- Email: [email protected]
Cultural differences in perceived social norms and social anxiety
ARTICLE IN PRESS
Behaviour Research and Therapy 44 (2006) 1187–1197 www.elsevier.com/locate/brat
Cultural differences in perceived social norms and social anxiety Nina Heinrichsa,, Ronald M. Rapeeb, Lynn A. Aldenc, Susan Bo¨gelsd, Stefan G. Hofmanne, Kyung Ja Ohf, Yuji Sakanog a
Institute of Psychology, Technical University of Braunschweig, Spielmannstr. 12a, 38106 Braunschweig, Germany b Macquarie University, Australia c University of British Columbia, Canada d University of Maastricht, The Netherlands e Boston University, USA f Yonsei University, Korea g Waseda University, Japan Received 12 April 2005; received in revised form 10 August 2005; accepted 6 September 2005
Abstract Cultural considerations in social anxiety are a rarely investigated topic although it seems likely that differences between countries in social norms may relate to the extent of social anxiety. The present study investigated individuals’ personal and perceived cultural norms and their relation to social anxiety and fear of blushing. A total of 909 participants from eight countries completed vignettes describing social situations and evaluated the social acceptability of the behavior of the main actor both from their own, personal perspective as well as from a cultural viewpoint. Personal and cultural norms showed somewhat different patterns in comparison between types of countries (individualistic/collectivistic). According to reported cultural norms, collectivistic countries were more accepting toward socially reticent and withdrawn behaviors than was the case in individualistic countries. In contrast, there was no difference between individualistic and collectivistic countries on individuals’ personal perspectives regarding socially withdrawn behavior. Collectivistic countries also reported greater levels of social anxiety and more fear of blushing than individualistic countries. Significant positive relations occurred between the extent to which attention-avoiding behaviors are accepted in a culture and the level of social anxiety or fear of blushing symptoms. These results provide initial evidence that social anxiety may be related to different cultural norms across countries. r 2005 Elsevier Ltd. All rights reserved. Keywords: Culture; Anxiety; Norms; Behavior; Blushing
Introduction The past decade has seen a growing recognition of the intricate interplay between culture and psychopathology. Some research has focused on psychopathologic manifestations of a specific disorder across cultures (Draguns & Tanaka-Matsumi, 2003) while in other studies, the focus has been on comparing Corresponding author. Tel.: +49 531 391 2808; fax: +49 531 391 8105.
E-mail address: [email protected] (N. Heinrichs). 0005-7967/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.brat.2005.09.006
ARTICLE IN PRESS 1188
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
disorder-typical symptoms across cultures (e.g., Kleinknecht, Dinnel, Kleinknecht, Hiruma, & Harada, 1997). In addition, Draguns and Tanaka-Matsumi (2003) concluded that, although culture has a considerable impact upon psychopathology, there is a lack of knowledge about ‘‘kinds of features or dimensions of culture’’ that ‘‘are implicated in generating the distinctive manifestations of disturbance of a given time and place’’ (p. 767). It is the purpose of the present study to compare social anxiety across cultures and to examine the association between social anxiety and one such cultural feature, social behavioral norms (could also say ‘‘norms for social behavior’’). In considering the relationship between culture and social anxiety, there are two competing hypotheses. First, it might be expected that countries1 that hold clear and stringent norms about socially appropriate behavior will be characterized by lower levels of social anxiety than countries in which social behavior is less norm governed. In cultures where unequivocal social norms are held by the group, individuals know precisely what is expected from them in a social situation, and this in turn may help to minimize social distress. On the other hand, the opposite may be true. Specifically, levels of social anxiety may be higher in countries in which social norms are clear and extensive than in countries with more relaxed social norms because the consequences of breaking these norms are considerably greater. Hofstede (1984, 2001) introduced the dimension of ‘‘Uncertainty avoidance’’. This is the extent to which members of a country can tolerate ambiguous situations. These are usually situations which are novel or unknown and may include some uncontrollability. This intolerance of ambiguity at the cultural level has been measured with the Uncertainty Avoidance Index and it is strongly related to anxiety. Thus, there is at least some evidence that a cultural phenomenon is related