Emotional Expressivity iii

Running head: EMOTIONAL EXPRESSIVITY

Channeling Our : A 30-Culture Study on

Emotional Expressivity, Emotionality, and Emotional Control

Wong, So-wan

A Thesis Submitted in Partial Fulfillment

of the Requirements for the Degree of

Master's of Philosophy

in

Psychology

• The Chinese University of Hong Kong

May 2004

The Chinese University of Hong Kong holds the copyright of this thesis. Any person(s) intending to use a part or whole of the materials in the thesis in a proposed publication must seek copyright release from the Dean of the Graduate School. utiiv^""""/M 丨卩P蘭i SYSTEf^^ Emotional Expressivity iii

Abstract of thesis entitled:

Three forms of , namely, nonverbal activity, interpersonal

movement behavior, and verbal activity, were distinguished under the broad term of

emotional expressivity. Using the hierarchical linear modeling (HLM) technique, a

model of emotional expressivity, with emotional intensity and emotional regulation

as predictors, was tested across the three forms of expression, across two ,

viz., and , and across 30 cultural groups. The data were employed from a

large-scale questionnaire s tudy, t he Intercultural S tudy on Emotional A ntecedents

and Reactions (ISEAR). Differentiated patterns on predicting the three forms of

emotional expressivity were found across the two emotions at the individual level,

suggesting that different forms of emotional expression should be treated separately.

Results showed that emotional intensity generally had a positive relationship with

emotional expressivity, while a generally negative relationship was found between

emotional control and emotional expressivity. Universal, as well as culturally varied

linkages were also found among the components. Three interactions with

cultures were discovered: between emotional control and verbal activity for the

emotion of anger, between emotional control and interpersonal movement behavior

for the emotion o f j oy, and b etween emotional control and verbal activity for the

emotion of joy. However, the affluence index, the citizen scores of emotional

intensity, and emotional control could not help unpackage the cultural differences in

the strength of linkages among the emotion components. Directions for future

research in cross-cultural study on emotional expressivity were suggested.

Submitted by WONG, SO-WAN for the degree of Master's of Philosophy in

Psychology at the Chinese University of Hong Kong in May 2004. Emotional Expressivity iii

論文摘要

在情感表達（emotional expressivity)這個槪念下，三種情感表達方式被區分出

來。它們分別是：非語言的活動，人際間的移動行爲，及語言表達。利用等級

制度的線性模型（HLM)技術測試了情感表達模型。在情感表達模型內，預測

變數包括情感強度及情感控制。模型測試橫跨了三十個文化，包括了兩種情

感，分別是喜樂和忿怒；及三種的表達方式。本硏究的資料來源自一個大型的

問卷調查-Intercultural Study on Emotional Antecedents and Reactions (ISEAR)�

在個人層面上，三種情感表達在兩種情感中有不同的預測模式。由此可推論出

不同的情感表達方式應被分開硏究。硏究結果顯示出一般而言，情感強度跟情

感表達有正面的關係，而情感控制跟情感表達有負面的關係。在各種的情感原

素中，有跨文化普遍以及跨文化差異的關係。三個與文化有交互作用被發現：

在忿怒情緒下，情感控制與語言表達的關係；在喜樂情緒下，情感控制與人際

間的移動行爲的關係；及在喜樂情緒下，情感控制與語言表達的關係。可是社

會的豐裕指標，人民情感強度及情感控制的平均指數未能解構(unpackage)那

些情感原素間有跨文化差異的關係的強度分別。本硏究報告爲未來有關情緒表

達的跨文化硏究作出了數項的建議。 Emotional Expressivity iii

Acknowledgement

"And what so poor a man as Hamlet is/ May do, to express his and

friending to you," says Hamlet. I am sharing the same kind of now, as I do

not know how I can express my and appreciation enough to all teachers,

friends, and family members, who have given me enormous amount of guidance and

support in the course of this two-year study.

My greatest gratitude goes to Prof. Michael Bond, who has been my

supervisor since my undergraduate studies. With his supervision and guidance, I have acquired professional growth, as well as personal growth, through these years.

He has never promised me a rose garden, but with his support and guidance, I have

learned and come this far.

My appreciation also goes to Prof. Wai Chan and Mr. Kin-Kit Li, who have had tremendous amount of patience in giving me statistical assistance and advice, as well as emotional support. Their help and support are essential to the completion of this thesis. I also want to express my appreciation to my other two thesis advisors,

Prof. Helene Fung, and Prof. K. T. Hau, who have given me valuable advice in improving the quality of this piece of work. I would like to thank all the members of the Department of Psychology at the Chinese University of Hong Kong. They are some of the most helpful and supportive people I have ever met, and have provided me with all kinds of support, including technical and social, in the course of my studies.

I am very grateful that I have a very supportive and caring family. My family is always here for me during the past quarter of a century. As typical Chinese, they support me in some very delicate, inconspicuous ways. My heartfelt gratitude to my parents, and to my brother. Emotional Expressivity iii

Last but definitely not least, is my appreciation and gratitude to Ms. Suet-yan

Tarn, who has been my best friend since my freshman year as an undergraduate

student. She shares my happy times, as well as my down times, and has given me

enormous emotional support and care in all these years.

No words can fully represent my appreciation enough. However, with my best effort, I have produced this piece of work, and this is one of the ways to show that I appreciate the guidance and support that are given to me. Emotional Expressivity iii

Table of Contents

Sections Pages

I Abstract

English Version i

Chinese Version ii

II Acknowledgement Hi

III Table of Contents v

IV Chapter 1: Introduction 1

Theories of Emotional Expressivity - Individual Level 4

Cross-Cultural Studies on Emotional Expressivity 11

Model of Emotional Expressivity of the Present Study 17

V Chapter 2: Method

Data Source 22

Data Analyses 25

VI Chapters： Results 27

VII Chapter 4: Discussion 32

Model of Emotional Expressivity 32

Future Direction 40

Concluding Notes 42

VIII References 44

IX Tables 54

X Figure 60

Model of Emotional Expressivity at the Individual Level.. 61

Dendogram of hierarchical cluster analysis of the three 62 forms of emotional expressions for the emotion of anger...

Dendogram of hierarchical cluster analysis of the three 63 forms of emotional expressions for the emotion of joy Emotional Expressivity iii

Channeling Our Feelings: A 30-Culture Study on

Emotional Expressivity, Emotionality, and Emotional Control

No man is the lord of anything, Though in and of him there be much consisting, Till he communicate his parts to others.

Shakespeare, Troilus and Cressida

The feeling of emotions is an internal process that one experiences. It is

through the expression of emotions via different channels, e.g. facial expression,

verbal and vocal expressions, as well as non-verbal behaviors, like body gesture and

movement, that we communicate our emotions to others. Due to its important

communicative function in human interaction, emotional expression has long been the

topic of investigation.

The study of emotions and emotional expression began with the emergence of

modem evolutionary biology, particularly the studies conducted by

(1872/1979; Scherer, 2003). Darwin's discrete emotion theory purposes the existence

of a small number of basic or fundamental emotions. It is argued that these basic

emotions are characterized by very specific response patterns in facial expression and

physiology (Ekman, 1972; Izard, 1971; Tomkins, 1962). For instance, from his

studies of facial expression, Ekman (1972) distinguished six basic emotions, namely,

, , ，anger, , and . Following this theory of discrete

emotion, research related to emotional expression explored the expression of emotion

in different modalities (i.e., different bodily channels that carry signals of emotions,

facial and vocal expressions in particular), and the relationship between expression and affective states (see Scherer, 2003, for a review).

Recognizing the importance of emotional expression in communication, another line of research focuses on deciphering the meaning of nonverbal behavior Emotional Expressivity iii

related to emotions (e.g., Mehrabian, 1970; Osgood, 1966). This approach assumes

that the perception of emotion is largely dependent on the observation of non-verbal

reactions of the interactional partner (Wallbott, Ricci-Bitti, & BSiminger-Huber，

1986), and at the same time, individuals can also regulate others' emotional input by

means of facial, gestural, and vocal expression (Rathbart & Berryderry，1981). It has

been suggested that the referents of nonverbal behavior, including facial and vocal

expression, posture and position, movement, as well as the implicit aspects of

verbalization, can be characterized in terms of a three-dimensional space, namely,

evaluation, potency/status, and responsiveness (Mehrabian, 1970).

A third line of research vis-a-vis emotional expression concentrates on the

individual differences in emotional expressivity. This approach is based on the

assumption that emotions do not compel individuals to act/react in particular ways:

they only incline individuals to act/react in certain ways (Gross & John, 1997).

Therefore, emotion only activates the action/response tendency, or the tendency to

express the internal feeling state of an individual. Individual differences in

expressivity can be attributed to differences in the threshold and the regulation of

emotional expression, as well as individual's personality.

In addition, the features of the social situation, i.e., the proximal environment,

may also play an important role in determining individuals' emotional expressivity.

These features include: the setting where the emotional episode takes place, e.g., public vs. private setting; the status of the interactional partners, e.g., superior, peer, or inferior. Furthermore, the broader social context, i.e., the distal environment, may also influence individuals' emotional display and contribute to cultural differences in individuals' emotional expressivity. The term, "cultural ", coined by

Ekman and Friesen (1969) denotes different socialization of emotional expression of Emotional Expressivity iii

individuals among cultural groups due to different features of the societies. These

cultural factors may be socioeconomic, e.g., economic development and affluence

level; and psychological, e.g., degree of individualism-collectivism. The value system

and the belief system endorsed by a cultural group may also influence such

socialization of emotional expression of individuals.

Different theories have been developed and various empirical studies have

been conducted in an attempt to discover the underpinnings of individual, as well as

cultural differences in individuals' emotional expressivity (e.g., Eisenberg & Fabes,

1992; Gross & John, 1997; Matsumoto, 1990; Scherer, 1982). The structure and

domains of emotional expressivity have also been explored in other studies (e.g.,

Gross & John，1997，1998; Trierweiler, Eid, & Lischetzke，2002). In addition, the

importance of emotional expressivity has been explored through studies which

investigated, for instance, its effect on psychological well-being (Gross & Levenson，

1997), and its impact on social interaction (Friedman & Riggo, 1981). Several

measures have also been developed in order to capture the concept of emotional

expressivity in individuals, e.g., the Berkeley Expressivity Questionnaire (Gross &

John, 1995)，the Emotional Expressivity Questionnaire (King & Emmons，1990)，the

Emotional Expressivity Scale (Kring, Smith, & Neale, 1994)，and the Affective

Communication Test (Friedman, Prince, Riggio, & Dimatteo’ 1990) (see also, Larsen

& Diener，1987; Snyder, 1987).

The present study aimed to develop and test a model of emotional expressivity

across cultures. In addition, whether the strength of linkages between the factors and emotional expressivity is universal or culturally different in the model was also investigated. Emotional Expressivity iii

Theories of Emotional Expressivity - Individual Level

In the study of emotional expressivity, the term has been conceptualized in

various ways. It can be operationalized in a very broad sense, so that it encompasses

"individual differences in the extent to which individuals can and do monitor their

self-presentation, expressive behavior, and nonverbal affective display" (Snyder, 1974,

p. 526-527). On the other hand, the term can be more narrowly defined, focusing on

the "individual differences in the extent to which people outwardly display their

emotions, regardless of valence (positive or negative), or channel (facial, vocal, or

gestural)" (Kring et al.，1994，p. 934).

