Green, 2005); the same may be true in a reminiscing context. Although
Laible (2004a, Laible & Song, 2006) found no significant associations between reminiscing style and emotion references, Thompson et al. (2003) examined the type of emotion reference made by mothers in more depth and found that high elaborative parents discussed emotion causes, outcomes, definitions, and associated emotional events to a greater extent than did low elaborative parents. In conversations that actively engage their children in discussion of emotions, high elaborative parents might therefore facilitate children’s emotion knowledge to a greater extent than might low elaborative parents.
To date, one researcher has directly examined the associations of parents’ reminiscing style and emotion references with children’s emotion knowledge; yet findings are inconsistent (see Laible, 2004a, 2004b, Laible & Song, 2006). Laible (2004b) asked the mothers of 30-month-old children to discuss with their child one instance in which their child had behaved well, and one instance in which their child had misbehaved. Mothers’ elaborative style during these reminiscing conversations was measured on a 5-point scale 61
(where 1 = low in elaboration and 5 = high in elaboration), and emotion references were counted. Children’s emotion knowledge was measured on the Denham Emotion Knowledge
Task (Denham, 1986) six months later. Findings showed that mothers’ high elaborative reminiscing when children were 30-months-old predicted children’s emotion knowledge at 36 months (also see Laible & Song, 2006), yet mothers’ emotion references during reminiscing did not. In a similar study, Laible (2004a) further found that the frequency of mothers’ positive but not negative emotion references during reminiscing predicted 3- to 5-year-old children’s emotion knowledge.
There are several possible explanations for these discrepant results. First, it is possible that the young age of the children in Laible’s (2004b) study (toddlers rather than preschoolers) may have meant that their understanding of emotions was not sufficiently developed to be associated with maternal emotion references. This is unlikely, however; although findings are mixed (e.g. Smiley & Huttenlocher, 1989), research suggests that children as young as 2 years are able to refer to a range of feeling states in both themselves and others (Dunn et al., 1987). Second, it is possible that total emotion references did not consistently emerge as a significant predictor of emotion knowledge for methodological reasons. It may be that discussion of emotion causes is particularly critical to children’s understanding of emotion knowledge, over and above emotion references. Alternatively, it may be that differences in children’s emotion knowledge are more easily identified if total emotion references per discussion, rather than proportions per utterance, are the focus of analysis. A number of researchers have argued that it is the absolute amount of emotional information that children derive from conversations, and not the extent to which these references dominate the conversation, that is important for child outcomes (e.g. Dunn et al.,
1991; Fivush et al., 2000; Reese et al., in press). Finally, it may be that by merely using language in the context of social interaction, children are implicitly learning about the nature 62 of representation (Pons et al., 2003; Thompson et al., 2003). This may particularly be the case during reminiscing, when the object of discussion (the past experience) is not physically present. For example, it is via parent-child discussion of the past that children may come to realise that mental representations are subjective, and try to understand why (Nelson &
Fivush, 2004; Thompson et al., 2003). According to this view, rich, high elaborative discussions of the past should more greatly facilitate an understanding of representation - including “objects of representation” such as emotion (Pons et al., 2003, p.352) - than should discussion low in elaboration, as the language interaction itself is extended. Thus, specific emotion-oriented reminiscing may not be necessary.
Using a concurrent design, the present study aimed to examine the associations between high elaborative reminiscing, emotion-rich reminiscing, language, and children’s emotion knowledge. Reminiscing conversations about four events in which the child had experienced happiness, sadness, anger, and fear were elicited from mother-child dyads, and the child subsequently discussed four separate events independently with an experimenter.
The conversations were coded for the total number of emotion state, cause, and consequence utterances made, as well as the total number of high elaborative utterances and the proportion of high versus low elaborative utterances made, to enable us to more precisely identify associations with emotion knowledge and language given the inconsistent findings in the extant literature. It was hypothesized that high elaborative parents would make more references to emotion, and, in particular, emotion causes, than would low elaborative parents, and would have children who also elaborated more and made more references to emotion.
Second, it was hypothesized that children of high elaborative and emotion-rich parents would display better emotion knowledge than would children of low elaborative parents. An alternative hypothesis, that language and elaborative reminiscing are themselves associated 63 with emotion knowledge (such that emotion-rich reminiscing is not essential), was also considered.
Method
Participants
Twenty-five parent-child dyads were recruited via local Sydney preschools, and via
Canberra- and Sydney-based parenting groups. Participants were predominantly Caucasian, and resided in middle class areas of Sydney and Canberra. All children were aged between 37 and 63 months of age (M = 48.72, SD = 6.13). Eight mother-daughter dyads, fifteen mother- son dyads, and two father-son dyads participated in the study1. Dyads who did not speak
English at home were excluded. There was no significant difference in age between boys and girls, t(23) = 0.82, p > .05.
Measures and Procedure
Two sessions were conducted between one day and one week apart. One female experimenter conducted all sessions, either in families’ homes or at the university. During session one, children participated in parent-child memory conversations and completed an emotion knowledge task. During session two, children participated in experimenter-child memory conversations and completed a language task. In five cases, parents were unavailable for a second day and the second session was therefore conducted on the same day as the first, after a suitable break2. Each session was audio-taped.
1 Although most reminiscing research has focused on mother-child and not father-child dyads, all parents in our study were primary caregivers. Results did not differ when father-child dyads were excluded from the sample.
2 The location of each session and the timing of the second session relative to the first did not influence the pattern of results.
64
Memory Conversations
Parents were asked to nominate on paper eight past experiences that were one-time occurrences and had been experienced by parent and child in the past one to four weeks. They were asked to nominate two experiences each in which their child had been happy, sad, angry, and scared (see Fivush et al., 2003; Liwag & Stein, 1995 for similar protocol). During the first session, the experimenter randomly selected four of the eight events nominated; one of each emotion. Parents were asked to discuss the four selected events with their child, as they normally would, whilst the experimenter left the room. No mention of emotions was made.
During the second session, the experimenter discussed the remaining four events with the child. Parents were present but were asked not to contribute to these discussions. The experimenter commenced the discussion using the prompt: ‘Your Mum told me that you
……………. the other day. Tell me everything you can remember about that’. The experimenter continued to use prompts such as ‘uh huh?’ and ‘tell me more!’ but did not provide the child with any further information about the event.
Children’s Emotion Knowledge
Children’s emotion knowledge was measured using Denham’s Emotion Knowledge
Task (Denham, 1986). The task has good reliability (Cronbach’s alpha = .95; Denham, 1986) and is used frequently with children aged 2-5 years (e.g. Cutting & Dunn, 1999; Denham et al., 1994; Laible, 2004a, 2004b). Before beginning, parents were given a questionnaire asking how they predicted their child would feel in each of eight scenarios.
Emotion recognition (affective labelling). Children were shown four cloth faces, each with a different emotional expression (happy, sad, angry, scared). They were asked to expressively and then receptively identify the expression. Alternative labels, such as ‘mad’ for angry and ‘frightened’ for scared were accepted. Children received a score of 2 for each 65 correct answer, a score of 1 for incorrect answers of the right valence, and a score of 0 for incorrect answers of the wrong valence. Total possible scores ranged from 0 to 16.
Emotion perspective taking. Dolls were used to enact 16 vignettes (see Appendix A).
The protagonist doll had a blank face to which children could affix the four cloth faces used for emotion recognition. In eight vignettes, the doll felt the way that most people would feel in that scenario (e.g. happiness at being given ice cream). The remaining eight vignettes depicted scenarios in which different emotions could be felt (e.g. upon seeing a big dog a child may feel happy or scared). Using the parent’s questionnaire responses, the doll enacted an emotion opposite to how the child would feel. For each vignette the experimenter used a tone of voice and facial expression consistent with the emotion depicted. Children were asked to nominate how the doll felt by choosing a face for the doll. Responses were scored as for the emotion recognition task. Total possible perspective-taking scores ranged from 0 to 32, and total possible emotion knowledge scores (emotion recognition + emotion perspective-taking) ranged from 0 to 48.
Children’s Language
Children’s language ability was measured using the Peabody Picture Vocabulary Test-
III (PPVT: Dunn & Dunn, 1997). The PPVT assesses single-word listening comprehension ability, considered important for participation in reminiscing exchanges. It shows high alternate forms, test-retest and internal consistency reliability (range .86-.98), and content, construct, internal, and criterion validity (see Dunn & Dunn, 1997). Each item consists of four line drawings. The experimenter used the prompt ‘point to [stimulus]’ to ask children to select from four line drawings the picture that best represents the stimulus word. Children began on the set designated as a start point for their age group. After 8 or more in a set of 12 were answered incorrectly, the test stopped and children were awarded a raw score of their last item minus the number of errors made. Standardized scores were used (M =100, SD = 15). 66
Coding
All parent-child and experimenter-child discussions were transcribed verbatim. Each utterance (subject-verb proposition) was then coded for reminiscing style, and again for content. Parents’ and children’s utterances were considered separately, and children’s utterances during parent-child and experimenter-child conversations were also coded separately. In order to calculate inter-rater reliability, the coding scheme was explained to a second coder. The second coder coded a small number of practice transcripts with the primary coder before independently completing 25% of transcripts. Inter-rater reliability of coding, calculated using Cohen’s Kappa, was .87 (range .64 – 1.00). Data analysis was based on codes assigned by the primary coder.
With regards to style, each utterance was coded as being high elaborative, low elaborative, or a confirmation (off-topic utterances were disregarded). These categories were mutually exclusive (see Fivush et al., 2003; Harley & Reese, 1999; Hudson, 1993; Peterson,
Jesso, & McCabe, 1999).
1. High elaborative utterances: open-ended wh- questions; ‘do you remember … ’
questions; and information statements (either spontaneous or in response to a
question). Tag questions (e.g., ‘we went to the zoo, didn’t we!’) were coded as
information statements + confirmations.
2. Low elaborative utterances: repetitions of information or information requests
already provided; and close-ended yes / no questions and answers.
3. Confirmations: agreements or disagreements with information provided in the
previous conversation turn (or requests for agreement, e.g. ‘isn’t that right, bub?’).
The degree to which a parent or child used an elaborative reminiscing style was determined by calculating their high elaborative utterances / (high + low elaborative utterances) as a proportion. The total number of elaborations used was also counted. Note that close-ended 67 yes-no questions were coded as low-elaborative (see Laible, 2004a; Laible & Song, 2006).
Although evidence suggests that yes/no questions might be a developmentally appropriate means of engaging very young samples in reminiscing (Fivush et al., 2006; Hudson, 1990), for the current sample - that is, a community sample of preschoolers – wh-questions were considered to be more highly elaborative and more likely to elicit recall than were their equivalent yes-no questions. Importantly, supplementary analyses found no difference in the overall pattern of results when yes-no questions were instead coded as being high elaborative; thus, this coding did not influence any conclusions drawn.
With regard to content, high and low elaborative utterances were each coded as being factual or emotional in content. Following Fivush et al. (2003), emotional utterances were further categorized as being attributions, causes, or consequences. As it is the overall emotion information that is likely to be of importance for child outcomes, and not the extent to which this content dominates the conversation (which may vary in length), total emotion utterances rather than proportions (emotional utterances / [emotional + factual utterances]) were considered (see Dunn et al., 1991; Fivush et al., 2000; Reese et al., in press).
Results
Preliminary Analyses and Sample Characteristics
First, a repeated-measures ANOVA was run to determine whether reminiscing conversations differed according to the emotion discussed. Mothers made more total utterances in conversations focused on happy events (M = 13.75, SD = 6.69) than in conversations focused on sad (M = 8.83, SD = 5.77), angry (M = 6.96, SD = 5.30), or scary (M
2 = 8.08, SD = 7.34) events, F(1, 23) = 3.31, p < .05, ŋp = .19. No other differences were significant, however: there was no difference in the number of total elaborations, proportion of high elaborative relative to low elaborative utterances, or emotion references of any kind
2 made by parents or children, Fs(1, 23) < .18, ps > .05, ŋp s < .11. Conversations about happy, 68 sad, angry, and scary events also showed the same pattern of association with children’s independent recall and emotion knowledge; thus, all analyses reported are collapsed across emotion.
Second, children’s contributions were examined. Children asked no questions during parent-child or experimenter-child conversations. By inference, therefore, all high elaborative utterances they made were information statements. Children also made very few low elaborative utterances during experimenter-child conversations (see Table 1), and all variation was therefore captured by their frequency of total elaborations rather than their high compared to low elaborative utterances (which was constrained).
Table 1 shows parents and children’s average number of total high elaborative utterances, factual utterances, and emotional utterances per conversation. Parents spoke more overall than did children, and children spoke more with their parents than they did with the experimenter. Averaged across events, the proportion of utterances in which an elaborative style was used during parent-child memory conversations was .74 for parents (SD = .14) and
.76 for children (SD = .14).
Table 1: Style and Content Utterances during Shared Reminiscing and Independent Recall
Parent Child Child (with parent) (with experimenter) Utterance M (SD) M (SD) M (SD) High elaborative 8.57 (3.54) 4.52 (2.05) 2.22 (2.62)
Low elaborative 1.34 (0.72) 0.94 (0.36) 0.02 (0.00)
Factual 6.60 (3.53) 3.67 (1.87) 2.04 (2.34)
Emotional 1.66 (1.66) 0.91 (0.90) 0.16 (0.34)
-Attribution 1.04 (1.09) 0.58 (0.66) 0.13 (0.23)
-Cause 0.35 (0.43) 0.27 (0.43) 0.11 (0.43)
-Consequence 0.26 (0.41) 0.03 (0.08) 0.00 (0.00)
69
Children’s scores on Denham’s emotion knowledge task and the PPVT-III language task were each within the normal range. Emotion knowledge scores ranged from 27 to 46 (M
= 39.56, SD = 4.29). They were moderately correlated with age, and this correlation approached significance, r(23) = .35, p < .10. Standardized language scores ranged from 83 to
123 (M = 101.04, SD = 11.25). They were moderately, but not significantly, related to emotion knowledge, r(23) = .29, p > .05; the lack of significance may indicate insufficient power (power analyses revealed only a 48% chance of detecting a moderate effect).
A one-way ANOVA was used to explore possible child gender differences in variables measuring parents’ elaborative style and content, children’s elaborative utterances, children’s emotion knowledge, and children’s standardized language scores. There were no significant differences in child gender on any variable, Fs(1, 23) < 0.71, ps > .05.
Correlation Analyses of Reminiscing Style and Content, and Children’s Language and
Emotion Knowledge
A correlation analysis was conducted to determine the associations between parents’ elaboration and emotion talk variables, and children’s elaboration, emotion knowledge, and language variables. Given the relatively wide range of ages included in the study, children’s age was partialled out. Furthermore, given the significant associations of children’s language with their emotion knowledge and their elaborative utterances during shared recall, the analysis was re-run with children’s age and children’s language both partialled out. Partialling out language made no difference to the pattern of results; thus, only results from the first analysis are reported. Power analyses revealed sufficient (80%) chance of detecting a large correlation of 0.5, yet only a 44% chance of detecting a moderate correlation of 0.2 and a 15% chance of detecting a small correlation of 0.1.
70
Associations between Parents and Children’s Memory Conversation Variables
Table 2 shows that parents’ elaborative style during parent-child conversations was strongly and significantly associated with children’s elaborative utterances during parent-child conversations. Parents’ total elaborations were not significantly associated with children’s total elaborations during either parent-child or experimenter-child conversations. Additional analyses revealed strong and significant associations between the content of parents and children’s parent-child conversation utterances: facts, r(23) = .53, p < .01, emotion attributions, r(23) = .81, p < .01, emotion causes, r(23) = .70, p < .01, and emotion consequences, r(23) = .45, p < .05. The content of parents’ utterances made during parent- child conversations was not associated with the content of children’s utterances made during experimenter-child conversations (p > .05 in each case).
Associations between Parents’ Style and Content Memory Conversation Variables
Table 2 shows that parents’ elaborative style during parent-child conversations was strongly and significantly related to their emotion cause utterances, but not to any other factual or emotional content variable. Additional analyses revealed that neither parents’ nor children’s elaborative style was related to the content of children’s parent-child or experimenter-child utterances (p > .05 in each case).
Associations between Memory Conversation Variables and Children’s Emotion Knowledge
Table 2 shows that parents’ elaborative style and emotion cause utterances were each strongly associated with children’s emotion knowledge. No other parent content utterances were associated with children’s emotion knowledge. It is possible, from this result alone, that parent’s elaborative style was associated with children’s emotion knowledge only because high elaborative parents made more emotion cause utterances, also shown to be associated with emotion knowledge, than did low elaborative parents. To investigate this
71
Table 2: Associations between Parents’ and Children’s Reminiscing Variables, Children’s Emotion Knowledge, and Language 2 3 4 5 6 7 8
Parent Reminiscing Variables
1 Elaborative style (proportion) 0.17 -0.09 0.15 0.04 0.45* 0.05 0.65**
2 Total high elaborative utterances - 0.66** 0.25 0.19 0.27 0.19 -0.21
3 Factual utterances - -0.19 -0.19 -0.11 -0.04 -0.46*
4 Overall emotional utterances - 0.93** 0.76** 0.69** 0.14
5 Emotion attribution utterances - 0.54* 0.46* 0.15 -
6 Emotion cause utterances - 0.59** 0.23
7 Emotion consequence utterances - -0.09 -
Child Reminiscing Variables
8 Elaborative style with parent (proportion) -
9 High elaborative utterances with parent
10 Independent recall to experimenter
Other Child Variables
11 Emotion knowledge
12 Language
*p < .05, **p < .01 72 possibility, follow up analyses investigated the correlation between parents’ elaborative style and emotion knowledge whilst controlling for all emotion references. This correlation was strong, r(23) = .64, p < .01.
Associations between children’s memory conversation variables and their emotion knowledge were minimal. With regards to children’s memory conversation variables, Table 2 shows that children’s total elaborative utterances during parent-child conversations were related to their emotion knowledge, but their proportion of high compared to low elaborative utterances during parent-child conversations and their total elaborative utterances during experimenter- child conversations were not.
Discussion
The present study aimed to examine the associations between mothers’ high elaborative reminiscing and emotion content, children’s reminiscing and independent recall, and children’s emotion knowledge. As hypothesised, the degree to which parents engaged in high relative to low elaborative utterances was directly related to the degree to which they discussed and explained emotion causes. Consistent with these findings, Thompson et al
(2003) also reported that high elaborative parents discussed emotion causes to a greater extent than did low elaborative parents (together with outcomes, definitions, and associated emotional events). Researchers have noted that high relative to low elaborative parents provide a fuller, more embellished account of the child’s experience (e.g. Fivush & Fromhoff,
1988). Given that the experiences discussed during reminiscing are typically personally significant and emotionally meaningful, causally relating the emotion states that characterise an experience to other information about the experience is one way in which this full and embellished account can be provided (Cervantes & Callanan, 1998; Fivush, 1993; Fivush et al., 2003). 73
Interestingly, there was no significant associations between parents’ elaboration and their emotion attribution or consequence utterances; consistent with the findings of Laible
(2004a, 2006) there were also no significant associations between elaboration and overall emotion references (note that emotion consequence utterances were very rare for all parents).
It is possible that this was due in part to the manner in which parents were asked to nominate events for discussion. Various researchers have speculated that low elaborative parents may see reminiscing tasks as a kind of memory test, and therefore focus strongly on aspects of an event they would like their child to remember rather than aspects of the event the child finds interesting (see Cleveland & Reese, 2005; Fivush & Fromhoff, 1988). Although parents in our study were not explicitly prompted to discuss emotions, they were asked to nominate events with an emotional theme. Given this, it is possible that parents lower on the elaboration continuum were more likely than parents higher on the elaboration continuum to use the emotional focus of the discussed events as a cue that memory for emotions was being tested in some way, and this may therefore have elevated the number of emotion references they made. This explanation cannot account for more elaborative parents’ greater use of emotion cause utterances, however. The findings suggest that the degree to which parents’ elaborate while reminiscing (that is, their reminiscing style) is associated with the way in which they discuss these emotions – that is, with a focus on why emotions occurred - over and above any potential associations between reminiscing style and raw frequency of overall emotion references.
The associations between parental reminiscing style and children’s contributions to the discussion were evident in several ways. The more a parent elaborated, the more likely it was that their child would both use more high compared to low elaborative utterances, and provide more event information overall. Similarly, parents who discussed the emotional aspects of the child’s experience more had children who also discussed the emotional aspects of the child’s 74 experience more. These findings are consistent with much other research suggesting that, by age 4 years, young children have begun to internalize both the style and content of their shared memory conversations (e.g. Kuebli et al., 1995; Nelson & Fivush, 2004; Reese, 2002a,
2002b; Welch-Ross et al., 1999). Associations between parents’ reminiscing style and content and children’s independent recall to an experimenter were not yet evident; however, past research suggests that children may not generalise shared recall skills to their independent recall until 5 or 6 years of age (Cleveland & Reese, 2005; Reese et al., 1993).
The degree to which parents used a high elaborative reminiscing style was directly associated with children’s emotion knowledge. The notion that emotion references might be a mediating variable, however, was only partially supported (note that because of the small sample size, formal tests for mediation could not be conducted). As discussed earlier a positive relationship was found between the degree to which parents elaborate during reminiscing, and the degree to which they explained emotion causes to their children during reminiscing. In turn, it was found that parents’ explanations of emotion causes, but not their discussion of emotion attributions or consequences, were strongly related to children’s emotion knowledge. Interestingly, findings from research investigating other contexts of parent-child interactions, such as shared book reading and everyday family interactions, similarly show that discussing the causes of emotions predicts children’s emotion knowledge more strongly than does other emotion talk (e.g. Brown & Dunn, 1996; Garner et al., 2005).
Why might emotion-cause reminiscing be more strongly associated with emotion knowledge than other emotion-oriented reminiscing? Psychological terms, such as emotion labels, typically refer to private and unobservable experiences (de Rosnay & Hughes, 2006).
Emotion attributions tell children only that an emotion state has occurred. Explanation of emotion causes, however, provides children with the opportunity to internalize socially 75 relevant information about why these various emotion states occur, and what situational factors might predict them (Cervantes & Callanan, 1998).
Importantly, emotion causes were not the only discussion variable directly associated with children’s emotion knowledge. Firstly, children’s language and emotion knowledge were moderately - yet not significantly - related. Associations between various aspects of language and emotion knowledge have also been found previously, suggesting that the lack of significance in the current study may reflect the small sample size and low power available to detect a moderate effect size (de Rosnay et al., 2004; Pons et al., 2003). Secondly, and consistent with Laible’s findings (Laible, 2004b; Laible & Song, 2006), the relationship between parents’ high elaborative utterances and children’s emotion knowledge remained strong, even once parents’ emotion references and children’s language were each controlled.
By participating in reminiscing exchanges with their parents, children may implicitly be learning about mental representations (Pons et al., 2003; Thompson et al., 2003). High elaborative reminiscing extends the language interaction more greatly than does low elaborative reminiscing, such that children are given greater opportunity to internalize representational information. By generalising their understanding of memory to representational skills other than memory, therefore, children may also come to develop a better understanding of emotions.
Given the correlational nature of the present study, caution must be used when attempting to determine causality (Dunn, 2006; Fivush, 1991). Although parents’ high elaborative style and emotion cause references during reminiscing might facilitate children’s emotion knowledge, it might also be the case that children with superior pre-existing emotion knowledge elicit more elaborative and emotion-cause-oriented reminiscing discussion from their parents. In a longitudinal study focused on the bi-directionality of mother-child reminiscing style, Reese et al. (1993) found that although children’s early contributions to 76 reminiscing conversations did not influence maternal reminiscing style, their contributions at
58 months predicted maternal reminiscing style at 70 months. Farrant and Reese (2000) found that children’s interest in participating in reminiscing conversations predicts mothers’ reminiscing style at as young as 19 months of age. Taken together, these findings show that child factors can influence the way in which parents reminisce with their children. It is unclear how children’s emotion knowledge would manifest itself in reminiscing conversations; however, Wang, Hutt, Kulkofsky, McDermott and Wei (2006) present evidence to suggest that children’s emotion knowledge may contribute to their independent autobiographical recall. Perhaps it is the case that children with superior emotion knowledge contribute more to reminiscing conversations with their parents, who respond by elaborating all elements of the experience being discussed (emotional and non-emotional) to a greater degree. The possibility and nature of a bi-directional relationship between parents’ reminiscing style and content on one hand, and children’s emotion knowledge on the other, should be elucidated in future research.
Over and above the associations between parent reminiscing and child emotion knowledge that emerged, the small sample size of the present study represented a significant limitation. Power was only sufficient to detect medium to large effects, meaning small and medium effects that may have emerged with a larger group of participants are necessarily overlooked. Replication with a larger sample is therefore important. Non-significant findings cannot be interpreted as a lack of effect until this replication has occurred.
In summary, the association between parents’ elaboration and emotion talk during reminiscing, and preschool-aged children’s elaboration, emotion references, and emotion knowledge was examined. Events were elicited in which happiness, sadness, anger, or fear featured. Despite the limitation of a small sample size, several significant findings emerged.
The more parents used a high elaborative reminiscing style when discussing emotional past 77 events with their child, the more likely they were to discuss emotions in a causal manner. In turn, parents’ causal discussion of emotions, parents’ high elaborative reminiscing style, and children’s language were all uniquely associated with children’s emotion knowledge. These findings suggest that there are multiple ways in which high elaborative and emotion-rich reminiscing benefits emotional development. 78
CHAPTER 3: EMOTION ORIENTED REMINISCING AND CHILDREN’S RECALL FOR
A NOVEL EVENT
Study 2 aimed to examine children’s recall for a novel, staged event following high elaborative, emotion-oriented reminiscing with an experimenter. As discussed in Chapter 1, research shows that although the emotional aspects of past events are included in reminiscing from about age 30 months, some parents refer infrequently to these aspects of the child’s experiences, whereas others - particularly securely-attached dyads - discuss the emotional features to a greater extent (e.g. Burch et al., 2004; Fivush & Haden, 1997; Newcombe &
Reese, 2004; Sales et al., 2003). As it is the emotions of an event that are thought to provide the event with meaning, and therefore make it worthy of recall (Alea & Bluck, 2003; Fivush,
1993; Pasupathi, 2003; Fivush et al., 1993), discussing these emotional features may arguably enhance children’s recall of the entire event.
In Study 1 (Chapter 2) it was shown that high elaborative parents discussed emotion causes during reminiscing with greater frequency than did low elaborative parents. Although replication is needed, given the small sample size, this research may help to clarify the mixed findings of past research; some of which shows that high elaborative mothers engage in more overall discussion of emotions (Laible, 2004b; Thompson et al., 2003), and some of which shows no association (Laible, 2004a; Laible & Song, 2006). Whatever the relationship between style and content, however, findings of naturalistic longitudinal research show that children of high-elaborative mothers provide more elaborative accounts of their experiences; similarly, mothers who engage in reminiscing about emotions have children who, over time, come include this same emotional information in their own reminiscing (Hudson, 1993; 79
Kuebli et al., 1995; Leichtman et al., 2000; Sales & Fivush, 2005; see Study 1 for similar findings using a concurrent correlational design).
Recent research has extended this strong body of naturalistic research by adopting an experimental paradigm in which the nature of the adult-child discussion has been manipulated and the event has been staged. Findings establish a causal relation between reminiscing style and children’s recall: children engaged in high elaborative reminiscing report more information about the experience than do children who participate in reminiscing that is low in elaboration or comprised of ‘empty’ uninformative talk (Conroy & Salmon, 2006;
McGuigan & Salmon, 2004). To date, however, no research has adopted an experimental paradigm to investigate the relationship between adult-child reminiscing about emotion and children’s recall of an experience – its emotional and non-emotional aspects. Moreover, as, in much extant research, elaboration and emotional content have been confounded.
