Cultural factors in social anxiety: A comparison of social phobia symptoms and Taijin Kyofusho

Journal of Anxiety Disorders, Vol. 11, No. 2, pp. 157-177, 1997 Copyright 0 1997 Elsevier Science Ltd Printed in the USA. All rights reserved 0887-6185/97 $17.00 + 00

Pergamon

PII SO887-6185(97)00004-2

Cultural Factors in Social Anxiety: A Comparison of Social Phobia Symptdms and Taijin Kyofusho’ RONALD

A.

KL.EINKNECHT, PH.D., DALE AND ERICA E. KLEINKNECHT, Western

Washington

NATSUKI Tokyo Woman’s

NOZOMI Western

L. DINNEL, PH.D., B.A.

Giver@

HIRUMA.

M.S.

Christian

Universi@

HARADA, Washington

B.A. University

Abstract-The

present study examined two forms of culturally-defined social anxiety: social anxiety or phobia, as defined by DSM-IV; (i.e., a (concern of public scrutiny or embarrassment) and Taijin Kyofusho (TKS), a Japanese form of social anxiety centered around concern for offending others with inappropriate behavior or offensive appearance. These versions of social anxiety are also examined in relation to culturallydetermined self definition as independent and interdependent. One hundred eighty-one U.S. students and 161 students enrolled in Japanese universities were administered scales to assess social anxiety and phobia and TKS symptoms and behaviors, as well as construal of self as independent or interdependent. Factor analyses of the three scales

We thank Michiko Yusa from the Department of Foreign Languages and Literature at Western Washington University for checking the accuracy of the translations of the instruments from English to Japanese and Richard Heimberg, Richard McNally, and Walter J. Lonner, for their comments on an earlier version of this manuscript. Correspondence and reprint requests should be sent to Ronald A. Kleinknecht, Department of Psychology, Western Washington University, Bellmgbam, WA 982259089. E-mail: KNBCHT@CC. WWU.EDU. ’ We have chosen to use the spelling, Tagin Kyofusho, and the acronym, TKS, to represent this condition after Kirmayer (1991). We recognize that other authom have used alternate spellmgs of this disorder, such as Taijin - Kyofu - sho (Kasahara, 1988; Md%Jly et al., 1990). Taijin-kyafu-sho (Reynolds, 1976) and tai-jin kyofu (‘kkahashi, 1989). 157

158

R A. KLEINKNECHT

ET AL.

used to assess social anxiety yielded three factors, each clearly corresponding to the respective scales and defining TKS and DSM-defined social anxiety. A case analysis indicated that there was an approximate 50% co-occurrence between high scorers on the TKS and social phobia scales. Multiple regression analyses resulted in a different set of predictors of TKS and SPS for the U.S. and Japanese respondents. Results were interpreted as suggesting that cultural variables can mediate the expression of social anxiety but that both forms of social anxiety can he found in each sample. 0 1997 Elsevier Science Ltd

The experience of intense anxiety or fear associated with social situations in which one might be viewed or scrutinized by others appears to be a universal phenomenon (e.g., Good & Kleinman, 1985; Mesquita & Frijda, 1992), and possibly extends down the phylogenic scale to lower primates and below (Mineka & Zinbarg, 1995). This social anxiety reaction, as with all anxiety responses, is elicited by some form of perceived threat to the person. It is likely, however, that there are numerous cultural variations, both in the expression of such anxiety and in the situations and contexts in which it is elicited. That is, there is likely to be cultural variation in the perception of what constitutes social threat. Such cultural variation in social anxiety, in part, is likely a function of how a given culture shapes the way in which its members define or construe the self as the object of social threat (Kirmayer, 1991; Takahashi, 1989). Among some cultures, particularly those of East Asia, such as the Japanese, the self is defined largely by one’s familial or social group, such that one’s self is an extension of that group (Markus & Kitayama, 1991; Marsella, 1985; Singelis & Sharkey, 1995). An accomplishment or social deviation reflects directly and foremost upon the group. Individualism, self-aggrandizement, or deviation from the group is not tolerated; “the nail that stands out gets pounded down” (Markus & Kitayama, 1991). In such cultures, often referred to as “collectivist” (e.g., Kim, Triandis, Kagistcibasi, Choi, & Yoon, 1994; Markus & Kitayama, 1991; Triandis, 1989, 1995), one’s sense of self is interdependent with one’s group. The individual is defined only with reference to his or her larger group and is but a part of a larger whole. In contrast to collectivist cultures that foster an interdependent construal of self, other cultures, such as the dominant white cultures of North America, tend to foster a definition of the self as mom individualistic or independent. In those cultures, individualism is encouraged and one receives praise for standing out from the crowd. Individualistic cuhures tend to promote an independent senseor construal of self. Indeed, one’s sense of self is defined by one’s unique, individual characteristics and abilities. The sense of self is developed through achieving independence ti-om others and is conceived of as an autonomous, bounded entity (Markus & Kitayama, 1991; Singelis, 1994). The person who always goes along with the crowd is said to have “no senseof self.” Thus, if one commits a socialfaux pas or misstep, it is one’s own self who is responsible. Thus, embarrassment and shame are brought primarily upon the individual.