Different theories and models of emotion have been developed which can help

understand individual differences in emotional expressivity. Three existing

theories/models of emotions are pertinent to the present study, namely, Scherer's

(1982，1984，1986, 2000) component process model of emotion, Gross and John's

(1997，1998, 2003) model of emotion process; as well as Eisenberg & Fabes (1992)

model of regulation in emotional and behavioral responding. The commonality among

these three theories/models of emotions is that they all highlight the importance of

emotional intensity and/or emotional regulation in determining individual's emotional

expressivity. Emotional intensity is one aspect of emotionality, which is defined as the

individual differences in thresholds of reaction, latency, intensity, and recovery time

(Rothbart, 1989). On the other hand, emotion regulation is defined as the successful

management of emotional to secure effective social functioning. The

regulatory process can be at the physiological, cognitive, and behavioral levels

(Dodge & Garber，1991; Rydel, Berlin, & Bohlin, 2003). In addition, emotional regulation can also be enhancement/facilitation, as well as inhibition/suppression of expressive behaviors (Zaalberg, Manstead, & Fischer，2004). Emotional Expressivity iii

Component Process Model of Emotion (Scherer,1982,1984,1986,2000a)

One of the recent approaches of emotion study is the componential-dynamic approach in which emotion is regarded as a process (e.g. Ellsworth, 1991; Frijda,

1986; Lazarus, 1991; Scherer, 1984; Smith & Ellsworth, 1985). Emotion is assumed to "consist of continuously changing configurations of component states, including cognitive and motivational processes" (Scherer, 2000a, p.71). This approach stands against the traditional structural-modular approach (e.g., Zajonc, 1980) in which emotion is conceptualized as an independent system from and motivation.

The component process modal of emotion proposed by Scherer (1982, 1984，

1986，2000a) comes from the componential-dynamic approach. In this model, the expression of emotion is one of the subsystems or components of emotion. The five subsystems are: the cognitive system (appraisal), the autonomic nervous system

(arousal), the motor system (expression), the motivational system (action tendencies), and the monitor system (feeling). The model predicts a component patterning process in which an individual appraises events based on a series of sequential evaluation checks (SEC, i.e., for novelty, intrinsic pleasantness, goal significance, coping potential, and norm compatibility). The result of each SEC modifies each of the five, interdependent subsystems. The cumulative results of the interactions of the subsystems give the pattern of emotional reaction, and the period of synchronization of the subsystems is regarded as an emotion episode. The episode is then mapped onto the cultural meaning structure, with the categorization and labeling of the process.

In this model, expressive behaviors, particularly motor expression (e.g., facial, vocal, postural, and body movement), are assumed to be determined by the dynamic changes in response to the SEC outcomes. The characteristics of the motor expression at a particular point in time represent the net result of the SECs of all the subsystems, Emotional Expressivity iii

and this processing is coined as the componential patterning model of affective

reaction (Scherer, 1984).

Thus, emotional expressivity of an individual depends on the synchronization

process in the component process model, in which appraisal is assumed to be the

driver of the emotion-constituting synchronization process. Emotional regulation,

which is defined as the control of triggered expressive behavior, is also regarded as an

important input in the synchronization process of the subsystems in the model. This

regulatory process can be voluntary, as well as automatic or occurring in a

spontaneous fashion (Scherer, 2000b), and the expressions are being consistently

monitored and changed as a result.

Gross and John 's (1997, 1998’ 2003) Model of Emotion Process

Emotional expressivity is referred as the typical styles of emotion-expressive

response of individuals in Gross and John's (1997, 1998, 2003) emotion process

model. According to the model, emotion begins with an evaluation of emotion cues,

which will then trigger a coordinated set of response tendencies (Gross & John, 2003).

The response tendencies can be experiential, behavioral, and physiological. However, emotional response tendencies may or may not be expressed as visible behavior. It is the modulatory output filter - emotional regulation - that determines whether and how response tendencies are expressed behaviorally. Thus, the activation of the emotional-response tendencies, and their subsequent modulation are assumed to be the two factors that are preeminent in determining individual differences in emotional expressivity.

Having developed this process model of emotion, Gross (1998) further elaborated the role of emotional regulation in relation to the emotional responses in the model. Emotional regulation is the attempt to influence which emotions one has, Emotional Expressivity iii

when one has them, or how these emotions are experienced or expressed (Gross,

1998). Emotions may be regulated in two major ways: antecedent-focused, and

response-focused. The former refers to the manipulation of the stimulus input to the

system before the emotion response tendencies have become fully activated. It may

take the form of reevaluating the situation so as to decrease its emotional relevance,

i.e., reappraisal. The latter refers to the employment of "strategies that intensify,

diminish, prolong or curtail ongoing emotional experience, expression, or

physiological responding" (Gross, 1998，p. 225). Suppression, i.e.，the conscious

inhibition of emotion-expressive behavior while emotionally aroused (Gross &

Levenson, 1997)，is one form of the response-focused emotion regulation strategy.

Related to the expression of emotions, it is hypothesized, as well as supported

by empirical findings (Gross, 1998)，that the use of antecedent-focused emotion

regulation strategies will decrease the activation of emotion response tendencies. Less

intense subjective emotional experience, as well as lesser physiological and fewer

expressive behaviors of emotions then result. On the other hand, since response-

focused emotion regulation strategies focus the regulatory efforts selectively on behavior, the use of such strategies will lead to lesser emotion-expressive behaviors, though the subjective feeling should remain the same.

Gross and John (1995, 1997) explored emotional expressivity further by investigating the structure of the concept. Emotional expressivity was operationalized as "the behavioral changes (e.g., facial, postural) that typically accompany emotion"

(Gross & John, 1997，p.435). It was examined across different channels, as well as across diverse emotional states. Emotional expressivity is regarded as a stable trait, and it was hypothesized, as well as supported by their empirical results, that it is a multifaceted concept. Three facets of expressivity were identified, namely, positive Emotional Expressivity iii

expressivity, negative expressivity, and impulse strength. The first two facets parallel

the two major dimensions of affects, i.e., positive and negative , found in

previous studies (Tellegen, 1985; Watson & Tellegen, 1985). The last one, on the

other hand, refers to the general strength of emotion-response tendencies. The three

facets of expressivity were found to be moderately correlated. Together, they form a

hierarchical model of emotional expressivity all under the concept of general

expressivity. It is assumed that emotion-expressive behaviors are determined by

personal traits of individuals, the immediate situations，as well as the interaction

between person and situation.

Model of Regulation in Emotional and Behavioral Responding (Eisenberg &

Fabes,1992)

Another model that can explain the underpinnings of emotional expression

came out of the study on individuals' emotional arousal and their styles of emotion

management by Eisenberg and Fabes (1992). In their model, two variables are

considered as important in influencing an individual's level of emotional arousal, viz.,

emotional intensity and self/emotional regulation.

Emotional intensity is conceptualized as "the stable differences in the strength

with which individuals typically experience emotions" (Eisenberg & Fabes，1992,

p.134-5). Individual differences in emotional intensity can be attributed to two factors.

The first factor is the typical magnitude that an individual experiences emotions,

which is similar to affective intensity defined by Larsen & Diener (1987). The other

factor is the threshold to relative intense levels of emotional responding, i.e., the ease with which an individual responds intensely. Emotional intensity is regarded as an aspect of temperament, so it is assumed to have considerable intraindividual consistency (Eisenberg & Fabes，1992; Rothbart & Derryberry，1981). Emotional Expressivity iii

Regarding emotional regulation, two types of regulation processes are

identified. The first type involves the regulation of individuals' emotional reactivity

through allocating and regulating internal states, i.e., shifting attention away

from stimuli, sustaining attention, activational control, and inhibitory control

(Rothbart & Derryberry，1981). The other type of emotion regulation involves the

change of cognitive and behavioral responses in order to manage the external or

internal demands which are appraised as taxing. This type of strategy is also referred

to as coping, which can be furthered differentiated as problem-focused, and emotion-

focused (Lazarus & Folkman，1984). The former includes mechanisms for changing

aspects of the emotion-laden situation, e.g., by displaying postural and facial cues

indicative of one's inner state, and by enacting behaviors that alter the situation. The

latter includes ways of modulating the degree of emotional arousal by altering one's

subjective emotional experience.

Eisenberg and Fabes，(1992) model of the role of regulation in emotional and

behavioral responding stipulates that social behavior, including emotional expression,

is a function of the interaction between emotional arousability and regulatory/coping

skills. In the model, aspects of the stimulus situation activate an individual's internal response systems, which then lead to the internal experience of emotion. At the same time, the evaluation of the social situation as potentially rewarding or punishing elicits the use of various forms of emotion regulation, i.e., regulation that more directly controls the experience of emotion. The experience of emotion then leads to cognitive responses, as well as behavioral responses, which include the self-regulatory process.

Self-regulation is defined as the expression, suppression, and shaping of behaviors that result from the experience of emotion. Thus, overt expression or inhibition of emotion, e.g., gestural and facial reactions, results. Emotional Expressivity iii

However, since emotional regulation can occur in terms of modulating the

emotional experience or the associated behavior or situation, it is difficult to separate

different levels of regulation in the model (Eisenberg & Fabes，1992). Therefore, the

different mechanisms used to modulate one's own emotional experience, the situation,

or one's own behavior, are not differentiated in the model. Yet, three types of

regulation are identified, namely, highly controlled (i.e., high inhibitory control),

optimal regulation (i.e., moderate use of inhibitory control), and underregulated (i.e.,

low use of inhibitory control).

Related to emotional expressivity, it is hypothesized that individuals with optimal regulation and high in emotional intensity are expressive. On the other hand, individuals with optimal regulation but lower in emotional intensity are only average in expressiveness. For individuals who are underregulated, those who are high in emotional intensity are assumed to be uncontrolled and active, as well as prone to reactive aggression (i.e., aggression based on emotional arousal). As of those who are low in emotional intensity, they are assumed to be involved with erratic behavior, and prone to proactive aggression (i.e., relatively nonemotional, goal-focused aggression).

Finally, the overcontrolled individuals with high emotional intensity are assumed to inhibit overt expression of emotions due to their restrained social style, though they are expected to be emotionally expressive in childhood. Those who are with low emotional intensity, on the other hand, are assumed to be nonexpressive, as well as inhibited and passive.

Although the three theories/models mentioned cannot totally capture the dynamic and recursive nature of emotion, nor the multifaceted evaluation and modulation processes in the emotion process (Gross, 1998)，they nonetheless provide Emotional Expressivity iii

means for understanding the underpinnings individual differences in emotional

expressivity.

Cross-Cultural Studies on Emotional Expressivity

The study of culture and emotion aims at discovering the cultural universality

and relativity in emotions and their expression. Charles Darwins's study on the

universal patterns of emotional expression across different cultures, published in The

Expression of Emotions in Man and Animals (1872/1979), was one of the first

systematic cross-cultural studies in emotions, as well as in the expression of emotions.

After a hiatus of in the area，cross-cultural emotion research gained the

limelight again with the studies on the universality of emotions, i.e., basic emotions,

and facial expression (Ekman, 1972, 1973; Izard, 1971). Since then, cross-cultural

emotion research has been dominated by the studies on basic emotions, and cross-

cultural studies of emotional behavior have mainly focused on facial expression

(Mesquita, Frijda, & Scherer，1997).

The central argument of the basic emotion theory is there are a number of

fundamental emotions, i.e. basic emotions, across cultures. In addition, different

emotions are closely and invariably linked to various components, e.g. experience,

facial expression, and physiological response. Related to emotional expression, it is

claimed that each basic emotion is marked by distinct and unique facial expressions.