It is unclear whether discussion of emotions per se, over and above the influence of reminiscing style, might enhance children’s recall. Investigation of this issue not only has theoretical implications, potentially expanding understanding of the pathways to children’s recall of their personal experiences, but may also inform clinical interventions aiming to optimize socio-emotional functioning (including the ability to discuss emotions and emotion knowledge) across the early years (see Wareham & Salmon, 2006, for review). It is increasingly recognized that training parents in high elaborative reminiscing can significantly benefit children’s autobiographical recall (Peterson et al., 1999), and that children’s emotion knowledge can be advanced by parental discussion of emotion (e.g., Havighurst, Harley,
Littlefield, Prior, & Gavidia-Payne, 2002, cited in Denham & Burton, 2003), but clarification of the specific effective elements of reminiscing is necessary.
There are several possible associations between emotional content in reminiscing – independent of elaboration - and children’s event recall. For example, just as talk about any 80 aspects of an experience can render those specific details more memorable, reminiscing about emotion may facilitate children’s recall just of those emotional aspects but not other aspects
(Bloise & Johnson, 2007). Indeed, reminiscing about the emotional aspects of the experience, at the expense of non-emotional aspects, may even impair recall of the latter aspects. This possibility is consistent with findings of research investigating the influence of selective post- event review on information not reviewed (Barnier, Hung, & Conway, 2004; Conroy &
Salmon, 2006), and suggests the possibility that only the specific type of emotional information discussed will later be recalled. Alternatively, to the extent that the emotional elements provide a meaningful interpretational framework that can ensure active and integrated processing of the emotional and non-emotional event components, emotional reminiscing may enhance recall of the entire event (Craik & Tulving, 1975; Laney, Campbell,
Heuer, & Reisberg, 2004; Liwag & Stein, 1995; Wang et al., 2006). Finally, it is possible that the association between emotional reminiscing and recall depends on the nature of the emotional content.
Much research shows that a strong degree of connectedness between the components of an event, for example, by virtue of logical rather than arbitrary connections, produces better recall (Bauer & Mandler, 1989; Conroy & Salmon, 2006). No research has yet examined the influence on recall of the type of emotion references made during reminiscing, yet, presumably, emotion cause-oriented reminiscing would have the same advantages for recall as would other causal discussion. Emotion-cause oriented reminiscing may therefore benefit recall by providing a potential avenue via which a causal framework can be established, irrespective of the specific emotion content being discussed. Alternatively, if emotional reminiscing of any kind is shown to benefit recall as is suggested above, a causal framework may enhance the process by which this occurs. In any case, discussion focusing on the causes of a child’s emotional experiences should serve to explicitly link the event’s components 81 more effectively than should non-causal discussion (Bird & Reese, 2006; Cervantes &
Callanan, 1998).
The overarching aim of the current study, therefore, was to investigate the influence of emotion-oriented reminiscing, over and above a high elaborative style, on 3- to 4- and 5- to 6- year-old children’s recall. This age range was of particular interest in that it spans significant advances in autobiographical remembering, use of emotion language, and emotion knowledge
(e.g. Denham, 1998; Fivush & Haden, 1997; Kuebli et al., 1995; Pons et al., 2004; Reese et al., 1993; Stein & Liwag, 1997). Although each of these abilities develops continuously; that is, there is no cut off point during the preschool years where a child is said to ‘gain’ autobiographical memory, emotion language, or emotion knowledge (see Wareham &
Salmon, 2006, for review), there was a 6 month gap between the oldest child in the ‘younger’ group (52 months) and the youngest child in the ‘older’ group (58 months). Thus, even within the preschool years, significant differences in the memory and emotion representation abilities of our younger and our older group might be expected to contribute to a difference in the impact of emotion-oriented reminiscing on recall.
An experimental design was used, in which a staged event was accompanied by an emotion-rich narrative (adapted from the “Visit to the Pretend Zoo” event used by McGuigan
& Salmon, 2004), and the style and content of subsequent adult-child reminiscing was manipulated. Specifically, children in three experimental conditions were engaged in high elaborative reminiscing following participation in the zoo event, but the emotional content differed. Thus, in one condition (emotion cause), the causes of the zoo animal’s emotions were discussed; in a second condition (emotion expression), the animals’ emotion expression was labelled and discussed; and in a third condition (no emotion), physical characteristics of the animal were discussed, with no emotional aspects. In a fourth control condition (minimal), rather than engagement in high elaborative reminiscing, the children were asked only closed 82 questions with scant provision of detail and no emotional content. Consistent with previous research by McGuigan and Salmon (2004), children’s recall of the experience was assessed after a two-week delay.
The second aim of the study was to investigate whether emotion knowledge and language would contribute to children’s recall of the event, as recently found by Wang et al.
(2006). Using a concurrent design, Wang et al. (2006) assessed 2.5- to 3.5-year-old children’s emotion knowledge together with their independent recall of autobiographical events. Over and above the influences of age, language skill, gender, and culture (Chinese or American); children’s emotion knowledge, as assessed by their judgements of the emotions of story characters and by their ability to generate appropriate causes of emotions, was found to predicted their autobiographical recall. Wang et al. (2006) suggested that, in helping children to understand and interpret the emotional meaning of an event, more advanced emotion knowledge might provide an organizational structure that facilitates all aspects of information processing: encoding, storage, and retrieval. Notwithstanding these important findings, this study did not focus on children’s recall of a single, known event, and replication is needed. In the current experiment, we investigated the unique contribution of emotion knowledge to children’s recall of the “Visit to the Pretend Zoo” event, with age and language ability also included as predictors.
In summary, the major aim of the present study was to examine experimentally the relative influences of high elaborative reminiscing, both with and without an emotion focus, on 3- and 5-year-old children’s independent recall of a novel, emotion-rich event (“Visit to the Pretend Zoo”) experienced two weeks prior to recall. It was anticipated that, over and above typical age differences in recall, children who reminisced about the emotional aspects of the experience (those in the emotion-cause and -expression conditions) would recall more emotional and non-emotional event information than would those in the no-emotion and 83 minimal conditions, but, given the potential importance of causal talk in highlighting the links and meaning of the event, children in the emotion-cause condition would have the greatest advantage. An alternative hypothesis was also considered, whereby reminiscing about the emotional aspects of the experience may enhance children’s recall of emotional event information only. With regards to age, a significant age x condition interaction was expected.
More specifically, differences between conditions were expected to be strongest in younger children: younger children possess less ability to discuss and represent emotion and other event information than do older children, meaning that they may rely more heavily on the emotion content provided by the experimenter during reminiscing (in the emotion-cause and emotion-expression conditions). Finally, it was expected that emotion knowledge would be positively predict children’s recall of emotional and non-emotional event information, over and above the impact of language skill and age. Language and age were also expected to contribute.
Method
Participants
One hundred and three children were recruited from local preschools and primary schools in Sydney. Written parental consent was obtained. The majority of the children were
Caucasian, and schools were located in middle-class areas. All children spoke English at home. Nineteen children were absent from school for one or more experimental sessions, and were excluded from analysis. The final sample consisted of 83 children, aged between 41 months and 77 months. The sample was divided into a younger (n = 37) and an older (n = 46) group. Children in the younger group ranged in age from 41 to 52 months (M = 46.24, SD =
3.02), and attended preschool. There were 13 boys and 24 girls. Children in the older group ranged in age from 58 to 77 months (M = 67.86, SD = 5.84), and attended primary school.
There were 30 boys and 16 girls. As there were more girls than boys in the younger group, it 84 was thought likely that this difference would reduce the likelihood of detecting differences between age groups, given findings that gender differences in autobiographical recall and socio-cognitive understanding tend to favour girls (e.g. Bloise & Johnson, 2007; Hughes &
Dunn, 2002; Kuebli et al., 1995).
Materials
Zoo Event Props
The zoo event consisted of five animal ‘stations’, with one animal at each. The giraffe, koala, lion, and monkey were each made from furry material and stuffing, with large 1.5m x
1.0m MDF wood backing painted green. A grey, 30cm tall soft toy ‘baby elephant’ was also used. Each animal had a different emotional expression, made from coloured felt. These expressions are described in Table 3. The zookeeper wore a khaki shirt and wide-brimmed hat, and children were also given a khaki wide-brimmed hat to wear. Nine smaller props were used, including a blanket, Duplo™ building blocks, a box of tissues, a pet food bowl, a small
Tupperware™ container full of cat biscuits, a bottle of water, a banana, a small desk chair, and a cushion.
Table 3: Emotional Expressions of Each Animal in the Zoo Event.
Animal Emotion Expression
Giraffe Worry Eyes open wide and forehead creased, mouth pursed
Koala Sadness Eyes and eyebrows drooped, mouth closed with ends turned down
Lion Anger Eyebrows furrowed in the middle, mouth pursed
Monkey Excitement Eyes open wide with eyebrows lifted, mouth open with ends turned up
Elephant Fear Eyes open wide with eyebrows lifted, mouth open and ends turned down
85
Procedure
Balanced across gender and school, children within each age group were randomly allocated to one of the four reminiscing conditions: emotion cause, emotion expression, no- emotion, or minimal.
Zoo Event
A female research assistant escorted each child to a classroom and explained that the child was going to visit the pretend zoo. The child was then introduced to ‘the zookeeper’, a second research assistant, who guided him or her through the event. At each animal station the zookeeper labelled the animal’s emotion, and stated why the animal felt that way (‘emotional information’). The child then participated in two activities, both of which were narrated, as listed in Table 4. For example, at the koala station the zookeeper said: “Koala is sad. She has no one to play with. We can’t have that. Let’s wipe away her tear with this [tissue]. Now we’ll play a quick game with her. Together we’ll build the highest tower we can, using these blocks.” Note that children were told the animals’ emotions (“Koala is sad”) and the causes of those emotions (“She has no one to play with”), but were required to make the causal inference themselves (Koala is sad because she has no one to play with). Table 5 shows the standardized scripts for the event and for reminiscing by condition, for one of the five animal stations making up the “zoo” event (the monkey). Children were not permitted to play with props or perform any actions other than those in the script. The event lasted approximately seven minutes.
Reminiscing Conversation
Two days after the event, the zookeeper reminisced about the zoo event with each child (see Table 5 for reminiscing about the monkey). This timing was consistent with the experimental research of McGuigan and Salmon (2004), who also examined 3- and 5-year- olds’ recall of a staged event following reminiscing. For children in the emotion-cause, 86
Table 4: Event Emotions, Causes, and Activities.
Emotion Cause Activity
Worry Searching for baby elephant Put blanket over her to aid sleep
Blow sleeping dust
Sadness No one to play with Build block tower
Wipe tear with tissue
Anger No one has fed him Put food in bowl
Tiptoe
Excitement It's his birthday Give him present (banana)
Do a dance
Fear He is lost Find him behind cushion
Hug him
emotion-expression, and no-emotion conditions, the zookeeper engaged the child in elaborative reminiscing by asking open-ended questions and providing detailed information.
Importantly, past research has shown that information provided to a child may differ in memorability from information they generate themselves (McGuigan & Salmon, 2004;
McNamara & Healy, 1995; Slamecka & Graf, 1978); thus, in all three high-elaborative conditions, information that the child was told and information that the child was required to produce themselves varied systematically between animals discussed (such that children were both told and generated an equal amount of information discussed across the event). In all three conditions, furthermore, non-emotional content was discussed in the same way (e.g.
“When we visited koala we found some blocks. We got all of the blocks, and then we played.
What did we build?”). Over and above the standardized non-emotional content, the type of emotional information discussed differed between conditions.
87
Table 5: Event and Reminiscing Scripts for ‘Visiting the Monkey’, by Condition.
During-Event Script Reminiscing Script
Emotion-cause Emotion-expression No-emotion
The fourth animal is the After the lion we visited After the lion we visited After the lion we visited monkey. Monkey is another animal. Which another animal. Which animal another animal. Which excited. Monkey’s birthday animal did we visit? did we visit? animal did we visit? is coming up!
Lets do the birthday After the lion we visited the After the lion we visited the After the lion we visited th dance. Put your hands on monkey! And the monkey monkey! And the monkey monkey! And the monkey your hips like this, and sing was very excited about was very excited; do you had very big arms; do you ‘Happy birthday to you, something; do you remember how his face remember where monkey you live in a zoo!’ Now remember what it was? looked? had his arms? your turn. Great, monkey was excited Great, we knew monkey was Great, monkey had his Great work! Now we need because it was his birthday! excited because he had a big arms up in the air! We did to give him this present we We did the happy birthday smile, and his eyes opened dance for the monkey, have for him. Put the dance for the monkey, and wide! We did a dance for the where we kicked our legs banana at his feet. kicked our legs out, and monkey, where we kicked our out, and then we gave him then we gave him a special legs out, and then we gave a special present. What Good, he’ll like that. present. What was our him a special present. What was our present? present? was our present? Right, it was a banana! Right, it was a banana! Right, it was a banana! Bananas are a great Bananas are a great Bananas are a great present present for monkey. It’s fu present for monkey. for monkey. It’s fun to get a to get a surprise, isn’t it? Birthdays are exciting, surprise, isn’t it? aren’t they? 88
In the emotion-cause condition, the animal’s emotion was labelled and the cause of the emotion was stated, for example, “Koala didn’t have anyone to play with, did she? How did that make koala feel?” (child told cause, generated emotion) and “Monkey was very excited about something; do you remember what it was?” (child generated cause, told emotion). In the emotion-expression condition, the animal’s emotion was labelled and the emotional expression displayed by the animal was described, for example, “Koala had a tear in her eye, and her mouth was turned down, wasn’t it? How did Koala feel?” (child told expression, generated emotion) and “Monkey was very excited; do you remember how his face looked?”
(child generated expression, told emotion). In the no-emotion condition, physical characteristics of the animals’ bodies, such as fur colour and texture, were described, for example, “Koala had grey fur all over her body, didn’t she? And how did her fur feel on your hand?” and “Monkey had very big arms; do you remember where monkey had his arms?”
(child told and generated additional non-emotional information). Note that although questions in all three high-elaboration conditions were open-ended, the specificity of the questions was considered necessary to maintain the fidelity of the manipulation (so that, for example, children in the emotion-expression or no emotion condition would not spontaneously produce information about emotion causes). Finally, in the minimal condition, children were asked only close-ended questions with little opportunity to contribute, and were provided only with non-specific event information low in elaboration, for example, rather than saying “After the sad koala, we visited the angry lion”, the zookeeper said, “Over there we saw the next animal”.
The emotion-cause, emotion-expression, and no-emotion scripts were 534, 530, and
532 words long respectively, and reminiscing lasted approximately seven minutes. The minimal script was 260 words long, and reminiscing lasted approximately five minutes. 89
Memory Interview
Two weeks after the event, children were each escorted to an empty classroom by a third researcher, and were interviewed about the event following a standardized memory interview script. Again, this timing was consistent with that of McGuigan and Salmon (2004).
In free recall, the experimenter asked the child to provide information about the event via an open-ended prompt: ‘I heard that you got to visit the pretend zoo and be the zoo helper the other day. That sounds like a lot of fun! Tell me everything you can remember about that’.
The child was encouraged to contribute any additional information via non-directive prompts
(‘Yes?’, ‘Uh-huh?’, and ‘Tell me some more about that!’). When it was clear that the child would provide no further information, the interviewer moved to prompted recall. The child was cued with each animal label (e.g. ‘Tell me about your visit to the giraffe’; or, if the child had already mentioned some detail about the giraffe during free recall, ‘Tell me some more about your visit to the giraffe’). Non-directive prompts were again provided to elicit additional information. In direct questioning, the child was asked three open-ended questions about each animal (15 questions in total). Note that questions that the child had already
‘answered’ during free or prompted recall were skipped, as only the additional recall of information that had not been mentioned during earlier stages of the interview was of interest.
The first question asked the child to name how the animal had felt, the second asked the child why the animal felt that way, and the third was a question about the non-emotional core activity that they participated in whilst visiting that animal (e.g. ‘What present did you give to monkey?’). Correct answers were confirmed and incorrect answers corrected at the end of each question. The duration of the memory interview was approximately 10-15 minutes.
Children’s Emotion Knowledge
Children of both age groups were administered an Emotion Cause Production Task as a measure of emotion knowledge. This task was considered more appropriate than Denham’s 90
Emotion Knowledge Task (as used in Study 1), as many children were older than the children in Study 1 and it was therefore expected that some might reach ceiling on the Denham Task
(see Hughes & Dunn, 1998). In contrast, the Emotion Cause Production Task has been used successfully with children aged 2.5- to 6-years: additionally, the task shows good inter-rater reliability (Cohen’s Kappa range .66 – 1.00) (see Denham & Zoller, 1991; Hughes & Dunn,
1998; Wang et al., 2006). Whilst it is limited by the unavailability of any formal measure of validity, researchers who have used the task in conjunction with other measures of emotion knowledge such as Denham’s Emotion Knowledge Task (1986) and an Emotion Judgement
Task (Wang, 2003) show that individual differences in children’s emotion knowledge are consistent across measures (See Hughes & Dunn, 1998; Wang et al., 2006).
In the task, each child was asked ‘Tell me some things that make you feel happy’; findings of Wang et al. (2006), indicate that asking children to respond to a prompt focused on themselves yields identical results to asking them to respond to prompts focused on others.
The child was encouraged to continue providing causes until it became clear that he or she could provide no further information. The child was then asked ‘Tell me some things that make you feel sad / angry / scared’ in turn. After each emotion prompt, the child was encouraged to continue providing causes until he or she could provide no further information.
Any unique, plausible emotion cause provided by the child was given a score of 1, and total scores were tallied. Two researchers independently scored 40% of responses. Inter-rater reliability, calculated using Cohen’s Kappa, was .84. Data analysis was based on scoring by the primary coder.
Children’s Language
As in Study 1, the PPVT was used to measure children’s receptive language ability
(Dunn & Dunn, 1997). See Study 1 (p. 65) for a description of this measure. 91
Coding
All memory interviews were transcribed verbatim. Information was coded according to whether it had been reported during free recall, prompted recall, or direct questioning. In prompted recall and direct questioning, only information that had not been mentioned during earlier stages of the interview was coded.
Emotional information consisted of emotions, emotion expressions, and emotion causes. An emotion was credited when a child correctly recalled the emotion that one of the five animals had felt during the event (e.g. “Koala was sad”). An emotion expression was credited when a child correctly recalled the facial expression of one of the five animals (e.g.
“He was smiling”), and an emotion cause was credited when a child correctly recalled the cause behind one of the five animals’ emotions. Note that although emotion causes were by definition causally related to emotion, and therefore constituted emotional event information, the provision of an emotion cause did not in itself necessarily imply that children recalled the causal link: a cause was credited whether or not the accompanying emotion had been stated.
For example, a child who mentioned both an emotion and the cause of the emotion (e.g.,
“Elephant was scared because he was lost”) was credited with the recall of both the emotion and the cause, whereas a child who recalled that “It was monkey’s birthday!”, but did not explicitly acknowledge that monkey felt excited for this reason, was credited with the recall of the emotion cause but not the emotion. Thus, emotions, emotion expressions, and emotion causes were mutually exclusive.
Non-emotional information consisted of animals, core activities, and additional details.
An animal was credited when a child freely recalled one of the five animals. Animals were not credited during prompted recall and direct questioning, as animal labels were used as prompts. A core activity was credited when a child recalled one (of the five) major activities of the event. For example, “Found the baby elephant!” received one animal point and one 92 core activity point. An additional detail was credited when a child correctly recalled non- emotional event information other than the animal or core activity categories, and included objects, non-core activities, and descriptors. For example, “I put food in the yellow bowl” received two object points (food, bowl), one non-core activity point (put), and one descriptor point (yellow). Animals, core activities, and additional details were also mutually exclusive.
Each item of incorrect emotional or non-emotional information was credited with one error point. Errors included confusions, in which a child recalled but incorrectly described actions and objects that had been part of the event (e.g. incorrectly attributing anger to the giraffe rather than the lion), and intrusions, in which actions or objects that had not been part of the event were introduced (see Fivush & Hamond, 1989).
A second researcher uninformed of the child’s reminiscing condition coded a small number of practice transcripts with the primary coder, and then independently coded 25% of transcripts from each condition. Inter-rater reliability of coding, calculated using Cohen’s
Kappa, averaged .86 for free recall (animals = 1.00, emotions = .91, causes = .83, core activities = .88, additional details = .70), and .85 for prompted recall (emotions = .88, causes
= .78, core activities = .95, additional details = .80). Reliability was 1.00 for each aspect of direct questioning. Data analysis was based on codes assigned by the primary coder.
Results
Regression and Preliminary ANOVA Analyses
Children’s performance on the Emotion Cause Production Task was significantly correlated with both their age, r = .47, p < .01, and standardized receptive language ability, r
= .41, p < .01. Simultaneous regression analyses were therefore used to examine the contributions of unique contribution of emotion knowledge, age, and language to recall. As can be seen in Table 6, age and emotion knowledge each predicted children’s free recall of emotional information; and age, emotion knowledge, and standardized language each 93 predicted their free recall of non-emotional information. Additionally, standardized language predicted the recall of emotional information during direct questioning, β = .26, t(78) = 2.37, p < .05. There were no further contributors to the recall of emotional or non-emotional information at any stage of the memory interview, βs < .13, ts(78) < 1.07, ps > .05. Note that one child participated in free recall but not prompted recall or direct questioning, as the school day was about to end. Therefore, N = 83 for free recall and N = 82 for all other recall.
Table 6: Summary of Simultaneous Regression Analyses for Variables Predicting the Recall of Emotional and Non-emotional Information Variable B SE B β
Emotional Information
Age 0.07 0.03 .30**
Emotion Knowledge 0.03 0.01 .25*
Language 0.01 0.01 0.11
Non-emotional Information
Age 0.14 0.07 0.22*
Emotion Knowledge 0.09 0.03 0.33**
Language 0.05 0.02 0.21*
Note. R2 = .26 for the recall of emotional information and R2 = .24 for the recall of non-emotional information. *p < .05. **p < .01.
Given the significant contributions of standardized receptive language and emotion cause knowledge to children’s recall of both emotional and non-emotional information, it was important to ensure no language or emotion knowledge differences between conditions.
Preliminary analyses confirmed that this was the case. Conditions did not differ by language,
2 F(3, 80) = 1.07, p > .05, ŋp = .04, with an average standardised language score of 95.74 (SD
= 13.88) across conditions. Furthermore, conditions did not differ by emotion knowledge,
2 F(3, 80) = 1.08, p > .05, ŋp = .04, with an average of 6.56 emotion causes produced (SD = 94
5.30) across conditions. Finally, conditions did not differ by gender, χ2(3, N = 84) = 0.24, p >
.05 (as results within age groups mirrored results across age groups, only the results across age groups are reported here). Thus, controlling for language and age in analyses examining the effects of emotional reminiscing on recall was not necessary.
Intrusion errors during recall were extremely rare and, on average, children made less than one confusion error overall (M = 0.82, SD = 0.19). Thus, no further analyses based on errors were conducted. All subsequent analyses were based on correct recall. Preliminary analyses revealed that when analyzed separately, the correct recall of emotions and emotion causes each showed an identical pattern of results to each other and to emotional recall overall
(recall of emotion expressions was extremely rare, M < 1.00, even in the emotion-expression condition). This was the case despite children not being required to explicitly state the emotional nature of an emotion cause for recall of the cause itself to be credited; the implicit association of this information to an emotion state was considered sufficient for the recall to be considered emotional (see coding). Likewise, when analysed separately, the correct recall of all components of non-emotional recall - animals, core activities, and additional details - showed an identical pattern of results to each other and to non-emotional recall overall. These results suggest that the categories of emotional recall (overall) and non-emotional recall
(overall) represent their subcategories well. Thus, only the results for correct emotional and non-emotional recall, and not the results for each subcategory, are reported here.
Effects of Emotional Reminiscing on Recall
Statistical Strategy
In light of significant associations between emotional and non-emotional information recalled in total recall, r(82) = .58, p < .01, and free recall, r(82) = .67, p < .01, a series of (2) 95 x 2 x 4 repeated measures ANOVAs were conducted3. The within subjects factor was type of information recalled (emotional or non-emotional), and the between subjects factors were age group (younger or older) and condition (emotion cause, emotion expression, no emotion, or minimal). Note that main effects for the type of information recalled were not of interest; a greater amount of non-emotional than emotional information was available for recall and it was expected, therefore, that children would recall more non-emotional than emotional information.
The influence of the experimental manipulations on children’s total correct recall (that is, free recall, prompted recall, and direct questioning, summed) was considered first, and thereafter, to identify whether reminiscing influenced all phases of the interview, each was analysed separately. Effect sizes were evaluated according to Cohen’s (1992) standard: .2 = small, .5 = moderate, .8 = large. Power analyses revealed sufficient power to detect large main effects and interactions (98% and 83% average chance of detection, respectively), together with moderate main effects (84%). Power to detect other effects was low: moderate interaction effects have a 48% chance of being detected, and small main and interaction effects have only a 17-24% chance of being detected. Table 7 presents descriptive statistics for each stage of the memory interview, separated into emotional and non-emotional recall.
Correct Total Recall
There was a large and significant main effect of emotion type and a small but
2 significant main effect of age, F (1, 74) = 22.34, p < .01, ŋp = .23, modified by a significant
2 interaction between information type and age, F (1, 74) = 9.28, p < .01, ŋp = .40. Not surprisingly, children reported more non-emotional than emotional information, F (1, 74) =
2 312.43, p < .01, ŋp = .81, and although older children recalled more information than did
3 A repeated measures ANCOVA was also conducted for each aspect of recall (free, prompted, or direct questioning), with language and emotion knowledge as covariates. Consistent with the lack of language and emotion knowledge differences between conditions shown in the preliminary analyses, however, each ANCOVA showed an identical pattern of results to the relevant ANOVA and is therefore not repeated here. 96
Table 7: Average Information Recalled by Condition and Age Group.
Condition
Emotion-
Information Type Emotion-cause expression No-emotion Minimal Mean
Younger (ages 3-4 years)
(n = 10) (n = 10) (n = 9) (n = 8)
Free Recall
Emotional 5.25 3.22 2.70 2.00 3.18
Non-emotional 10.63 7.89 7.50 5.70 7.78
Prompted Recall
Emotional 2.50 1.89 0.60 0.20 1.22
Non-emotional 2.88 1.89 2.20 1.90 2.19
Direct Questioning
Emotional 1.88 1.33 2.00 1.50 1.68
Non-emotional 1.63 1.78 1.00 0.90 1.30
Total 24.77 18.00 16.00 12.20 17.35
Older (ages 5-6 years)
(n = 12) (n = 9) (n = 12) (n = 12)
Free Recall
Emotional 6.00 4.89 3.58 2.96 4.32
Non-emotional 13.17 13.11 8.67 8.75 10.78
Prompted Recall
Emotional 1.75 1.60 0.75 0.67 1.17
Non-emotional 3.17 3.20 2.25 2.42 2.74
Direct Questioning
Emotional 2.25 0.89 2.58 2.21 2.06
Non-emotional 1.33 1.44 1.91 1.58 1.58
Total 27.67 25.13 19.74 18.59 22.65 97 younger children for both emotional, F (1, 74) = 7.28, p < .01, and non-emotional information, F (1, 74) = 21.67, p < .01, this difference was greater for non-emotional information.