CULTURAL

FACTORS

IN SOCIAL

ANX’iETY

159

Each of these two culturally-defined and elaborated views of the self can be seen as potential objects of social threat. However, these different views are thought to underlie differences in the form of social anxiety found in the two types of cultures. Further, these two culturally-shaped views of self are not mutually exclusive. Rather, any one person can have both independent and interdependent self-definitions that vary in strength of development and emphasis. Therefore the respective senses of self may vary in cognitive accessibility as a function of one’s cultural conditioning as well (Singelis & Sharkey, 1995; Trafimow, Triandis, & Goto, 1991; Triandis, 1995). Thus, to the extent that these two types of self-construal relate to different forms of social anxiety, some of each should be identifiable in each culture, although one form should predominate in a given culture. Our study examines these two forms of social anxiety in relation to their respective culturally-prescribed systems of self-construal as independent and interdependent (Markus & Kitayama, 1991; Singelis, 1994; Singelis & Sharkey, 1995). In so doing, we examine cultural variations in two forms of social anxiety: that defined by DSM and by Taijin Kyofusho, a Japanese variant of social anxiety and phobia. Social phobia as a diagnostic entity, separate from specific phobias, was initially described by Marks (1969) and subsequently became incorporated into DSM-III (APA, 1980) as a discrete diagnosis. The essential feature of social phobia is marked fear in social or performance situations in which one might be embarrassed or in which others might judge one to be odd or different (APA, 1994). Since being designated as a formal diagnostic category, research on social phobia has greatly expanded our knowledge of its descriptive and diagnostic features and its treatments (Brown, Heimberg, & Juster, 1995; Heimberg, Holt, Schneier, Spitzer, & Liebowitz, 1993; Heimberg, Hope, Dodge, & Becker, 1990; Heimberg, Liebowitz, Hope, & Schneier, 1995; Liebowitz, Gorman, Fyer, & Klein, 1985). Further, in contrast to earlier versions of the diagnostic manual, DSM-IV explicitly acknowledges that culture affects symptom presentation and points out that DSM social phobia has similarities to TKS, noting that it is: “A culturally distinctive phobia in Japan, in some ways resembling social phobia in DSM-IV.” (p. 849). Although it is now well recognized that cultural variation affects the expression of social phobia (APA, 1994; Kirmayer, 1991; Prince & Tcheng-Laroche, 1987), research into cultural influences relative to its etiology, phenomenology, and expression has been minimal (e.g., Good & Kleinman, 1985; Kirmayer, 1991; Kleinknecht, Dinnel, Tanouye-Wilson, & Lonner, 1994; Leung, Heimberg, Holt, & Bruch, 1994). One impediment to this cross-cultural research in psychopathology is the recognition that any diagnostic or assessment system will be inherently culture-bound to some degree and thus, category correspondence across cultures will never be perfect (Dana, 1993). Nonetheless, to more fully understand each system, it would seem important to explore in depth the differences and similarities of diagnostic categories across cultures.

160

R A. KLEINKNECHT

ET AL.