However, research with this approach only focuses on "the potential for certain

emotions, rather than on their practice in the sense of prevalence or significance"

cross-culturally (Mesquita, Frijda, & Scherer，1997, p. 259).

Cross-cultural emotion research adopting the latter approach includes those

investigating cultural variations in different emotion components, as well as their relations, in the emotion process (e.g., Scherer & Wallbott，1994; Wallbott, Ricci-Bitti, Emotional Expressivity iii

& Banninger-Huber，1986; see also, Scherer，1997). Another line of research with the

same approach focuses on discovering the cultural differences in the display rules of

emotions (e.g., Matsumoto, 1990; Matsumoto, Takeuchi, Andayani, Kouznetsova, &

Krupp, 1998). Thus, cultural differences in the norms of appropriateness/

inappropriateness in the expression of emotion in different situations are investigated.

Universality and Cultural Variation of Emotional Response Patterning

In their large scale, cross-cultural study on emotions, involving eight European

countries including Israel, Wallbott and his colleagues (Wallbott et al., 1986)

discovered differences in the nonverbal emotional expressive behaviors across the

four emotions studied, viz., joy, sadness, fear，and anger. It was then concluded that

emotions “seem to be characterized by relatively distinct patterns of non-verbal reactions" (Wallbott et al, 1986，p. 107). The nonverbal reactions modalities studied

included speech (speech quality), voice, face, body-part movements, and whole body movements. Results of the study revealed that, for instance, joy was characterized by

'laughing and smiling', 'movements towards another person', and 'expansive movements'. On the other hand, anger was characterized by 'changes in movement quality' and 'changes in speech quality'.

However, the differences in reaction patterns of nonverbal behaviors of the emotions were found to be, to a large extent, independent of country differences. Thus, it was suggested that the distinct combination of components of the emotional reactions of a given emotion was largely uninfluenced by cultural difference. Country differences were found only in terms of the quantity of nonverbal reactions, and the amount of control reported. Using the same database, Cosnier and his colleagues

(Cosnier, Dols, & Fernandez，1986) found significant country differences among the types of verbal expressions, i.e., the "nothing" response, exclamation (e.g., word, Emotional Expressivity iii humming, and vocal emblems), and discussion (e.g., sentence, complete expression, and discussion), for the four emotions studied.

Similar findings were discovered in a subsequent study involving 37 countries and seven emotions (i.e., joy, fear, anger, sadness, disgust, , and ; Scherer

& Wallbott, 1994). Subjective feelings, including intensity and control attempts, as well as motor expression patterns, viz., approach behavior, nonverbal behavior, paralinguistic behavior, and verbal behavior, were studied. It was discovered that there were strong and consistent differences on the reaction patterns across the seven emotions; however, those differences were independent of country differences.

Cultural variations were only found in the absolute level, but not the patterning, for the emotion feelings and responses.

Dimensions of Culture and Cultural Variations in the Expression of Emotion

Cultural dimensions and cultural display rule. The term "cultural display rule" was coined by Ekman and Friesen (1969) in their studies of universality and cultural differences of facial expression. It denotes the cultural differences in the ecology of emotional expressions, which results in the socialization of individuals in a particular culture to suppress, attenuate, and/or enhance facial expressivity or other expressions in different social contexts (Mesquita et al., 1997). Thus, these are "learned, culturally determined rules that govern the display of emotion depending on social circumstance" (Matsumoto, 1990, p. 196).

Some early cross-cultural research on emotional expression and cultural display rules compared the emotional expressive behaviors of individuals from different cultural groups (usually two). However, "culture", the Individualism-

Collectivism dimension in particular, was used as the "cure-all" concept to explain the cultural differences found in emotional expression. In order to explain cultural Emotional Expressivity iii differences in behaviors, including emotional expressive behaviors, elements of culture, including psychological dispositions, have to be extracted and related to the behavioral aspects of cultures (Smith & Bond, 1999).

Thus, going beyond the "first generation" approach in cross-cultural psychology, recent research in the area has evolved into a stage in which different dimensions of cultures are taken into account and measured. Research using this approach attempts to "unpackage" the cultural effects on psychological variables, in which unpackaging refers to "the identification of specific, psychological dimensions of culture that may account for between-country differences in the variable of interest" (Matsumoto, 2001，p. 184). It is assumed that cultures have "core cultural ideas" (Markus & Kitayama, 1994，p. 341)，and they vary with respect to the ideas they make salient, which in turn shape behaviors. Gudykunst and Ting-Toomey's

(1988) study, as well as Matsumoto's series of studies on cultural display rules (e.g.,

Matsumoto, 1990; Matsumoto et al., 1998) are some of the research that follows this research approach. The merit of this approach is that it helps unpackage the cultural differences in emotional expressions by uncovering the underpinnings of the psychological functioning, as well as overt behaviors, of individuals of different cultures.

Gudykunst and Ting-Toomey's (1988) study employed the expressive data from Wallbott and Scherer's (1986a) cross-cultural study of emotions (involving eight

European countries, including Israel). Ecological-level correlational analyses were conducted between the expressive behavioral scores, which were generated by collapsing data across emotions, and Hofstede's cultural dimension scores (Hofstede,

1980). The four expressive behavioral scores were nonverbal vocal, nonverbal nonvocal (e.g., reactions that include use of body), verbalization, and verbal control Emotional Expressivity iii scores. Two statistical significant findings were found, with Individualism predicting levels of nonverbal expressions and verbalization.

Matsumoto's (1990) study examined the cultural similarities and differences with American and Japanese participants on cultural display rules. Photos of six emotions were used as facial stimuli, and display rules were defined as the appropriateness of displaying emotions through facial expressions in different social situations, involving people of varying intimacy and status. The hypotheses of the study were based on Hofstede's (1980) cultural dimensions of Individualism-

Collectivism and Power Distance, as well as Triandis and his colleagues' (Triandis,

Botempo, Villareal, Asai, & Lucca，1988) postulation about the relationship between

Individualism-Collectivism and the ingroup-outgroup distinction. Individuals of individualistic cultures were hypothesized to express more negative and less positive emotions to the ingroup, as well as to high status individuals, relative to individuals of collectivistic cultures, and vice versa. The findings showed that the Americans participants rated the display of negative emotions as less inappropriate than the

Japanese participants in ingroups. On the other hand, the Japanese participants rated it more appropriate than the American participants to display negative emotions in outgroups, as well as towards lower status individuals.

In a subsequent study with four cultures (South Korea, America, Russia, and

Japan), Matsumoto and his colleagues (Matsumoto et al.，1998) examined cultural differences in display rules across seven emotions in four social contexts in a questionnaire study. Display rules were defined as the expressing, deamplifying, amplifying, neutralizing, qualifying, masking, or having other types of emotional expression in a given social circumstance. Cultural differences were found for display rules, and were attributed to and correlated with the cultural differences in Emotional Expressivity iii

Individualism-Collectivism of the cultural groups measured at the individual level, viz., idiocentricism and allocentricism, in the study.

Measuring culture with socioeconomic variables. Another approach to explain cultural differences of the different components of the emotion process is to use other

"hard" cultural variables as "context variables" (van de Vijver & Leung, 1997, p. 142) to unpackage those cultural differences. In the study of emotions, socioeconomic, geo-political, and demographic variables, as well as climate, and religion indices have been used as context variables in various studies (e.g., Basabe, Paez, Valencia,

Gonzalez, Rime, & Diener, 2002; Scherer, 1997; Wallbott & Scherer，1988) to help uncover the underpinnings of the emotion process at the cultural level.

In their study which explored the cultural differences in emotional hedonic level, Basabe and his colleagues (Basabe et al.，2002) correlated national mean self- ratings of affect balance and subjective well beings with socio-economic factors, i.e., the human development index (HDI), climate, as well as cultural factors. Results showed that HDI was significantly related to subjective well-being and affect balance in their cultural-level analysis.

Related to the subjective experience of emotion, Wallbott and Scherer (1988) discovered from their cross-cultural study with 30 nations that the gross national product (GNP) was significantly, but negatively related to characteristics of the emotional experience across seven emotions. It was then suggested that poorer countries are associated with more attributions of emotions to less controllable causes, more intense emotional experiences, and stronger emotional expression. On the other hand, richer countries are associated with more attributions to self，less intense emotional experience, and less emotional expression. Emotional Expressivity iii

In his study of the appraisal patterns of antecedents of emotions of seven emotions across 37 countries, Scherer (1997) used geopolitical region, as well as other

"key cultural variables" (Scherer, 1997，p. 913)，including, socioeconomic and demographic indices (viz., affluence, urbanism, and number of cinema visits), climate indices, as well as religion, to explain the cultural differences in the level of specific appraisal variables. It was assumed that these cultural variables might play important roles in the emotion process as they affect social representations and the pattern of daily life of individuals (Scherer, 1997). The results showed that of all the cultural variables used in the study, geopolitical region was the only variable that could significantly predict cross-cultural differences in appraisal tendencies. The cultural differences between geopolitical regions were still significant after the other key cultural variables were controlled.

Various studies have demonstrated different cultural dimensions can be used as context variables to unpackage cultural differences in different components of the emotion process. It should be noted that the differences being unpackaged are usually quantitative differences among cultural groups. However, another kind of cultural difference, namely, the differences in the strength of linkages or the size of correlations among variables across cultures, can also be investigated (see Diener &

Diener, 1995).

Model of Emotional Expressivity of the Present Study

Individual Level

Recognizing the importance of emotional intensity and emotional regulation in determining individuals' emotional expressivity, the present study proposed an additive model of emotional expressivity at the individual level. Emotional Expressivity iii

Insert Figure 1 about here

It was proposed that two factors, emotional intensity and emotional regulation, contributed to individual differences in emotional expressivity. Emotional intensity was defined as the self-reported magnitude of subjective feeling; while emotional regulation was defined as the self-reported attempt to inhibit the expression of emotion. It was hypothesized that:

Hi： Emotional intensity has a positive relationship with emotional expressivity.

H2： Emotional regulation has a negative relationship with emotional

expressivity.

Therefore, more intense feeling of emotional experience would lead to more expression of emotion. At the same time, the more attempts for regulation of emotions would lead to lesser expression of emotion.

Two emotions were included in the present study, one positive, i.e., joy, and one negative, i.e., anger. In addition, three forms of emotional expression, namely, nonverbal behavior, interpersonal movement behavior, and verbal activity, were distinguished under the broader term of emotional expressivity. Comparisons of the influence of emotional intensity and emotional regulation on emotional expression across emotions, and across modalities for expressions were then made possible.

However, no specific hypotheses were made regarding the similarity/difference of the effects of emotional intensity and emotional regulation on the three forms of emotional expressions across the two emotions. Emotional Expressivity iii

Cultural Level

A culture can be defined as "a relatively organized system of shared meanings" (Smith & Bond, 1999, pp. 39). Although nations are not necessarily unitary systems and monocultural, national cultures can still be developed because individuals within a nation are bound by the same sets of social institutions and normative . Therefore, nation was used as the unit for the cultural level analyses in the present study.

The present study aimed to unpackage the cultural differences in the strength of linkages among the variables, if any, at the cultural level, with three context variables, viz., affluence, citizen scores of emotional intensity, and the citizen scores of emotional control of the respective emotions.

The use of the affluence index as a predictor was based on Berry's (1976) ecocultural framework, as well as by Berry and his colleagues' subsequent work on the taxonomy and the relationships between ecosocial indices and psychological variables (Georgas & Berry, 1995; Georgas, van de Vijver, & Berry, 2004).