Of particular interest, however, there was a moderate, significant main effect of
2 condition, F (3, 74) = 16.69, p < .01, ŋp = .40. Tukey tests showed that children in the emotion-cause condition (M = 8.92, SD = 1.85) reported more information than did children in the emotion-expression condition (M = 7.28, SD = 2.59), and children in the emotion- expression condition reported more information than did those in the no-emotion (M = 5.66,
SD = 2.24) and minimal (M = 4.94, SD = 1.99) conditions, who did not differ. No other
2 effects were significant (all Fs < 1.93, all ps > .05, all ŋp < .07).
These findings show that across age groups, children who engaged in emotion- oriented reminiscing (emotion-cause, emotion-expression conditions) reported more information about the zoo event than did those for whom there was no emotional content (no- emotion, minimal conditions, and this was particularly the case for children who reminisced about emotion causes. Older children also recalled more event information than did younger children.
Correct Free Recall
As for total recall, children recalled more non-emotional than emotional information,
2 F(1, 75) = 307.46, p < .01, ŋp = .81. There were also small but significant main effects of age,
2 2 F (1, 75) = 15.06, p < .01, ŋp = .17, and condition, F(3, 75) = 8.93, p < .01, ŋp = .26,
2 modified by a significant age x condition interaction, F(3, 75) = 2.76, p < .05, ŋp = .10.
Moreover, significant interactions were found for age x information type, F(1, 75) = 6.47, p <
2 2 .05, ŋp = .08, and condition x information type, F(3, 75) = 3.63, p < .05, ŋp = .13, however, the 3-way interaction of age, condition, and information type was not significant, F (3, 75) =
2 0.78, p < .01, ŋp = .03. 98
To investigate the age x condition interaction, differences between conditions were analysed separately for younger and for older children (see Figure 1). For younger children, the main effect of condition approached significance, F (3, 36) = 2.81, p < .06. Tukey tests showed that younger children in the emotion-cause condition freely recalled more than did children in the minimal condition, but there were no other significant differences between conditions. For older children, the main effect of condition was significant, F (3, 42) = 8.55, p
< .01. Tukey tests showed that the amount of information recalled by older children in the emotion-cause and emotion-expression conditions did not differ, and was greater than that recalled by children in the no-emotion and minimal conditions, who also did not differ. That is, relative to children in the no-emotion and minimal conditions, older children who participated in emotion-oriented reminiscing (emotion-cause, emotion-expression conditions) recalled more information, whereas for younger children, this was the case only for children in the emotion-cause condition.
To investigate the interactions of age and condition with information type, emotional and non-emotional information were examined separately. Note, however, that these analyses are likely to be constrained by the relatively limited amount of emotional information (M=
1.07, SD = 1.50) relative to non-emotional information (M= 6.48, SD = 3.57) recalled in free recall. Older children recalled more emotional and non-emotional information than did younger children, but this difference was greater for non-emotional than for emotional information, F (1, 80) = 8.37, p < .01. Moreover, although children who engaged in emotional reminiscing (emotion expression, emotion cause condition) recalled more information
(emotional, non-emotional) than did children in the other conditions (no-emotion, minimal), this was greater for non-emotional than emotional information, F (3, 79) = 3.71, p < .05. That is, more non-emotional information was recalled by older than by younger children and by children who engaged in emotional rather than non-emotional reminiscing. 99
14
12 Younger
10 Older
8
6 Recalled Information Recalled 4
2
0 Minimal No Emotion Emotion Emotion Cause Expression
Condition
Figure 1. Overall number of correct items (emotional and non-emotional, combined) freely recalled.
100
Additional Recall during Prompted Recall and Direct Questioning
For prompted recall, there were a significant main effects of information type, F (1,
2 2 74) = 21.65, p < .01, ŋp = .23, and of condition, F (3, 74) = 5.18, p < .01, ŋp = .17, but no
2 other significant main effects or interactions (all Fs < 1.58, all ps > .05, all ŋp < .03). Tukey tests showed that children in the emotion-cause condition recalled more information than did those in the no-emotion and minimal conditions, who did not differ, with the amount of information recalled by children in the emotion expression condition not differing from that recalled by children in any of the other conditions (see Table 7). For direct questions, there
2 were no significant effects, (all Fs < 3.21, all ps > .05, all ŋp = .04).
Summary
For total recall, children who engaged in emotional reminiscing reported more information than those who did not, but there was a particular memorial benefit when reminiscing included the causes of emotion. When the phases of the interview were examined separately, the most substantial age and condition differences emerged in free recall. Older children who engaged in emotion-oriented reminiscing (emotion cause, emotion expression conditions) recalled more information than those engaged in non-emotional reminiscing (no emotion, minimal conditions). In contrast, the recall of the younger children differed from the minimal condition only when reminiscing included emotion causes (emotion-cause condition). All children recalled greater amount of non-emotional than emotional information, but this was particularly so for the older age group and those in the emotion-oriented reminiscing conditions. No further interactions with information type were observed, suggesting that age and condition influenced recall for all aspects of the event: emotional and non-emotional. 101
Discussion
The major finding of the current study was that, consistent with expectation, children of both age groups who were engaged in emotion-oriented reminiscing (emotion-cause and – expression conditions) recalled more information about the “Visit to the Pretend Zoo” event than did children whose reminiscing included no emotional content (no-emotion and minimal conditions). This same pattern was also found for the recall of emotional and non-emotional information considered separately, suggesting that the type of information recalled was not important. That children in the emotion-cause and -expression conditions subsequently recalled more emotional information than did children who participated in reminiscing that did not include emotional information is perhaps unsurprising. These children were exposed to emotional information on two occasions; during the event, which, like many naturally- occurring experiences, was characterised by emotion states and their causes (together with subsequent activities); and during reminiscing, which entailed emotion-rich retrieval-cues.
There was, therefore, ample opportunity for the second exposure (reminiscing) to reinstate the first (the event) in memory. Much research shows that, across the lifespan, "the memory- preserving characteristics of [reinstatement] are both powerful and incontrovertible" (Howe,
O’Sullivan, & Marche, 1992, p. 248).
Particularly interesting, however, was the finding that children who were engaged in emotion-oriented reminiscing also reported a greater amount of non-emotional information.
This was despite the fact that reminiscing about the non-emotional aspects of the event was almost identical for children in all but the control condition. One possible explanation of this pattern relates to cognitive load. Specifically, as the reminiscing engaged in by children in the no-emotion condition included peripheral new non-emotional information that had not been discussed during the event (for example, the monkey’s raised arms), the resultant cognitive demands may have impaired processing of the non-emotional information common to all 102 three elaborative reminiscing conditions (e.g., Gathercole, Pickering, Ambridge, & Wearing,
2004; van Merrïenboer & Sweller, 2005). It is unlikely that this explanation accounts for the findings completely, however. Children in the no-emotion condition were able to visually encode this new information throughout the event (such as the monkey’s raised arms), despite the fact that it was not discussed. Reminiscing was therefore their second - albeit weaker – exposure. Further, and more compellingly, during reminiscing children in the emotion- expression condition were similarly presented with new information (the animals’ expressions) that had not been discussed during the event, yet they still manifested superior recall of non-emotional information relative to those in the no-emotion condition.
A second and more likely explanation for the superior recall of non-emotional information by children in the emotion-cause and –expression conditions is that emotional reminiscing served the functions of providing a meaningful goal and coherent structure for the event, which facilitated processing of all its aspects (Bird & Reese, 2006; Brown & Craik,
2000; Fivush et al., 2003; Trabasso & Stein, 1997). Thus, that the monkey was “excited because it was his birthday” (emotion-cause condition), or “we knew (he was excited) because he had a big smile” (emotion-expression condition), potentially enabled explanatory linkages to be made with other aspects of the event to a greater extent than did “monkey had very big arms… up in the air” (no-emotion condition). Moreover, the emotion-rich cues in the standardized reminiscing script provided maximal opportunity for the child to retrieve and reinstate in memory information about the event in a coherent, interconnected “story”.
Consistent with this latter possibility, Liwag and Stein (1995) reported that preschool children instructed to use emotion-rich retrieval cues recalled more central and associated information about an emotional experience, with a greater degree of structure, than did those who used cues that entailed less or no emotion content. 103
In addition, it may be that highlighting the emotional aspects of the event most likely rendered the reminiscing more attentionally engaging and interesting, and this, in combination with the provision of a meaningful goal and coherent structure, contributed to better encoding and retrieval of both emotional and non-emotional material (Compton et al., 2003; McDaniel,
Waddill, Finstad, & Bourg, 2000).
The findings of superior recall of all aspects of the event following emotion-rich reminiscing are consistent with those of other recent research. For example, Laney, Heuer, and Reisberg (2004) found that adults demonstrated superior recall of all types of information
– central, peripheral, and visual detail - from a narrated slide sequence when it had a
“thematically-induced emotional” (as opposed to neutral) theme. Thematically-induced arousal occurs in response to everyday emotional situations of relevance to one’s own life and goals (Laney et al., 2003), and is arguably the response induced by the event in the current study.
Across all stages of the interview, children in the emotion-cause condition provided significantly more information than those in all other conditions. In other words, both emotional and non-emotional information was particularly well recalled when the emotions were couched in an explicitly causal framework during reminiscing. The discussion of causes provided a logical structure for the event, including an explanation of how and why a given emotion was related to the remaining actions, objects, and descriptions, and, as has been shown by much research, a logical structure has a strongly facilitative effect on recall (Bauer
& Mandler, 1989; Cervantes & Callanan, 1998; Conroy & Salmon, 2006). Whilst emotion talk and causal talk are necessarily confounded in the emotion-cause condition, the results nonetheless show that causal talk about emotion – whatever the mechanism of effectiveness – is particularly powerful. 104
Interestingly, our findings suggested that younger children were more reliant on an explicit causal framework than were their older counterparts; indeed, for measures of free recall, providing the cause of the emotion was necessary for the younger children’s recall to exceed that of their young peers in the no emotion and minimal conditions. Much research shows that when children are very young, parents heavily scaffold reminiscing conversations by providing most of the narrative structure and content (Eisenberg, 1985; Hudson, 1990).
Whilst children aged 3- to 4-years have certainly begun to internalise the memory skills necessary to participate actively in reminiscing, the relative structure and content they are able to contribute to an event representation develops gradually and is less than a 5- or 6-year-old may contribute: it is not until 58 months, for example, that children’s contributions may influence both maternal structure and their own subsequent contributions to shared recall
(Reese et al., 1993). Perhaps, therefore, our younger children aged 3- to 4-years were unable to infer the causal structure of the event without some degree of explicit scaffolding from the experimenter. For older children, information about emotion expression may have been sufficient to infer the causal relations between the emotion and associated event components
(Wenner, 2004; see also Russell, 1990).
Contrary to the findings of other experimental studies comparing high elaborative and minimal styles of reminiscing, children in the no-emotion condition failed to recall significantly more information than did children in the minimal condition (Conroy & Salmon,
2006; McGuigan & Salmon, 2004). In the current study, and in contrast to the earlier experimental and naturalistic work, children from all conditions were provided with an elaborative, emotion rich narration of the event as it occurred. Thus, even children in the minimal condition had at least one opportunity to encode the event in a rich and full manner, with likely benefits to memory (see Boland et al., 2003; Tessler & Nelson, 1994). Greater, significant differences in recall between children in the no-emotion condition and the minimal 105 condition may have been found had the event also been accompanied by minimal or empty narration.
Notwithstanding the benefits of using an experimental paradigm to investigate the impact of emotional reminiscing on recall, corresponding limitations should be noted. For example, in naturalistic contexts, children’s recall is enhanced when parents show sensitivity to their child and focus on the aspects of the event that interest them (Fivush et al., 2006;
Cleveland & Reese, 2005). Moreover, reminiscing typically focuses on the parent’s and child’s own emotional reactions to an event, as in Study 1, as it is these emotional reactions that make an event personally meaningful to the child (Bird & Reese, 2006; Fivush, 1993).
Given the standardized event and the scripted nature of the reminiscing in the current study, neither of these conditions could be fulfilled. The findings are, therefore, likely to underestimate the influence of high elaborative, emotion-rich reminiscing on children’s recall.
It is noteworthy then, that a clear causal relationship between emotion-oriented reminiscing and children’s recall of the event nonetheless emerged.
Indeed, in a number of ways, the findings extend research investigating the associations between reminiscing, its emotional content, and children’s recall. Importantly, the experimental paradigm enabled us to tease apart the effect of reminiscing content and style, showing that emotion talk, particularly within a causal framework, aids children’s recall of the emotional and non-emotional aspects of an emotionally salient event, above and beyond the impact of reminiscing not focused on emotion. Findings of much other research indicate that children come to internalise their parents’ reminiscing style and content over time (e.g. Hudson, 1993; Kuebli et al., 1995; Leichtman et al., 2000). Although Study 1 suggests that high elaborative reminiscing is associated with discussion of emotion causes but not other emotion references, and past research investigating the relationship between a high elaborative reminiscing style and emotional reminiscing presents mixed findings (see Laible, 106
2004a, 2004b; Laible & Song, 2006; Thompson et al., 2003; Wareham & Salmon, 2007); the present findings nonetheless suggest that parents who frequently reminisce with their children about emotions using a high elaborative style are boosting their child’s autobiographical recall via two pathways (style and content). Moreover, it is likely that these parents will have children who come to show better recall for both discussed and undiscussed events.
The second focus of our study was to investigate the contributions of children’s emotion knowledge to their recall of the event, together with the contributions of age and language. Consistent with our expectations, the independent measure of emotion cause knowledge contributed significantly to children’s free recall of both emotional and non- emotional aspects of the “Visit to the Pretend Zoo” event, over and above the contributions of both age and language. In discussing their own findings that children aged 2.5 to 3.5 years who had a greater level of emotion understanding recalled more detail in their independent reports of a past experience, Wang et al. (2006) suggested that one important function of emotion knowledge is to enable children to interpret the emotional meaning of the experience and to perceive its personal relevance, with consequent positive implications for its memorability. That, in the current study, children’s understanding of emotion causes was associated with superior recall of both emotional and non-emotional information suggests that more sophisticated emotion knowledge may permit more coherent organization of the entire experience.
There are a number of possible directions for future research. First and foremost, the emotion-cause condition is confounded in that it contains two potentially beneficial elements: emotion talk, and causal talk. Whilst the benefits of emotional reminiscing of any kind are evident, it is unclear whether the additional advantage of emotion cause-oriented reminiscing is specific to reminiscing about emotion causes - perhaps by enhancing the process by which emotions provide an event with a meaningful goal and coherent structure - or would be 107 expected following causal reminiscing of any kind. An experimental study that includes a causal-only (no emotion) condition may allow stronger conclusions to be made about the respective roles of emotion talk and causal talk. Causal talk about mental states not involving emotion could perhaps also be considered, to further disentangle the influence of emotion from other causal talk.
Additionally, it would be interesting in future research to examine the influence on recall of self- versus experimenter-generated reminiscing within each condition, for both emotional and non-emotional information. This could be investigated in two ways. First, whilst children both generated and were provided an equal amount of information across the event, the type of information generated and provided could be counterbalanced between animals. This would allow the recall for generated versus provided information to be examined for each animal, without the possibility that one animal was more memorable than others. Second, children’s correct responses to the questions asked in the three high- elaboration conditions during reminiscing could be examined. In the present study reminiscing was conducted just two days after children had participated in the event
(following McGuigan & Salmon, 2004). Although recall during shared reminiscing was not formally measured, children were generally able to recall the information they were asked about the event and thus participate fully in their reminiscing conversations. By extending the delay between the event and reminiscing and then coding the amount of correct information that children generate during reminiscing, individual differences may be observed within each condition. Following past research, information that children were able to supply themselves during reminiscing (that is, self-generated information) may be more memorable than might information that children were unable to correctly generate during reminiscing and were thus provided by the experimenter (see McGuigan & Salmon, 2004; McNamara & Healy, 1995; 108
Slamecka & Graf, 1978). Nonetheless, given young children’s difficulties with emotions, interactions with age and information type may emerge.
In conclusion, it was found that high elaborative, emotion-oriented reminiscing between children and an experimenter, particularly when couched in an explicitly causal framework, enhances the children’s recall of an event above and beyond the impact of reminiscing not focused on emotion. Emotions not only provide the event with a meaningful goal and coherent structure; they may also make the reminiscing conversation more salient and interesting such that encoding and recall are enhanced. Moreover, it was also found that children’s pre-existing emotion cause knowledge and language contribute to children’s memory for the event. Further research aimed at clarifying the exact causal or emotional (or both) mechanism by which emotion-cause reminiscing is beneficial is recommended.
Nonetheless, these findings extend those of a robust body of naturalistic and longitudinal research that high elaborative reminiscing shapes children’s autobiographical memory, by demonstrating the significant influence of a focus on emotion.
109
CHAPTER 4: TRAINING MOTHERS IN HIGH ELABORATIVE, EMOTION-RICH
REMINISCING: FACILITATING AUTOBIOGRAPHICAL MEMORY AND EMOTION
KNOWLEDGE
Taken together, the findings of Study 1 and 2 highlight the importance of high elaborative, emotion oriented reminiscing (and, in particular, emotion-cause oriented reminiscing) for children’s autobiographical recall and emotion knowledge development.
There are a myriad of ways in which the development of these skills are important. As discussed in Chapter 1, theorists suggest that autobiographical memory is intimately connected to children’s concept of self, guides and directs current behaviour, and, particularly when focused on emotion, serves many fundamental social functions including the maintenance of intimacy (Alea & Bluck, 2003; Cockroft & Reese, in press; Pillemer, 1992;
Wang & Conway, 2004). When autobiographical memories are impoverished or incomplete, individuals may be deprived of both a rich source of past learning to guide current behaviour and a means of effective social communication (see Wareham & Salmon, 2006, for review).
Of equal importance, research shows that emotion knowledge at preschool predicts empathic, pro-social behaviour with peers and family members, positive peer status, and teachers’ evaluations of social competence at kindergarten (Denham et al., 2003; Hughes & Dunn,
1998; Izard et al., 2001). Less advanced understanding of emotions in preschoolers is associated with internalizing and externalizing problems throughout childhood, such as angry and aggressive behaviour at preschool, impaired social relationships, and feelings of anxiety, depression, and social isolation (Denham et al., 2003; Izard et al., 2001; Fine, Izard, Mostow,
Trentacosta, & Ackerman, 2003). Given the importance of young children’s autobiographical memory and emotion knowledge, Study 3 aimed to facilitate their development. 110
Interventions that target individual differences in the style and content of parent-child interactions may be one way in which autobiographical memory and emotion knowledge development can be enhanced. Notwithstanding the well established role that parents’ high elaborative reminiscing style plays in children’s development of autobiographical memory, and recent evidence suggesting that high elaborative reminiscing also facilitates children’s emotion knowledge, however, few studies have attempted to train parents to use a high elaborative style when reminiscing. In those studies that have been conducted, training has focused on the high elaborative style but not emotion content, and outcome measures have included memory but not emotion knowledge.
There are two reasons to suggest that interventions targeting high elaborative, emotion-oriented reminiscing may be effective in enhancing the development of both autobiographical memory and emotion knowledge. First, preschool represents a time when development of key cognitive and socio-emotional skills is still occurring, and when individual differences are most prominent. According to Dadds (2002), intervention programs are best carried out before patterns of cognitive and emotional development are well established. Second, extensive research on the effectiveness of early intervention aiming to boost (typically disadvantaged) children’s cognitive, academic, and social outcomes indicates that programs that actively involve the child, as parent-child interactions do, provide more enduring benefits than those that rely on intermediary routes such as parent training only (see
Ramey & Ramey, 1998, 2004, for reviews).
To date, studies aimed at training mothers to use a high elaborative style have been successful. In the first reminiscing intervention of its kind, Peterson et al. (1999) allocated 20 economically disadvantaged mothers to a reminiscing or a control group. Fourteen mothers completed the study. Reminiscing mothers were successfully trained to spend more time reminiscing with their 43-month-old preschoolers, ask more open-ended and context eliciting 111 questions, and encourage longer narratives through the use of back-channel responses (where previously discussed elements of an event are restated and drawn together). Their children subsequently produced memory narratives with more context-setting descriptions about where and especially when the described experiences took place. Nonetheless, low participant numbers, a large proportion of drop-outs, and large variability between participants made strong conclusions difficult.
In a subsequent investigation of the high elaborative style during ongoing experiences,
Boland et al. (2003) trained mothers of preschoolers to use elements of the elaborative style whilst participating in a play camping activity. The high elaborative elements, including ‘wh’ questions, linking the event to the child’s prior knowledge, follow-ins (encouraging aspects of the conversation that the child is interested in), and praise, were outlined in a pamphlet and demonstrated on videotape. Mothers who received the training incorporated more of these elements into their discussions with their children than did untrained mothers, and children of trained mothers were subsequently able to recall more of the camping activity after both 1 day and 3 week delays than were children of untrained mothers. Children’s language skill did not affect the degree to which training was effective, yet also contributed to recall.
Finally, in the only reminiscing intervention to date to include a large sample and long follow-up, Reese and Newcombe (2007) trained 100 mothers to ask their 19-month-old toddlers open-ended questions and confirm their contributions during shared reminiscing conversations. Two months later, trained mothers asked more open-ended questions than did mothers on a waitlist control. Fifteen months later, trained mothers asked more open-ended questions and also made more confirmations and statement elaborations, irrespective of their original reminiscing style. Children of trained mothers provided more memory elaborations during shared recall than did children of control mothers, and, if they were initially high in self-recognition, independently provided more memory elaborations to an experimenter. 112
Reese and Newcombe (2007) caution that trained mothers also provided more low-elaborative repetitions after 15 months than did control mothers, as they often repeated the open-ended questions they were trained to ask. Although repetitions may in some contexts indicate sensitivity to children’s developmental level (e.g. for very young children, or for children who are less outgoing and active, repetitions may keep them on task and involved: Lewis 1999), a wealth of findings show that repetitions do not support preschool children’s recall as high elaborative utterances do (see Fivush et al., 2006, for review). Reese and Newcombe (2007, p.30) state that “in future elaboration interventions with parents, we recommend that researchers highlight to an even stronger degree the possible deleterious effects of repetitive questions on children’s memory and participation”.
These three interventions complement concurrent and longitudinal paradigms in suggesting that elements of an elaborative discussion style have a causal influence on children’s memory reports. Moreover, the findings show that a high elaborative style can be both identified and taught. Nonetheless, in the case of both Peterson et al.’s (1999) and Reese and Newcombe’s (2007) reminiscing interventions, only trained mothers and not waitlist control mothers were encouraged to spend additional time interacting with their child.
Children of trained mothers may therefore have developed better reminiscing skills not due to an increase in reminiscing quality, but simply because they were given greater opportunity to practice reminiscing skills with their mothers (or, most likely, both quantity and quality may be important). Furthermore, in both Boland et al.’s (2003) and Peterson et al.’s (1999) intervention training studies, trained mothers were not only asked to use an elaborative style of questioning and commenting, but were also encouraged to follow their child’s lead.
Untrained mothers were given no such encouragement. Interestingly, Cleveland and Reese
(2005) recently found that ‘autonomy supportive’ elements of reminiscing, in which a parent supports and follows the child’s interests rather than controlling the interaction, are 113 statistically independent of the structural elements of elaboration, such as open-ended questioning. Nonetheless, both predict children’s recall. Although Boland et al.’s (2003) and
Peterson et al.’s (1999) interventions did not directly train parents in autonomy support, simply encouraging parents to follow their child’s lead may have had a similarly beneficial impact on recall. Control conditions that encourage attention and sensitivity towards the child’s interests and abilities in the same way as do training conditions are necessary to isolate the impact of different elaborative style on children’s autobiographical recall.
The present study aimed to enhance children’s cognitive and socio-emotional development using reminiscing training. This intervention was innovative in three respects.
First, mothers in the reminiscing condition were trained not only to use a high elaborative style when reminiscing (by asking wh-questions instead of yes/no questions, and providing detailed descriptions instead of repeating questions), but also to discuss the emotional aspects of their shared past events. This combination of high elaborative style and emotion content was expected to strongly facilitate children’s development of autobiographical memory, over and above the development of autobiographical memory in control children. Second, emotion knowledge was included as an outcome variable, along with autobiographical memory. High elaborative, emotion-oriented reminiscing was also expected to facilitate children’s emotion knowledge development.
Third, control mothers were trained in child-directed play; a clinical therapy based on attachment theory (Eyberg et al., 2001). During child-directed play, mothers are trained to follow their child’s lead, play at their child’s pace, and praise their child’s efforts without criticising, making corrections, or suggesting changes to the play (e.g. Greco, Sorrell, &
McNeil, 2001). Child-directed play is itself a powerful intervention, and has formed the basis of empirically supported interventions aiming to promote appropriate child behaviour and improve parental responsiveness and sensitivity (e.g., Eyberg et al., 2001; Hood & Eyberg, 114
2003; Kottler & McMahon, 2004; see Greco et al., 2001). For example, Eyberg et al. (2001) trained parents of young children with conduct problems in both child-directed interaction
(aimed at enhancing positive parent-child interactions) and parent-directed interaction (aimed at reducing non-compliant behaviour). Improvements were maintained across two years.
Although child directed play does not relate to reminiscing style or emotional content, it does control for the potential influence of additional attention and associated sensitivity that might contribute to the reminiscing intervention’s effectiveness. Mothers in the reminiscing and control conditions thus spent an equal amount of time with the experimenter and an equal amount of time practicing with their children, whilst both training programs encouraged mothers to be sensitive to their children’s interests and abilities.
Method
Participants
Participants were 80 mother-child dyads initially aged between 37 and 59 months (M
= 45.85, SD = 5.86). Mothers were recruited via advertisements placed in a free parenting magazine distributed throughout Sydney. Mothers from a wide variety of occupations and suburbs responded, and the majority were Caucasian. Dyads were screened to ensure that
English was the first language and spoken at home. Nineteen dyads did not complete the second assessment session; 8 of these did not continue after the first assessment and thus did not participate in training. Non-continuing participants (referred to hereafter as ‘non- continuers’) were even across condition (χ2 = 0.83, p > .05). Seventeen dyads that completed the second assessment session did not complete the third assessment session, and again, non- continuers were even across condition (χ2 = 0.04, p > .05). Given the high number of non- continuers, causes of not continuing were analyzed. Thirty-five percent of non-continuers chose not to continue due to maternal commitments (e.g. returning full time work, pregnancy), 31 % were either uninterested in the project or unable to be contacted, 21% 115 moved interstate or overseas, and 13% of dyads did not continue for reasons pertaining to the child (e.g. diagnosis of severe anxiety, non-compliance). In sum, 72 dyads were initially allocated to and participated in training, and 44 dyads completed all stages of the program. Of the 44 dyads who completed all stages, there were 13 boys and 10 girls in the reminiscing condition (M = 45.00, SD = 3.92 at the first time-point), and 10 boys and 11 girls in the control condition (M = 46.16, SD = 4.86 at the first time-point). There were no significant age, t(39) = 0.85, p > .05, or gender, χ2 = 0.21, p > .05, differences between condition.
Measures and Procedure
Dyads in both the reminiscing and the control conditions individually participated in seven training and assessment sessions over a period of eight months. These consisted of an initial pre-training assessment session (A1), four training sessions (T1-T4), a post-training assessment session (A2), and a six-month follow-up assessment session (A3). The timing of these sessions is shown in Table 8. All sessions were conducted at the University of New
South Wales. One experimenter, initially blind to condition, conducted A1 and the four training sessions. To ensure that experimenters remained unaware of the condition to which participants had been assigned, different experimenters conducted A2 and A3. Experimenters were counterbalanced across training and assessment sessions.