Social anxiety or phobia in Japan has been described since the 1920s. The Japanese conceptualization of social phobia is called nijin Kyojiuho (TKS), and is a form of Shinkeishitsu, or nervous character or temperament as initially described by Shoma Morita (Good & Kleinman, 1985; Reynolds, 1976). Taijin Kyofusho literally means symptoms (sho), of fear (kyofu), experienced in situations where people have face-to-face contact (taijin; Kirmayer, 1991; Takahashi, 1989). TKS is described as an obsession of shame, manifest by morbid fear of embarrassing or offending others by blushing, emitting offensive odors or flatulence, staring inappropriately, improper facial expressions, a blemish, or physical deformity (Takahashi, 1989). This fear of offending or bringing shame upon others in public results in social avoidance. The key factor in social avoidance is the fear of disrupting group cohesiveness by making others uncomfortable (Kasahara, 1988; Prince & Tcheng-Laroche, 1987). Typically, social avoidance revolves around circumstances that may require an expression of familiarity, such as talking to colleagues or to a group of close friends (Takahashi, 1989). The Japanese social phobic suffers from an extreme fear of inappropriate social interaction or appearance with familiar people. The cultural basis of the fear and avoidance then is that the behavior of one member of an in-group will bring shame or embarrassment upon the group as a whole. TKS sufferers become obsessed with the thought that they might offend others and thereby bring shame upon their social or familial group. TKS has been called a culture-bound form of social anxiety, unique to Japan (Kasahara, 1988; Kirmayer, 1991) where it is a formal psychiatric diagnostic category (APA, 1994). However, others have reported similar conditions elsewhere, including Korea (Lee, 1987) various European societies (Prince & Tcheng-Laroche, 1987), and the U.S. (McNally, Cassidy, & Calamari, 1990), suggesting that while TKS might be culturally elaborated and shaped, it is not unique to a particular culture. TKS resembles social phobia in that those who suffer from it experience anxiety in social situations in which they can be observed by others. Further, the TKS sufferer experiences anxiety and embarrassment and takes steps to cover up appearances or to avoid such social situations altogether (Kirmayer, 1991; Takahashi, 1989). For example, they will purchase facial creams that cover up blushing that might offend associates (Takahashi, 1989). As with social phobia in Western cultures, the typical age of TKS onset in Japan is in adolescence and early adulthood (Takahashi, 1989). Although epidemiologic data are not available, clinical data from Japan indicate that more males than females have the TKS condition (Kirmayer, 1991; Takahashi, 1989). Therefore, the gender distribution appears to differ from that of social phobia in the U.S., where the prevalence of females is somewhat greater than that of males (Chapman, Mamruzza, & Fyer, 1995; Rapee, 1995). There are clear similarities between TKS and social phobia as defined in DSM-N in that the major defining features of both are fear and avoidance associated with social interactions and performances lest some action or ap-

CULTURAL

FACTORS

IN SOCIAL

ANXIETY

161

pearance be judged inappropriate by observers. However, a major differentiating feature of these two forms of social anxiety is the culturally-defined meaning and behavioral consequences attached to appearing different or inappropriate. Specific symptoms and cognitions that define and maintain these disorders arise from and differ according to the social expectations of the particular culture (Good & Kleinman, 1985; Kirmayer, 1991; Kleinknecht et al., 1994). That is, TKS sufferers fear that they will offend or embarrass others, while social phobia sufferers fear that they will embarrass themselves. This distinction is critical to understanding differences between these two forms of social anxiety and derives from a culturally-determined socialization as it affects ones’ definition or construal of the nature of “self.” The Japanese conceptualization of TKS encompasses much of social fear and phobia similar to that described in DSM-IV, but also extends further. TKS is conceptualized as running a wide range of severity from the highly prevalent but mild social concerns of adolescence, through social phobia, to inordinate concern with bodily features in which the person obsesses on some imagined or exaggerated physical defect [e.g., body dysmorphic disorder (APA, 1994)]. In extreme cases, the construct of TKS extends to the level of delusional disorder - somatic type, depending on the tenacity and bizarreness of one’s belief that his or her actions or characteristics are offending or damaging others (Kasahara, 1988; Kirmayer, 1991; Takahashi, 1989). Thus, the two social anxiety conditions share some features, although differences are apparent as well. In a previous study we began developing a scale to assess TKS and to examine its relationship to social phobia as a function of acculturation, assessed by the Suinn-Lew Asian Self-identity Acculturation Scale (Suinn, RickardFigueroa, Lew, & Vigil, 1987) among Japanese Americans living in Hawaii (Kleinknecht et al., 1994). Our findings were partially consistent with cultural predictions. Acculturation to U.S. culture was inversely associated with TKS scores but revealed no relationship with scores on a social phobia scale. Further, there was no gender differential associated with TKS scores whereas females scored higher on a social phobia scale. TKS and social phobia scores were correlated .62, suggesting that these two forms of social anxiety share a common construct. In the present investigation, we extend this line of inquiry to examine these two forms of social anxiety in U.S. and Japanese university students as a function of their self-construal as independent or interdependent.