According to the ecocultural framework (Berry, 1976)，ecological and sociopolitical contexts are linked to culture and then linked to human behavior. Adopting this framework in cross-cultural research, ecological and sociopolitical contexts are used as independent variables and their influence on the psychological variables can be investigated. Thus, interpretation of psychological variables at either the individual level or the country level with specific ecological and social elements is made possible (Georgas et al., 2004). In one of their studies, Georgas and his colleagues

(2004) explored the context-behavior relationships between ecosocial indices and psychological variables within such a ecocultural framework. It was discovered that affluences was a potent predictor and it was related to the psychological variables Emotional Expressivity iii used in the study. Therefore, in the present study, affluence was used as a cultural level context variable.

Affluence was defined as the degree of socioeconomic development of a nation. It was hypothesized that:

H3： The strength of linkages between emotional regulation and emotional

expression are stronger in more affluent cultures, while the strength of

linkages are weaker in less affluent cultures.

This assumption was based on Elias’ (1939/1994) theory of civilization. It is argued in the theory that as a society changes from simple to complex, individuals are compelled to regulate their affective impulse, emotions, and actions in a more consistent manner. The "civilizing" change of behavior of individuals includes the moderation of spontaneous emotions, the tempering of affects (Wong & Bond, 2003), together with the lengthening of the chains of social action and interdependence.

The underlying of this theory is that as a society civilizes, the social functions become more differentiated, and individuals in the society become more functionally dependent on one another. Thus, people are required to attune their conduct to that of others, and their actions must be organized more strictly and accurately, if they are to fulfill their multiple social functions. As the social fabric becomes more intricate, individuals' psychological make-up is also changed as a result. They have to assert more self-control on their affective impulses, emotions, and actions. Based on this theory, it was argued in the present study that there was stronger connection between emotional regulation and the emotional expressivity of individuals in more civilized, and more affluent cultures, while vice versa was true for the less civilized, and less affluent cultures. Emotional Expressivity iii

The citizen scores of emotional intensity and emotional control for the two emotions were also used as the cultural level context variables in the analyses. Citizen scores of emotional intensity and emotional control of the emotion of joy and anger of a cultural group were defined as the average score of the male average and female average scores of the given construct for the given emotion. The use of these variables was based on the assumption that there was a linear relationship between the levels of the citizen scores of the emotion construct and the importance of that emotion construct, in predicting the strength of relationship between the emotion construct and different forms of emotional expression at the cultural level. Thus, the higher the citizen score of emotional intensity of a culture, the more important the citizen score of emotional intensity in predicting the strength of relationship between emotional intensity and emotional expression, and vice versa. Similarly, the higher the citizen score of emotional control of a culture, the more important the citizen score of emotional control in predicting the strength of relationship between emotional control and emotional expression, and vice versa. Thus, it was hypothesized that:

H4： The strength of linkages between emotional intensity/emotional regulation

and emotional expression are stronger in cultures with higher citizen

scores in the construct, while the strength of linkages are weaker cultures

with lower citizen scores in the construct. Emotional Expressivity iii

Chapter 2: Method

Data Source

The data used in the present study were extracted from a large-scale intercultural study, the Intercultural Study on Emotional Antecedents and Reactions

(ISEAR), on the self-reported emotional experiences of individuals for seven emotions, viz., joy, anger, fear, sadness, disgust, shame, and guilt, with a total subject pool of 2,921 participants from 37 countries on five continents. The participants were university students (55% female and 45% male), with a mean age of 21.8 years old

(Scherer, 1997; Scherer & Wallbott，1994; Scherer, Wallbott, & Summerfield，1986;

Wallbott & Scherer，1988).

Selection of Cultures

As the number of participants varied greatly from culture to culture in the

ISEAR dataset, in the present study, only those cultures with at least 20 male and 20 female participants were included in the analyses. Thirty of the thirty-seven cultures met the above criterion, so the responses of the participants from these thirty cultures

{N= 2,574) were used in the present study (see Table 1 for the list of cultures included and the breakdown of participants by gender).

The ISEAR Questionnaire

In the questionnaire, the respondents were asked to recall a situation in which he or she had recently experienced an emotion of the kind indicated, and for which he or she vividly remembered both the circumstances and his or her reactions. The respondents were first required to describe the situation that brought forth the emotion in an open-ended format. Then, they answered a number of structured questions related to the situation, their emotional experiences, their reactions, and their appraisals of the situations. A complete, detailed description of the method and the Emotional Expressivity iii

procedure used in the ISEAR study can be found in Scherer and Wallbott (1994) and

Scherer (1997).

The present study focused on two of the seven emotions included in the

ISEAR database, namely, anger and joy. Therefore, only responses related to these

two emotions were analyzed. The responses in the questionnaire pertinent to the

present study including those related to the participant's subjective feeling states and

his/her expressive reactions during the reported emotion arousing situations that the

participant selected him/herself.

Subjective feeling state. Participants' self-reported intensity/magnitude of the

emotions and the length/duration of the emotions were included in the analyses. For

emotional intensity, participants were asked: how intense was this feeling, and responses were made on a four-point scale, ranging from "not very intense" to "very

intense". For emotional length, participants were asked: how long did you feel the

emotion, and responses were also made on a four-point scale, from "a few minutes",

"an hour", "several hours", to “a day or more".

Expressive behavior. Three expressive behavioral variables were included in the present study, namely, nonverbal activity, interpersonal movement behavior, and verbal activity. The nonverbal activity was a composite variable composed by tallying the number of responses checked in that category by the respondents in the questionnaires. Six expressive reactions were included in the nonverbal activity category: laughing/smiling, crying/sobbing, other facial expression changes, screaming/yelling, other voice changes, and changes in gesture; and the score for this variable ranged from zero to six. Interpersonal movement behavior was a variable computed by combining two responses: withdrawing from people/things (-1), and moving towards people/things (+1)，with zero being assigned when no movement Emotional Expressivity iii category was endorsed by the participant. The score for this variable was therefore on a three-point scale, ranging from -1 to +1. Finally, the verbal activity variable was composed by combining four responses: silence (0), short utterance (1)，one or two sentences (2), and lengthy utterance (3); so this variable was analyzed as an interval scale on a four-point scale, with scores ranging from zero to three.

The emotional control variable was also included in the present study. It was a measure of the self-reported attempt to regulate the emotional expressions in the situation by the participant. Participants were asked: did you try to hide or control your feelings so that nobody would know how you really felt; and responses were made on a three-point scale, ranged from "not at all" to "very much".

Cultural Level Variable

Affluence. The affluence index (Georgas, Van de Vijver, & Berry, 2004) was an empirically derived indicator of economic development of a nation. It comprised

GNP per capita, daily calorie supply per capita as a percentage of requirements, percentage of consumption of commercial energy per capita, percentage of the population employed in agriculture (negative), percentage of the population employed in industry (negative), percentage of the population employed in services, and electricity consumption per capita in kilowatt hours of a given nation. This index can be taken as an indicator of the economic development and the degree of affluence of a cultural group. The scores on the affluence index were available for all the 30 cultural groups included in the present study and they are listed in Table 1.

Citizen scores of emotion constructs. The citizen scores of emotional intensity and emotional control for the emotion of anger and joy were the averages of the male average scores and the female average scores of the given construct in a given culture.

The four citizen scores for each of the 30 cultures studied in the present study, viz., Emotional Expressivity iii

citizen scores of emotional intensity of anger and joy, as well as the citizen scores of

emotional control of anger and joy, are listed in Table 1.

Data Analyses

Multilevel Analyses

The focus of the present study was to investigate the effects of emotional

experience and emotional control on emotional expressivity for the emotions, anger

and joy, at both individual and cultural levels. It also aimed to explore if the degree of

affluence of different cultures helped unpackage the cultural variations of those

relationships, if found. The data in the present study was of a hierarchical nature such

that individual participants (level-1) were nested within cultures (level-2). Given the

present requirement of multilevel analyses, the hierarchical linear modeling (HLM)

technique was adopted. HLM has the ability to model within-level and between-level

relationships simultaneously, i.e., identifying the relationships within each of the

micro (level-1) units, and identifying if the within-unit relationships vary among the

macro (level-2) units (Bryk & Raudenbush, 1992). In the present study, with the use

of HLM, the effects of emotional intensity and emotional control on emotional

expressivity at the individual level, and how these relationships varied across cultures

were examined. HLM takes into account the individual-level error in estimating the

cultural-level coefficient, and partitions variances associated with variables at

different levels. Therefore, HLM ensures an appropriate and accurate assessment of

the relationships among the variables at the two levels. In the present study, HLM

analysis was conducted separately for each of the expressive behavioral variables for

each emotion studied.

Variables Emotional Expressivity iii

In the analyses for each of the two emotions, anger and joy, three dependent variables (level-1 outcome variables) were included, viz., nonverbal activity, interpersonal movement behavior, and verbal activity. In order to eliminate the effect of emotional length on the dependent variables in the HLM analyses, residual scores

(unstandardized) were generated for each of the six expressive behavioral variables.

The residual scores were calculated by partialing out the effect of emotional length on each of the outcome variable. Two independent variables were included at the individual level (level-1 predictors), namely, emotional intensity and emotional control. At the cultural level, the affluence index, as well as the citizen scores of emotional intensity and/or emotional control of the respective emotion were used as the level-2 predictor in the model when an interaction with culture was found for the level-1 variables in the present study. Emotional Expressivity iii

Chapter 3: Results

Preliminary Analyses

Correlational analysis was first conducted among the three emotional expressive behavioral variables for each of the two emotions. A pooled sample with equal number of male and female participants from each culture were used in order to ensure equal representation of participants of all cultures (see Bond, 1988). As the minimum number of either male or female participants of all cultures was 20，a pooled sample with 1，-200 participants was created. The results showed that the three expressive behaviors were only weakly correlated for each of the two emotions (Table

2); therefore, they were treated separately in the subsequently analyses of the study.

In addition, hierarchical cluster analysis with the three forms of emotional expression, controlling the citizen average of emotional intensity, for each of the two emotions, i.e., anger and joy, was performed.

Insert Figures 2 and 3 about here

The two dendograms revealed that the clustering of cultures for both the emotions did not following any discemable pattern based on geographic or ethnic groupings of the cultures. Thus, the results implied that other cultural variables, rather than geographic and ethnic groupings of cultures, should be used in order to understand the patterns of emotional expression for the emotion of anger and joy for the 30 cultures studied in the present study. Moreover, comparing the clustering of cultures of the two emotions, the clusters formed for anger were different from those for joy. It was suggested that the cultural variations in the emotional expression of anger were different from the Emotional Expressivity iii cultural variations in the emotional expression of joy, which again implied that the two emotions should be treated separately regarding emotional expressivity.

HLM Analyses

The results of the two-level model are shown in Table 3 for the emotion of anger, and Table 4 for the emotion of joy. The results indicate the extent to which the three types of emotional expressivity, viz. nonverbal activity, interpersonal movement behavior, and verbal activity, could be predicted by emotional intensity and emotional control for each of two emotions, anger and joy, at the individual level (fixed effect).

In addition, they also showed whether each individual-level relationship varied significantly across the 30 cultures (random effect).