Table 8: Assessment and Training Timeline
Session Break Session
Pre-training Assessment (A1) 1 week Training (T1)
Training (T1) 1 week Training Follow-up (T2)
Training Follow-up (T2) 1 week Training Follow-up (T3)
Training Follow-up (T3) 2 weeks Training Follow-up (T4)
Training Follow-up (T4) 1 week Post-training Assessment (A2)
Post-training Assessment (A2) 6 months Follow-up Assessment (A3) 116
Assessments
A1, A2, and A3 assessment tasks measured mother-child and experimenter-child memory conversations, children’s emotion knowledge, and children’s receptive and expressive language. Children’s emergent literacy, children’s narrative skill, and mothers’ expressed emotion were also measured, but these tasks are not described here. Sessions were videotaped and audio-taped, and took approximately 90 minutes. Dyads were given a 30 minute break halfway through the session, and were also given smaller breaks whenever the child showed signs of tiredness. During each break, dyads were encouraged to leave the room and either go for a walk or have a snack before returning.
Memory Conversations
Mothers were asked to nominate on paper four past events that were one-time occurrences and had been experienced by parent and child in the past one to four weeks.
Unlike Study 1, in which mothers were asked to nominate everyday happy, sad, angry, and scary events (such that the emotional nature of each event was highlighted), no mention of emotions was made: this adjustment was considered necessary in order to observe mothers’ natural frequency of emotion references during the discussion of everyday events. The experimenter randomly selected two of the events nominated by the mother. Mothers were then asked to discuss the selected events with their child, following the procedure outlined in
Study 1 (p.63). Following mother-child memory conversations the experimenter then discussed with the child the remaining two events nominated by the mother, also following the procedure outlined in Study 1 (p.63).
Children’s Emotion Knowledge
Children’s emotion knowledge was measured in two ways. Firstly, all children were administered Denham’s Emotion Knowledge Task (Denham, 1986). This measure was also 117 used in Study 1 and is described on page 64. Secondly, children were given an Emotion
Cause Production Task. This measure was also used in Study 2, and is described on page 89.
Children’s Language
The Preschool Language Scale – fourth edition (PLS4) was used to assess both receptive and expressive language ability (Zimmerman, Steiner, & Pond, 2002). It shows high test-retest, internal consistency, and inter-rater reliability (range .90-.99), and good content, structural, and criterion validity (Zimmerman et al., 2002). The expressive scale, used to measure expressive language ability, consists of 68 items and the auditory scale, used to measure receptive language ability, consists of 62 items. Following standardised instructions, children began on age appropriate items and, once having correctly completed three items in a row, continued until they made seven errors in a row. Raw scores were converted into receptive, expressive, and full-scale age norms (M = 100, SD = 15).
Training
Dyads were randomly allocated to training condition. The reminiscing and control training programs were designed to each encourage attending to the child and control for the amount of training provided to the mother, but to differ in content. Mothers in the reminiscing condition were encouraged to frequently reminisce with their child using the ‘WDE’ style: asking the child open-ended wh-questions, using detailed descriptions that describe the event and build on the child’s descriptions, and discussing emotions and their causes. Mothers in the control condition were encouraged to frequently play with their child using child directed play: attending to the child fully, allowing the child to direct the play, and playing at the child’s pace. Both programs advocated praise and child sensitivity.
During T1, mothers watched a training video describing the interaction style - WDE reminiscing or child directed play, according to condition - and were led through a training booklet by the experimenter. They then practiced the interaction style with their child, and 118 were given feedback (for dyads in the control condition who practiced child directed play, a dolls house, puppets, building blocks, and a magnetic balls and sticks game were provided).
Finally, they were asked to practice the interaction style with their child once a day for 5 minutes, and to record one practice on audiotape (provided). The session took one hour.
During T2, T3, and T4, the mother and experimenter listened to the practice tape recorded during the week, and the experimenter gave feedback. Mothers then practiced with their child on the spot, integrating the initial feedback, and were given further feedback. They were asked to continue practicing once a day for 5 minutes and, for T2 and T3 only, to record another practice to be used during their next training. The sessions each took 30 minutes.
Following T4, mothers were telephoned at four spaced intervals (duration: one week, two weeks, one month, two months) to encourage continued participation.
Training Video
The reminiscing video showed a mother discussing two past events with her child.
Voice-over narration drew attention to the WDE elements: wh-questions, detailed descriptions, and emotion talk. The control video showed parents playing with their children.
Voice-over narration was used to draw attention to the child directed play elements: attending to the child, allowing the child to direct play, playing at the child’s pace. Each training video was approximately 18 minutes long.
Training Booklet
A four-page booklet was created for each program and given to mothers to keep (see
Appendix C). The booklet outlined the key elements being trained, practice guidelines, and
‘troubleshooting’ practice tips that outlined how change non-targeted behaviour into targeted components of reminiscing / child directed play (e.g. rather than repeating a question, provide more detail or answer yourself; rather than making suggestions, praise aspects of the play that you like). In addition, the reminiscing booklet included a transcript of a high elaborative, 119 emotion-oriented parent-child past event discussion. No example was provided in the control booklet.
Assessment Coding
As in Study 1, mother-child and experimenter-child reminiscing conversations during assessment sessions were transcribed verbatim. Each utterance (subject-verb proposition) was coded once for style and once for content. Parents’ and children’s utterances were coded separately. In addition, children’s utterances were coded once during parent-child conversations and once during experimenter-child conversations. A second coder practiced coding a small number of transcripts with the primary coder and then independently coded
25% of transcripts. Inter-rater reliability, calculated using Cohen’s Kappa and based on the
25% of transcripts completed by both the primary and second coder, was .82 (range .69 – .90) for parent-child conversations and .89 (range .63 – .100) for experimenter-child conversations.
Style was coded using the same coding scheme as in Study 1 (p.65). According to this scheme, each utterance was coded as high elaborative, low elaborative, or a confirmation.
Note that whilst this scheme closely followed others (e.g. Fivush et al., 2003; Harley &
Reese, 1999; Hudson, 1993; Peterson et al., 1999), yes-no questions were coded as low elaborative. As in Study 1, analyses revealed no difference between the pattern of results for yes-no questions individually and low elaborative utterances as a whole, and no significant difference in the findings whether yes-no questions were coded as high or low elaborative.
Thus, this coding decision did not influence the conclusions that were drawn.
Following Study 1, furthermore, elaboration was measured in two ways. First, the total number of high elaborative and low elaborative utterances made by each participant was counted (e.g. Reese & Newcombe, 2007). Second, the degree to which a parent or child used high elaborative utterances as a proportion of overall talk (high + low elaborative utterances) 120 was also calculated (e.g. Haden & Fivush, 1996; see also Low & Durkin, 2001). Due to few low elaborative utterances being made, however, the proportion of high elaborative talk made was high across all time-points and conditions (Ms > .80) and was therefore not considered further.
With regard to content, utterances were each coded as being factual, emotional, or preferential in content. Preference references included positive and negative evaluations of the event (e.g., “It was good, wasn’t it?”) that were not explicit emotion references.
Following Fivush et al. (2003), who coded emotion attributions, causes, and consequences; emotional utterances were further categorized as being attributions, behaviors, causes, outcomes, or resolutions. This scheme was similar to, but more detailed than, that used in
Study 1 (in which emotion references were coded as attributions, causes, or consequences only). As in Study 1, total emotion utterances rather than proportions (emotional utterances /
[emotional + preferential + factual utterances]) were considered (see Dunn et al., 1991;
Fivush et al., 2000; Reese et al., in press).
Results
There were no significant differences between reminiscing or control dyads on any variable at A1, Fs(1, 71) < 1.46, ps > .05. Of the initial 80 mothers, 19 did not continue to A2
(8 following A1, 11 during or following training), and an additional 17 did not continue to
A3. Thus, a total of 72 mothers were allocated to and commenced training. There were no significant differences in the numbers of non-continuers or causes of not continuing as a function of condition, χ2s < 0.83, ps >.05. Furthermore, there were no significant differences on previous assessments between those that were subsequently excluded and those that completed the next assessment, Fs < 2.73, ps > .05.
Given the lack of systematic differences in frequency or cause of non-continuers between conditions, the ‘last observation carried forward’ (LOCF) method was used to 121 account for missing data in the most conservative way possible. This method is used in randomised control trials to account for dataset attrition (e.g. Åkerblad, Bengtsson, von
Knorring, & Ekselius, 2006; Kinon et al., 2006; Powers, Bannon, Eubanks, & McCormick,
2007; Tandon et al., 2006; Williamson et al., 2006). In accordance with the LOCF method, excluded dyads’ last score on each variable was also used as their score on those variables for the remaining assessment sessions; thus, there is an assumption that no development or change has occurred (e.g. Hennen, 2003; Houck et al., 2004). The method must be used with caution, as it is inherently biased towards finding no effect or low effect size in groups that do not initially differ (Hennen, 2003; Houck et al., 2004). Nonetheless, this limitation is also an advantage: any significant effects found are therefore reflections of true group differences as a function of treatment or training. To compare reminiscing and control dyads, the LOCF method was only used for those dyads that had been allocated to and commenced training, n =
72.
Mothers’ Reminiscing Style and Content, Children’s Reminiscing Style and Content, and
Children’s Emotion Knowledge
Using the LOCF method and an alpha level set at .05, a series of 2 x (3) repeated measures General Linear Models were used to examine differences between dyads in the reminiscing and the control condition across the duration of the study. Power analyses revealed sufficient power to detect both large and moderate effects (>99% and 91% chance of detection, respectively), yet insufficient power to detect small effects (33% chance of detection). Findings are first reported for mothers’ reminiscing style and content, then children’s shared and independent recall, and finally, children’s emotion knowledge.
Children’s age-standardized language scores were high (M = 119.45, SD = 10.77) but did not
2 differ as a function of time, F(1, 71) = 1.19, p > .05, ŋp = .04, condition, F(1, 71) = 0.08, p > 122
2 2 .05, ŋp < .01, or time x condition, F(2, 70) = 0.72, p > .05, ŋp = .03; thus, controlling for language was considered unnecessary.
Mothers’ Shared Reminiscing Style and Content
Mothers’ reminiscing style. With regards to style, it was hypothesised that mothers in the reminiscing condition would make more high-elaborative utterances when reminiscing than would mothers in the control condition, but would not increase their low elaborative utterances. To examine this, mothers’ total high and total low elaborative utterances were analysed. Total high elaborative utterances consisted of open-ended questions (wh-questions,
‘do you remember’ questions) and information statements (memory elaborations), summed.
Total low elaborative utterances consisted of close-ended questions and repetitions, summed.
Interestingly, although total high and total low elaborative utterances were considered to be of primary importance, it is noteworthy that the pattern of results for each high elaborative and each low elaborative component individually did not differ from the pattern for total high elaborative utterances (summed) and total low elaborative utterances (summed), respectively.
For example, the pattern for wh-questions was the same as for total high elaborative utterances, and the pattern for yes-no questions was the same as for total low elaborative utterances.
2 There were significant main effects of time, F(1, 71) = 5.11, p < .01, ŋp = .07, and
2 condition, F(1, 71) = 18.51, p < .01, ŋp = .22, on mothers’ high elaborative utterances,
2 modified by a significant time x condition interaction, F(2, 70) = 16.33, p < .01, ŋp = .20. As shown in Table 9, mothers in the reminiscing condition made significantly more high elaborative utterances at A2 and A3 than they did at A1, yet mothers in the control condition continued to use the same number of high elaborative utterances as at A1. There was also a significant main effect of time on mothers’ low elaborative utterances, F(1, 71) = 5.90, p <
2 2 .01, ŋp = .08, but no main effect of condition, F(1, 71) = 2.16, p > .05, ŋp = .03, and no 123
Table 9: Mothers’ and Children’s Mean High and Low Elaborative Utterances
Shared Recall Independent Recall
Mother Child Child
Condition M (SD) M (SD) M (SD)
High Elaborative Utterances
Reminiscing Training
Pre-training (A1) 18.55 (8.99) 8.84 (4.69) 4.22 (4.09)
Post-training (A2) 29.46 (16.18) 13.11 (7.92) 4.97 (4.47)
6-month follow-up (A3) 28.11 (17.76) 14.76 (8.98) 6.21 (5.28)
Control Training
Pre-training (A1) 16.59 (9.49) 10.30 (7.82) 3.77 (3.69)
Post-training (A2) 13.80 (7.44) 8.81 (5.33) 4.81 (5.17)
6-month follow-up (A3) 13.65 (6.03) 9.38 (4.81) 5.08 (3.93)
Low Elaborative Utterances
Reminiscing Training
Pre-training (A1) 4.83 (3.13) 2.70 (2.39) n/a
Post-training (A2) 3.85 (2.56) 2.34 (1.95) n/a
6-month follow-up (A3) 3.92 (3.25) 2.71 (2.44) n/a
Control Training
Pre-training (A1) 4.55 (3.38) 2.86 (2.66) n/a
Post-training (A2) 3.12 (3.58) 2.15 (2.62) n/a
6-month follow-up (A3) 2.52 (1.74) 1.75 (1.62) n/a
2 interaction, F(2, 70) = 1.07, p > .05, ŋp = .02. All mothers decreased their use of low elaborative comments across time.
Mothers’ reminiscing content. With regards to content, it was hypothesised that children in the reminiscing condition would make more total emotion references when reminiscing than would children in the control condition. Total emotion references; that is, total emotion attributions, behaviours, causes, and outcomes, summed, were calculated. Interestingly, and 124 with the exception of emotion resolutions (which were rarely mentioned), the pattern of results for each emotion reference component individually reflected the pattern of results for total emotion references.
2 There were significant effects of time, F(1, 71) = 14.34, p < .01, ŋp = .18, and
2 condition, F(1, 71) = 29.19, p < .01, ŋp = .31, on mothers’ total emotion references, modified
2 by a significant time x condition interaction, F(2, 70) = 19.74, p < .01, ŋp = .24. As shown in
Table 10, mothers in the reminiscing condition made significantly more emotion references at
A2 and A3 than they did at A1, yet mothers in the control condition did not. There were no
2 significant main effects of time, F(1, 71) = 0.30, p > .05, ŋp = .01, or condition, F(1, 71) =
2 0.21, p > .05, ŋp = .03, on mothers’ total preference references, and no significant time x
2 condition interaction, F(2, 70) = 0.04, p > .05, ŋp < .01 (M = 3.22, SD = 3.32).
Table 10: Mothers’ and Children’s Emotion References during Shared Recall
Shared Recall
Mother Child
Condition M (SD) M (SD)
Reminiscing Training
Pre-training (A1) 1.16 (2.64) 0.58 (1.64)
Post-training (A2) 8.72 (8.37) 2.76 (3.61)
6-month follow-up (A3) 7.46 (7.74) 3.49 (4.21)
Control Training
Pre-training (A1) 1.12 (2.23) 0.71 (1.91)
Post-training (A2) 0.68 (1.39) 0.35 (1.12)
6-month follow-up (A3) 0.65 (1.20) 0.30 (0.94)
125
Children’s Shared Reminiscing Style and Content
Children’s reminiscing style. With regards to reminiscing style, it was hypothesised that children in the reminiscing condition would come to recall more and thus make more high-elaborative utterances when reminiscing than would children in the control condition, but would not increase their low elaborative utterances. Children’s total high elaborative utterances (prompted + unprompted information statements) and total low elaborative utterances (yes-no responses + repetitions) were analysed. High and low elaborative questions were extremely rare, and therefore did not contribute to total high or total low elaborative utterances. It is noteworthy that, although total high and total low elaborative utterances were of primary consideration, the pattern of results for each high elaborative and low elaborative component did not differ from the pattern of results for total high elaborative utterances and total low elaborative utterances, respectively. In the case of high elaborative utterances, for example, both prompted and unprompted information statements showed the same pattern of results, when examined separately, as did total high elaborative utterances (prompted and unprompted information statements, summed).
2 There were significant time, F(1, 71) = 4.18, p < .05, ŋp = .06, and condition, F(1, 71)
2 = 4.68, p < .05, ŋp = .06, main effects on children’s high elaborative utterances during shared
2 recall, modified by a significant condition x time interaction, F(2, 70) = 8.96, p < .01, ŋp =
.11. As shown in Table 2, children in the reminiscing condition increased their use of high elaborative utterances across time, whereas children in the control condition used the same number of high elaborative utterances at A2 and A3 as they had at A1. There was no
2 2 significant time, F(1, 71) = 1.86, p > .05, ŋp = .02, or condition, F(1, 71) = 0.41, p > .05, ŋp
= .01, main effect on children’s low elaborative utterances, and no significant time x
2 condition interaction, F(2, 70) = 1.56, p > .05, ŋp = .02. 126
Children’s reminiscing content. With regards to reminiscing content, it was hypothesised that children in the reminiscing condition would come to make more emotion references during shared reminiscing than would children in the control condition. Total emotion references; that is, total emotion attributions, behaviours, causes, and outcomes, summed, were calculated. Whilst emotion references overall were of primary interest, it is also noteworthy that the pattern of results for each emotion reference component individually
(e.g. emotion causes) did not differ from the pattern of results for total emotion references
(summed).
2 There were significant time, F(1, 71) = 9.21, p < .01, ŋp = .12, and condition, F(1, 71)
2 = 19.55, p < .01, ŋp = .22, main effects on children’s total emotion references, modified by a
2 significant condition x time interaction, F(2, 70) = 13.81, p < .01, ŋp = .17. As shown in
Table 10, children in the reminiscing condition increased their emotion references between
A1 and A2, and maintained this difference at A3. Children in the control condition did not
2 increase their emotion references. There were no significant time, F(1, 71) = 0.91, p > .05, ŋp
2 = .01, or condition, F(1, 71) = 0.00, p > .05, ŋp = .00, main effects on children’s preference
2 references, and no significant condition x time interaction, F(2, 70) = 0.09, p > .05, ŋp = .01
(M = 1.37, SD = 1.85).
Children’s Independent Reminiscing Contributions
In addition to predictions that children in the reminiscing condition would come make more high elaborative utterances and more emotion references during shared reminiscing than would children in the control condition, it was also predicted that they would make more high elaborative utterances and emotion references during independent reminiscing with an experimenter. As experimenters provided children with open-ended prompts exclusively, high elaborative information statements (prompted or unprompted) and low elaborative repetitions were the only were the only utterances that children could make in response. Low elaborative 127 repetitions were extremely rare (M < 1.00), as was emotion and preference content of any kind (M < 1.00); thus, these variables were not analysed statistically.
There was a significant time main effect on children’s independent information
2 statements, F(1, 71) = 5.11, p < .01, ŋp = .07, but no significant condition main effect, F(1,
2 2 71) = 0.45, p > .05, ŋp < .01, or time x condition interaction, F(2, 70) = 0.47, p > .05, ŋp <
.01. As shown in Table 2, high elaborative utterances increased across time but were not affected by condition.
Children’s Emotion Knowledge
Children in the reminiscing condition were predicted to show higher emotion knowledge after training than were children in the control condition. To investigate, emotion knowledge was measured in two ways: using Denham’s Emotion Knowledge Task, and using an Emotion Cause Production Task. There was a significant main effect of time on children’s
2 Denham Emotion Knowledge Task scores, F(1, 71) = 23.58, p < .01, ŋp = .25, but no
2 significant main effect of condition, F(1, 71) = 0.71, p > .05, ŋp = .01, and no significant time
2 x condition interaction, F(2, 70) = 1.01, p > .05, ŋp = .01. Over time, children’s scores on the
Denham task increased (M = 40.21, SD = 4.47 at A1, M = 41.92, SD = 3.39 at A2, and M =
43.54, SD = 3.28 at A3). Importantly, scores on the Denham task at the 6-month follow-up seemed constrained. Of the 72 participants included in the LOCF method, 45% scored 45 or above out of 48 at A3 (M = 43.17; SD = 3.81). Results are even more striking when only the
44 participants who completed A3 are considered: in this sub-sample 65% scored 45 or above out of 48, and variability was low (M = 45.00; SD = 1.62). The possibility of ceiling effects should therefore be considered.
There was also a significant main effect of time for children’s emotion cause
2 knowledge scores, F(1, 71) = 7.21, p < .01, ŋp = .15. Whilst there was no significant main
2 effect of condition, F(1, 71) = 1.54, p > .05, ŋp = .04, effects were nonetheless modified by a 128
2 significant time x condition interaction, F(2, 70) = 4.70, p < .05, ŋp = .10. As shown in Figure
2, children in the reminiscing condition showed slightly poorer emotion cause knowledge at
A1 and A2 than did children in the control condition; however, they showed significantly greater emotion cause knowledge at A3.
Comparison of Results using the LOFC Method and the Final Sub-sample Only
In addition to the above analyses, which used the LOCF method to substitute missing values with non-continuers’ last observed scores (i.e. with a conservative assumption that no subsequent change has occurred), analyses were also conducted for the final sub-sample of 44 participants who completed all stages. The pattern of results for these 44 participants replicated the pattern of results using the LOCF method, with one exception. Using the LOCF method, a significant time x condition interaction was observed for children’s high elaborative utterances during shared recall, whereby children in the reminiscing condition, but not the control condition, increased their high elaborative utterances over time (p < .05). When considering the final sample of 44 participants only, the same general pattern of means was observed, however this interaction was only marginally significant, F(2, 42) = 2.97, p = .06,
2 ŋp = .08.
Supplementary Analysis of the Association between Mothers’ Reminiscing and Children’s
Independent Recall
Given that condition did not significantly impact children’s independent memory contributions, we thought it possible that the six month duration between training and follow- up was not sufficient time for children in the reminiscing condition to internalise the memory skills practised during high-elaborative reminiscing with their mothers. We therefore conducted a supplementary analysis to examine concurrent and time-lagged associations between mothers’ utterances during shared reminiscing and children’s independent recall, controlling for children’s age and language and age (note that the same pattern of results 129
Control Training 7 Reminiscing Training 6 n
5
4
3
2 Emotion Cause Productio Cause Emotion 1
0 Pre-training Post-training 6 month follow- (A1) (A2) up (A3) Assessment
Figure 2. Children’s emotion-cause knowledge as a function of time and training. 130
emerged whether condition was controlled for or ignored; thus, condition is ignored in the results reported here). Several significant associations emerged (see Table 11). First, mothers’ high elaborative utterances at A1 were associated with children’s independent memory contributions at A2, and mothers’ high elaborative utterances at every time-point (that is, A1,
A2, and A3) were each associated with children’s independent memory contributions at A3.
Second, mothers’ low elaborative utterances at A1 also predicted children’s independent memory contributions at A2 and A3; however, low elaborative utterances at A2 and A3 were not associated with children’s independent memory at any stage.
Table 11: Associations between Mothers’ Reminiscing and Children’s Independent Recall to an
Experimenter
Mothers' Reminiscing Utterances
High Elaborative Low Elaborative
A1 A2 A3 A1 A2 A3
Independent Recall at A1 .15 .07 .01 .10 .16 .11
Independent Recall at A2 .35** 0.18 .19 .25* .01 .05
Independent Recall at A3 .38** .38** .27* .40** .18 .09
*p < .05, **p < .01
Discussion
The aim of the present study was to train mothers to increase their high elaborative utterances and emotion references during reminiscing, and to observe the effects of training on children’s autobiographical memory and emotion knowledge. A powerful control training was used to disentangle the influence of the high elaborative style from sensitivity and attention variables. Consistent with the hypotheses, findings showed that mothers in the reminiscing condition increased both their high elaborative utterances and their emotion 131 references after training, and maintained the differences across the following six months.
Children in the reminiscing condition made more high elaborative utterances and emotion references during shared recall than did children in the control condition, and, after six months, showed superior emotion cause knowledge to children in the control condition.
That mothers can be trained to increase their use of a high elaborative style supports findings from Boland et al. (2003) and Peterson et al. (1999), who found that during discussions of ongoing and past events respectively, mothers increased their long-term use of the specific elements of the high elaborative style that they had been trained to use. Reese and
Newcombe’s (2007) findings are also largely supported. They found that training was effective in increasing mother’s use of open-ended questions and confirmations over a period of 25 months, from when children were 19 months of age. Unexpectedly, however, repetitions were also increased, due largely to the repetition of open-ended questions. In the present study mothers were encouraged not to repeat the same question, but, if a child was unresponsive, to respond to their own question with additional information. This explicit instruction appeared effective, in that high elaborative utterances targeted by the intervention increased significantly at both A2 and A3, but low elaborative utterances did not. That mothers can also be trained to increase their emotion references during reminiscing extends the findings of existing intervention training programs (e.g. Boland et al., 2003; Reese & Newcombe, 2007,
Peterson et al., 1999), by focusing on emotion rich content.
A large body of observational research demonstrates that children come to discuss past events, with respect to both style and content, in a way that is similar to that of their parents
(Hudson, 1993; Kuebli et al., 1995; Leichtman et al., 2000, Reese et al., 1993; Sales &
Fivush, 2005). The present study provides experimental evidence to support this research.
Mothers in the reminiscing condition not only made more high elaborative utterances and emotion references during shared reminiscing at A2 and A3 than did mothers in the control 132 condition; their children also made more high elaborative information statements and emotion references than did children in the control condition.
Notwithstanding the significant impact of reminiscing training on children’s contributions to shared recall with their mother, there were no differences between reminiscing and control children’s independent recall to an experimenter at either A2 or A3.
Thus, although reminiscing children had begun to use a similar reminiscing style and content as their parents during shared recall, they may not yet have fully internalised this style or content (which was undoubtedly new to at least some dyads). This possibility is supported by previous research. In both observational and experimental work it has been established that for children as young as age 3, high elaborative adult-child reminiscing about a particular event benefits children’s subsequent independent recall for that same event more greatly than does low elaborative or empty reminiscing (Conroy & Salmon, 2006; Leichtman et al., 2000;
McGuigan & Salmon, 2004), and in Study 2, high elaborative, emotion-oriented reminiscing enhanced children’s memory to a greater extent than did high elaborative reminiscing alone.
Nonetheless, it is not typically until age 5 or 6 that children fully internalise these representational skills, such that they are better able to independently encode and retrieve unshared events (e.g. Cleveland & Reese, 2005; Harley & Reese, 1999; see Reese &
Newcombe, 2007, for findings that this process might occur earlier in those children who show high self-awareness but not those who show low self-awareness). Whilst children in our sample were aged 4 to 5.5 years at the time of their final assessment, many of these children would be exposed to the high elaborative style for only 6 months during training, rather than from the time late in infancy at which parents first begin to reminisce with their children.
Should the high-elaborative style of mothers’ in the reminiscing condition be sustained for longer than the 6 months measured in this study, perhaps differences in children’s 133 independent recall for unshared events will emerge between children in the reminiscing and the control conditions.
Although no independent recall differences between the reminiscing and control conditions were observed, a small supplementary analysis - in which condition was ignored - suggested that mothers’ natural reminiscing style was associated with independent recall.
Mothers’ general talkativeness (that is, high and low elaborative utterances) at A1 was associated with children’s independent recall at A2 and A3; high elaborative utterances at A2 and A3 were associated with independent recall at A3. The findings of this supplementary analysis suggest two things. First, although reminiscing training was successful in enhancing the degree to which mothers elaborated and referred to emotions, strong individual differences remained across conditions. Second, that children did at this age show evidence of internalising natural elements of their mothers’ reminiscing style, that is, elements which they had likely been exposed to across childhood, provides further support for the argument above that perhaps children in the reminiscing condition simply needed more time for independent memory skills related to condition to emerge.