METHOD Purticipants

A total of 418 students volunteered to participate in the present study. Two hundred forty-two participants were students enrolled in undergraduate psychology courses at Western Washington University and received course credit for participation. Because social anxiety is most prevalent among young adults,

162

R. A. KLEINKNECHT

ET AL.

we decided to restrict the analysis to an age range of 18 to 26 years old. Data from 28 respondents from the U.S. were eliminated from the analysis as a result of this age restriction. In addition, the data of 33 U.S. respondents were eliminated from the analysis because they did not provide complete information on the measures. Thus, a total of 18 1 U.S. respondents (115 females, 66 males) with a median age of 19 comprised the final participant pool. One hundred seventy-six participants enrolled in universities in Tokyo (Tokyo Woman’s Christian University, Teikyo University, and Tokyo Economics College) participated for extra credit points in their courses. Japanese students were enrolled in either English classes or business classes at the undergraduate level. Data of 15 Japanese respondents were eliminated from the analysis because they did not provide complete information on the measures. A total of 16 1 Japanese respondents (89 females, 72 males) comprised the final participant pool. Japanese students ranged in age from 18 to 24 years with a median age of 20.

Measures

Five scales were completed by each respondent: Mattick and Clarke’s Social Phobia Scale (SPS; Mattick & Clark, 1989), Mattick and Clarke’s Social Interaction Anxiety Scale (SIAS; Mattick & Clarke, 1989), Singelis’ SelfConstrual Scale (SCS; Singelis, 1994), Modigliani’s Embarrassability Scale (ES, Modigliani, 1966), and a Zbijin Kyofisho Scale (TKS) modified by the principal authors based on the research of Takahashi (personal communication). Natsuki Hiruma from the Department of English of Tokyo Woman’s Christian University translated the Social Phobia, Social Interaction Anxiety, Embarrassability, and Taijin Kyojksho Scales into Japanese. Michiko Yusa, Professor Department of Foreign Languages and Literature at Western Washington University confirmed the accuracy of the translations in terms of content and construct equivalence (Dana, 1993). Theodore Singelis from Chico State University provided a Japanese translation of the Self-Construal Scale. These instruments were placed in a packet in counterbalanced order. Since the order of the instruments did not result in reliably different scores on any of the measures, the analyses reported below were collapsed across order of presentation. At the end of the packet, respondents were asked to provide demographic information including gender, age, and country/countries from which they held legal citizenship. Social Phobia Scale (SPS) and Social Interaction Anxiety Scale (SZAS). Reliability and validity measures for the Social Phobia Scale and the Social Interaction Anxiety Scale were conducted at the same time. Thus, results are reported together here. The SPS and the SIAS each contain 20 items. The SPS assessesanxiety in situations in which the individual may be observed by others

CULTURAL

FACTORS

IN SOCIAL

ANXIETY

163

(e.g., “I get nervous that people are staring at me as I walk down the street”). The SIAS assessesanxiety in social situations (e.g., “I am unsure whether to greet someone I know only slightly”). Respondents for both scales are instructed to rate each statement on a 5-point scale as it applies to them (0 = not at all characteristic or true of me, 4 = extremely characteristic or true of me). Mattick and Clarke (1989) reported Cronbach’s alphas from five samples (243 DSM-III-diagnosed social phobics, 48 1 college students, 3 15 community volunteers, and two smaller samples of agoraphobics and simple phobics) ranging from .89-.94 for SPS and .88-.93 for SIAS. Test-retest correlations for each scale exceeded .90 at intervals of up to 13 weeks. Construct validity was demonstrated by social phobics scoring higher on the SPS and the SIAS than normal groups or agoraphobics. In addition, significant correlations were found between SPS and SIAS and several measures of social interaction or performance anxiety. Heimberg, Mueller, Holt, Hope, and Liebowitz (1992) provide additional evidence of the validity and reliability of the SPS and SIAS. Self-Construal Scale (KS). The Self-Construal Scale consists of 24 items intended to measure respondents’ beliefs about the relationship between the self and others and the degree to which they see themselves as separate from others or connected with others. Thus, the SCS has two subscales: Independent Self-Construal and Interdependent Self-Construal. The Independent SelfConstrual subscale consists of 12 items that measure the degree to which self emphasizes separateness from social context (e.g., “I enjoy being unique and different from others in many respects”). The Interdependent Self-Construal subscale consists of 12 items that measure the degree to which self emphasizes involvement with the social context (e.g., “Even when I strongly disagree with group members, I avoid an argument”). A confirmatory factor analysis from two different samples provided adequate support for the two-factor structure of the SCS (Singelis, 1994). Construct validity was demonstrated by comparing Asian Americans to Caucasian Americans on each subscale. Asian Americans were characterized as interdependent and Caucasian Americans were characterized as independent (Singelis, 1994). In addition, reliability was demonstrated utilizing Cronbach’s alpha for two samples yielding reliability coefficients of .73 and .74 for the Interdependent scale and .69 and .70 for the Independent scale (Singelis, 1994). Embarrassability Scale (ES). The Modigliani