Emotional intensity, emotional control, and emotional expressivity. For the emotion of anger, emotional intensity consistently had a positive relationship with two of the three types of expressive behaviors, nonverbal activity ((3 = .196,/? < .001) and verbal activity ((3 = .069，p < .05). Emotional control consistently had a significant negative relationship with each of the expressive behaviors (nonverbal activity, (3 =-

.105,/? < .01; interpersonal movement behavior, p = -.124,/? < .001; and verbal activity, (3 = -.448, p < .001). For nonverbal activity and interpersonal movement behaviors, their significant relationships with emotional intensity and emotional control did not vary across the 30 cultures. The absence of cultural variation in these relationships was demonstrated by the non-significant chi-square values in the level-2 analyses. For verbal activity, its significant positive relationship with emotional intensity was consistent across cultures; however, the significant negative relationship it had with emotional control varied as a function of culture. This interaction with culture was indicated by the significant chi-square value,义2(29) = 46.754,p < .05. A closer examination of the beta coefficients of emotional control on verbal activity for Emotional Expressivity iii

the 30 cultures (see Table 5) revealed that the coefficients were negatives for all 30

cultures, ranging from -.001 to -.839.

For the emotion of joy, there were significant positive relationships between

emotional intensity and nonverbal activity, as well as between emotional intensity and

interpersonal movement behavior (P = .165,/? < .01; p = .048, p < .001, respectively).

On the other hand, there were significant negative relationships between emotional

control and each of the two expressive behaviors (P = -.149,< .001 for nonverbal

activity; p = -.084,< .01 for interpersonal movement behavior). In the case of

nonverbal activity, the significant positive relationship with emotional intensity and

the significant negative relationship with emotional control were consistent across the

30 cultures, with non-significant chi-square values in the level-2 analyses.

Regarding interpersonal movement behavior, the significant positive relationship with emotional intensity was also consistent across cultures; however, the

significant negative relationship with emotional control varied as a function of culture, with a significant chi-square, x\29) = 60.54,/? < .01, in the level-2 analysis. The beta

coefficients of emotional control on interpersonal movement behavior were negative

for 18 of the 30 cultures, while the other 12 were positive, ranging from -.448 to .179

(Table 5).

Finally, in the case of verbal activity, it was only significantly related to emotional control in a negative direction (p = -.371，/? < .001), and this relationship varied across cultures as indicated by a significant chi-square value in the level-2

analysis, /{29) = 46.65, p < .05. An examination of the beta coefficients for all the 30

cultures revealed that except for one culture, the coefficients were all negative for

emotional control on verbal activity, ranging from .290 to -1.106 (Table 5). Emotional Expressivity iii

Thus, hypothesis 1 and 2 were generally supported as emotional intensity was positively related to the emotional expressive behaviors, and emotional regulation was negatively related to the emotional expressive behaviors across the two emotions in most cases. However, differences were also found in the patterns on how the two variables, i.e., emotional intensity and emotional regulation, predicted each of the three emotional expressive behaviors for each of the two emotions.

Affluence, citizen scores of the emotion constructs, and emotional expressivity.

Across the two emotions, three significant interactions with cultures were found in the analyses. Interestingly, all of them were related to emotional control, viz., emotional control and verbal activity for anger, emotional control and interpersonal movement behavior for joy, and emotional control and verbal activity for joy. Additional analyses were conducted to explore the cultural moderating effects for those interactions.

The Affluence Index and the citizen scores of emotional control of the respective emotion were included as the level-2 predictor variable in the HLM analyses. The aim of the analyses was to investigate if the degree of affluence and the citizen averages of emotional control of the cultural groups could unpackage the differences in the strength of linkages between emotional control and different types of emotional expressivity for different emotions across the 30 cultures. The citizen scores of emotional control were included in the analyses, but not the citizen scores of emotional intensity, because only emotional control was found to be related to the interactions found in the level-1 analyses.

The results showed that the Affluence Index and the citizen scores of emotional control of the respective emotion were not significantly related to emotional control in any of the three conditions (Table 6). Hence, they had no Emotional Expressivity iii moderating effect on the relationships between emotional control and emotional expressivity across the two emotions.

Including the Affluence Index and the citizen scores of emotional control of the respective emotion in the models did not reduce a significant amount of the chi- square from the initial models. More specifically, for the emotion of anger, between emotional control and verbal activity, the chi-square value was only slightly reduced and still significant, from ^ (29) = 46.65, p < .05, to i (27) = 46.38, p < .05.

Similarly, for the emotion of joy, between emotional control and interpersonal movement behavior, the chi-square value was only reduced slightly and was still significant, from 义(29) = 60.54, p < .01，to ^ (28) = 59.85, p < .01. Finally, between emotional control and verbal activity for joy, the chi-square value had only a small reduction in size and was still significant; it changed from ^ (29) = 46.65, p < .05, to i (27) = 41.60, p < .05. Therefore, hypothesis 3 and 4 were not supported as affluence and the citizen scores of emotional control could not help explain the cultural differences in the strength of linkages among the variables at the individual level. Emotional Expressivity iii

Chapter 4: Discussion

...it charges me in manners the rather to express myself

Shakespeare, Twelfth Night

The present study aimed to develop a model of emotional expressivity, and explore the patterns of relationships among the emotion components across expressive modalities, across emotions, as well as across cultures. Two emotions, one positive, joy, and one negative, anger, were studied vis-a-vis the expression of emotion. It was especially suitable to incorporate these two emotions in the study of emotional expression because they are regarded as active, (Babda & Wallbott,

1986; Wallbott & Scherer，1986b). Previous research suggested that joy and anger were "talkative" emotions, which were characterized by "numerous discussions and rare cases of absence of speech" (Cosnier et al., 1986, p. 120). In addition, these two emotions were also found to be more likely to be accompanied by many nonverbal expressions (Scherer & Wallbott, 1994).

Model of Emotional Expressivity

The findings of the present study generally confirmed existing theories/models of emotions that emotionality and emotional control can each help predict individuals' emotional expressivity. Specific to the present study, for the emotions of joy and anger, emotional intensity, which was defined as an aspect of emotionality and as the magnitude of the subjective experience of emotion, was found to have a positive relationship with the expression of emotion across the two emotions. On the other hand, emotional control was found to have negative relationship with the expression of emotion across the two emotions. Emotional control was defined as the inhibition of the overt expressive behavior aroused by the internal experience, and it was similar to the emotion-related behavior regulation defined in other studies (see e.g., Eisenberg, Emotional Expressivity iii

Fabes, Gutherie, & Reiser，2000). Thus, for the emotions of joy and anger, the more intense the experience of emotion, the more the expression of emotion of an individual; and the more the attempt to control the expression of emotion, the less the expression of emotion of an individual. In addition, these relationships were generally universal across the 30 cultures studied.

However, differential patterns on the combination of the two emotion components, i.e., emotional intensity and emotional control, in predicting different

forms of emotional expression for each of the two emotions studied were also found.

The findings suggested that different forms of expression of emotions should be

studied separately in order to understand the full picture of emotional expression of

different emotions.

Universal Patterns of Emotionality, Emotional Control, and Emotional Expressivity

The findings of the present study revealed three universal patterns among the

emotion components in relation to emotional expressivity. For nonverbal activity,

both emotional intensity (positive) and emotional control (negative) helped predict the

amount of this expressive behavior, and this applied to the emotions of both joy and

anger. Regarding interpersonal movement behavior, only emotional control (negative)

could predict the amount of this expressive behavior in the anger emotion.

Cultural Variations of the Patterns of Emotionality, Emotional Control, and

Emotional Expressivity

Three interactions with culture were found among the emotion components in

predicting emotional expressivity. When predicting the amount of interpersonal

movement behavior for the emotion of joy, there was a universal positive relationship

between emotional intensity and the behavioral outcome. However, the relationship

between emotional control and the expressive behavior varied as a function of culture. Emotional Expressivity iii

Although the overall relationship between emotional control and interpersonal movement behavior was negative across the 30 cultures, a closer examination of the beta weights of each of the cultures revealed that not only the strength of the relationship between the two variables was different across cultures, but the directions

of the relationships between the two variables were cross-culturally different. For

some cultures, there was a strong inverse relationship between emotional control and

interpersonal movement behavior, in which a stronger attempt to control the behavior

led to a lesser degree of moving towards, or even moving away from, the interactional

partner. However, in some other cultures, there was a strong positive connection

between emotional control and the degree of interpersonal movement behavior, in

which a stronger attempt to control the behavior still led to a higher degree of moving

towards the interactional partner. Yet, in some other cultures, the connection between

emotional control and movement behavior of individuals was much weaker. The

cultures with the highest positive beta weights were the United States, Switzerland,

and Costa Rica, while the cultures with the highest negative beta weights were

Germany, the Netherlands, and India. However, these cultures did not form into

groups based on geographic or ethnic groupings, nor other cultural dimensions, e.g.,

individualism or collectivism, which might help understand the underpinnings of this

interaction between interpersonal movement behavior and emotional control for the

emotion of joy.

The other two interactions with culture found in the present study were related

to verbal activity. For the emotion of anger, although there was a universal, positive

relationship between emotional intensity and verbal expression, the relationship

between emotional control and verbal expression was negative and culturally variable.

Regarding the emotion of joy, only emotional control helped predict the amount of Emotional Expressivity iii verbal expression (negative), and this relationship also varied as a function of culture.

An examination of the beta weights across the 30 cultural groups revealed that the cultural differences lay in the differences in the strength of relationships, i.e., the size of the beta weight, between emotional control and verbal activity. Therefore, in some cultures, there was a close connection between emotional control and verbal activity, in which a stronger attempt to control the behavior led to less verbal expression of emotion. However, in other cultures, there were weaker connections between verbal activity and the attempt to control such expressive behavior. For the interaction between verbal activity and emotional control for anger, the cultures with highest negative beta weights were China, Guatemala, and Mexico; while the cultures with lowest negative beta weights were Portugal, Greece, and Austria. On the other hand, for the interaction between verbal activity and emotional control for joy, the cultures with highest negative beta weight were Venezuela, Mexico, and Australia; while the cultures with lowest negative beta weights were Sweden, Italy, and Finland. Again, the groupings of the cultures did not fall into discemable patterns based on geographic or ethnic groupings, nor other cultural dimensions, which implied that further analyses could be conducted in order to understand the underpinnings of the interactions.

In the present study, two interesting findings were revealed. The first one was that emotional control always helped predict the amount of expression of emotion for all three expressive modalities, and for both emotions. Regarding emotional intensity, however, there were situations when it did not help predict the expression of emotion.

Thus, emotional control was found to be a useful predictor for the expression of emotion in both positive emotion, joy, and negative emotion, anger. Emotional Expressivity iii

Another interesting finding of the present study was that all the interactions found with culture were related to emotional control. This finding was in line with the suggestion that regulation processes are responsive to cultural variations in emotional phenomena (Mesquita & Frijda，1992). Since what is considered socially desirable or undesirable behavior differs in different cultures, there are cultural variations in the sociocultural norms with respect to the expression of emotion, and the control of emotional expression (Hochschild, 1983; Mesquita & Frijda，1992). More specifically, it is the strength of linkages between emotional control and emotional expression that varies from culture to culture. Therefore, regardless of the intensity of the emotional experience, the cultural variation in emotional expression is exercised in strength of relationship between the need to control the expression of emotion and the actual expression.