to anxiety and well-being on an individual level. Little attention has been focused on the question of which countries hold more stringent social norms and expectations. However, another related concept that has been given a considerable degree of attention in cross-cultural research is the notion of individualism/collectivism (Hofstede, 1984, 2001; Singelis, 1994). In collectivistic cultures, harmony within the group is the highest priority and individual gain is considered to be less important than improvement of the broader social group. Thus, in collectivistic countries it is likely that more overt social norms will exist to maintain social harmony. In contrast, in individualistic societies, individual achievements and success receive the greatest reward and social admiration. Hence, social norms may be less frequent and less overt since fitting into society is a less valued concept. Lucas, Diener, Grob, Suh and Shao (2000) demonstrated that social contacts serve different purposes in individualistic versus collectivistic cultures. In individualistic cultures, individual feelings and thoughts determine behavior more directly. In collectivistic cultures, harmony within the group is the highest priority, and norms and role expectations impact on behavior considerably. Thus, in collectivistic cultures more rules and guidelines for social behavior should exist that make social slips more obvious than in individualistic cultures. In continents and countries like Asia, South America, or Pacific Islands, strict social rules are supposed to be provided about what behavior is appropriate in certain social situations (e.g., Argyle, 1986). If an individual deviates from these social rules, they are threatened by sanctions, such as exclusion from the group. It is therefore highly relevant for individuals in such countries that their social behavior is evaluated as appropriate and positive (Suh, Diener, Oishi & Triandis, 1998). Further, norms are strong predictors of life satisfaction in collectivistic but not individualistic countries (Suh et al., 1998). A concept that is related to social anxiety is embarrassment (APA, 1994). Again, it is possible that embarrassment and, more specifically, the fear of embarrassment is related to social norms in a culture. Singelis and Sharkey (1995) have suggested that it is easier to embarrass individuals from South-East Asia because more rules for social behavior exist there. Asian individuals should therefore be more concerned about their social behavior because social deviations are easier to detect. Other authors have also suggested that embarrassment is more common in collectivistic cultures because it is induced by external sanctions whereas
1
Although often done, considering countries as the unit of analysis is certainly too global because cultures are not equal to countries (Triandis, 1995). However, since nationality is easier to assess than culture and prevalence rates usually have been reported by countries and not by cultures, we use these terms interchangeably for the purpose of present study. The reader who is interested in the differences between these two concepts and the implication for genuine cross-cultural research is referred to Triandis (1995).
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1189
guilt and self-blame are more common in individualistic cultures because they are induced by internal sanctions (Scherer, Wallbott & Summerfield, 1986; Wallbott & Scherer, 1995). Thus, there is some evidence and considerable conjecture regarding different social norms between collectivistic societies including South-East Asian, and South American, and individualistic societies such as most Western countries. To our knowledge, no studies have yet been undertaken specifically comparing perceived individual and social norms between such societies. However, at least one study has assessed levels of social anxiety between countries (Kleinknecht et al., 1997). Although comparisons between societies are difficult to conduct due to possible differences in interpretation of questionnaire items, these authors reported means on social interaction anxiety of 24.6 (11.6) for participants from the USA and 29.8 (12.6) for Japanese participants. Close to one half standard deviation difference is actually quite large across a population. Therefore, it is possible that these data are consistent with the suggestion that stronger social rules in a society lead to greater levels of social anxiety. But these results clearly need to be replicated in a larger sample. In sum, it is not yet clear whether cultural factors may shape levels of social anxiety. There is little clear evidence relating to levels of symptoms of social anxiety or embarrassment across cultures but at least some evidence has suggested possible higher levels of social anxiety and a higher social significance of embarrassment in collectivistic relative to