Embarrassability Scale consists of 26 items that measure general susceptibility to embarrassment (e.g., “Suppose you slipped and fell in a public place, dropping a package of groceries”). Respondents are instructed to rate each statement as it applies to them on a seven-point scale (1 = not at all embarrassing, 7 = very embarrassing). Cronbath’s alpha for tbe ES have been found to range from .87 to .89 (Cupach & Metts, 1992; Miller, 1987; Modigliani, 1966; Singelis & Sharkey, 1995). This scale has been demonstrated to covary with a variety of measures including:

164

R. A. KLEINKNEKHT

ET AL.

social anxiety, extraversion, public self-consciousness, and fear of negative evaluatation (Singelis & Sharkey, 1995; Zebb, Meyers, & Beck, 1994). Taijin Kyofusho Scale (TKS). The Taijin Kyofusho Scale is a revision of a previous version (Kleinknecht et al., 1994). The present version consists of 31 items that were modified from an extensive questionnaire developed by Takahashi (personal communcation). Items for inclusion in this scale were among those that Takahashi found most highly discriminated TKS patients in Japan from nonpatients. Others were constructed by the authors to be consistent with descriptions of TKS’s definitional symptoms (e.g., Kirmayer, 1991; Takahashi, 1989). Items reflected respondents’ concerns that they would do something or present an appearance that would offend or embarrass others (e.g., “I am afraid that when talking with others my trembling head, hands, and/or feet will offend them”) Respondents are instructed to rate each statement as it applies to them on a 7-point rating scale (1 = exactly true, 7 = totally false). Cronbach’s alphas were .93 and .92 for U.S. and Japanese participants, respectively, in the present study. Procedure Respondents were given a packet containing measures listed above in counterbalanced order followed by a demographic page. They were informed that the purpose of the study was to determine how people rated themselves in different social situations. Respondents were also informed that the questionnaire contained many descriptions or problems of “self” and were encouraged to respond to every item. They were then instructed to indicate on the space provided to the left of each statement a rating that indicated the degree to which the statement described them. They were informed that the rating scale would appear at the top of each page, that it may not be the same scale from page-to-page, and encouraged to read the scale at the top of a page before responding to any of the statements on that page. United States respondents were given the option to complete the questionnaire at that point or to complete it within a 5-day time frame. Japanese respondents were asked to complete the survey within a 5-day time frame. Analyses Our main analysis of interscale relationships is correlational, conducted across scales within each cultural group. Cross-cultural comparisons are made only in terms of patterns of correlations or variable distributions. This restriction was imposed to avoid possible methodological problems, such as culturally-based response sets that would lead to a lack of conceptual equivalence of absolute scores in the two cultures (Lonner & Ibrahim, 1996). That is, without knowing whether absolute scores from one culture are conceptually

CULTURAL

TABLE 1 GENDER DIFFERENCES M THEMEASURES OF S~CIALANXIETY FOR U.S. RESPONDENTS Measure Self-Construal: Independent Male Female Total Self-Construal: Interdependent Male Female Total Taijin Kyofisho Male Female Total Embarrassability Male Female Total Social Interaction Anxiety Male Female Total Social Phobia Male Female ‘Total