In addition, the finding also coincided with the concept of cultural display rule fEkman and Friesen, 1969), which suggests that as there are culturally determined rules that govern the display of emotion, and that individuals are socialized to suppress, attenuate, and/or enhance their emotional expressions in different social situations (Mesquita et al., 1997). Thus, in the present study, the effect of culture came into play when linking emotional control and emotional expression of individuals across cultures. The differences in the strength of linkages between emotional control and emotional expression found can be interpreted as a form of implicit display rule of emotions. Thus, in cultures where there are rules regarding the expression of emotion of a particular emotion, there is a stronger relationship between the level of emotional control and emotional expression. On the other hand, in cultures where there is an absent of such display rule, the linkage between emotional control and the expression of emotion is much weaker. Emotional Expressivity iii

Compared with emotional control, emotional intensity did not vary as a function of culture when comparing its effect on the expression of emotion across cultures in the present study. It may be due to the fact that emotional intensity is less susceptible to modification by the social environment (Eisenberg & Fabes, 1992), and therefore less culturally variable. Moreover, when comparing the two emotions, there were more culturally varied linkages between emotional control and the expression of emotion for the emotion of joy than anger, while there were more universal linkages between emotional control and the expression of emotion for the emotion of anger. It was suggested that since the expression of anger is generally regarded as socially destructive, the exercise of control over the expression is more universal as a result.

However, for the emotion of joy, there are less universal norms related to restraining the expression of such emotions; therefore, more cultural variations were found.

Unpackaging Cultural Variations

Having discovered the three interactions with culture regarding emotional control and expressive behaviors, the present study aimed to unpackage those cultural variations, i.e., the differences in the strength and direction of relationships between the emotion components, with context variables, namely, affluence and citizen scores of emotional control. However, both indicators failed to explain the cultural variations in the direction and strength of linkages between emotional control and interpersonal movement behavior in joy; as well as the strength of linkages between emotional control and verbal activity in both emotions.

This result was somewhat counter-intuitive, because it was suggested by

Cohen (2001) that examining culture "as a functional response to the environment,

ecology, and economy must be the starting point" (p. 465) when studying culture;

therefore, socioeconomic and ecosocial variables (Georgas & Berry, 1995) appeared Emotional Expressivity iii to be the possible candidates in explaining cultural differences. This remains a puzzling finding, and a question to be answered as to what contributed to the cross- cultural differences in the strength (as well as direction) of linkages among the emotion components.

Characteristics of the Present Study

The data of the present study came from a large scale questionnaire study, the

Intercultural Study on Emotional Antecedents and Reactions (ISEAR). Participants self-selected the emotion episodes of the emotions being studied, and reported on the physiological responses, expressive responses, control attempts, as well as appraisals accompanying the episodes. The use of questionnaire in emotion study of this kind was justified by Scherer and Wallbott (1994; see also Wallbott & Scherer，1989) who purported that it is "preferable to have access to real, and often intimate, emotions, through verbal report on recalled emotion experiences in anonymous questionnaires"

(Scherer & Wallbott，1994，p. 312), than not studying emotion episodes in real life at

all. In addition, some components of the emotion process, e.g., subjective feeling state

and attempt to regulate emotion, can only be assessed through self-report. However, it

should be noted that regarding emotional expression, the data were the subjective recall of emotional expressive behaviors, including nonverbal reactions, interpersonal

movement behavior, and verbal behaviors, of participants. Therefore, it was the self-

perceived reactions and behaviors, instead of the objective observations of reactions

and behaviors, that were being studied (Wallbott, et al., 1986).

Another issue of the present study was the equivalence of emotion terms used

across cultures in the ISEAR study. It was argued that some presumably equivalent

emotion words from different languages are seldom truly equivalent, e.g., in terms of

modal intensity, range of meaning, or frequency of usage. Therefore, emotion words Emotional Expressivity iii that are semantically equivalent in one way are often nonequivalent in other ways

(Mesquita & Frijda, 1992). However, on the other hand, Russell (1991) suggested that

"there is great similarity in emotion categories across different cultures and languages" (p. 444). Thus, emotion terms like, joy and anger, have lexical equivalents in the translations in different languages, as well as linguistic equivalents, i.e., words that are similar in meaning, across language (Mesquita et al, 2000). In addition, another empirical finding that supports the equivalence of the emotion terms in the

ISEAR study was from Scherer's (1997) study in emotion-antecedent appraisal using the ISEAR database. It was found that there were universal patterning of appraisal profiles for the seven emotions studied, including joy and anger, across the 37

cultures studied. It was suggested that for each of the two emotions, the translations of

the term were comparable and carried similar meaning, i.e., were equivalent, across

different languages.

One criticism of the present study could be the generalizability of the findings.

The respondents of the ISEAR database were all college students; therefore, they

were selected samples whom were highly modernized rather than nationally

representative samples. As suggested by Scherer (1997)，the results generated from

the database should be limited to modem mass societies due to the nature of the

respondents. However, when cultural variations were still found given this

characteristic of the samples, they should be given ample attention as those cultural

differences remained salience without being “too affected by rural-urban, or other

extraneous factors" (Scherer, 1997，p. 916). Emotional Expressivity iii

Future Direction

Unpackaging Cultural Variations in Linkage Differences

One of the characteristics of the present study was the use of a large-scale intercultural database, the ISEAR database, with data from participants from 30 countries on five continents. As the data were of a multilevel structure, i.e., individual level and cultural level, the hierarchical linear model was useful in "explicating the effects on a dependent variable of individual- and culture-level variables as well as their interactions" (Van de Vijver & Leung, 1997，p. 126). The present study aimed to unpackage the strength of linkages or the size of correlations among the variables, which was coined as a study characteristic of the "third stage of cross-cultural psychology" (Bond, 2004). However, it was demonstrated by the findings that affluence level and the predominant religion of cultural groups could not help explain the cultural differences in the strength of linkages found among the emotion components vis-a-vis emotional expression. Yet, since different forces in the cultural setting, including ecology, values, norms, and beliefs, can also influence emotional practice (Mesquita et al., 1997)，other cultural dimensions or core culture ideas

(Markus & Kitayama，1994)，e.g., independent/ interdependent tendency of cultures

(Markus & Kitayama, 1991), value dimensions (Schwartz, 1994)，and social axioms

(Leung & Bond, in press), can be used as context variables in future unpackaging studies.

Unpackaging Cultural Variation in Level Differences

One of the aims of the present study was to unpackage the strength of linkages among the emotion components if interactions with cultures were found, while the cultural difference in the absolute level of emotional expression was not the focus of the study. As the ISEAR data were derived from participants' self-selection of Emotional Expressivity iii emotion-arousing episodes, the cultural variations of the level of emotional expression might be more susceptible to the cultural variations in the selection of emotion- arousing episodes. Hence, the findings might be open to other plausible rival hypotheses. Yet, the unpackaging of the cultural difference at the level of emotional expression with socioeconomic or ecocultural context variables is still deemed to be an interesting study. Again, affluence and religion indices can be used as context variable to unpackage those cultural differences.

A preliminary correlational analysis with the citizen scores (average scores of male and female participants of each culture) of the six expressive behaviors across the two emotions and the affluence index revealed some significant relationships between emotional expression and the affluence level of the culture. More specifically, it was shown that nonverbal activity and interpersonal movement behavior in the emotion of anger, as well as verbal activity in the emotion of joy, were significantly related to the affluence index, r{29) = .49，r(29) = .43, and r(29) = .46, respectively,/?

< .05 in all cases.

Another approach of the unpackaging studies that can be used in future cross- cultural research in emotional expression is the inclusion of personality variables at the individual level as the context variable. This approach was adopted by Matsumoto

& Kupperbusch (2001) in their study of the influence of idiocentrism-allocentrism on emotional expression and experience.

Effect of Context on Emotional Expression

Another line of study of the expression of emotion in future research should take situation factors into consideration. It was suggested by a previous study that the expression and control of emotion varied according to the social interactions, e.g., in social settings (dyad or small group), or non-social settings (alone or in a large, Emotional Expressivity iii anonymous mass of people, as well as the familiarity of the interactional partner

(Ricci-Bitti & Scherer，1986). In the studies of facial expressions, it was also found that facial behavior during emotional was affected by social context, e.g., type of audience (Fridlund, 1991; Zaalberg et al.，2004).

In the present study, emotional control was defined as an individual' attempt to inhibit the expression of emotion. However, the regulation of emotional expression not only entails inhibition, but may also include the facilitation of expressive responses (Zaalberg et al., 2004). The use of inhibition or expansive strategies for emotional expression may depend on the emotion concerned, as well as the social situation. Therefore, the nature of the emotion-arousing situation may have an effect on the strength and form of emotional expression, as well as the attempt to regulate such expressions. In order to understand the full picture of the underpinnings of cultural display rule, including the socialization of emotion responding and regulation, as well as the influence of cultural models on the relative salience of different emotional expressive behaviors (Mesquita, 1999), social and situation factors should be taken into account in future cross-cultural research in the expression of emotion. In his review of cross-cultural research in emotion, Matsumoto (2001) mentioned about

the need to incorporate context in the cross-cultural study of judgment of emotion in

future research. However, this advice may also be applicable vis-a-vis the studies of

emotional expression.

Concluding Note

An important function of emotional expression is to communicate one's

emotional state to his/her interactional partner(s), thereby influencing others' behavior

as a result. An investigation of the underpinnings of different forms of emotional

expression in different emotions, as well as the universal/culturally varied patterning Emotional Expressivity iii of emotional expression, can help understand the interpersonal and intercultural functioning of people. Bearing this aim, it is hoped that the present study can help shed some light in future research in this area of cross-cultural study of emotional expressivity. Emotional Expressivity iii

References

Babad, E. Y., & Wallbott, H. G. (1986). The effects of social factors on emotional

reaction. In. K. R. Scherer, H. G. Wallbott, & A. B. Summerfield (Eds.),

Experiencing emotion: A cross-cultural study (pp. 69-83). Cambridge:

Cambridge University Press.

Basabe, N.，Paez, D.，Valencia, J., Gonzalez, J. L.，Rime, B, & Diener, E. (2002).

Cultural dimensions, socioeconomic development, climate, and emotional

hedonic level. Cognition and Emotion, 16, 103-125.

Berry, J. W. (1976). On cross-cultural comparability. International Journal of

Psychology, 4’ 119-128.

Bond, M. H. (1988). Finding universal dimensions of individual variation in

multicultural studies of values: The Rokeach and Chinese value surveys.

Journal of Personality and Social Psychology, 55, 1009-1015.

Bond, M. H. (2004, August). The third stage of cross-cultural psychology: Some

personal prescriptions for our future. In C. Kagit9iba§i (Chair), Past

presidents ‘ reflections: Past, present and future of cross-cultural psychology.

Symposium conducted at the 17th Congress of the International Association of

Cross-Cultural Psychology, Xian, China, PRC.

Byrk, A. S.，& Raudenbush，S. W. (1992). Hierarchical Linear Models: Applications

and Data Analysis Methods, Newbury Park, CA: Sage Publications.

Cohen, D. (2001). Cultural variation: Consideration and implications. Psychological

Bulletin, 127, 451-471.

Cosnier, J.，Dols, J. M. F., & Fernandez，A. J. (1986). The verbalization of emotional

experiences. In K. R. Scherer, H. G. Wallbott, & A. B. Summerfield (Eds.), Emotional Expressivity iii

Experiencing emotion: A cross-cultural study (pp. 117-128). Cambridge:

Cambridge University Press.

Darwin, C. (1979). The expression of emotions in man and animals. London, UK:

Julian Friedmann. (Originally published in 1872).

Diener, E.’ & Diener，M. (1995). Cross-cultural correlates of life satisfaction and self-

esteem. Journal of Personality and Social Psychology, 68, 653-663.

Dodge, K. A., & Garber，J. (1991). Domains of emotion regulation. In J. Garber, & K.

A. Dodge (Eds.), The development of emotion regulation and dysregulation

(pp. 3-11). Cambridge, England: Cambridge University Press.