The present study was concerned not only with the development of autobiographical memory, but also with the development of emotion knowledge. As predicted, after six months, children in the reminiscing condition showed superior emotion-cause knowledge to children in the control condition. These findings provide support for the observational research of Study 1, in which emotion cause references and a high elaborative reminiscing style were each associated with emotion knowledge on Denham’s Emotion Knowledge Task
(see Laible, 2004b, Laible & Song, 2006, for findings that high elaborative reminiscing predicted emotion knowledge yet emotion-oriented reminiscing did not). Observational findings that mothers’ emotion references in ongoing contexts predict children’s emotion knowledge are extended (e.g. Denham et al., 1994; Dunn, et al., 1991). Like other discussion 134 of emotion, emotion-oriented reminiscing provides children with a linguistic representation of their emotional experiences (O’Kearney & Dadds, 2005). In addition, reminiscing may allow children to reflect upon and internalize this emotional information, including information about causes and consequences, whilst they are somewhat distanced from the original emotion and no longer in a state of emotional arousal (e.g. Denham & Burton, 2003; Fivush et al.,
2000). High elaborative reminiscing is not only more likely to enhance this process of reflection, but may, via the explicit discussion of memory, also enhance children’s more general understanding of mental representation (Pons et al., 2003; Wareham & Salmon,
2006). Further research is required to disentangle the high elaborative style from emotion content and experimentally determine whether style, content, or, as the above observational studies suggest, both, are the key mechanism or mechanisms by which emotion cause knowledge is facilitated.
Interestingly, although children in the reminiscing condition manifested superior emotion cause knowledge on the Emotion Cause Production Task, when compared to children in the control condition, there was no difference between conditions on Denham’s Emotion
Knowledge Task. There are two possible explanations for this finding. First, it may be that emotion cause knowledge in particular, and not emotion recognition or emotion situation knowledge (as measured by the Denham task), was targeted by the intervention. This explanation seems unlikely, however: Study 1 showed that a high elaborative style and emotion cause references during reminiscing each uniquely predict emotion knowledge as measured using the Denham Emotion Knowledge Task. Second, children may have been just outside the upper range of the Denham task at the final assessment. Nearly all children were aged above 4 years, with many above 5 years, at the final assessment (in comparison, the average age of children in Study 1 was 48 months). The Denham task, on the other hand, is typically used with children between the ages of 3 and 5 (e.g. Cutting & Dunn, 1999; Denham 135 et al., 1994). At the latter time-point approximately half the participants scored full or almost full marks (at least 45 out of 48), with low variability, suggesting a ceiling effect. In contrast, children may learn about emotion causes across the preschool years: past research has identified individual differences in both 2.5- to 3.5 and 5- to 6-year-old children (e.g.
Denham, 1998; Denham & Zoller, 1991; Hughes & Dunn, 1998; Pons et al., 2004; Wang et al., 2006).
As standardized language skill did not differ as a function of condition, it was unnecessary to control for language in other analyses. Thus, both shared recall and emotion cause knowledge were facilitated by high elaborative, emotion-oriented reminiscing over and above the impact of language. It is possible that language did not underpin this development as, by preschool, children already possess the language for both external interaction with others and internal mental representation. This may particularly be the case with the current sample, as the PLS results show that children in both conditions were on average more than one standard deviation above the mean for their age at all assessment points, and for both expressive and receptive language. The results suggest that by late in the preschool years, social interaction may surpass language development as a key contributor to development.
Further research is required to replicate these findings, and to more comprehensively chart the relative importance of each contributor across the early childhood years.
The success of the intervention in enhancing the shared recall and emotion cause knowledge of children in the reminiscing condition, relative to children in the control condition, may have important implications. Many researchers have suggested that more effort is needed to translate the findings from basic research into primary prevention programs aimed at facilitating child development (Izard et al., 2001). Via the facilitation of autobiographical memory and emotion cause knowledge, training mothers in high elaborative, emotion-oriented reminiscing may contribute to children’s development of self, 136 ability to guide current cognitive and behavioural responses, and effective social communication (Alea & Bluck, 2003; Cockroft & Reese, in press; Pillemer, 1992; Wang &
Conway, 2004), together with their pro-social behaviour, peer status, and resistance to internalising and externalising problems (Denham et al., 2003; Hughes & Dunn, 1998; Izard et al., 2001, Fine et al., 2003).
Notwithstanding the broad success of the intervention, however, issues of retention must be addressed. Similar to Peterson et al.’s (1999) reminiscing intervention in which only
14 of 20 dyads returned for assessment after training, many dyads also failed to complete the current study. As these participants did not differ from those who remained in the study at each time-point, there is no reason to suspect that they would be differentially affected by training. Furthermore, there was little difference in the pattern of results when using the
LOCF method to provide a conservative estimate of effects with a sample of all 72 participants who commenced training, and when analysing only the 44 participants who completed the study. Nonetheless, the number of participants who did not complete the program indicates potential limitations in the extent to which such a program may be useful: only those who complete the program can be expected to benefit from it.
Nock and Kazdin (2005) identify two reasons participants fail to attend or adhere to treatment in clinical intervention programs: motivation for treatment is not commensurate with treatment demands, or barriers to participation are experienced (e.g. access to transport).
Given that many mothers in the sample worked at least part time, the intensive and longitudinal nature of the assessments may therefore have contributed to the low retention rates. Mothers were recruited from across Sydney and some travelled up to two hours each way, with young children, to attend each of the seven sessions at the university. In addition, assessments were quite long at 90 minutes. Thus, each session could take up to 5.5 hours, door-to-door. To attempt to reduce the rate of participants who did not complete the program, 137 a less encompassing assessment could be trialled to reduce the demands on parents.
Furthermore, as in other interventions, families could be visited in their homes. Consistent with this suggestion, Reese and Newcombe (2007) visited participants in their homes over a period of almost two years, and had very few participants who did not complete their reminiscing training study. Finally, programs that enhance parental engagement and compliance in clinical populations, such as Nock and Kazdin’s (2005) ‘Participation
Enhancement Program’, could be integrated with reminiscing training. The Participation
Enhancement Program takes only 5-15 minutes in addition to normal parent-child therapy sessions. Parents are provided with regular information about the importance of participation, and are encouraged to develop plans for overcoming barriers to treatment. Participants who receive this brief intervention in addition to treatment for conduct disorders were shown to have greater treatment motivation, attendance, and adherence, according to both therapist and parent reports (Nock & Kazdin, 2005). The motivational aspect of this training may be particularly important with community samples where no problems (such as child conduct) have been identified, and where, therefore, parents’ motivation for participation may be less than if their child had an identifiable problem that the program aimed to treat.
In summary, after training, mothers who were instructed to engage in high elaborative, emotion-rich reminiscing not only increased their high elaborative utterances and emotion references during reminiscing, relative to mothers who engaged in child-led play, they also sustained these differences over a period of six months. Furthermore, children of mothers in the reminiscing condition increased their own high elaborative and emotional memory contributions during shared recall, relative to children of mothers in the control condition. Although children of mothers in the reminiscing condition did not yet show significantly better independent recall than children of mothers in the control condition, and did not differ in their understanding of some aspects of emotion, they did after 6 months 138 show superior emotion cause knowledge. Whilst retention strategies must be addressed, the present study nonetheless provides an avenue by which efforts can be made to improve children’s long-term cognitive and socio-emotional development.
139
CHAPTER 5: AN INTEGRATION OF FINDINGS, AND THEIR IMPLICATIONS FOR
CHILDREN’S COGNITIVE AND SOCIO-EMOTIONAL DEVELOPMENT
The overarching aim of this thesis was to understand better the ways in which high elaborative and emotion-rich reminiscing contributes to preschool-aged children’s development of autobiographical memory and emotion knowledge. To achieve this aim, three studies were conducted. The aim of Study 1 was to examine associations between parent-child reminiscing about emotional events (children were aged 3.5- to 5-years), children’s autobiographical memory, and, in an extension of previous studies, children’s emotion knowledge. A concurrent correlational design was employed. The aims of Study 2 were twofold. First, the study aimed to manipulate high elaborative reminiscing to include causal emotion talk, non-causal emotion talk, or no emotion content, and to measure younger (3- to
4- years) and older (5- to 6-years) children’s subsequent memory for a staged event. This experimental design was important, as it allowed causal inferences to be made. Second, the study aimed to examine the extent to which pre-existing emotion knowledge predicts the recall of emotional and non-emotional information. The aims of Study 3 were to train mothers to engage in high elaborative, emotion-oriented reminiscing about everyday events with their
3- to 5-year-old children, and to measure the effectiveness of this training in facilitating children’s development of autobiographical memory and emotion knowledge across 6 months. Mothers were randomly allocated to either a reminiscing condition or a control condition. That Study 3 also used an experimental design to manipulate reminiscing style and content allowed inferences to be made about a direction of effect from mothers reminiscing to children’s memory and emotion knowledge. Note that reminiscing is a language-based activity: therefore, to examine the specific influence of high elaborative, emotion-oriented 140 reminiscing on memory and emotion knowledge development, it was considered important to control for children’s general language skill across studies. Thus, language was measured in each study. Wherever it associated with the findings in any way, it was partialled out.
Findings of the Current Research
Study 1 replicated findings that children of high elaborative parents recall more information during shared recall (Fivush & Fromhoff, 1988; Hudson, 1990, 1993; Reese &
Brown, 2000; Reese & Fivush, 1993; Reese et al., 1993), and extended these findings into the emotional domain. High relative to low elaborative parents did not make more emotion attributions, references to emotion consequences, or overall emotion references to their 3.5- to
5-year-old children, but did more often discuss the causes of emotions. In turn, parents’ explanations of emotion causes and their use of an elaborative reminiscing style were each uniquely associated with children’s emotion knowledge, as assessed using Denham’s Emotion
Knowledge Task (Denham, 1986). Although Study 1 highlights the importance of the way in which parents discuss emotions, the sample size was small (N = 25) and, like much research investigating parent-child reminiscing, the concurrent correlational design meant that a direction of effect was unable to be established (e.g. Fivush & Fromhoff, 1988; Hudson,
1990, 1993; Leichtman et al., 2000; Reese & Brown, 2000; Reese & Fivush, 1993; Tessler &
Nelson, 1994). Studies 2 and 3 therefore used larger sample sizes and employed experimental methodology: Study 2 manipulated the content of adult-child (as opposed to parent-child) talk, whereas Study 3 randomly allocated mothers to one of two training conditions, reminiscing or control.
Study 2 provided causal evidence that younger (3- to 4-years) and older (5- to 6-years) children who reminisced with a high-elaborative experimenter about emotions, and, in particular, emotion causes, independently recalled more total emotional and non-emotional information about a staged event than did children who reminisced with the high-elaborative 141 experimenter about non-emotional aspects of the event. During free recall, interactions with age were observed: high elaborative reminiscing that included labelling of emotions and discussion of their causes was especially important for the younger children, for whom the non-causal discussion of emotions - labelling emotions and their expressions - produced the same minimal recall as did reminiscing about non-emotional aspects of the event. For older children, any reminiscing that included the discussion of emotions, causal or non-causal, was beneficial. In addition to these experimental findings, in which a direction of effect from adult reminiscing to child event representation was established, Study 2 also found that children’s pre-existing emotion cause knowledge, measured using the Emotion Cause Production Task
(e.g. Denham & Zoller, 1991; Hughes & Dunn, 1998; Wang et al., 2006), predicted recall of both emotional and non-emotional aspects of an event.
Drawing the findings of Studies 1 and 2 together, in Study 3 mothers were successfully taught to increase both their high elaborative utterances and their emotion references – including emotion cause references - during reminiscing with their 3- to 5-year- old children. Differences in reminiscing style and content between mothers in the reminiscing training condition and a group of mothers engaged in a control condition (child directed play) were maintained over 6 months. Furthermore, whilst ceiling effects in both conditions were observed on Denham’s Emotion Knowledge Task (1986), children of reminiscing mothers not only increased and maintained their own high elaborative memory utterances and emotion references during shared recall, relative to children of control mothers; they also showed better emotion knowledge on the Emotion Cause Production Task after 6 months (e.g.
Denham & Zoller, 1991; Hughes & Dunn, 1998; Wang et al., 2006). This study therefore provides clear experimental evidence of a direction of effect from parents’ reminiscing style and content to children’s emotion cause knowledge and recall. Whilst a bidirectional association between parents’ reminiscing and children’s development is likely, whereby 142 sensitive mothers also adjust their reminiscing style and content in response to their children’s emotion knowledge and recall abilities, the direction of effect does not run from child to parent only.
Together, the present studies highlight the influence on cognitive and socio-emotional development of the ways in which parents talk about everyday experiences with their young children. Findings suggest, firstly, that parent-child reminiscing is an important context for the development of autobiographical memory and emotion knowledge, and secondly, that individual differences in reminiscing style and content determine the extent of this development. Specifically, when children engaged in high-elaborative, emotion-oriented reminiscing with an experimenter, they subsequently recalled more emotional and non- emotional information about a specific staged event than did children engaged in non- emotional reminiscing (Study 2). Children whose parents demonstrated the use of a high- elaborative, emotion-oriented reminiscing style and content (whether naturally, as in Study 1, or as the result of training, as in Study 3) showed greater ongoing emotion knowledge and recall for discussed events. These findings are important for highlighting and building on those aspects of parent-child interaction that facilitate representational development.
Children’s Internalisation of Reminiscing Style and Content
It is a key tenet of the sociocultural theory that children’s cognitive development is facilitated by their participation in social interactions with more knowledgeable others; this is particularly so when the skill required to participate fully lies somewhere between children’s current level of development and their potential when facilitated by a capable adult (Nelson,
1996; Rogoff, 1990; see Vygotsky, 1978, for the ‘zone of proximal development’).
Importantly, internalisation is highlighted as the critical mechanism of development. By regularly participating in these language-based social interactions with others, it is theorised that children come to ‘internalise’ the shared processes and skills that are being scaffolded, 143 practiced and extended over time (Nelson, 1996; Rogoff, 1990; Vygotsky, 1978). In other words, via regular engagement of these skills in a social context, children come to transfer these skills from the social to the individual plane (Rogoff, 1990; Vygotsky, 1978).
For evidence that internalisation has occurred, children must be observed to engage in a skill or behaviour independently (see Rogoff, 1990). To examine children’s internalisation of their parents’ reminiscing content and style, therefore, children’s independent recall to an experimenter was examined in Studies 1 and 3 - in which parent-child as opposed to experimenter-child reminiscing was the primary focus. The only evidence for internalisation was found in Study 3, the reminiscing training. In this study, supplementary analyses revealed that, independent of condition, mothers’ high elaborative utterances during each stage of the experiment - pre-training, post-training, and during a six-month follow-up - were associated with children’s independent recall at the latter two stages. Nonetheless, there was no significant association between parents’ elaborative style and children’s independent recall in
Study 1 (in which only concurrent associations were measured), and no significant differences in the independent recall of children in the reminiscing and control conditions at any point in
Study 3.
There are several explanations for these mixed results. First, associations between parents’ reminiscing and children’s independent recall may be significant in Study 3 but not in Study 1 due to the relatively older age of the children in Study 3. Children in Study 1 averaged only 48 months of age at the single assessment point, with a standard deviation of 6 months, whereas many children in Study 3 were aged above 5 years at the six-month follow- up. Internalisation is a gradual and continuous process; thus, the children in Study 1 may still have been in the process of developing independent memory skills (Rogoff, 1990). This explanation is consistent with past research: whilst 4-year-old children of high relative to low elaborative parents can recall more about events that have been richly scaffolded during 144 shared reminiscing (e.g. Leichtman et al., 2000; Kuebli et al., 1995; see Nelson & Fivush,
2004, for review), they may not be able to independently encode and retrieve events not scaffolded by an adult until age 5 or 6 (e.g. Cleveland & Reese, 2005; Harley & Reese, 1999).
Reese and Newcombe (2007) show that parents who have been trained to use a high elaborative style have children who show superior independent recall, relative to controls, at as young as 44 months; however, this is only the case if those children also have high and not low levels of self-awareness. As self awareness was not measured in this thesis, it is possible that independent recall skills were in fact emerging earlier in children with high self- awareness, but these independent recall scores were constrained by the scores of those children low in self-awareness.
Second, the relatively short length of the intervention training study may explain why there was no difference in the independent recall of children in the reminiscing and control conditions in Study 3. Although many children were aged above 5 at the six-month follow-up, and analyses revealed that training was successful in enhancing the degree to which mothers in the reminiscing condition used high elaborative and emotion-rich utterances, many mothers may only have been using these elements of style and content for six months (i.e., since training). Children therefore had a quite short period of time in which to practice and internalize independent recall skills. At Reese and Newcombe’s (2007) final time-point, in comparison, in which differences between the independent recall of reminiscing and control children were observed (if children were also high in self-awareness), mothers had been trained for almost two years.
In summary, evidence of internalisation was found when children were at least 5 years of age on average, and when parents’ natural reminiscing style and not a relatively short-term training was taken into account. Should the length of the reminiscing training study (Study 3) 145 be extended, past research suggests that differences in independent recall between children in the reminiscing and control conditions may emerge.
The Influence of Reminiscing on Recall for a Single, Discussed Event
Whilst research examining the association between parents’ reminiscing and children’s independent recall is important for understanding children’s internalisation of autobiographical memory skills, research examining children’s recall for a single, known event following reminiscing with an experimenter can also inform our understanding of this process. According to sociocultural theory, the internalisation of memory and other representational skills begins first with the regular practice of these skills as they are scaffolded and extended by more knowledgeable adults during social exchanges (Nelson,
1996; Rogoff, 1990). Preschool-aged children typically come to internalise their parents’ reminiscing style and content, and not that of other adults, because at this age their parents are the foremost socialisers in their lives (e.g. Nelson, 1996; Nelson & Fivush, 2004).
Researchers can nonetheless simulate parents’ style or content and therefore investigate how the process of internalization might occur, and, more specifically, which characteristics of reminiscing when internalised over time will best facilitate children’s event representation skills. If exposure to a particular reminiscing style or content facilitates children’s encoding and subsequent independent recall for the event discussed, whether by virtue of eliciting a more complete memory report from the child during reminiscing, for example, or by highlighting particular content, then the internalisation of this style and content over time is likely to facilitate children’s independent encoding and recall for undiscussed events in much the same way. Children should not only have more richly encoded representations of the specific event being recalled, for example, but more retrieval cues for future episodes that are the same or similar; a greater understanding that reminiscing is important; skill in representing experiences in a structured, coherent narrative form; and a greater understanding 146 of the event information, such as emotions, temporal information, and other contextualising information, that it is important to attend to (e.g. Fivush, 1993; Fivush & Fromhoff, 1988;
Nelson, 1996; Nelson & Fivush, 2000; Peterson et al., 1999; Vygotsky, 1978).
A large body of observational concurrent and longitudinal research demonstrates that children come to discuss both the style and the content of past events in the same way as do their parents (e.g. Hudson, 1993; Kuebli et al., 1995; Leichtman et al., 2000, Reese, Haden, &
Fivush, 1993; Sales & Fivush, 2005). In Studies 1 and 3, these findings were replicated: in the case of the latter study, the intervention training study, causal support is provided. Parents in the intervention study who were taught to make more high elaborative utterances and emotion references during shared reminiscing had children who also came to make more high elaborative memory contributions and emotion references during shared recall. These findings do not provide any evidence of internalisation per se, as children may simply be responding to their parents’ input. High elaborative parents ask a greater frequency of open-ended, emotion- oriented memory prompts, whereas low elaborative parents typically provide closed, repetitive, or scant prompting. Nonetheless, findings show that by participating in a shared social interaction scaffolded by their parents, children are able to practise reminiscing using particular style or content in the same way as do their parents. This practice is theorised to be an important precursor to internalisation (Rogoff, 1990; Vygotsky, 1978).
Contrary to past research comparing high elaborative and minimal styles of reminiscing (e.g. Conroy & Salmon, 2006; McGuigan & Salmon, 2004), and to the observational and causal findings of Studies 1 and 3 respectively, findings of Study 2 - in which children reminisced with an experimenter about a staged zoo event in one of four ways
(emotion cause, emotion expression, no emotion, minimal) - showed no difference in recall between those in the no emotion condition who participated in high elaborative reminiscing about non-emotional aspects of the zoo, and those in the minimal condition who participated 147 in low elaborative, ‘empty’ reminiscing about non-emotional aspects of the zoo. At face value, this finding could be interpreted as suggesting that a high elaborative style does not affect children’s representation of the specific event or, by inference, their more general autobiographical memory development across time. Importantly, however, as part of the experimental design all children in Study 2 were provided with high elaborative, emotion rich narration of the event as it occurred. Children therefore had at least one opportunity to encode the event in a rich and full manner, with likely benefits to memory (see Boland, Haden, &
Ornstein, 2003; Ornstein & Haden, 2006; Tessler & Nelson, 1994).
It is unclear whether during event talk or past event talk has a greater effect on memory representations. Spaced rather than massed exposures have been shown in learning studies to be optimal for encoding and retrieval (Bahrick, 2000; Dempster, 1996), and there are strong theoretical reasons to suggest that past event talk might also have a greater influence on memory than might during-event talk: past event talk serves to reinstate the event, such that children are granted two spaced exposures overall (the event itself and reminiscing; McGuigan & Salmon, 2004). Furthermore, the very small body of existing research addressing this question shows some support for this theory, yet replication is necessary (McGuigan & Salmon, 2004; but see Ornstein & Haden, 2006). Irrespective of the relative benefits of during- compared to post-event talk, however, findings suggest that both during- and post-event talk enhance recall of an event to a greater extent than do no talk or
‘empty’ talk (e.g. McGuigan & Salmon, 2004; Ornstein & Haden, 2006; see also Conroy &
Salmon, 2006). Furthermore, despite the fact that all children in Study 2 had participated in emotion-rich dialogue as the event occurred, differences in recall emerged between children who reminisced about emotions and those who did not: high elaborative, emotion-oriented reminiscing facilitated recall of all information, including emotional information, to a greater extent than did any other style-content combination. Thus, as for shared parent-child 148 reminiscing (in Studies 1 and 3), there was evidence that children were using the information provided during reminiscing to structure their recall of the event discussed.
Associations between Parents’ Reminiscing Style and Content
The relationship between a high elaborative reminiscing style and emotional reminiscing has not been clearly established in previous research, with some research showing that high elaborative mothers also engage in more discussion of emotions (Laible, 2004b;
Thompson, Laible, & Ontai, 2003), and others suggesting no association (Laible, 2004a;
Laible & Song, 2006). Study 1 attempted to clarify these findings by examining not only overall emotion references, but also different types of emotion references (i.e. attributions, causes, or consequences). Parents who used a high elaborative style were subsequently found to discuss and explain emotion causes - but not other aspects of emotion - with greater frequency than were low elaborative parents. That is, high and low elaborative parents used a similar number of emotion references for all categories except emotion causes, but differed with respect to emotion cause references only.
Interestingly, key sociocultural researchers such as Fivush, Reese, and Haden have suggested that whereas high elaborative parents may see reminiscing conversations as a social experience and aim to collaboratively recreate shared past events, low elaborative parents seem instead to approach reminiscing conversations as tests of their children’s memory abilities and aim to promote maximum independent memory performance in their children
(e.g. Reese et al., 1993). It is therefore possible, given that parents in Study 1 were asked to talk about events in which happiness, sadness, anger, and fear featured, that low elaborative parents viewed the task as a test of memory for emotions and therefore increased their overall rate of emotions references in an effort to prompt emotional responses from their children.
Thus, although no significant differences in emotion references were found between high and low elaborative parents in the current study, the possibility that low relative to high 149 elaborative parents refer to emotions less frequently in everyday reminiscing conversations not focused specifically on emotions should not be discarded.
Whilst no study has directly examined the possibility that some parents instructed to discuss emotional events might adjust the content of their reminiscing accordingly, Cleveland,
Reese, and Grolnick (2007) provide recent experimental support for the notion that the instructions given to parents may influence parents’ motivation and behaviour during reminiscing. Parents were assigned to either an ‘outcome-oriented’ condition or a ‘process- oriented’ condition, and were asked to discuss with their child (mean age = 46 months) a staged, standardised event that the children had participated in earlier that day. Outcome- oriented parents were told that their child would later be tested on their recall of the event, whereas process-oriented parents were told that the child’s personal perspective was of interest. Although outcome-oriented and process-oriented parents did not differ in the degree to which they used a high-elaborative style, outcome-oriented parents were found to be more controlling than were process-oriented parents (in turn, controlling relative to non-controlling parents had children who were less engaged in an independent memory assessment 2 weeks later).
That the high elaborative reminiscing style was still associated with the way in which emotions were discussed in Study 1, despite the potential prompting that the instructions to discuss happy, sad, anger-oriented, and scary events may have provided, is particularly salient. Given that the experiences discussed during reminiscing are typically both personally significant and emotionally meaningful, causally relating the emotion states that characterise an experience to other information about the experience may allow high elaborative parents to provide a full, embellished account of their children’s experiences (Cervantes & Callanan,
1998; Fivush, 1993; Fivush et al., 2003; Fivush & Fromhoff, 1988). 150
The Importance of Emotion Causes for Autobiographical Memory and Emotion Knowledge
As discussed in Chapter 1 and throughout this thesis, the inclusion of emotional content in past event narratives has received particular attention in recent years. From a theoretical perspective, not only may one’s self concept shaped by the events of one’s past; autobiographical recall with others also serves a range of fundamental social functions (Alea
& Bluck, 2003; Fivush et al., 2003; Pasupathi, 2003; Wang, 2004; Welch-Ross et al., 1999).
Autobiographical memory is therefore concerned primarily with episodes that have personal meaning, and it is the emotional aspects of these episodes make them worth of remembering to both ourselves and others (Fivush, 1993; Fivush et al., 2003). Nonetheless, not all emotion discussions are equal. In the present research, several themes emerged relating to high elaborative reminiscing about emotion causes, above and beyond other emotion references.
Not only were emotion causes found in Study 1 to be discussed more frequently by high elaborative mothers than low elaborative mothers, they also played a particularly important role in children’s emotion knowledge and autobiographical recall across studies.
In Study 1, parents’ discussion of four emotion-oriented past events with their children was uniquely associated with children’s emotion knowledge in two ways. First, children aged
3.5 to 5 years whose parents used a high elaborative reminiscing style to reminisce showed superior emotion knowledge relative to their aged peers, even once emotion references were statistically controlled. Second, parents’ high elaborative style and emotion cause references – but not other emotion references - were each uniquely associated with emotion knowledge on the Denham Emotion Knowledge Task (1986). This finding was consistent with findings related to ongoing events and activities: although emotion talk in a variety of contexts predicts children’s concurrent and future emotion knowledge (e.g. Denham et al., 1997; Denham et al.,
1994; Dunn et al., 1991), this is particularly the case for emotion cause talk (Brown & Dunn,
1996; Garner et al., 1997; Martin & Green, 2005). 151
Study 3 provided tentative causal support for the findings of both Study 1 and previous research, showing that emotion causes are important. A training program was established in which mothers in the reminiscing condition were taught to reminisce using a high elaborative style and emotion-rich content. Mothers in the reminiscing condition subsequently increased both their high elaborative utterances and their use not just of emotion attributions, but also causes, behaviours, outcomes.
That mothers in Study 3 increased their use of many different types of emotion references after training suggests that they were not simply referring once to an emotion and then moving on (as would be consistent with an increase in emotion attributions, or labels, only: for example, “Remember when you fell off the slide? You were sad. And what colour was the slide?”). Rather, they appear to have reconstructed the ‘story’ of the emotion in as complete a manner as possible: describing how the emotion manifested itself in behaviour, why it occurred, and what happened afterwards as a result of the emotion (e.g. “Remember when we went to the park? Something happened that made you sad. What made you sad?