165

FACTORS IN SOCIAL ANXETY

n

M

SD

66 115 181

4.91 4.68 4.11

0.67 0.64 0.66

66 115 181

4.64 4.63 4.63

0.69 0 61 0 67

66 115 181

78.97 81.95 80.86

27.10 26.81 26.88

66 115 181

99.24 109.17 105.55

23.90 22.04 2Z.17

66 115 181

23.41 25.22 24.56

10.45 1”.28 I, 11.65

66 115 181

15.02 17.78 16.77

9.03 10.22 9.87

AND BEHAVIOR

df

t

P

179

2.20

.029

179

0.12

,902

179

0.72

,415

179

2.83

,005

179

1.Ol

,316

179

1.83

,069

equivalent in another culture, direct comparisonsare difficult to interpret (Dana, 1993) or are meaninglessand may be seenas pejorative (Good & Kleinman, 1985). For example, a Social Phobia score of 30 in the U.S. samplemay not have the samemeaningasa 30 on the translatedversion in the Japanesesample. Both similarities and differences are difficult to interpret. However, comparisonsof patterns of relationshipsamong variables in each group can be instructive concerning how, in this case,self-construalrelatesto type of social anxiety.

RESULTS United Statesand Japaneserespondentscompleted the six measuresof social anxiety and behavior - self-construal (SC; independent and interdependent), Ezijin Kyojbho (TKS), embarrassability,social -interaction anxiety (SIA), and social phobia (SP). Means and standarddeviations for each measureand for each sample,separatedby gender, can be seenin Tables 1 and 2.

166

R A. KLEINKNECHT

TABLE GENDER

DIFFERENCES

Measure Self-Construal: Independent Male Female Total Self-Construal: Interdependent Male Female Total Tagin Kyofisho Male Female Total Embarrassability Male Female Total Social Interaction Anxiety Male Female Total Social Phobia Male Female Total

2

IN THE MEASURES FOR JAPANESE

n

ET AL.

OF SOCV\L

ANXIETY

AND BEHAVIOR

RESWNDENTS

M

SD

72 89 161

4.83 4.72 4.77

0.73 0.58 0.65

72 89 161

4.35 4.49 4.43

0.80 0.65 0.72

72 89 161

89.85 91.08 90.53

30.94 27.57 29.04

72 89 161

104.17 113.34 109.24

19.79 15.22 17.95

72 89 161

31.14 28.66 29.77

13.64 11.71 12.63

72 89 161

19.93 18.64 19.22

14.71 11.44 12.98

df

t

P

159

1.04

,298

159

1.21

,230

159

0.27

,790

159

3.32

,001

159

1.24

‘217

159

0.63

‘532

Gender Differences To determine if there were gender differences on any of the social anxiety and behavior measuresfor respondentsfrom the U.S. and Japan, independent samplest testswere conducted. To account for familywise error, the alpha was adjustedto .Ol. For both U.S. and Japaneserespondents,females scoredhigher on embarrassabilitythan males(seeTables 1 and 2). All other gender comparisonsin each country failed to yield reliable differences. Factor Analysis of the Social Anxiety Scales Since this was the first examination of this version of the TKS scale, we submittedits items to exploratory factor analysisto examine whether the items formed a discriminable scaleand therefore could be conceptualized asassessing the construct. To compareTKS items to those from the SPS and SJAS, we also included in the analysis the 40 items from those scales.The 71 items were submitted to a Principal Componentsanalysis.The Scree test indicated that the

CULTURAL

FACTORS

IN SOCIAL

167

AhXIETY

most parsimonious solution would include three factors, accounting for 37% of the variance. Since the three scales were known to be correlated, we used an oblique rotation procedure (Oblim; SPSSX, 1988:~ The two national samples were initially factored separately. Observation of the resultant pattern matrices suggested that the solutions for the two samples were highly similar. Accordingly, to provide greater stability and lessen the effect of outliers afforded by a larger sample size, we report the combined analysis using all 342 participants. The pattern matrix of the rotated solution can be seen in Table 3. The first factor is clearly composed of items from the TKS scale. As can be seen, the majority of the TKS items load substantially and exclusively on this factor. Factor II is composed of items from SPS and Factor III contains items forming the SIAS. The substantial majority of items loaded on their respective single scales with loadings greater than .40. However, there are a few notable exceptions in that some items such as TKS Items 3 and 20, or SPS Item 7, show complex structures, loading on one of the other scales. Relationship Correlations