Dodge, K. A., & Garber, J. (1991). Domains of emotion regulation. In J. Garber, & K.

A. Dodge (Eds.), The development of emotion regulation and dysregulation

(pp. 3-11). Cambridge, England: Cambridge University Press.

Eisenberg, N., & Fabes’ R. A. (1992). Emotion, regulation, and the development of

social competence. In M. Clark (Ed.), Review of personality and social

psychology (Vol. 14; pp. 119-150). Newbury Park, CA: Sage.

Eisenberg, N., Fabes, R. A., Guthrie, 1. K., & Reiser，M. (2000). Dispositional

emotionality and regulation: Their role in predicting quality of social

functioning. Journal of Personality and Social Psychology, 78, 136-157.

Ekman, P. (1972). Universals and cultural differences in facial expression of emotion.

In J. R. Cole (Ed.). Nebraska Symposium on Motivation (pp. 207-281).

Lincoln, NE: University of Nebraska Press.

Ekamn, P. (1973). Cross-cultural studies of facial expressions. In P. Ekman (Ed.),

Darwin and facial expression (pp. 169-222). New York: Academic Press.

Ekman, P., & Friesen，W. V. (1969). The repertoire of nonverbal behavior: Categories,

origins, usage, and coding. Semiotica, 1, 49-98. Emotional Expressivity iii

Elias, N. (1994). The civilizing process: The history of manners, and state formation

civilization (Translated by E. Jephcott). Cambridge: MA: Blackwell. (Original

work published 1939).

Ellsworth, P. C. (1991). Some implications of cognitive appraisal theories of emotion.

In K. Strongman (Ed.), International review of studies on emotion (pp. 143-

161). New York: Wiley.

Fridlund, A. J. (1994). Human facial expression: An evolutionary view. San Diego,

CA: Academic Press.

Friedman, H. S.，& Riggio，R. E. (1981). Effect of individual differences in nonverbal

expressiveness on transmission of emotion. Journal of Nonverbal Behavior, 7,

33-45.

Friedman, H. S., Prince, L. M., Riggio, R. E.，& Dimatteo，M. R. (1990).

Understanding and assessing nonverbal expressiveness: The affective

communication test. Journal of Personality and Social Psychology, 39, 333-

351.

Frijda, N. H. (1986). The emotions. Cambridge: Cambridge University Press.

Georgas, J, & Berry’ J. W. (1995). An ecocultural taxonomy for cross-cultural

psychology. Cross-Cultural Research, 29, 121-157.

Georgas, J., van de Vijver, F. J. R., & Berry，J. W. (2004). The ecocultural framework,

ecosocial indices, and psychological variable sin cross-cultural research.

Journal of Cross-Cultural Psychology, 35, 74-96.

Gross, J. J. (1998). Antecedent- and response-focused emotion regulation: Divergent

consequences for experience, expression, and physiology. Journal of

Personality and Social Psychology, 74, 224-237. Emotional Expressivity iii

Gross, J. J.，& John, O. P. (1995). Facets of emotional expressivity: Three self-report

factors and their correlates. Personality and Individual Differences, 19, 555-

568.

Gross, J. J.，& John, O. P. (1997). Revealing feelings: Facets of emotional

expressivity in self-reports, peer ratings, and behavior. Journal of Personality

and Social Psychology, 72, 435-448.

Gross, J. J., & John, O. P. (1998). Mapping the domain of expressivity: Multimethod

evidence for a hierarchical model. Journal of Personality and Social

Psychology, 74, 170-191.

Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation

processes: Implication for affect, relationships, and well-being. Journal of

Personality and Social Psychology, 85’ 348-362.

Gross, J. J., & Levenson, R. W. (1997). Hiding feelings: The acute effects of

inhibiting negative and positive emotion. Journal of Abnormal Psychology,

106, 95-103.

Gudykunst, W., & Ting-Toomey，S. (1988). Culture and affective communication.

American Behavioral Scientist, 31, 384-400.

Hofstede, G. H. (1980). Culture's consequences: International differences in work-

related values. Beverly Hills: Sage Publication.

Hochschild, A. R. (1983). The managed heart. Berkeley, CA: University of California

Press.

Izard, C. E. (1971). The face of emotion. New York: Appleton Century Crofts.

King, L, & Emmons, R. (1990). Conflict over emotional expression: Psychological

and physical correlates. Journal of Personality and Social Psychology, 58,

864-877. Emotional Expressivity iii

Kring, A. M., Smith, D. A., & Neale, J. M. (1994). Individual differences in

dispositional expressiveness: Development and validation of the emotional

expressivity scale. Journal of Personality and Social Psychology, 66, 934-949.

Larsen, R. J., & Diener, E. (1987). Affect intensity as an individual difference

characteristic: A review. Journal of Research in Personality, 21, 1-39.

Lazarus, R. S., & Folkman，S. (1984). Stress, appraisal, and coping. New York:

Springer.

Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press.

Leung, K.，& Bond, M. H. (in press). Social axioms: A model for social beliefs in

multi-cultural perspective. Advances in Experimental Social Psychology.

Markus, H. R.，& Kitayama, S. (1991). Culture and the self: Implications for

cognition, emotion, and motivation. Psychological Review, 98, 224-253.

Markus, H. R.，& Kitayama, S. (1994). The cultural shaping of emotion: A conceptual

framework. In S. Kitayama, & M. R. Markus (Eds.), Emotion and culture.

Empirical studies of mutual influences (pp. 339-351). Washington, DC:

American Psychological Association.

Matsumoto, D. (1990). Cultural similarities and differences in display rules.

Motivation and Emotion, 14, 195-214.

Matsumoto, D. (2001). Culture and emotion. In D. Matsumoto (Ed.), Handbook of

culture and psychology (pp. 171-194). New York: Oxford University Press.

Matsumoto, D.’ & Kupperbushch，C. (2001). Idiocentric and allocentric differences in

emotional expression, experience, and the coherence between expression and

experience. Asian Journal of Social Psychology, 4, 113-131. Emotional Expressivity iii

Matsumoto, D.，Takeuchi, S., Andayani, S.，Kouznetsova, N., & Krupp，D. (1998).

The contribution of individualism-collectivism to cross-national differences in

display rules. Asian Journal of Social Psychology, 1, 147-165.

Mehrabian, A. (1970). A semantic space for nonverbal behavior. Journal of

Consulting and Clinical Psychology, 35, 248-275.

Mesquita, B. (1999). Emotions as dynamic cultural phenomena. In R. Davidson, H.

Goldstein, & P. Rozin (Eds.), The handbook of affective sciences (pp. 871-

890). New York: Oxford University Press.

Mesquita, B.’ & Frijda, N. H. (1992). Cultural variation in emotions: A review.

Psychological Bulletin, 112, 179-204. Mesquita, B.，& Frijda, N. H. (1992).

Cultural variation in emotions: A review. Psychological Bulletin, 112, 179-

204.

Mesquita, B, Frijda, N.，& Scherer，K. (1997). Culture and emotion. In J. W. Berry, P.

R. Dasen, & T. S. Saras wthi (Eds.), Handbook of cross-cultural psychology;

Volume 2, Basic processes and human development (2nd ed., pp. 255-297).

Boston: Allyn & Bacon.

Osgood, C. E. (1966). Dimensionality of semantic space for communication via facial

expression. Journal of Psychology, 7，1-30.

Ricci-Bitti, P, & Scherer, K. R. (1986). Interrelations between antecedents, reactions,

and coping response. In. K. R. Scherer, H. G. Wallbott, & A. B. Summerfield

(Eds.), Experiencing emotion: A cross-cultural study (pp. 69-83). Cambridge:

Cambridge University Press.

Rothbart, M. K. (1989). Temperament in childhood: A framework. In G. A.

Kohnstamm, J. E.，Bates, & M. K. Rothbart (Eds.), Temperament in childhood

(pp. 59-73). New York: Wiley. Emotional Expressivity iii

Rothbart, M. K.，& Derryberry，D. (1981). Development of individual differences in

temperament. In M. E. Lamb, & A. L. Brown (Eds.), Advances in

developmental psychology (Vol. I，pp. 37-86). Hillsdale, NJ: Erlbaum.

Russell, J. A. (1991). Culture and the categorization of emotion. Psychological

Bulletin, 110, 426-450.

Rydell, A., Berlin, L., & Bohlin, G. (2003). Emotionality, emotional regulation, and

adaptation among 5- to 8-year-old children. Emotion, 5, 30-47.

Scherer, K. R. (1982). Emotion as a process: Function, origin, and regulation. Social

Science Information, 21, 555-570.

Scherer, K. R. (1984). Emotion as a multicomponent process: A model and some

cross-cultural data. In P. Shaver (Ed.) Review of Personality and Social

Psychology, Vol. 5. (pp.37-63). Beverly Hills, CA: Sage.

Scherer, K. R. (1986). Vocal affect expression: A review and a model for future

research. Psychological Bulletin, 99, 143-165.

Scherer, K. R. (1997). The role of culture in emotion-antecedent appraisal. Journal of

Personality and Social Psychology, 73, 902-922.

Scherer, K. R. (2000a). Emotion as episodes of subsystem synchronization driven by

nonlinear appraisal processes. In M. D. Lewis & I. Granic (Eds.) Emotion,

development, and self-organization: Dynamic systems approaches to

emotional development (pp. 70-99). New York/Cambridge: Cambridge

University Press.

Scherer, K. R. (2000b). Emotional expression: A royal road for the study of behavior

control. Control of human behavior, mental processes, and consciousness (pp.

227-244). Hillsdale, NJ: Lawrence Erlbaum Associates. Emotional Expressivity iii

Scherer, K. R. (2003). Vocal communication of emotion: A review of research

paradigms. Speech Communication, 40, 227-256.

Scherer, K. R.，& Wallbott，H. G. (1994). Evidence for universality and cultural

variation of differential emotion response patterning. Journal of Personality

and Social Psychology, 66, 310-328.

Schwartz, S. H. (1994). Beyond individualism/collectivism: New cultural dimensions

of values. In U. Kim, H. C. Triandis, C. Kagit9iba§i, S. Choi, & G. Yoon

(Eds)., Individualism and collectivism: Theory, method, and application (pp.

85-119). Thousand Oaks, CA: Sage.

Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion.

Journal of Personality and Social Psychology, 56, 339-353.

Smith, P., & Bond, M. H. (1999). Social psychology across cultures (2 ed.). Needham

Heights, MA: Allyn & Bacon.

Snyder, M. (1974). Self monitoring of expressive behavior. Journal of Personality

and Social Psychology, 30, 526-537.

Snyder, M. (1987). Public appearance/Private realities. New York: W. H. Freeman.

Tellegen, A. (1985). Structures of mood and personality and their relevance to

assessing , with an emphasis on self-report. In A. H. Tuman & J. D.

Maser (Eds.), Anxiety and the anxiety disorders (pp. 681-706). Hillsdale, NJ:

Erlbaum.

Tomkins, S. S. (1962). Affect, imagery, consciousness. The positive affects (Vol 1).

New York: Springer.

Triandis, H. C.，Botempo, R.，Villareal, M. J.，Asai, M.，& Lucca, N. (1988).

Indiviualism and collectivism: Cross-cultural perspectives on self-in-group

relationships. Journal of Personality and Social Psychology, 54, 323-338. Emotional Expressivity iii

Trierweiler, L. I.，Eid, M, & Lischetzke, T. (2002). The structure of emotional

expressivity: Each emotion counts. Journal of Personality and Social

Psychology, 82, 1023-1040.