…Right, you fell off the slide! You ran over to me with tears pouring down your face, and I gave you a biiig hug. How did you feel then?”). These increases appeared to be specific to training, as mothers in the reminiscing condition showed no increase in their use of preference terms such as ‘like’ and ‘enjoy’ that were not directly targeted in the training video or during training practice sessions, and thus used these terms at the same rate as did mothers in the control condition. The extended discussion of emotions by reminiscing mothers is more likely to facilitate development than is the mere mention of emotion labels (or preferences); causal aspects of the discussion may be a critical mechanism by which the development of representational skills such as autobiographical memory skills and emotion knowledge skills occurs. Consistent with this possibility, after 6 months children of reminiscing mothers, relative to children of control mothers, showed superior shared recall and superior 152 performance on an Emotion Cause Production Task (e.g. Denham & Zoller, 1991; Hughes &
Dunn, 1998; Wang et al., 2006). Although emotion recognition and emotion situation knowledge, as measured using the Denham task, did not also develop to a greater extent in the reminiscing condition than in the control condition over 6 months, this knowledge is typically acquired at an earlier age than is emotion-cause knowledge (e.g. Denham, 1986,1998; Pons,
Harris, & de Rosnay, 2004) and, given the high scores attained by children in each it seemed likely that ceiling effects had been reached. These findings suggest that high elaborative, emotion-oriented reminiscing with a causal focus is an effective tool with which to enhance at least some aspects of children’s emotional development.
Emotion causes also emerged as an important component in children’s autobiographical recall. In Study 2, children participated with an experimenter in one of four predetermined ways of reminiscing about a staged zoo event. This experimental methodology enabled the effect of reminiscing content and style to be teased apart and style controlled, with all but the minimal (control) condition discussing the event in a high-elaborative style.
High elaborative, emotion-oriented reminiscing of any kind aided children’s recall of the emotional and non-emotional aspects of an emotionally salient event more greatly than did high elaborative reminiscing that did not focus on emotions.
Given that emotions are personally meaningful elements of an event, emotion-oriented reminiscing of any nature may provide not only specific reinstatement cues but also an overall event goal and coherent “story” (Bird & Reese, 2006; Brown & Craik, 2000; Fivush et al.,
2003; Trabasso & Stein, 1997). This meaningful framework may facilitate children’s integration and processing of both emotional and non-emotional event components (Laney et al., 2004; Liwag & Stein, 1995; Wang et al., 2006). Consistent with this possibility, Liwag and Stein (1995) found that preschool children instructed to use emotion-oriented retrieval cues recalled more about an emotional experience, with a greater degree of structure, than did 153 those who used non-emotional cues. It is also possible that emotion-oriented discussion of any kind simply provides a more interesting and attentionally engaging reminiscing experience, which in turn may contribute to both encoding and retrieval (Compton et al., 2003; McDaniel et al., 2000). Notwithstanding the benefits of both emotion-cause and emotion-expression reminiscing for recall, however, this effect was stronger for emotion-cause reminiscing. Note that the importance of causal reminiscing may depend somewhat on age: only in the emotion- cause condition did younger children aged 3- to 4-years freely recall more than in the minimal condition, whereas older children aged 5- to 6-years freely recalled just as much in the emotion-expression condition. For both age groups, however, emotion cause talk benefited total recall more greatly than any other condition.
Emotion cause explanations do not simply tell children that an emotion state has occurred; they provide children with explicit and socially relevant information about why these various emotion states occur, what situational factors might predict them, and both how and why they are related to non-emotional event actions, objects, and descriptions (Cervantes
& Callanan, 1998). By nature, emotion cause explanations may therefore contribute to a stronger, more logical structure than do emotion references alone (Cervantes & Callanan,
1998). Children’s internalisation of the rich emotional information contained within these extended, logically structured discussions is the most likely mechanism by which emotion- cause discussions contribute to children’s rapid emotion knowledge development during preschool. Over and above emotions, furthermore, much research shows that a strong degree of connectedness between the components of an event produces better recall: that emotion- cause reminiscing provides a strong logical structure and not arbitrary connections between event components is significant in explaining how such reminiscing also facilitates autobiographical recall (Bauer & Mandler, 1989; Conroy & Salmon, 2006). 154
In the absence of a cause-only condition, in which emotions are not discussed, Study 2 was unable to clarify whether causal talk of any kind would benefit children’s memory to the same degree as did emotion-cause talk, or whether it was the discussion of emotion causes in particular that was beneficial. Clarifying this exact mechanism of effectiveness provides good opportunity for future research. Furthermore, it was unclear why the discussion of emotion causes, over and above other emotion discussion, was more important for younger (3- to 4- years) than older (5- to 6-years) children’s free recall of the emotional aspects of the staged zoo event in Study 2. It is perhaps the case that older children, having most likely internalised memory skills to at least some degree, are able to contribute to any emotion-oriented scaffolding provided and build on this scaffolding to internally construct a causal, emotion- oriented structure for the event. Younger children, however, may not yet have internalised these memory skills to the same degree and therefore rely on the external scaffolding provided by their conversational partner when the task is open ended – as is the case in free recall (Wenner, 2004; see also Russell, 1990).
Theoretical Implications for the Sociocultural Theory and the Development of Emotion
Knowledge
Implications for the Sociocultural Theory
Influenced by Vygotsky (1978), sociocultural theorists propose that autobiographical memory development occurs within social exchanges between parent and child (e.g. Nelson,
1993a, 1993b; Reese & Fivush, 1993; see Chapter 1). The present research provides support for the sociocultural theory and for Vygotsky’s seminal notion of cognitive development
(1962, 1978; Rogoff, 1984; 1990). Moreover, the present research also broadens this cognitive research into the socio-emotional domain and demonstrates that emotion knowledge also develops via language-based interactions. Vygotsky suggests that by regularly participating in language-based social exchanges with more knowledgeable others, children 155 come to internalise the skills that are being scaffolded, practiced and extended (Rogoff, 1990).
Whilst children’s independent recall skills for undiscussed events had not yet been fully internalised in Study 1, shared recall in both Study 1 and Study 3 was enhanced by parents’ high elaborative, emotion-rich reminiscing. Furthermore, in both Studies 1 and 3, children of high elaborative, emotion-rich parents showed superior emotion knowledge to other children.
Study 2 differed from Studies 1 and 3 in that experimenter-child and not parent-child reminiscing was the focus; furthermore, emotions discussed were of others (the zoo animals) and not the child’s own. Like Studies 1 and 3, however, Study 2 still presented evidence that children’s event representations were facilitated via social exchange. Using an experimental methodology, it was established that emotion oriented reminiscing, and particularly emotion- cause reminiscing, best facilitates children’s subsequent independent recall for a single discussed event (and, by inference, more general autobiographical memory development over time).
To date, research conducted within the sociocultural framework has focused primarily on associations between parents’ and children’s narrative style and content, particularly during reminiscing, and the influence these have on autobiographical memory development
(see Reese, 1995; Reese & Cleveland, in press, for recent exceptions). That emotion knowledge was also facilitated by emotion content and, in particular, a high elaborative style, therefore has important implications. Specifically, the findings suggest that it is not just autobiographical memory that is facilitated by reminiscing, but rather, may be representational skills in general. This possibility is consistent with the suggestions of both
Pons and colleagues and Thompson and colleagues, who each state that when the object of discussion (the past experience) is not present, children may implicitly learn about the nature of representation merely by participating in discussion with their parents (Pons et al., 2003;
Thompson et al., 2003). 156
Importantly, it should be highlighted that interactions in the present thesis were differentially effective in promoting children’s autobiographical memory and emotion knowledge. Vygotsky (1978) states that interactions in which development is promoted within the zone of proximal development (Vygotsky, 1978), that is, with scaffolding appropriate to the child’s skills (relative to those with less appropriate scaffolding), best facilitate the practice and internalization of further socio-cognitive skills. The present research helps clarify which interactions can best be considered to provide this scaffolding. Specifically, it was via high elaborative or emotion-rich reminiscing – particularly within a causal framework - that parents were able to build on the information that their preschool-aged children contributed, encourage children to extend their recall skills, and provide children with an emotional framework in which the importance of the events recalled is highlighted. In sum, therefore, parents who engaged in high elaborative, emotion rich reminiscing may guide their children in how best to represent both their experiences and their understanding of emotions in an organised and coherent manner (Nelson, 1996; Nelson & Fivush, 2004; Fivush, 1991).
Implications for the Development of Emotion Knowledge
According to the emotion coaching hypothesis, parents’ discussion and explanation of emotion directly facilitates their emotion knowledge development (Denham, 1998; Denham et al., 1994). The emotion coaching hypothesis is thus broadly consistent with the central premise of sociocultural theory in that both advocate the role of language-based parent-child interaction in children’s cognitive and representational development. Whilst the emotion coaching hypothesis focuses on the role of the parent, however, the sociocultural theory suggests that both parents’ and children’s roles in their language-based interactions are central to development (Fivush et al., 2006; Symons, 2004). A sizable body of literature supports an association between parents’ discussion of emotion with their children, particularly emotion causes, and children’s emotion knowledge: contexts in which this association have been 157 observed include family interactions, shared book-reading, and discussions of ongoing events
(Brown & Dunn, 1996; Denham et al., 1994; Dunn et al., 1991; Gardner et al., 1997; Martin
& Green, 2005). The present thesis supports these findings, extending them to a reminiscing context and highlighting the importance of the way in which emotions are discussed. In Study
1, parents’ emotion cause references during reminiscing were associated with children’s emotion knowledge, yet total emotion references (also including emotion attributions and consequences) were not. Furthermore, a high elaborative reminiscing style was also uniquely associated (also see Laible, 2004a, 2004b). Study 3 provided causal support for both the emotion coaching hypothesis and sociocultural theory, demonstrating that mothers can be taught to use high-elaborative, emotion-rich reminiscing, and that their children subsequently show superior emotion cause knowledge when compared to the children of control mothers.
There are several explanations for the role that reminiscing may play, relative to other contexts, in facilitating emotion knowledge development. Theorists highlight that fact that reminiscing typically occurs when children are no longer in a state of emotional arousal, which might otherwise serve to distract both adult and child from reflecting upon emotional information (Bretherton et al., 1986; Denham & Burton, 2003; Fivush et al., 2000).
Furthermore, reminiscing may allow children the opportunity to learn about the nature of representation, including representation of emotion (Pons et al., 2003; Thompson et al., 2003).
High elaborative reminiscing may extend the exchange (and, therefore, the process of reflection) more greatly than might low elaborative reminiscing, such that children are given greater opportunity to internalize representational information. To investigate the most effective context in which to enhance emotional development, future research could build on previous studies by experimentally manipulating and comparing the style and emotional content of during- versus post-event talk, and comparing emotion knowledge development over time. 158
To further investigate the development of emotion knowledge in a reminiscing context, associations between recall and emotion cause knowledge were examined in Study 2.
As predicted, emotion knowledge was significantly associated with children’s recall of the staged, emotionally salient “zoo” event (theory of mind, however, was not related to recall).
These findings mirror those of Wang et al. (2006), who also found a significant association between children’s (aged 2.5- to 3.5-years) emotion knowledge and their independent reports of a past experience. Wang et al. (2006) suggest that emotion knowledge enables children to interpret the emotional meaning of an experience and to perceive its personal relevance, with consequent positive implications for its memorability. In other words, there is a direction of effect from emotion knowledge to recall.
A second explanation for the association between emotion knowledge and recall is that, rather than (or, most likely, as well as) a direction of effect from emotion knowledge to recall, both may be associated simply because they are each representational skills. A child with good representational ability may express this ability in multiple ways; one who is good at recall will also understand emotions well. If this were the case, however, language would be expected to facilitate the association between the two. Vygotsky describes language as a tool of representation (1978). Language is not only externally related to representational skill, for example, by allowing parent-child interactions in which development is facilitated; it is also internally related in that, according to Vygotsky, it constitutes the representation itself
(Bosacki & Moore, 2004; Carpendale et al., in press; Rogoff, 1990; Vygotsky, 1978).
Nonetheless, language ability did not appear to play any role in the association between emotion knowledge and recall. These findings indicate that a general representational ability was not sufficient to explain the association between the two; rather, a direction of effect from emotion knowledge to recall (and, possibly, a bidirectional relationship) should be considered in future work. 159
Applied Implications of the Present Research
Studies 1 and 2 of this thesis are important for showing that high elaborative and emotion-rich reminiscing facilitates children’s recall of the event discussed, and is also associated with children’s emotion knowledge. In turn, Study 3 aimed to facilitate children’s memory and emotion knowledge skills using a reminiscing training intervention. This intervention was expected to be particularly effective in facilitating development, as, in contrast to both explanation and demonstration (which do not actively involve the child), parents and other adults were able to scaffold their child’s skills as they develop (Rogoff,
1984, 1990; Wersch & Stone, 1979). Past interventions show that parents can be trained to adopt the specific elements of a high elaborative style that they are taught (Boland et al.,
2003; Peterson et al., 1999; Reese & Newcombe, 2007). Not only does Study 3 replicate these results, albeit with a small final sample, but findings that mothers can also be successfully trained to increase their emotion references during reminiscing also extend the findings of these existing intervention training programs by suggesting that both style and content can be taught to and learnt by mothers.
Reminiscing training may represent an effective way of enhancing children’s cognitive and socio-emotional development: results of Study 3 showed that after 6 months, children in the reminiscing condition produced more memory elaborations and emotion references during shared recall, and also showed superior emotion-cause knowledge, when compared to children in the control condition. Of course, there are still issues to address: in the current training, for example, emotion content and high elaboration are confounded. Thus, it is unclear whether high elaboration, emotion content, or, perhaps most likely, both, contribute to each outcome variable. A manipulation of reminiscing training that teases these two elements apart and directly compares them would clarify this issue. Although replication of this study with a larger sample, more precise manipulation of reminiscing training content, 160 and a more comprehensive battery of emotion knowledge measurements is necessary before firm conclusions can be drawn, these early findings are promising. Specifically, for those children in the reminiscing condition who completed the study, these findings may have important implications for current and future functioning.
As highlighted in Chapter 1 and Chapter 4, autobiographical memory is theorised to contribute to a coherent ‘autobiographical life story’, to directive and problem solving functions (whereby successful past deeds can be drawn upon as a source of guidance), and, when shared with others, to facilitate social functions (Alea & Bluck, 2003; Bluck, 2003;
Cockroft & Reese, in press; Fivush et al., 2003; Pasupathi, 2003; Pillemer, 2003; Webster &
Cappeliez, 1993). Emotion-rich reminiscing is also important, as by definition it provides children with a linguistic representation of their emotional experiences. Within a secure attachment relationship, it is argued that this linguistic representation may allow children to report any emotional difficulties they have to others (see Wareham & Salmon, 2006, for review). Finally, research shows that the growth of emotion knowledge is essential to the development of emotion regulation, and also allows children to predict others’ emotions: children who show better emotion knowledge also show more empathy and prosocial behaviour, and have a positive peer status (Bridges et al., 2004; Denham et al., 1994; Denham et al., 2003; Hughes & Dunn, 1998; Izard et al., 2001; Kuebli et al., 1995; Thompson et al.,
2003).
Importantly, reminiscing training may not only benefit the development of autobiographical memory and emotion cause knowledge in a normal population, it may also be the first step towards enhancing the resilience of ‘at risk’ children. Research shows that interactions between mothers and their ‘at risk’ children tend to be somewhat impoverished when compared to interactions between control dyads, with potential implications for children’s development (e.g. Shipman & Zemen, 1999; Suveg et al., 2005; see Cicchetti & 161
Lynch, 1995, for review). In turn, clinical and ‘at risk’ children and adolescents often show poorer autobiographical memory and emotion representation skills – including the use and specificity of emotion language, and emotion knowledge - than do community samples (e.g.
O’Kearney & Dadds, 2005; Shipman & Zemen, 1999; Southam-Gerow & Kendall, 2002;
Vrielynck, Deplus, & Philippot, 2007). Given the association between reminiscing quality and children’s development in community samples, high elaborative, emotion-oriented reminiscing training may also increase ‘at risk’ children’s resistance to cognitive and emotional problems.
To implement reminiscing training in both the community and in clinical or ‘at risk’ groups, however, better retention strategies must be implemented. The degree to which the intervention is generalisable to the population was limited in Study 3 by a large number of non-continuing participants. The high proportion of non-continuing participants in Study 3 is not unprecedented for a reminiscing training study (see Peterson et al., 1999), and did not appear to have a significant bearing on the outcome of the intervention: nonetheless, participants who fail to complete the program cannot be expected to benefit from it. Issues of retention may be particularly important for clinical or ‘at risk’ populations where child difficulties have already been identified, for two reasons. First, interventions aimed at clinical populations typically experience high rates of participants who do not complete the program
(Nock & Kazdin, 2005). Second, the benefits of a reminiscing intervention in which autobiographical memory and emotion knowledge are facilitated may be greater for clinical or
‘at risk’ populations than for community populations: any disruption to emotional, social, or cognitive functioning at one developmental stage has significant repercussions for current and future functioning (Cicchetti & Lynch, 1995; Newman et al., 1996). To maximise the number of participants who complete reminiscing training, Chapter 4 suggests that it may be important to reduce participation demands on participants (p.134). Thus, assessment sessions 162 should be reduced in length, and home visits for all sessions (assessment and training) should be considered (see Reese & Newcombe, 2007). Finally, programs that specifically target intervention compliance, such as Nock and Kazdin’s (2005) ‘Participation Enhancement
Program’, could be integrated with reminiscing training (Nock & Kazdin, 2005).
Methodological Issues
The Coding of Close-Ended Questions
It should be noted that throughout this thesis, close-ended questions were considered low elaborative. Close-ended questions may be used by parents of very young children as a developmentally sensitive way to entice the child into a conversation to which he or she may be less capable of providing the answer to an open-ended question (Fivush et al., 2006;
Hudson, 1990). By the preschool years, however, open-ended questions are a more effective means of engaging children in the process of reminiscing (Farrant & Reese, 2000; Fivush et al., 2006; Haden et al., 2006). As discussed in Chapter 1, coding regarding close-ended questions is mixed: although many researchers code such utterances as high elaborative and code only repetitions as low elaborative (e.g. Farrant & Reese, 2000; Leichtman et al., 2000;
Reese & Brown, 2000; Wang & Fivush, 2005), citing the possible information that close- ended questions can add to a conversation, others code both close-ended questions and repetitions as low-elaborative (e.g. Laible, 2004a; Laible & Song, 2006). Given these differences, and in order to ensure that the decision to code close-ended questions as low elaborative was appropriate, individual elements of each reminiscing style were also examined. The results from these analyses revealed a good fit between close-ended questions and the low elaborative reminiscing category. In Studies 1 and 3, where parent-child reminiscing was examined, results for close-ended questions reflected those of other low elaborative components - namely, repetitions - and of the low elaborative category as a whole.
High elaborative elements of reminiscing were both associated with and causally related to 163 the outcome measures of shared recall and emotion knowledge, whereas low elaborative elements, including close-ended questions, were not. Although these results cannot be generalized to children at different developmental stages (e.g. younger children or those with developmental delays), they do indicate that coding was appropriate for the preschoolers targeted by the research in this thesis. Importantly, coding yes-no questions as low elaborative did not influence the conclusions that were drawn in either Study 1 or Study 3.
The Coding of High Elaborative Style: Proportions or Totals?
In Studies 1 and 3, in which parent-child reminiscing was examined, both elaborative style as a proportion (high elaborative utterances / high + low elaborative utterances), and total high elaborative utterances were calculated. Each method is consistent with previous research: in Reese and Newcombe’s (2007) recent intervention, for example, total high elaborative utterances and total low elaborative utterances were considered separately (see also Farrant & Reese, 2000; Wang & Fivush, 2005), whereas in previous sociocultural studies, elaborative style as a proportion has also been used (e.g. Haden & Fivush, 1996; see
Low & Durkin, 2001, for a scheme coding the number of high elaborative utterances made per low elaborative utterance). Total utterances provide the only way of capturing the overall quantity of information and high elaborative prompting a child is exposed to, whereas proportions correct for general talkativeness and may better account for the potentially negative impact on memory of repetitions (Reese & Fivush, 1993; Reese & Newcombe,
2007); each, therefore, presents different but potentially relevant information. Irrespective of these differences, studies using each method have found similar results, namely, that children of high elaborative parents come to internalise the high elaborative style across time, and therefore produce superior shared and independent recall to children of low elaborative parents (e.g. Fivush & Fromhoff, 1988; Hudson, 1993; Leichtman et al., 2000; Reese &
Fivush, 1993). 164
For Study 1, in which parents were asked to discuss emotional past events with their children, large to very large associations were found between parents’ and children’s elaborative style as a proportion of overall utterances, as well as between parents’ elaborative style, parents’ emotion cause utterances, and children’s emotion knowledge. Moderate associations were found between parents’ elaborative style as a proportion of overall utterances and children’s total recall. A similar pattern emerged when parents’ total elaborations were measured rather than their elaborative style as a proportion of overall utterances; however associations were not as strong - most were moderate to large and not large to very large. Nonetheless, the results of this study suggest that similar conclusions can be drawn irrespective of the method used to determine elaboration. In Study 3, results were slightly different. Specifically, it was found that proportions of elaboration were extremely high at all time-points and in both reminiscing and control conditions (perhaps due to the highly educated nature of those mothers who participated). Thus, the extent to which elaboration as a proportion of total utterances could be enhanced via training was greatly constrained. Total elaborations, which are uncapped, were instead used when comparing the extent to which elaboration is increased post training in the reminiscing and the control conditions. As each method of measuring elaboration is able to provide somewhat different information (as seen in Study 3), yet each is used interchangeably in the literature, both methods should continue to be used simultaneously in order to best capture the important influences on young children’s development across different ages and in different contexts.
The Assessment of Emotion Knowledge
Given the varying ages of the preschoolers examined in this thesis, the speed with which preschoolers’ emotion knowledge develops, and the relative dearth of research examining parent-child reminiscing and child emotion knowledge together, within a sociocultural context, two different methods of assessing emotion knowledge were used. 165
Denham’s Emotion Knowledge Task was used in Studies 1 and 3 (see Denham, 1986). This task measures emotion recognition and emotion situation knowledge. It has been used frequently with children aged up to 5 years, and is therefore the most obvious choice of measurement tool in that it will allow for the comparison of the present results with previous research (e.g. Cutting & Dunn, 1999; Denham et al., 1994; Laible, 2004a, 2004b; although see Hughes & Dunn, 1998, for evidence of ceiling effects at 54 months). Furthermore, an
Emotion Cause Production Task (e.g. Denham & Zoller, 1991; Hughes & Dunn, 1998; Wang et al., 2006) was used in Studies 2 and 3. This task taps a different type of emotion knowledge
(the understanding of emotion causes) than does the Denham Task, and shows greater flexibility in the ages for which it is suitable. Specifically, it has been used with cohorts ranging from 2- to 6-years-old (e.g. Denham, 1998; Denham & Zoller, 1991; Hughes &
Dunn, 1998; Pons, Harris, & de Rosnay, 2004; Wang et al., 2006). Whilst the task is limited in that only reliability and not validity information is available, researchers who have used the task in conjunction with other measures of emotion knowledge, including both the Denham
Task (1986) and an Emotion Judgement Task (Wang, 2003), indicate good agreement between the different measures (see Hughes & Dunn, 1998; Wang et al., 2006). An examination of high-elaborative, emotion-oriented reminiscing using both the Denham task and the Emotion Cause Production task was therefore considered the most comprehensive and age-sensitive approach.
In Study 1, parents and children discussed four emotional past events. Children were then administered the Denham’s Emotion Knowledge Task. Children’s emotion knowledge was associated with parents’ high elaborative reminiscing style and emotion cause content. In
Study 3, however, in which a group of mothers were taught to reminisce with their children in a high elaborative, emotion-oriented manner yet no subsequent increase in children’s performance on the Denham Task – over and above that of control children - was found. In 166 this latter study, there was some suggestion that children in both conditions may have been just outside the upper range of the Denham Task at the final assessment, when differences between conditions would be most likely to emerge. All children were above 48 months of age, and many were above 60 months; in comparison, children in Study 1 were 48 months of age on average. Nonetheless, children in the reminiscing condition did after 6 months show superior emotion cause knowledge to children in the control condition. These findings are consistent with those of Hughes and Dunn (1998), who also used both the Denham Emotion
Knowledge Task and an Emotion Interview (see Dunn & Hughes, 1998) in a longitudinal investigation of the associations between emotion understanding and theory of mind. The
Emotion Interview used by Hughes and Dunn is similar to the Emotion Cause Production
Task used in this thesis, and involved showing children a picture-card with a happy, sad, angry, or scared face on it, and then, after children had correctly identified the emotion, asking them four questions for each face: “What kind of things make you feel this way?”;
“Can you give me an example of a time you felt this way?”; “Lets pretend you saw [friend’s name] looking this way – why do you think he/she might be looking like that?”; and “Let’s pretend you saw mum looking this way – why do you think she might be looking like that?”
While children continued to show individual differences in the Emotion Interview across time-points, they were close to ceiling on the Denham task at 54 months of age and were thus not administered this task at 60 months of age.
Although it is possible that parents’ natural reminiscing style and content (as measured in Study 1) is associated with children’s emotion recognition and situation knowledge, and any changes to this style and content (as taught and measured in Study 3) are associated with children’s emotion cause knowledge, it seems more likely that high elaborative, emotion- oriented reminiscing facilitates the development of multiple emotion knowledge skills, depending on children’s age and current abilities. Pons and colleagues identify nine 167 components of emotion knowledge, ranging from emotion recognition to the understanding of how emotions may be affected by morality (Pons et al., 2004; see Denham, 1998, for a separate list of components). The Test of Emotion Comprehension (TEC), developed by Pons and colleagues, measures these components in children aged 3 to 12 (e.g. Pons & Harris,
2005; Pons et al., 2002, 2004). Using the TEC, it has been established that individual differences in emotion knowledge are consistent across components, thus representing a general characteristic (Pons et al., 2002a, 2003). These findings are important: if high elaborative, emotion-oriented reminiscing contributes to the development of multiple emotion knowledge skills, or perhaps generalised emotion knowledge, and if these skills continue developing until adolescence, then reminiscing style and content may continue to be important throughout childhood.
Summary and Conclusion
In conclusion, Study 1 found a significant association between high elaborative reminiscing and discussion of emotion causes; moreover, both were independently associated with children’s emotion knowledge. In Study 2, high elaborative, emotion-oriented reminiscing between children and an experimenter - particularly when couched in an explicitly causal framework - enhanced children’s recall of a staged event more greatly than did high elaborative but non-emotional reminiscing. Children’s pre-existing emotion knowledge also predicted both emotional and non-emotional recall. Finally, in Study 3, mothers were successfully trained to increase and maintain their use of high elaborative utterances and emotion references during reminiscing. Children subsequently increased their own high elaborative and emotional memory contributions during shared recall, and, after 6 months, showed superior emotion cause knowledge. Together, these studies provide both naturalistic and experimental support for the important roles that reminiscing style and content play in facilitating children’s autobiographical memory and emotion knowledge 168 development. Consistent with the sociocultural theory, the quality of reminiscing style and content was important: high elaborative, emotion-cause-oriented reminiscing facilitated development better than did low elaborative or non-emotional reminiscing. Whilst the role of high elaborative reminiscing in autobiographical memory development is well established, the importance of emotion content in memory development, the facilitation of emotion knowledge via high-elaborative reminiscing, and the findings that both a high elaborative style and emotion-oriented content can be taught are each significant contributions to the extant literature. 169
REFERENCES
Adams, S., Kuebi, J., Boyle, P.A., & Fivush, R. (1995). Gender differences in parent-child
conversations about past emotions: A longitudinal investigation. Sex Roles, 33, 309-323.