individualistic cultures. In addition, work is needed to determine whether social behavioral norms contribute to the hypothesized differences in social anxiety across countries. The first aim of the present study was to examine the extent to which social reticence vs. expressions of individuality are accepted or rejected by individuals in a given country. The second aim was to compare levels of social anxiety symptomatology and fears of blushing as an indicator of fear of visible embarrassment between countries. Finally, we put these measures together and aimed to investigate how the level of acceptance of social behavior relates to levels of social anxiety, and blushing. We expected that individuals from different countries would vary in the extent to which they accept social reticence or expression of individuality. We predicted that individuals from collectivistic countries would be more accepting toward the expression of social reticence and less accepting toward the expression of individuality on a personal as well as on a perceived cultural level. We also predicted that individuals from collectivistic countries would report greater social anxiety and fear of embarrassment than individuals from individualistic countries. And finally, we expected social anxiety and fear of blushing to be positively related with acceptance of social reticence on a personal as well as on a perceived cultural level. Method Participants Participants were in total 909 undergraduates in psychology at eight different universities with each university representing one country, including Australia (Macquarie University, Sydney; n ¼ 64), Canada (University of British Columbia, n ¼ 141), Germany (Technical University of Braunschweig, n ¼ 106), Japan (Waseda University, n ¼ 126), Korea (Yonsei University, n ¼ 159), the Netherlands (University of Maastricht, n ¼ 104), Spain (Universidad Complutense Madrid, n ¼ 109) and USA (Boston University, n ¼ 100). Inclusion criteria for the study were: (1) the participant and both parents were born in the respective country, and (2) aged between 18 and 22 years. Attempts were made across countries to collect relatively similar proportions of each age to minimize any differences in age. Similarly, researchers in each country attempted to collect about 75% female participants because more psychology students are female than male. Organization of participant countries Selection of countries for participation in the study was based on convenience. Each researcher in the study contributed data from one country and volunteered their participation based on interest. We classified the countries into two groups (collectivistic and individualistic) based upon a continuum of individualism/collectivism described by Hofstede (1984, 2001; see also Arrindell et al., 2004). In a work-goal related study, Hofstede (1984) identified four dimensions of culture including the dimension
ARTICLE IN PRESS 1190
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
individualism–collectivism. In order to simplify analyses, we conceptualized the countries in the present study in to two groups based on Hofstede’s data placing countries along this dimension. The individualistic group included USA, Australia, Canada, the Netherlands, and Germany. The collectivistic group included Japan, Spain and Korea. Age and gender differed significantly across country group, tð907Þ ¼ 11:0, po:001, and w2 ð1Þ ¼ 4:2, po:05, respectively. Students in individualistic countries (M ¼ 19:5, SD ¼ 1:3) were significantly younger than in collectivistic countries (M ¼ 20:3, SD ¼ 1:0). The sample from individualistic countries also had a slightly higher proportion of females (75%) than the one from collectivistic countries (69%). Measures Social behavior vignettes We developed specifically for the current study 19 hypothetical vignettes describing the social behavior of individuals in a variety of social situations. Some of the vignettes were developed to tap social assertive, attention-seeking social behavior (e.g., ‘‘You are sitting in a crowded bus. The man in front of you suddenly gets up and shouts loudly across the bus to a second person that they both will get off at the stop after the next.’’). Others were written to tap socially withdrawn, attention-avoiding social behavior (e.g., ‘‘You are sitting in a math class. The lecturer writes a problem on the board and asks if anybody can solve the problem. You can see that the woman sitting next to you has already worked out the problem but she does not step forward.’’). Vignettes were contributed from the whole research group. The lead authors (NH and RR) began to collect vignettes based upon their initial conception of the construct of interest (i.e., societal norms that may ultimately have an influence on levels of social anxiety). The remaining authors later added vignettes and comments. The construct under investigation developed