Van de Vijver, F., & Leung, K. (1997). Methods and data analysis for cross-cultural

research. Thousand Oaks, CA: Sage.

Wallbott, H. G.，Ricci-Bitti, P., & Banninger-Huber，E. (1986). Non-verbal reactions

to emotional experiences. In K. R. Scherer, H. G. Wallbott, & A. B.

Summerfield (Eds.), Experiencing emotion: A cross-cultural study (pp. 98-

116). Cambridge: Cambridge University Press.

Wallbott, H. G, & Scherer，K. R. (1986a). The antecedents of emotional experiences.

In. K. R. Scherer, H. G. Wallbott, & A. B. Summerfield (Eds.), Experiencing

emotion: A cross-cultural study (pp. 69-83). Cambridge: Cambridge

University Press.

Wallbott, H. G.，& Scherer，K. R. (1986b). How universal and specific is

emotional experience? Evidence from 27 countries on five continents. Social

Science Information, 25, 763-795.

Wallbott, H. G.，& Scherer, K. R. (1988). Emotion and economic development - Data

and speculations concerning the relationship between economic factors and

emotional experience. European Journal of Social Psychology, 18, 267-273.

Wallbott, H. G.，& Scherer，K. R. (1989). Assessing emotion by questionnaire. In R.

Plutchik & H. Kellerman (Eds.), Emotion: Theory, research, and experience

(Vol. 4, pp. 55-82). NY: Academic Press.

Watson, D.’ & Tellegen，A. (1985). Toward a consensual structure of mood.

Psychological Bulletin, 54, 1063-1070. Emotional Expressivity iii

Wong, S. & Bond, M. H. (2003). Societal predictors of emotional recency, length, and

intensity across the citizens of 30 nations. (Manuscript submitted for

publication).

Zaalberg, R.，Manstead, A. S. R.，& Fisher，A. H. (2004). Relations between emotions,

display rules, social motives, and facial behavior. Cognition and Emotion, 18,

183-207.

Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American

Psychologist, 2, 151-176.

J I i I Table 1 List of Cultures, the Number of Participants, Affluence Index, and Citizen Scores of Emotion Constructs Culture Total Number Male Female Affluence Citizen Scores of Emotional Intensity Citizen Scores of Emotional Control of Participants Participants Participants Index Anger Joy Anger Joy Australia 117 37 80 1.44781 2.88 2.91 1.71 1.31 Austria 69 28 41 1.23254 3.24 3.31 1.44 1.25 Botswana 79 54 25 .05927 3.27 3.24 1.63 1.22 Bulgaria 73 31 42 .94868 2.96 3.25 1.63 1.36 Chile 65 32 33 -.07138 2.69 2.91 1.44 1.23 China 79 48 31 -.79308 3.00 3.04 1.34 1.33 Costa Rica 58 29 29 -.18209 2.88 3.22 1.67 1.44 Finland 76 25 51 1.56966 2.74 2.82 1.58 1.27 Germany 117 45 72 1.70337 2.92 3.21 1.68 1.19 Greece 66 34 32 .40504 3.12 3.40 1.38 1.14 Guatemala 45* 21 23 -.62007 2.82 3.03 1.58 1.25 Hong Kong 81* 43 37 .76452 2.81 2.98 1.79 1.32 India 68 36 32 -.76210 2.71 3.25 1.78 1.57 \ Italy 98 50 48 .58672 3.05 3.14 1.49 1.33 Japan 214 97 117 1.32790 2.85 2.89 1.61 1.34 Malawi 75 38 37 -.92142 3.26 3.20 1.65 1.37 Mexico 139 64 75 .25444 3.00 3.16 1.58 1.22 Netherlands 69 24 45 1.24107 3.20 3.05 1.60 1.25 � New Zealand 60 22 38 1.04481 3.15 3.00 1.58 1.31 g Nigeria 77 50 27 -.73876 2.80 3.18 1.66 1.25 g. Poland 87 43 44 .49168 2.95 3.16 1.62 1.20 g Portugal 88 21 67 .28204 2.89 3.31 1.56 1.31 p Spain 78 39 39 .66848 3.03 3.12 1.43 1.21 W Sweden 84 37 47 1.95395 3.08 3.08 1.48 1.26 ^ Switzerland 80 22 58 1.67313 3.24 3.09 1.42 1.22 ^ United States 69 40 29 2.28317 3.10 3.15 1.73 1.20 Venezuela 73 28 45 .27453 3.10 3.14 1.55 U5 Yugoslavia 80 40 40 .46141 3.09 3.38 1.68 1.48 乂 Zambia 111* 72 38 -.30890 3.16 3.17 1.67 1.41 Zimbabwe 99 52 47 -.26134 3.35 3.35 1.54 1.26 * For Guatemala, Hong Kong, and Zambia, one participant from each of these cultures was not specified of his/her gender. Table 2

Correlation among the Expressive Behaviors for Anger and Joy

Anger Joy Expressive behaviors Nonverbal activity Interpersonal Movement Nonverbal activity Interpersonal Movement Behavior Behavior

Nonverbal activity - “

Interpersonal -026 _ .250*** - Movement Behavior

Verbal Activity .274*** .187*** .116*** | —_

*n< .05; **/?< .01; ***j9< .001. w

C/3 w � Table 3

The Effect of Emotional Intensity and Emotional Control on Emotional Expressivity (Anger)

Fixed Effect Random Effect Expressive behaviors Coefficient SE t-ratio (df= 29) Variance SD ；^ {df= 29) Nonverbal activity Emotional Intensity .196 .027 7.210*** .0042 .065 33.179 Emotional Control -.105 .030 -3.515** .0004 .021 17.761

Interpersonal Movement Behavior Emotional Intensity .003 .014 .271 .0020 .044 42.873 Emotional Control -.124 .017 -7.473*** .0020 .045 38.210

w Verbal Activity | Emotional Intensity .069 .026 2.629* .0037 .061 32.319 | Emotional Control -.448 .036 -12.436*** .0172 .131 46.754* ^ X *p< .05; **p< .01; ***j9< .001. ^

—• � On Table 4

The Effect of Emotional Intensity and Emotional Control on Emotional Expressivity (Joy)

Fixed Effect Random Effect Expressive behaviors Coefficient SE t-ratio (df= 29) Variance SD / {df= 29) Nonverbal activity Emotional Intensity .165 .023 7.096*** .0018 .042 22.810 Emotional Control -.149 .037 -4.044*** .0014 .038 19.577

Interpersonal Movement Behavior Emotional Intensity .048 .012 3.895** .0005 .023 19.696 Emotional Control -.084 .030 -2.798** .0148 .121 60.539**

Verbal Activity § Emotional Intensity -019 .023 -.802 .0018 .043 30.097 | Emotional Control -.371 .049 -7.565*** .0305 .175 46.647* w *p< .05; **/7< .01; ***/?< .001. 1

—• Emotional Expressivity 58

Table 5

List of Beta Coefficients for Emotional Control and Emotional Expressivity

Culture Anger Joy Control & Control & Control & Verbal Activity Interpersonal Verbal Activity Movement Behavior

Australia -.539 -.224 -.918 Austria -.036 -.203 -.311 Botswana -.436 .036 -.404 Bulgaria -.667 .009 -.486 Chile -.497 .002 -.452 China -.810 -.012 -.233 Costa Rica -.229 .262 -.434 Finland -.350 -.214 -.030 Germany -.234 -.448 -.169 Greece -.116 .078 -.473 Guatemala -.839 -.029 -.659 Hong Kong -.315 -.130 -.242 India -.327 -.319 -.425 Italy -.438 -.006 -.170 Japan -.437 -.028 -.255 Malawi -.320 -.192 -.262 Mexico -.779 .013 -.733 Netherlands -.315 -.408 -.331 New Zealand -.609 -.096 -.654 Nigeria -.523 -.151 -.256 Poland -.358 .009 -.656 Portugal -.001 .101 .290 Spain -.538 -.208 -.416 Sweden -.706 -.148 -.112 Switzerland -.569 .139 -.266 United States -.749 .179 -.304 Venezuela -.340 -.307 -1.106 Yugoslavia -.366 .081 -.166 Zambia -.482 -.004 -.167 Zimbabwe -.579 .024 -.603 M=-A50 M=-.073 M=-.380 Table 6

The Effect of Affluence and the Citizen Score of Emotional Control on Emotional Control on Emotional Expressivity

Fixed Effect Random Effect Expressive behaviors of Anger Coefficient SE t-ratio {df = 27) Variance SD /(£//= 27) Verbal Activity Emotional Control 0.018 0.135 46.389* Affluence -0.004 0.041 -0.110 Citizen Score of Emotional Control -0.003 0.355 -0.009

Expressive behaviors of Joy Interpersonal Movement Behavior Emotional Control 0.127 0.016 59.853" Affluence -0.020 0.033 -0.619 W Citizen Score of Emotional Control -0.093 0.418 -0.222 | O

Verbal Activity W Emotional Control 0.175 0.031 41.598* | Affluence 0.051 0.053 0.956 | H-» • Citizen Score of Emotional Control 0.770 0.479 1.609 � *p< .05; **p< .01; ***p< .001. 'O Emotional Expressivity iii

Figure Caption

Figure 1. Model of emotional expressivity at the individual level

Figure 2. Dendogram of hierarchical cluster analysis of the three forms of emotional

expressions for the emotion of anger

Figure 3. Dendogram of hierarchical cluster analysis of the three forms of emotional

expressions for the emotion of joy Emotional Expressivity iii

Emotional Intensity Z X

Emotional Expressivity

r\Emotional Control v S

Figure 1. Model of emotional expressivity at the individual level Rescaled Distance Cluster Combine CASE 0 5 10 15 20 25 Label Num + + + + + +

Malawian 16 j Portuguese 22 —• Hong Kongese 12 Japanese 15 —•I Yugoslav 28 Indian 13 ~• Greek 10 Dutch 18 —I Zambian 29 Mexican 17 — Pole 21 」 Guatemalan 11 ~ Venezuelan 27 Botswanan 3 — Zimbabwean 30 一 Costa Rican 7 」 Chinese 6 Chilean 5 —i Italian 14 1 Swede 24 」 1 “ 3 Spanish 23 O Bulgarian 4 白 Finn 8 —r 1 已 ^ • 。C I W Swiss 25 Austrian 2 ”| 1 ^ American 26 」 绞 New Zealander 19 —i~| Nigerian 20 」 • ^ Australian 1 German 9 Figure 2. Dendogram of hierarchical cluster analysis of the three forms of emotional expressions for the emotion of anger g Rescaled Distance Cluster Combine CASE 0 5 10 15 20 25 Label Num + + + + + +

Botswanan 3 — Swede 24 Guatemalan 11 」 Costa Rican 7 1 Nigerian 20 Austrian 2 1~ Swiss 25 1 —I Venezuelan 27 1 German 9 ^ Chilean 5 —i Pole 21 」 1 Malawian 16 ~|~ Mexican 17 * ~ Bulgarian 4 1 Chinese 6 —i Yugoslav 28 Zimbabwean 30 」 New Zealander 19 —i “ American 26 1 frj Australian 1 」 3 Spanish 23 —. —' g^ Zambian 29 — £3 P Italian 14 —— 二 (Tl Dutch 18 」 X Finn 8 • ^ Greek 10 -i—i Japanese 15 」 1 j^, r-h Hong Kongese 12 “ 1 V： Indian 13 Portuguese 22 Figure 3. Dendogram of hierarchical cluster analysis of the three forms of emotional expressions for the emotion of joy gs _

_ CUHK Libraries •••III QOmMMblD