Åkerblad, A., Bengtsson, F., von Knorring, L., & Ekselius, L. (2006). Response, remission
and relapse in relation to adherence in primary care treatment of depression: A 2-year
outcome study. International Clinical Psychopharmacology, 21, 117-124.
Alea, N., & Bluck, S. (2003). Why are you telling me that? A conceptual model of the social
function of autobiographical memory. Memory, 11(2), 165-178.
Anderson, J.R., & Reder, L.M. (1979). An elaborative processing explanation of depth of
processing. In L.S. Cermak & F.I.M. Craik (Eds.), Levels of processing in human memory
(pp. 385-404). Mahwah, NJ: Erlbaum.
Astington, J.W., & Jenkins, J.M. (1999). A longitudinal study of the relations between
language and theory-of-mind development. Developmental Psychology, 35, 1311-1320.
Bahrick, H.P. (2000). Long-term maintenance of knowledge. In E. Tulving & F.I.M. Craik
(Eds.), The Oxford handbook of memory (pp. 347-362). London: Oxford University Press.
Barnier, A.J., Hung, L., & Conway, M.A. (2004). Retrieval-induced forgetting of emotional
and unemotional autobiographical memories. Cognition & Emotion, 18, 457-477.
Bartlett, F.C. (1932). Remembering: A study in experimental and social psychology.
Cambridge: Cambridge University Press.
Bauer, P.J, & Mandler, J.M. (1989). One thing follows another: Effects of temporal structure
on 1- to 2-year-olds' recall of events. Developmental Psychology, 25, 197-206.
Bauer, P., & Wewerka, S.S. (1995). One- to two-year-olds’ recall of events: The more
expressed, the more impressed. Journal of Experimental Child Psychology, 59, 475-496. 170
Beals, D.E. (2001). Eating and reading: Links between family conversations with
preschoolers and later literacy. In D.K. Dickinson and P.O. Tabors (Eds.), Beginning
language with literacy: Young children learning at home and at school (pp.75-92).
Baltimore, ML: Paul H. Brookes Publishing Co.
Bennett, D.S., Bendersky, M., & Lewis, M. (2005). Antecedents of emotion knowledge:
Predictors of individual differences in young children. Cognition & Emotion, 19, 375-396.
Bird, A., & Reese, E. (2006). Emotional reminiscing and the development of an
autobiographical self. Developmental Psychology, 42, 613-626.
Bjorklund, D.F. (2004). Introduction: Special issue: Memory development in the new
millennium. Developmental Review, 24, 343-346.
Bloise, S.M., & Johnson, M.K. (2007). Memory for emotional and neutral information:
Gender and individual differences in emotional sensitivity. Memory, 15, 192-204.
Bluck, S. (2003). Autobiographical memory: Exploring its functions in everyday life.
Memory, 11, 113-123.
Boland, A.M., Haden, C.A., & Ornstein, P.A. (2003). Boosting children’s memory by training
mothers in the use of an elaborative conversational style as an event unfolds. Journal of
Cognition and Development, 4, 39-65.
Bosacki, S.L., & Moore, C. (2004). Preschoolers’ understanding of simple and complex
emotions: Links with gender and language. Sex Roles, 50, 659-675.
Bridges, L.J., Denham, S.A., & Ganiban, J.M. (2004). Definitional issues in emotion
regulation research. Child Development, 75, 340-345.
Bretherton, I., Fritz, J., Zahn-Waxler, C., & Ridgeway, D. (1986). Learning to talk about
emotions: A functionalist perspective. Child Development, 57, 529-548. 171
Brophy, M., & Dunn, J. (2002). What did Mummy say? Dyadic interactions between young
‘hard to manage’ children and their mothers. Journal of Abnormal Child Psychology, 30,
103-112.
Brown, S.C., & Craik, F.I.M. (2000). Encoding and retrieval of information. In E. Tulving &
F.I.M. Craik (Eds.), The Oxford handbook of memory (pp.93-107). New York: Oxford
University Press.
Brown, J.R., & Dunn, J. (1996). Continuities in emotion understanding from three to six
years. Child Development, 67, 789-802.
Burch, M.M., Austin, J., & Bauer, P.J. (2004). Understanding the emotional past: Relations
between parent and child contributions in emotionally negative and nonnegative events.
Journal of Experimental Child Psychology, 89, 276-297.
Carpendale, J. I. M., Lewis, C., Susswein, N., & Lunn, J. (in press). Talking and thinking: The
role of speech in social understanding. In A. Winsler, C. Fernyhough, & I. Montero
(Eds.). Private speech, executive function, and the development of verbal self-regulation.
Cambridge: Cambridge University Press.
Carver, L.J., & Bauer, P.J. (1999). When the event is more than the sum of its parts:
Individual differences in 9-month-olds long-term ordered recall. Memory, 2, 147-174.
Cervantes, C.A., & Callanan, M.A. (1998). Labels and explanations in mother-child emotion
talk: Age and gender differentiation. Developmental Psychology, 34, 88-98.
Cicchetti, D., & Lynch, M. (1995). Failures in expectable environment and impact on
individual development: The case of child maltreatment. In Cicchetti, D., & Cohen, D.
(Eds.), Developmental Psychopathology (pp.32-71). New York: Wiley.
Cleveland, E.S. & Reese, E. (2005). Maternal structure and autonomy support in
conversations about the past: Contributions to children’s autobiographical memory.
Developmental Psychology, 41, 376-388. 172
Cleveland, E.S., Reese, E., & Grolnick, W.S. (2007). Children’s engagement and competence
in personal recollection: Effects of parents’ reminiscing goals. Journal of Experimental
Child Psychology, 96, 131-149.
Cockroft, A., & Reese, E. (in press). Emotional reminiscing and the development of an
autobiographical self. Developmental Psychology.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.
Compton, R.J., Banich, M.T., Mohanty, A., Milham, M.P., Herrington, J., Miller, G.A., Scalf,
P.E., Webber, A., & Heller, W. (2003). Paying attention to emotion: An fMRI
investigation of cognitive and emotional Stroop tasks. Cogntive, Affective, and Behavioral
Neuroscience, 3, 81-96.
Conroy, R., & Salmon, K. (2006). Talking about parts of a past experience: The impact of
discussion style and event structure on memory for discussed and nondiscussed
information. Journal of Experimental Child Psychology, 95, 278-297.
Conway, M.A., Pleydell-Pearce, C.W. (2000). The construction of autobiographical memories
in the self-memory system. Psychological Review, 107, 261-288.
Courage, M.L., & Howe, M.L. (2004). Advances in early memory development research:
Insights about the dark side of the moon. Developmental Review, 24, 6-32.
Craik, F.I.M., & Tulving, E. (1975). Depth of processing and the retention of words in
episodic memory. Journal of Experimental Psychology: General, 104, 268-294.
Cutting, A.L., & Dunn, J. (1999). Theory of mind, emotion understanding, and family
background: Individual differences and interrelations. Child Development, 70, 853-865.
Dadds, M.R. (2002). Early intervention approach for children and families at risk for
psychopathology. In T. Patterson (vol ed). Comprehensive Handbook of Psychotherapy,
Vol 2. (pp.51-72). New York: Wiley. 173 de Rosnay, M., & Hughes, C. (2006). Conversation and theory of mind: Do children talk their
way to socio-cognitive understanding? British Journal of Developmental Psychology, 24,
7-37. de Rosnay, M., Pons, F., Harris., P.L., & Morrell, J.M.B. (2004). A lag between
understanding false belief and emotion attribution in young children: relationships with
linguistic ability and mothers’ mental state language. British Journal of Developmental
Psychology, 22, 197-218.
Dempster, F.N. (1996). Distributing and managing the conditions of encoding and practice. In
E.L. Bjork & R.A. Bjork (Eds.), Memory. Handbook of perception and cognition (2nd ed.,
pp. 317-344). San Diego, CA: Academic Press.
Denham, S.A. (1986). Social cognition, prosocial behavior, and emotion in preschoolers:
Contextual variation. Child Development, 57, 194-201.
Denham, S.A. (1998). Emotional development in young children. New York: Guilford Press.
Denham, S.A., & Auerbach, S. (1995). Mother-child dialogue about emotions and
preschoolers’ emotional competence. Genetic, Social, and General Psychology
Monographs, 121, 313-337.
Denham, S.A., Blair, K.A., DeMulder, E., Levitas, J., Sawyer, K., Auerbach-Major, S., &
Queenan, P. (2003). Preschool emotional competence: Pathway to social competence?
Child Development, 74, 238-256.
Denham, S.A., & Burton, R. (2003). Social and Emotional Prevention and Intervention
Programming for Preschoolers. New York: Kluwer Academic / Plenum Publishers.
Denham, S.A., Mitchell-Copland, J., Strandberg, K., Auerbach, S., & Blair, K. (1997).
Parental contributions to preschoolers’ emotional competence: Direct and indirect effects.
Motivation and Emotion, 21, 65-86. 174
Denham, S.A., & Zoller, D. (1991). “When my hamster died, I cried”: Preschoolers’
attributions of the causes of emotions. Journal of Genetic Psychology, 152, 371-373.
Denham, S., Zoller, D., & Couchod, E. (1994). Socialization of preschoolers’ emotion
understanding. Developmental Psychology, 30, 928-936.
Dinnel, D., & Glover, J.A. (1985). Advance organisers: Encoding manipulations. Journal of
Educational Psychology, 77, 514-521.
Dodici, B.J., Draper, D.C., & Peterson, C.A. (2003). Early parent-child interactions and early
literacy development. Topics in Early Childhood Special Education, 23, 124-136.
Dunn, J. (2006). A discussion of the Merrill-Palmer Quarterly special issue (developmental
issues). Merrill-Palmer Quarterly, 52, 151-158.
Dunn, J., Bretherton, I., & Munn, P. (1987). Conversations about feeling states between
mothers and their young children. Developmental Psychology, 23, 132-139.
Dunn, J., Brown, J., Beardsall, L. (1991). Family talk about feeling states and children’s later
understanding of others’ emotions. Developmental Psychology, 27, 448-455.
Dunn, L.M., & Dunn, L.M. (1997). Peabody Picture Vocabulary Test – Third Edition. Circle
Pines, MN: American Guidance Service.
Dunn, J., & Hughes, C. (1998). Young children’s understanding of emotions within close
relationships. Cognition and Emotion, 12, 171-190.
Eder, R.A. (1989). The emergent personologist: The structure and content of 3 ½ -, 5 ½ -, and
7 ½ -year-olds’concepts of themselves and other persons. Child Development, 60, 1218-
1228.
Eisenberg, A.R. (1985). Learning to describe past experiences in conversation. Discourse
Processes, 8, 177-204.
Eisenberg, N., Smith, C.L., Sadovsky, A., & Spinrad, T.L. (2004). Effortful Control.
Relations with emotion regulation, adjustment, and socialization in childhood. In R.F. 175
Baumeister & K.D. Voks (Eds.), Handbook of Self-Regulation; Research, Theory, and
Applications (pp. 259-282). New York: The Guilford Press.
Eyberg, S.M., Funderburk, B.W., Hembree-Kigin, T.L., McNeil, C.B., Querido, J.G., &
Hood, K.K. (2001). Parent-child interaction therapy with behaviour problem children:
One and two year maintenance of treatment effects in the family. Child & Behavior
Therapy, 23, 1-20.
Farrant, K., & Reese, E. (2000). Maternal style and children’s participation in reminiscing:
Stepping stones in children’s autobiographical memory development. Journal of
Cognition and Development, 1, 193-225.
Fine, S.E., Izard, C.E., Mostow, A.J., Trentacosta, C.J., & Ackerman, B.P. (2003). First grade
emotion knowledge as a predictor of fifth grade self-reported internalizing behaviors in
children from economically disadvantaged families. Development & Psychopathology, 15,
331-342.
Fivush, R. (1989). Exploring sex differences in the emotional content of mother child
conversations about the past. Sex Roles, 20, 675-691.
Fivush, R. (1991). The social construction of personal narratives. Merrill-Palmer Quarterly,
37, 59-82.
Fivush, R. (1993). Emotional content of parent-child conversations about the past. In C.A.
Nelson (Ed.), Memory and affect in development: The Minnesota Symposia in Child
Psychology (pp.39-78). Hillsdale, NJ: Erlbaum.
Fivush, R., Berlin, L.J., Sales, J., Mennuti-Washburn, J., & Cassidy, J. (2003). Functions of
parent-child reminiscing about emotionally negative events. Memory, 11, 179-192.
Fivush, R., Brotman, M.A., Buckner, J.P., & Goodman, S.H. (2000). Gender differences in
parent-child emotion narratives. Sex Roles, 42, 233-253. 176
Fivush, R., & Fromhoff, F.A. (1988). Style and structure in mother-child conversations about
the past. Discourse Processes, 11, 337-355.
Fivush, R., & Haden, C.A. (1997). Narrating and representing experience: Preschoolers’
emerging autobiographical recounts. In P. van den Broek, P.J. Bauer, & T. Bourg (Eds.),
Developmental spans in event comprehension and representation: Bridging fictional and
actual events (pp. 169-198). Hillsdale, NJ: Erlbaum.
Fivush, R., Haden, C., & Reese, E. (2006). Elaborating on elaborations: The role of maternal
reminiscing style in cognitive and socioemotional development. Child Development, 77,
1568-1588.
Fivush, R., & Hamond, N.R. (1989). Time and again: Effects of repetition and retention
interval on 2 year olds’ event recall. Journal of Experimental Child Psychology, 47, 259-
273.
Fivush, R., Pipe, M., Murachver, T., & Reese, E. (1997). Events spoken and unspoken:
Implications of language and memory development for the recovered memory debate. In
Conway, M.A. (Ed.), Recovered Memories and False Memories (pp.34-62). Oxford:
Oxford University Press.
Garner, P.W., Carlson Jones, D., Gaddy, G., & Rennie, K.M. (1997). Low-income mothers'
conversations about emotions and their children's emotional competence. Social
Development, 6, 37-52.
Gathercole, S.E., Pickering, S.J., Ambridge, B., & Wearing, H. (2004). The structure of
working memory from 4 to 15 years of age. Developmental Psychology, 40, 177-190.
Gelman, R., & Gallistel, R. (1978). The child’s understanding of number. Cambridge, MA:
Harvard University Press. 177
Gopnik, A., & Astington, J.W. (1988). Children’s understanding of the representational
change and its relation to the understanding of false-belief and the appearance-reality
distinction. Child Development, 59, 26-37.
Greco, L.A., Sorrell, J.T., & McNeil, C.B. (2001). Understanding manual-based behavior
therapy: Some theoretical foundations of parent-child interaction therapy. Child & Family
Behavior Therapy, 23, 21-36.
Grolnick, W.S., Fodi, A., & Bridges, L. (1984). Maternal control style and the mastery
motivation of one-year-olds. Infant Mental Health Journal, 5, 72-82.
Haden, C.A., & Fivush, R. (1996). Contextual variation in maternal conversational styles.
Merrill-Palmer Quarterly, 42, 200-227.
Haden, C.A., Ornstein, P.A., Eckerman, C.O., & Didow, S.M. (2001). Mother-child
conversational interactions as events unfold: Linkages to subsequent remembering. Child
Development, 72, 1016-1031.
Han, J.J., Leichtman, M.D., & Wang, Q. (1998). Autobiographical memory in Korean,
Chinese, and American children. Developmental Psychology, 34, 701-713.
Harley, K., & Reese, E. (1999). Origins of autobiographical memory. Developmental
Psychology, 35, 1338-1348.
Harris, P.L. (1999). Acquiring the art of conversation. In M. Bennett (Ed.), Developmental
Psychology: Achievements and prospects (pp. 89-105). Philadelphia: Psychology Press.
Harris, P.L. (2006). It’s probably good to talk. Merrill-Palmer Quarterly, 52, 158-169.
Harris, P.L., de Rosnay, M., & Pons, F. (2005). Language and children’s understanding of
mental states. Current Directions in Psychological Science, 14, 69-73.
Hayne, H., & Herbert, J. (2004). Verbal cues facilitate memory retrieval during infancy.
Journal of Experimental Child Psychology, 89, 127-139. 178
Hennen, J. (2003). Statistical methods for longitudinal research on bipolar disorders. Bipolar
Disorders, 5, 156-168.
Hertsgaard, L., & Matthews, A. (1993). The ontogeny of memory revisited: Commentary on
Nelson and Fivush. In C.A. Nelson. (Ed.), Memory and affect in development (pp. 79-86).
Hillsdale, England: Lawrence Erlbaum Associates, Inc.
Hood, K.K., & Eyberg, S.M. (2003). Outcomes of parent-child interaction therapy: Mothers’
reports of maintenance 3- to 6 years after treatment. Journal of Clinical Child and
Adolescent Psychology, 32, 419-429.
Houck, P.R., Mazumdar, S., Koru-Sengul, T., Tang, G., Mulsant, B.H., Pollock, B.G., et al.
(2004). Estimating treatment effects from longitudinal clinical trial data with missing
values: Comparative analyses using different methods. Psychiatry Research, 129, 209-
215.
Howe, M.L., & Courage, M.L. (1997). The emergence and early development of
autobiographical memory. Psychological Review, 104, 499-523.
Howe, M.L., Courage, M.L., & Edison, S.C. (2003). When autobiographical memory begins.
Developmental Review, 23, 471-494.
Howe, M. L., O'Sullivan, J. T., & Marche, T. A. Toward a theory of the development of long-
term retention. (1992). In M. L. Howe, C. J. Brainerd, & V. F. Reyna (Eds.), Development
of long-term retention (pp. 245-255). New York: Springer-Verlag.
Hudson, J.A. (1986). Memories are made of this: General event knowledge and the
development of autobiographical memory. In K. Nelson (Ed.), Event knowledge:
Structure and function in development (pp. 97-118). Hillsdale, NJ: Erlbaum.
Hudson, J. A. (1990). The emergence of autobiographical memory in mother-child
conversation. In R. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young
children (pp. 166-196). Cambridge: Cambridge University Press. 179
Hudson, J.A. (1993). Reminiscing with mothers and others: Autobiographical memory in
young two-year-olds. Journal of Narrative and Life History, 3, 1-32.
Hudson, J.A., Gebelt, J., Haviland, J., & Bentivegna (1992). Emotion and narrative structure
in young children’s personal accounts. Journal of Narrative and Life History, 2, 129-150.
Hughes, C., & Dunn, J. (1998). Understanding mind and emotion: Longitudinal associations
with mental state talk between young friends. Developmental Psychology, 34, 1026-1037.
Hughes, C., & Dunn, J. (2002). ‘When I say a naughty word’. A longitudinal study of young
children’s accounts of anger and sadness in themselves and close others. British Journal
of Developmental Psychology, 20, 515-535.
Hughes, C., Jaffee, S.R., Happe, F., Taylor, A., Caspi, A., & Moffitt, T.E. (2005). Origins of
individual differences in theory of mind: From nature to nurture? Child Development, 76,
356-370.
Izard, C., Fine, S., Schultz, D., Mostow, A., Ackerman, B., & Youngstrom, E. (2001).
Emotion knowledge as a predictor of social behaviour and academic competence in
children at risk. Psychological Science, 12, 18-23.
Jenkins, J.M., Turrell, S.L., Kogushi, Y., Lollis, S., & Ross, H.S. (2003). A longitudinal
investigation of the dynamics of mental state talk in families. Child Development, 74, 905-
920.
Kinon, B.J., Lipkovich, I., Edwards, S.B., Adams, D.H., Ascher-Svanum, H., & Siris, S.G.
(2006). A 24-week randomized study of olanzapine versus ziprasidone in the treatment of
schizophrenia or schizoaffective disorder in patients with prominent depressive
symptoms. Journal of Clinical Psychopharmacology, 26, 157-162.
Kleinknecht, E., Beike, D.R. (2004). How knowing and doing inform an autobiography:
Relations among preschoolers' theory of mind, narrative, and event memory skills.
Applied Cognitive Psychology, 18, 745-764. 180
Kochanska, G., Aksan, N., & Nichols, K.E. (2003). Maternal power assertion in discipline
and moral discourse contexts: commonalities, differences, and implications for children’s
moral conduct and cognition. Developmental Psychology, 39, 949-963.
Kopp, C.B. (1989). Regulation of distress and negative emotions: A developmental view.
Developmental Psychology, 25, 343-354.
Kottler, J., & McMahon, R.S. (2004). Compliance and noncompliance in anxious, aggressive,
and socially competent children: The impact of the “child’s game” on child and maternal
behavior. Behavior Therapy, 35, 495-512.
Kuebli, J., Butler, S., & Fivush, R. (1995). Mother-child talk about past emotions: Relations
of maternal language and child gender over time. Cognition and Emotion, 9, 265-283.
Lagatutta, K.H., & Wellman, H.M. (2002). Differences in early parent-child conversations
about negative versus positive emotions: Implications for the development of
psychological understanding. Developmental Psychology, 38, 564-580.
Laible, D. (2004a) Mother-child discourse in two contexts: Links with child temperament,
attachment security, and socioemotional competence. Developmental Psychology, 40,
979-992.
Laible, D. (2004b). Mother-child discourse surrounding a child’s past behavior at 30 months:
Links to emotional understanding and early conscience development at 36 months.
Merrill-Palmer Quarterly, 50, 159-181.
Laible, D., & Song, J. (2006). Constructing emotional and relational understanding: The roles
of affect and mother-child discourse. Merrill Palmer Quarterly, 52, 44-69.
Laible, D.J., & Thompson, R.A. (2000). Mother-child discourse, attachment security, shared
positive affect, and early conscience development. Child Development, 71, 1424-1440.
Laney, C., Campbell, H.V., Heuer, F., & Reisberg, D. (2004). Memory for thematically
arousing events. Memory & Cognition, 32, 1149-1159. 181
Laney, C., Heuer, F., & Reisberg, D. (2003). Thematically-induced arousal in naturally-
occurring emotional memories. Applied Cognitive Psychology, 17, 995-1004.
Leichtman, M.D., Pillemer, D.B., Wang, Q., Koreishi, A., & Han, J.J. (2000). When baby
Maisy came to school: Mothers’ interview styles and preschoolers’ event memories.
Cognitive Development, 15, 99-114.
Leichtman, M.D., Wang, Q., & Pillemer, D.B. (2003). Cultural variations in interdependence
and autobiographical memory: Lessons from Korea, China, India, and the United States.
In R. Fivush & C.A. Haden (Eds.), Autobiographical Memory and the construction of a
narrative self: Developmental and cultural perspectives (pp. 73-97). Mahwah, NJ:
Lawrence Erlbaum Associates.
Lewis, K.D. (1999). Maternal style in reminiscing: Relations to child individual differences.
Cognitive Development, 14, 381-399.
Liwag, M.D., & Stein, N.L. (1995). Children’s memory for emotional events: The importance
of emotion-related retrieval cues. Journal of Experimental Child Psychology, 60, 2-31.
Lohmann, H., & Tomasello, M. (2003). Language and social understanding: commentary on
Nelson et al. Human Development, 46, 47-50.
Low, J., & Durkin, K. (2001). Individual differences and consistency in maternal talk style
during joint story encoding and retrospection: Associations with children’s long-term
recall. International Journal of Behavioural Development, 25, 27-36.
MacDonald, S., Uesiliana, K., & Hayne, H. (2000). Cross-cultural and gender differences in
childhood amnesia. Memory, 8, 365-376.
Martin, R.M., & Green, J.A. (2005). The use of emotion explanations by mothers: Relation to
preschoolers’ gender and understanding of emotions. Social Development, 14, 229-249. 182
McCabe, A., & Peterson, C. (1991). Getting the story: A longitudinal study of parental styles
in eliciting oral personal narratives and developing narrative skill. In A. McCabe & C.
Peterson (Eds.), Developing narrative structure (pp. 217-253). Hillsdale, NJ: Erlbaum.
McDaniel, M.A., & Einstein, G.O. (1989). Material appropriate processing. Educational
Psychology Review, 1, 113-145.
McDaniel, M.A., Waddill, P.J., Finstad, K., & Bourg, T. (2000). The effects of text-based
interest on attention and recall. Journal of Educational Psychology, 92, 492-502.
McGuigan, F., & Salmon, K. (2004). The time to talk: The influence of the timing of adult-
child talk on children's event memory. Child Development, 75, 669-686.
McGuigan, F., & Salmon, K. (2006). The influence of talking on showing and telling: Adult-
child talk and children’s verbal and nonverbal event recall. Applied Cognitive Psychology,
20, 365-381.
McNamara, D.S., & Healy, A.F. (1995). A procedural explanation of the generation effect:
The use of an operant retrieval strategy for multiplication and addition problems. Journal
of Memory & Language, 34, 399-416.
Meins, E. (1997). Security of attachment and the social development of cognition. East
Sussex, UK: Psychology Press.
Meins, E., & Fernyhough, C. (1999). Linguistic acquisitional style and mentalising
development: The role of maternal mind-mindedness. Cognitive Development, 14, 363-
380.
Meltzoff, A.N. (1985). Immediate and deferred imitation in fourteen- and twenty-four-month
old infants. Child Development, 56, 62-72.
Melzoff, A.N. (1995). What infant memory tells us about infantile amnesia: Long term recall
and deferred imitation. Journal of Experimental Child Psychology, 59, 497-515. 183
Miller, P.J. (1994). Narrative practices: Their role in socialization and self construction. In U.
Neisser & R. Fivush (Eds.), The remembered self: Construction and accuracy in the self-
narrative. New York: Cambridge University Press.
Nelson, C.A. (1995). The ontogeny of human memory: A cognitive neuroscience perspective.
Developmental Psychology, 31, 723-738.
Nelson, K. (1993a). Events, narratives, memory: What develops? In C.A. Nelson (Ed.),
Memory and affect in development: The Minnesota symposia on child psychology, volume
26 (pp. 1-24). Hillsdale, NJ: Lawrence Erlbaum Associates.
Nelson, K. (1993b). The psychological and social origins of autobiographical memory.
Psychological Science, 4, 1-8.
Nelson, K. (1996). Language in cognitive development: Emergence of the mediated mind.
Cambridge: Cambridge University Press.
Nelson, K., & Fivush, R. (2004). The emergence of autobiographical memory: A social
cultural developmental theory. Psychological Review, 111, 486-511.
Newcombe, R., & Reese, E. (2004). Evaluations and orientations in mother-child narratives as
a function of attachment security: A longitudinal investigation. International Journal of
Behavioural Development, 28, 230-245.
Newman D.L., Moffitt, T.E., Caspi, A., Magdol, L., Silva, P., & Stanton, W. (1996).
Psychiatric disorder in a birth cohort of young adults: prevalence, comorbidity,
significance, and new cases from ages 11 to 21. Journal of Consulting and Clinical
Psychology, 64, 552-562.
Nock, M., & Kazdin, A.E. (2005). Evaluation of a brief randomized controlled trial for
increasing participation in parent management training. Journal of Consulting and
Clinical Psychology, 73, 872-879. 184
O’Kearney, R., & Dadds, M.R. (2005). Language for emotions in adolescents with
internalizing and externalizing disorders. Development and Psychopathology, 17, 529-
548.
Ornstein, P.A., & Haden, C.A. (2006, July). Adult-child interactions and the development of
children’s mnemonic skills. Symposium conducted at the meeting of the International
Conference on Memory – 4, Sydney, Australia.
Ornstein, P.A., Haden, C.A., & Hedrick, A.M. (2004). Learning to remember: Social-
communicative exchanges and the development of children’s memory skills.
Developmental Review, 24, 374-395.
Pasupathi, M. (2003). Emotion regulation during social remembering: Differences between
emotions elicited during an event and emotions elicited when talking about it. Memory,
11, 151-163.