Between

Self-Constural

and Social AnxietyflKS

in Zero-Order

The SC variables and the social anxiety variables were examined in zeroorder correlations. These relationships are shown separately for U.S. and Japaneserespondentsin Tables 4 and 5, respectively. As can be seen, SelfConstrual as Independent is inversely correlated with all measuresof social anxiety for both U.S. and Japaneserespondents.That is, more independent self-construal is associatedwith lessreported social anxiety. This SC-Independent/social anxiety relationship appearsstronger in the U.S. samplefor SIA, SP, and embarrassabilitywhile its relationship with Taijin Kyofusho is stronger in the Japanesesample,although thesedifferences did not reach statistical significance. The SC as Interdependent was moderately positively correlated with SIA (r = .18), SP (r = .28), and TKS (r = .28) in the U.S. sample, but was not significantly correlated with any of thesemeasuresamongthe Japaneserespondents. Further, the differences between correlations between the two groups were significantly different. That is, SC-Interdependent X TKS and SCInterdependent X SP correlations, were significantly larger in the U.S. sample than they were in the Japanesesample(Z= 2.54 and 2.63, respectively, p < .05). Regression

Analyses

on TKS and SP

To examine whether culture, asit atTectsself-definition, mediatesthe form of social anxiety, we tested whether Taijin Kyofusho had a different setof predictor variables for U.S. and Japanesesamples.Stepwise regressionanalyseswere conducted using the measures of independent and interdependent selfconstruals, embarrassability, social interaction anxiety, and social phobia. For

168

R. A. KLEWKNECHT

FACTOR STRUCTURE (PA~ERN

ET AL.

TABLE 3 MATRIX) OF SOCIAL ANXIETY SCALES

Item

Factor I

Factor II

Factor III

.39 .35

.02 .13

-.12 .Ol

.26 .52

.46 .21

-.07 -.13

.62

.08

.Ol

.59

-.Ol

.I2

.47 .47

-.I3 .14

.ll -.04

.40

.I6

.03

TKS Scale

1. I am afraid that I may unintentionally hurt other’s feelings. 2. I tend to stop doing what I want to do if others think I am not doing it right. 3. I cannot really feel relaxed even when I chat with my friends. 4. When I see others, some times I am afraid that my looks might leave a bad impression on them. 5. Because I perceive myself as having a displeasing appearance. it bothers me to present myself to other people. 6. I am afraid that when talking with others my trembling voice will offend them. 7. Sometimes I stiffen or blush when I am with my friends. 8. I cannot help thinking how my eyes look when someone looks me in the eye. 9. I get more nervous when I see someone I know than when I see a stranger. 10. I am afraid that when talking with others my trembling head, hands and/or feet will offend them. 11. I am afraid that my presence will offend others. 12. When I talk with others, I feel ugly and fear that I bore them. 13. Because I perceive myself as being very awkward, it bothers me to present myself to other people. 14. Sometimes I cannot laugh when I talk with another person because I become very anxious and my face stiffens. 15. I am afraid my family will find out that something is wrong with me and that will trouble them. 16. At a hair dresser’s shop, I cannot stand for the hair dresser to look me in the face. 17. I feel small and feel like apologizing to others. 18. I am afraid I will blush in front of other people and as a result offend them. 19. I do not know where I should look when I talk with others. 20. I cannot really feel relaxed when I chat with strangers. 21. When I talk with my friends, I am afraid that they might point out my faults. 22. When I am with others, I sometimes feel that I am stupid and feel sorry for them for being with me. 23. I am afraid that when talking with others my stiff facial expressions will offend them. 24. I am afraid that my sweating or having nervous perspiration will offend other people. 25. I am afraid that my body odors will offend other people.