Perner, J., Leekam, S.R., & Wimmer, H. (1987). Three-year-olds’ difficulty with false belief.
British Journal of Developmental Psychology, 5, 125-137.
Peterson, C. (2002). Children's long-term memory for autobiographical events.
Developmental Review, 22, 370-402.
Peterson, C., Jesso, B., & McCabe, A. (1999). Encouraging narratives in preschoolers: An
intervention study. Journal of Child Language, 26, 49-67.
Peterson, C., & McCabe, A. (1994). A social interactionist account of developing
decontextualised narrative skill. Developmental Psychology, 30, 937-948.
Peterson, C., & Slaughter, V. (2003). Opening windows into the mind: mothers’ preferences
for mental state explanations and children’s theory of mind. Cognitive Development, 18,
399-429. 185
Pillemer, D.B. (1992). Remembering personal circumstances: A functional analysis. In E.
Winograd & U. Neisser (Eds.). Affect and accuracy in recall: Studies of "flashbulb"
memories (pp. 236-264). New York: Cambridge University Press.
Pillemer, D.B. (2003). Directive functions of autobiographical memory: The guiding power of
the specific episode. Memory, 11, 93-202.
Pons, F., & Harris, P.L. (2005). Longitudinal change and longitudinal stability of individual
differences in children’s emotion understanding. Cognition and Emotion, 19, 1158-1174.
Pons, F., Harris, P.L., & de Rosnay, M. (2004). Emotion comprehension between 3 and 11
years: developmental periods and hierarchical organisation. European Journal of
Developmental Psychology, 1, 127-152.
Pons, F., Lawson, J., Harris, P.L., & de Rosnay, M. (2003). Individual differences in
children’s emotion understanding: Effects of age and language. Scandanavian Journal of
Psychology, 44, 347-353.
Povinelli, D.J. (1995). The unduplicated self. In P. Rochat (Ed.). Advances on psychology Vol.
112. The self in infancy: Theory and Research (pp.161-192). Amsterdam: North Holland-
Elsevier.
Povinelli, D.J., Landau, K.R., & Perilloux, H.K. (1996). Self-recognition in young children
using delayed versus live feedback. Evidence of a developmental asynchrony. Child
Development, 67, 1504-1554.
Powers, P.S., Bannon, Y., Eubanks, R., & McCormick, T. (2007). Quetiapine in Anorexia
Nervosa patients: An open label outpatient pilot study. International Journal of Eating
Disorders, 40, 21-26.
Price, H.L., Connolly, D.A., & Gordon, H.M. (2006). Children’s memory for complex
autobiographical events: Does spacing of repeated instances matter? Memory, 14, 977-
989. 186
Ramey C., & Ramey, S.L. (1998). Early intervention and early experience. American
Psychologist, 53, 109-120.
Ramey, C., & Ramey, S.L. (2004). Early learning and school readiness: Can early
intervention make a difference? Merrill-Palmer Quarterly, 50, 471-491.
Reese, E. (1995). Predicting children’s literacy from mother-child conversations. Cognitive
Development, 10, 381-405.
Reese, E. (2002a). A model of the origins of autobiographical memory. In J.W. Fagen & H.
Hayne (Eds.), Progress in infancy research, Volume 2. (pp.215-260). Mahwah, NJ:
Lawrence Erlbaum Associates.
Reese, E. (2002b). Social factors in the development of autobiographical memory: The state
of the art. Social Development, 11, 124-142.
Reese, E., & Brown, N. (2000). Reminiscing and recounting in the preschool years. Applied
Cognitive Psychology, 14, 1-17.
Reese, E., Cockroft, A., & Tripp, G. (in press). Children’s self esteem and moral self: Links to
parent-child conversations about emotion. Social Development.
Reese, E., & Farrant, K. (2003). Social origins of reminiscing. In R. Fivush & C.A. Haden
(Eds.), Autobiographical memory and the construction of a narrative self: Developmental
and cultural perspectives (pp. 29-48). Mahwah, NJ: Erlbaum.
Reese, E., & Fivush, R. (1993). Parental styles of talking about the past. Developmental
Psychology, 29, 596-606.
Reese, E., Haden, C., & Fivush, R. (1993). Mother-child conversations about the past:
Relationships of style and memory over time. Cognitive Development, 8, 403-430.
Reese, E., Haden, C., & Fivush, R. (1996). Mothers, fathers, daughters, sons: Gender
differences in autobiographical reminiscing. Research on Language and Social
Interaction, 29, 27-56. 187
Reese, E., & Newcombe, R. (2007). Training mothers in elaborative reminiscing enhances
children’s autobiographical memory and narrative. Child Development, 78, 1153-1170.
Roediger, H.L., III. (1990). Implicit memory: Retention without remembering. American
Psychologist, 45, 1043-1056.
Rogoff, B. (1990). Apprenticeship in Thinking: Cognitive Development in Social Context.
New York: Oxford University Press.
Rogoff, B., & Gardner, W. (1984). Adult guidance of cognitive development. In B. Rogoff &
J.Lave (Eds.), Everyday cognition: Its development in social context (pp. 95-116).
Cambridge, MA: Harvard University Press.
Rudek, D.J., & Haden, C.A. (2005). Mothers’ and preschoolers’ mental state language during
reminiscing over time. Merrill – Palmer Quarterly, 51, 523-549.
Ruffman, T., Slade, L., & Crowe, E. (2002). The relation between children’s and mothers’
mental state language and theory-of-mind understanding. Child Development, 73, 734-
751.
Russell, J.A. (1990). The preschooler’s understanding of the causes and consequences of
emotion. Child Development, 61, 1872-1881.
Sales, J.M., & Fivush, R. (2005). Social and Emotional Functions of Mother-Child
Reminiscing About Stressful Events. Social Cognition, 23, 70-90.
Sales, J.M., Fivush, R., & Peterson, C. (2003). Parental reminiscing about positive and
negative events. Journal of Cognition and Development, 4, 185-209.
Salmon, K., & Pipe, M. (2000). Recalling an event one year later: The impact of props,
drawing and a prior interview. Applied Cognitive Psychology, 14, 99-120.
Schank, R.C., & Abelson, R. (1977). Scripts, plans, goals, and understanding: an inquiry into
human knowledge structures. Mahwah, NJ: Erlbaum. 188
Schneider, W., & Bjorklund, D.F. (1998). Memory. In W. Damon (Ed.), Handbook of Child
Psychology: Volume 2: Cognition, Perception, and Language (pp. 467-521). Hoboken,
New Jersey: John Wiley and Sons, Inc.
Shipman, K. & Zeman, J., (1999). Emotional understanding: a comparison of physically
maltreating and nonmaltreating mother–child dyads. Journal of Clinical Child Psychology
28, 407–417.
Siegal, M., & Beattie, K. (1991). Where to look first for children’s knowledge of false beliefs.
Cognition, 38, 1-12.
Siegler, R.S. (2004). Turning memory development inside out. Developmental Review, 24,
469-475.
Slamecka, N.J., & Graf, P. (1978). The generation effect: Delineation of a phenomenon.
Journal of Experimental Psychology: Human Learning and Memory. 46, 592-604.
Simcock, G., & Hayne, H. (2002). Breaking the barrier? Children fail to translate their
preverbal memories into language. Psychological Science, 13, 225-231.
Simcock, G., & Hayne, H. (2003). Age-related changes in verbal and nonverbal memory
during early childhood. Developmental Psychology, 39, 805-814.
Smiley, P., & Huttenlocher, J. (1989). Young children’s acquisition of emotion concepts. In
C. Saarni & P.L. Harris (Eds.), Children’s understanding of emotion (pp. 27-49). New
York: Cambridge University Press.
Southam-Gerow, M.A., & Kendall, P.C. (2002). Emotion regulation and understanding:
Implications for child psychopathology and therapy. Clinical Psychology Review, 22, 189-
222.
Squire, L.R. (1987). Memory and brain. New York: Oxford University Press. 189
Squire, L.R. (1995). Biological foundations of accuracy and inaccuracy in memory. In D. L.
Schacter (Ed.), Memory distortions: How minds, brains, and societies reconstruct the past
(pp. 197-225). Cambridge, MA: Harvard University Press.
Suveg, C., Zeman, J., Flannery-Schroeder, E., & Cassano, M. (2005). Emotion socialization
in families of children with an anxiety disorder. Journal of Abnormal Child Psychology,
33, 145-155.
Symons, D.K. (2004). Mental state discourse, theory of mind, and the internalization of self-
other understanding. Developmental Review, 24, 159-188.
Tandon, R., Marcus, R.N., Stock, E.G., Riera, L.C., Kostic, D., Pans, M., et al. (2006). A
prospective, multicenter, randomized, parallel-group, open-label study of aripiprazole in
the management of patients with schizophrenia or schizoaffective disorder in general
psychiatric practice: Broad Effectiveness Trial With Aripiprazole (BETA). Schizophrenia
Research, 84, 77-89.
Tenenbaum, H.R., Visscher, P., Pons, F., & Harris, P.L. (2004). Emotional understanding in
Quechua children from an agro-pastoralist village. International Journal of Behavioral
Development, 28, 471-478.
Tessler, M., & Nelson, K. (1994). Making memories: The influence of joint encoding on later
recall by young children. Consciousness and Cognition, 3, 307-326.
Thompson, R.A. (2006). Conversation and developing understanding: Introduction to the
special issue. Merrill-Palmer Quarterly, 52, 1-16.
Thompson, R.A., Laible, D.J., & Ontai, L.L. (2003). Early understandings of emotion,
morality, and self: Developing a working model. In R.V. Kail (Ed.), Advances in child
development and behavior, Vol. 31 (pp. 137-171). San Diego, CA: Academic Press.
Trabasso, T., & Stein, N.L. (1997). Narrating, representing, and remembering event
sequences. In P. van den Broek, P.J. Bauer, & T. Bourg (Eds.), Developmental spans in 190
event comprehension and representation: Bridging actual and fictional events (pp. 237-
270). Mah Wah, N.J: Lawrence Erlbaum and Associates.
Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson (Eds.),
Organization of memory (pp. 382-403). New York: Academic Press.
Tulving, E. (1983). Elements of episodic memory. New York: Oxford University Press. van Merriënboer, J.J.G; & Sweller, J. (2005). Cognitive load theory and complex learning:
Recent developments and future directions. Educational Psychology Review, 17, 147-157.
Vrielynck, N., Deplus, S., & Philippot, P. (2007). Overgeneral autobiographical memory and
depressive disorder in children. Journal of Clinical Child and Adolescent Psychology, 36,
95-105.
Vygotsky, L.S. (1962). Thought and language. Oxford: Wiley.
Vygotsky, L.S. (1978). Mind in society: The development of higher psychological processes.
Cambridge, MA: Harvard University Press.
Wang, Q. (2003). Emotion situation knowledge in American and Chinese preschool children
and adults. Cognition & Emotion, 17, 725-746.
Wang, Q. (2004). The emergence of cultural self-constructs: Autobiographical memory and
self-description in European American and Chinese children. Developmental Psychology,
40, 3-15.
Wang, Q. (2006). Relations of maternal style and child self-concept to autobiographical
memories in Chinese, Chinese immigrant, and European American 3-year-olds. Child
Development, 77, 1794-1809.
Wang, Q., & Conway, M.A. (2004). The stories we keep: Autobiographical memory in
American and Chinese middle-aged adults. Journal of Personality, 72, 911-938. 191
Wang, Q., & Fivush, R. (2005). Mother-child conversations of emotionally salient events:
Exploring the functions of emotional reminiscing in European-American and Chinese
families. Social Development, 14, 473-495.
Wang, Q., Hutt, R., Kulkofsky, S., McDermott, M., & Wei, R. (2006). Emotion situation
knowledge and autobiographical memory in Chinese, immigrant Chinese, and European
American 3-year-olds. Journal of Cognition and Development, 7, 95-118.
Wang, Q., Leichtman, M.D., & Davies, K.I. (2000). Sharing memories and telling stories:
American and Chinese mothers and their 3-year-olds. Memory, 8, 159-178.
Wareham, P., & Salmon, S. (2006). Mother-child reminiscing about everyday experiences:
Implications for psychological interventions in the preschool years. Clinical Psychology
Review, 26, 535-554.
Webster, J.D., & Cappeliez, P. (1993). Reminiscence and autobiographical memory:
Complementary contexts for cognitive aging research. Developmental Review, 13, 54-91.
Welch-Ross, M. (1997). Mother-child participation in conversation about the past:
Relationships to preschoolers’ theory of mind. Developmental Psychology, 33, 618-629.
Welch-Ross, M.K., Fasig, L.G., & Farrar, M.J. (1999). Predictors of preschoolers' self-
knowledge: Reference to emotion and mental states in mother-child conversation about
past events. Cognitive Development, 14, 401-422.
Wellman, H.M., & Bartsch, K. (1988). Young children’s reasoning about beliefs. Cognition,
31, 239-277.
Wellman, H.M., Cross, D., & Watson, J. (2001). Meta-analysis of theory-of-mind
development: The truth about false belief. Child Development, 72, 655-684.
Wenner, J. A. (2004). Preschooler's comprehension of goal structure in narratives. Memory,
12, 193-202. 192
Whitehurst, G.J., & Lonigan, C.J. (1998). Child Development and Emergent Literacy. Child
Development, 69, 848-872.
Williamson, D., Brown, E., Perlis, R.H., Ahl, J., Baker, R.W., & Tohen, M. (2006). Clinical
relevance of depressive symptom improvement in bipolar I depressed patients. Journal of
Affective Disorders, 92, 261-266.
Zimmerman, I.L., Steiner, V.G., & Pond, R.E. (2002). Preschool language scale: Fourth
edition. San Antonio, TX: The Psychological Corporation.
193
APPENDIX A: DENHAM’S EMOTION KNOWLEDGE TASK VIGNETTES4
Stereotypical Vignettes
1. TOM/JANE: “Hi! I’m Tom/Jane. Here is my sister/brother. Ah! She/he gave me some ice
cream. YUM, YUM!!” (HAPPY)
2. TOM/JANE: “We are walking home.” SIBLING: “I am going to push you down!!”
TOM/JANE: “Ow!! it hurts!! OWW!!” (SAD)
3. TOM/JANE: “I just finished building this tower, and I feel really good about
it. Doesn’t it look good?” SIBLING: “No! I think it looks yuckky. I’m going to knock
it down!” CRASH!! (MAD)
4. EXPERIMENTER: “Shhhh!! Nancy/Johnny is asleep.” TOM/JANE: “Ooh, I am
dreaming. There is a tiger chasing after me!! OH NO!!” (SCARED)
5. TOM/JANE: “Here comes Mummy. Mummy is going to take me to the zoo.” MOTHER:
“Come on, Tom/Jane. Let’s go see the animals.” TOM/JANE: “Oh, I love the
elephants. Here we go! Bye, bye!” (HAPPY)
6. TOM/JANE: “I am going to go ride my Bike. Where is it? Someone took it! It’s gone!
Someone stole it.” (SAD)
7. EXPERIMENTER: “Tom/Jane is all alone.” TOM/JANE: “It’s really dark in here.
There’s no one around. 00OOOOO.” (SCARED)
8. TOM/JANE: “I don’t want to eat cabbage!!” MOTHER: “You have to eat it, and that’s that!” TOM/JANE: “Ugh! No! No!” (MAD)
4 Adapted from Denham (1986) 194
Non-stereotypical Vignettes
1. EXPERIMENTER: “Here come Tom/Jane and her/his Mummy”.TOM/JANE: “We are coming to school…”
A. “…I like it here--We have so much fun!” (HAPPY)
B. “…I don’t like it here. I miss my mommy. Don’t go, Mommy!” (SAD)
2. TOM/JANE: “We’re going to the airport. Mommy is going on a trip…”
A. “…It’s really fun to see all the planes. WOW!” (HAPPY)
B. “…I don’t want mommy to go. Don’t go!!” (SAD)
3. TOM/JANE: “Hi, Mummy. What are you cooking?”
A. MOTHER: “[participant’s favourite food]” TOM/JANE: “Ugh! Yuck! I won’t eat
it!” (MAD)
B. MOTHER: “[participant’s least favourite food]” TOM/JANE: “Yum, yum. That
sounds great!!” (HAPPY)
4.MOTHER: “Come in for dinner Tom/Jane!” TOM/JANE: “I am swinging…”
A. “…But I’m hungry &Mommy’s food is godl I will go in. Okay, Mommy.”
(HAPPY)
B. “…I wanna swing. I wanna stay outside!! No, no I won’t come in!” (MAD)
5. TOM/JANE: “Here comes a big dog…”
A. “… He looks mean; his teeth are big.” (SCARED)
B. “… He looks nice; his big teeth are smiling at me.” (HAPPY)
6. TOM/JANE: “We are going to the swimming pool; it’s a hot day…”
A. “... the pool is so much fun! The water feels good!” (HAPPY)
B. “…I don’t like this water! It’s too deep! I don’t want it on my face-Let me out of
here!” (SCARED)
7. TOM/JANE: “We are playing blocks. We’re building a house.” SIBLING: “I’m going to 195 play with Jimmy, and you can’t come. So there!”
A. MAD [depicted by voice, expression]
B. SAD [depicted by voice, expression]
8. MOTHER: “We are going to get some ice cream at the ice cream store, but you have to stay home. Bye, Bye.”
A. MAD [depicted by voice, expression]
B. SAD [depicted by voice, expression] 196
APPENDIX B: MEMORY QUESTIONS USED DURING THE DIRECT QUESTIONING
PHASE OF STUDY 2
1. Now I’ve just a got a couple more questions to ask you, because you’re remembering
so well! How did the giraffe feel at the start?
2. Why did the giraffe feel worried?
3. Yeah, great work. And what did you do to help giraffe sleep?
4. And how did the koala first feel?
5. He felt sad, why did he feel sad?
6. And what did you play with the koala?
7. And how did the lion first feel?
8. Angry. Why did he feel angry?
9. And how did you have to walk around lion?
10. And how did the monkey first feel?
11. He felt excited, didn’t he? Why did he feel excited?
12. And what present did you give to him?
13. Yeah! And how did the elephant first feel when you first saw him?
14. He felt scared! Why did the elephant feel scared?
15. And you found him, where did you find him? 197
APPENDIX C: INTERVENTION TRAINING SCHEDULES USED IN STUDY 3
Reminiscing Training
Reminiscing Training Session 1 Guidelines (approximately 1 hour)
General Introduction (approximately 5 mins) • Presenter(s) welcome parent(s), introduce themselves, the program and its aims (i.e. looking at the effect of the interaction style on the child’s language, emotion knowledge and memory for past event). • Briefly outline the structure of the training session as: o Overview/ Introduction o Video o Role plays o Take home booklet and discussion of monitoring/ follow up activities
1. Overview/ Intro of Style (approximately 5-10 mins) • Emphasise that parents will learn a new style of talking in detail with their child about past events. • State that we have chosen to teach the style because research shows it could have long term benefits for their child’s memory, understanding of emotions and language development. • Introduce the name of the style and briefly go through components: wh-questions, details, and emotion talk. • The style is different to what parents might learn about being short and concise with children. They need to say more to the child than they normally would. • Mention that the focus is on encouraging the parents to do most of the talking whilst also setting up the conversation to encourage the child’s contribution. • Encourage parents not to worry if the child does not say much. The focus is on training the parents, not the child.
2. Video (approximately 5-10 minutes) • Set aside approximately 5-10 mins for some questions at end. 198
• Make sure the features of W-D-E are emphasised when answering any questions.
3. Role plays (approximately 10-15 minutes) • Ask parents to practise the style for 2 past events that they experienced with their child. Provide some suggestions if they are stuck e.g., “talk about the trip to the ….. last …” or “talk about one event from the weekend…”. • Trainer to give feedback and support to the parent about their use of the style. Encourage them to use the key components of the W-D-E style and point out where they were using it and where they might have used it more. • Questions or comments with the parent after role plays. • Emphasise that “practice makes perfect” and the aim is to try and incorporate the style into everyday life.
4. Take home booklet and follow up (15-20 minutes) • Slowly go through the practice booklet and expand on each point. • Discuss recording, monitoring and follow up. Parents to tape record 1 conversation with their child over the next 7-10 days. • Final questions or comments. • Make time with parent for the first follow up session where they bring their tapes back in and discuss with you.
Reminiscing Training Follow-up Session Guidelines (approximately 1/2 hour)
Session 1. • Check how past week has gone. • Listen to tapes. Discuss tapes. Provide encouragement, feedback and fine tuning referring to the booklet they received regarding the W-D-E style. • Re-emphasise via further role play if appropriate. • Discuss requirements for further practice over the next 1-2 weeks and make time for second (final) follow up session. Further practice to involve recording another 4 conversations about a past event using the W-D-E style.
Session 2. 199
• Review week (s). • Listen to and discuss tapes. Provide encouragement, feedback and fine tuning.
Session 3. • Review week (s). • Listen to and discuss tapes. Provide encouragement, feedback and fine tuning. • Make plans/ time slot for A2 post-treatment assessment at clinic in 1-2 weeks time. 200
Control Training
Control Training Session Guidelines (approximately 1 hour)
General Introduction (approximately 3 mins) • Presenter(s) introduce themselves, the program and its aims (i.e. investigating the impact of child directed play on parent child interactions). • Discuss that research suggests that there is a relationship between child direct play and parental sensitivity towards the child’s thoughts, feelings and behaviours. • Briefly outline the structure of the session as: o Overview/ Introduction of the style o Video to demonstrate key aspects of the style o Demonstration/ practice activities o Discussion of monitoring/ follow up activities
1. Overview/ Intro of Style (approximately 5-10 mins) • Parents will learn a style of interacting with their child during play. • Reiterate that we have chosen to teach the style because research shows that it might help increase your sensitivity to your child and their compliance • Discuss that we want parents to concentrate on increasing: o Their observations of their child during their play o The level of attention they give to their child And reducing: o The number of directions/ instructions they give to their child during play o Any attempts to make play go faster o Any power struggles during play (i.e., wanting play to go according to their ideas or preferences rather than following their child’s lead) • Briefly go through each of the key aspects of child directed play below providing examples where possible.
201
2. Video and discussion during video (approximately 15-20 minutes) • Show “child directed play” video. Stop after each vignette to reiterate to parents what is helpful or unhelpful about the style of play in each scene. Provide a simple statement about each scene. E.g., “Notice how the father in that scene did not respond to the child. As a result, you could see that the child became frustrated and bored.” • Make sure the features of the style are emphasised when answering any questions during the video.
3. Role plays and feedback • Ask parents to spend some time now practicing the style with their child as you observe them. • Encourage them to use the key components of “child directed play” and point out where they are using them and where they might use them more. • Questions or comments and re-emphasise the key aspects. • Emphasise that “practice makes perfect” and the aim is to try and incorporate the style into everyday life.
4. Take home practice guide and follow up (approximately 10 minutes) • Go through the hand out and expand on/ discuss each point. • Parents to practise the style at least once every day for 5-10 minutes. • Ask parents to tape record 4 attempts. • Make a time for first follow up session 7-10 days later.
Control Training Follow-up Session Guidelines (approximately 1/2 hour)
Sessions 1 and 2. • Check how past week has gone. • Listen to tapes. Discuss tapes. Provide encouragement, feedback and fine tuning referring to the booklet they received regarding “child directed play”. • Re-emphasise via further role play if appropriate. • Discuss requirements for further practice over the next 1-2 weeks i.e., parents to record 4 more attempts. Make time for next follow up session.
202
Session 3. • Review week (s). • Listen to and discuss tapes. Provide encouragement, feedback and fine tuning. • Make plans/ time slot for A2 post treatment assessment at clinic in 1-2 weeks time.
203
APPENDIX D: INTERVENTION TRAINING BOOKLETS PROVIDED TO PARENTS IN
STUDY 3
Reminiscing Training Booklet THE W-D-E STYLE PRACTICE GUIDE
The W-D-E style
• WH-questions What? Where? When? Why? And How?
• Detailed descriptions - Describe who was there and what they did - Describe objects, colours, shapes and sizes - Describe locations
• Emotions - Label emotions your child experienced - Describe the behaviours that indicated the emotion - Focus on what caused the emotion, and its consequences
204
Reminiscing Training Booklet Cont. (page 2)
PRACTICE GUIDELINES
• Pick a “big event” you experienced together.
• Be prepared. Think about the details. Perhaps write them down before you discuss them.
• Start your conversation with a “Do you remember…?” question.
• Follow the W-D-E guide from page 1.
• Keep away from using questions that only lead to “yes/ no” answers.
• Don’t worry if you feel uncomfortable using the style at first. That’s normal!
• Practise at least once a day.
• The more you practice the easier it will become
205
Reminiscing Training Booklet Cont. (page 3)
HELPFUL PRACTICE TIPS
• Pick practice times when your child is likely to be alert and attentive.
• Try and practice when your child is alone with you.
• Praise your child’s contributions to the conversation with comments like “good remembering” or “well done”.
• If your child changes the topic, follow their lead and keep using the style with the new topic.
• Stop if your child loses interest. You can come back to the style at another time.
• Don’t worry if your conversations aren’t that long.
206
Reminiscing Training Booklet Cont. (page 4)
EXAMPLE
Mother (M): Remember when we went to the aquarium on Sunday and there were lots of people there? We walked down a dark corridor. What did we see there? Child (C): Fishes! M: Yeah, that’s right we saw fishes. What kinds of fishes did we see? C: Big, big, big! M: They were very big. What were their names? C: DK M: What about my favourite kind of big fish. There was a big grey one with big teeth. It looked mean and ugly. C: Sssshak. M: Yes, it was a shark, that’s great remembering! C: Swimming M: Yes, the shark was swimming above our heads. Remember when we walked in the tunnel under the water and the big sharks were swimming around us and you were scared. You held my arm very tight and you turned your back to the tank. C: Uh huh. M: What made you scared? C: Mmm M: You thought the shark might bite you, did you? C: Yeah. Ouch! M: Yes, ouch! That would have hurt. What about when we first came in the aquarium and we looked down and there were lots of black and white birds in the water? C: Ducks! M: Good try, they walked a bit like ducks didn’t they? But they weren’t ducks. Remember how they had little suits on. They were black and white and you thought they walked funny. They were called Penguins 207
Control Training Booklet “CHILD DIRECTED PLAY”: PRACTICE GUIDE
The Key Components
• Observe and encourage your child Give your full attention to your child. Encourage and praise them as they play. Respond to them positively as they include you in their play.
• Follow your child’s lead Allow your child to be in charge in the play. Try not to direct your child or to organise their play for them. Even if you can think of ways to enhance their play, let them decide what to do and how they’d like to do it.
• Play at your child’s pace Follow your child’s pace during play. Do not set the pace of the play yourself. Give them time to explore and develop an activity.
208
Control Training Booklet Cont. (page 2)
PRACTICE GUIDELINES
• Pick a time when your child is likely to able to spend about 10 minutes playing with you.
• Be prepared. Think about the components of “child directed play” before you use them.
• Follow the “child directed play” guide from page 1.
• Keep away from prompting your child as they play.
• Don’t worry if you feel uncomfortable using the style at first. That’s normal!
• Practise at least once a day.
• The more you practice the easier it will become
209
Control Training Booklet Cont. (page 3)
HELPFUL PRACTICE TIPS
• Pick practice times when your child is likely to be alert and attentive.
• Try and practice when your child is alone with you.
• Praise your child’s contributions to the play with comments like “great idea” or “Wow! That sounds great”.
• If your child changes their mind during play, follow their lead and keep using the key components of “child directed play”.
• Stop if your child loses interest. You can come back to the style at another time.