64

-.ll

.16

.66 .69 .63

.18 .19 .31

-.03 -.09 -.03

.69

.07

.08

.45

.09

-.Ol

.49

.I0

.06

.67 .72

.12 -.04

-.09 -.04

.39 .12 .58

.39 .41 .I4

-.13 .08 .02

.66

.I8

-.09

.73

-.05

-.Ol

.67

-.20

.21

.42

-.Ol

.18

CULTURAL

FACTORS

TABLE

IN SOCIAL

169

ANXIETY

3--CONTINUED

Item 26. I am afraid that my staring at other people’s body parts will offend them. 27. I am afraid that I will release intestinal gas in the presence of others and offend them. 28. I am afraid that eye to eye contact with other people will offend them. 29. When I talk to strangers, I am afraid that they might point out my faults. 30. I am afraid that my physical appearance will in some way offend others. 31. Sometimes I stiffen or blush when I am with strangers. Social Interaction Anxiery Scale I. I get nervous if I have to speak with someone in authority (teacher, boss). 2. I have difficulty making eye-contact with others. 3. I become tense if I have to talk about myself or my feelings. 4. I find difficulty mixing comfortably with the people I work with. 5. I find it easy to make friends of my own age. 6. I tense-up if I meet an acquaintance in the street. 7. When mixing socially I am uncomfortable. 8. I feel tense if I am alone with just one person. 9. I am at ease meeting people at parties 10. I have difficulty talking with other people. 11. I find it easy to think of things to talk about. 12. I worry about expressing myself in case I appear awkwarrt. 13. I find it difficult to disagree with another’s point of view. 14. I have difficulty talking to an attractive person of the opposite sex. 15. I find myself worrying that I won’t know what to say in social situations. 16. I am nervous mixing with people I don’t know well. 17. I feel I’ll say something embarrassing when talking. 18. When mixing in a group, I find myself worrying I will be ignored. 19. I am tense mixing in a group. 20. I am unsure whether to greet someone I know only slightly. Social Phobia Scale 1. I become anxious if I have to write in front of other people. ‘2.1 become self-conscious when using public toilets. 3. I can suddenly become aware of my own voice and of others listening to me. 4. I get nervous that people are staring at me as I walk down the street. 5. I fear I may blush when I am with others. 6. I feel self-conscious if I have to enter a room where others are already seated. 7. I worry about shaking or trembling when I’m watched by other people. 8. I would get tense if 1 had to sit facing other people on a bus or a train. 9. I get panicky that others might see me faint, or be sick Dr ill.

Factor I

Factor II

Factor III

.53

-.Ol

.18

.39

-.16

.20

.55 .66

-.16 -.05

.20 .04

.77 .49

-.04 .Ol

.02 .17

.07

.13

.34

-as .19 .17 .28 .Ol .33 a4 .21 .13 .30 -.02 -.12 .29 .26 .04 .ll .Ol .34 .17

-.05 .06 .09 .oo .I7 -.03 .ll -.19 .06 .20 .03 .I7 .08 .13 .17 -.04 .34 .Ol .31

.49 .31 .46 .40 .55 .39 .59 .44 .47 .31 .63 .62 .22 .51 .59 .59 .39 .36 .41

-.09 .24 .04

.41 .48 .50

.29 -.15 .09

.12 04 .02

.52 .43 .43

.02 -.17 .14

.ll .08 -.08

.41 .52 .40

.03 .12 .oo

(continued on next page)

170

R. A. KLEINKiWCHT

TABLE

ETAL.

3--CONTINUED

Item

Factor I

Factor II

Factor III

.09 04 .22

.67 .48 .39

.Ol -.I3 .17

.14 .13 .13 -.17 .27 .12 .05 a0

.34 .45 .46 .68 .28 .40 .49 .42

.22 .14 .22 .22 .31 .23 .I9 .22

10. I would find it difficult to drink something if in a group of people. 11. I am worried people will think my behavior odd. 12. It would make me feel self-conscious to eat in front of a stranger at a restaurant. 13. I would get tense if I had to carry a tray across a crowded cafeteria. 14. I worry I’ll lose control of myself in front of other people. 15. I worry I might do something to attract the attention of other people. 16. When in an elevator, I am tense if people look at me. 17. I can feel conspicuous standing in a line. 18. I can get tense when I speak in front of other people. 19. I worry my head will shake or nod in front of others. 20. I feel awkward and tense if I know people are watching me.

the U.S. sample, only social phobia and social interaction anxiety contributed to the prediction of TKS scores (see Table 6). These two variables accounted for 54% of the variability in the TKS scores of U.S. respondents. However, social interaction anxiety scores, social phobia scores, and independent self-construal contributed to the prediction of TKS scores for the Japanese sample accounting for 55% of the variability in the TKS scores (see Table 7). We also were interested in whether the set of predictor variables of social phobia would be different for U.S. and Japanese samples. We conducted stepwise regression analyses using the measures of independent and interdependent self-construals, embarrassability, social interaction anxiety, and Taijin Kyofisho for each sample. For the U.S. respondents, Taijin Kyofisho, social interaction anxiety, and embarrassability contributed to prediction of social phobia (see Table 8). These three factors accounted for 59% of the variability TABLE 4 CORRELATIONS FOR EACH OF THE SOCIALANXIETY AND BEHAVIOR MEASURES FOR U.S. RESFQNDENT~

Self-Construal: Independent (SC: Ind) Self-Construal: Interdependent (SC: Int) Taijin Kyofusho (TKS) Embarrassability (Emb) Social Interaction Anxiety @IA) Social Phobia (SP) *&I < .05; **p