gradually across this process of collecting vignettes and discussing them among the authors. Each author was required to consider the intended themes of the vignettes with respect to relevance to his/her country (e.g., in terms of activities involved) and to add more vignettes to the initial core. Participants were asked to read each vignette and provide a response to each of three different questions: the respondent’s personal judgment about how appropriate the displayed behavior was (individual norms for social adequacy), the respondent’s degree of positivity/negativity toward the vignette’s main actor (social evaluation), and how appropriate the main actor’s behavior would be in the respondent’s culture (cultural norms for social adequacy). Each question was answered on a 6-point Likert-type scale (0–5). Social anxiety (Social Interaction Anxiety Scale, SIAS, Mattick & Clarke, 1998) The SIAS assesses fears of general social interaction. Social interaction anxiety as measured with the SIAS refers to distress when meeting and interacting with other people. The central concern surrounds fears of being inarticulate, boring, sounding stupid, not knowing what to say or being ignored. The scale has high levels of internal consistency (Cronbach’s alpha in a sample of undergraduates .88 and in patients with social phobia .93, and test-retest-reliability r ¼ :92 for 4 and 12 weeks, respectively) and established validity (e.g., high convergent and satisfactory divergent validity with other anxiety and mood measures). In the present study, internal consistencies varied across countries from .85 (Germany, Spain) to .93 (Japan). Fear of blushing (Blushing Questionnaire, B-Q; Bo¨gels & Reith, 1999) The B-Q measures a fear of blushing. Six questions surrounding the topic of blushing are scored by means of a visual analogue scale (0-100). A blushing phobia is usually assessed if the mean score of the six items exceeds 50. Cronbach’s alpha for this questionnaire is .95. In the present study, internal consistencies varied across countries from .83 (USA) to .91 (Korea). Translation of measures For the five non-English speaking countries in the present study, the principal researcher for each site translated the vignettes and the blushing questionnaire into the native language of their respective country. Back-translation was only conducted in Germany due to the lack of financial support for this study.
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1191
Procedure Participants were recruited via a notice placed on university information boards or announcements in classes. In most countries (Australia, USA, Canada, Korea, Germany in part), undergraduates received course credit for their voluntary participation. Students were tested in class groups (in Australia, USA, Korea, Spain, Japan, Germany in part, The Netherlands in part) or individually (Canada, Germany in part, The Netherlands in part). Upon arrival, they were asked to read an information sheet about the study. Subsequently, they completed the package of questionnaires. Participation was anonymous and each respective university ethics committee approved the methodology where required. Results Development of vignettes to assess perceived social norms2 We originally developed 19 vignettes describing different behaviors displayed in social situations. Three vignettes were dropped based on consensus due to lack of convergence with the others, leaving 16 vignettes. Analyses of responses to the three questions relating to each vignette indicated high correlations between the first two questions (individual norms for social adequacy and social evaluation, r ¼ :80). Responses to these two items for each vignette were averaged because they were not sufficiently independent for separate dimensions. We were therefore left with two dimensions across the 16 vignettes: the respondent’s personal view of the appropriateness of the behavior (which we labeled Personal Norm) and the extent to which the respondent believed their culture would view the behavior as appropriate (which we labeled Cultural Norm)3. Both scales consisted of 16 items with each scale yielding a range from 0 to 80. Lower scores indicate a preference for attention-avoiding social behaviors while higher scores indicate a preference for attention-seeking behavior in the personal and the cultural norm. Cronbach’s alpha for the Cultural Norm Scale was good for a cross-cultural instrument (a ¼ :71; range across countries: .38–.65), the Personal Norm Scale, however, was less than satisfactory (a ¼ :40, range across countries: .30–.53). The Personal and Cultural Norm scales were not significantly correlated in the entire sample, rð909Þ ¼ :05, ns. When running separate analyses for country groups, individualistic countries demonstrate the same zero correlation (r ¼ :04, ns), whereas a significant association in collectivistic countries was found (r ¼ :13, po.02). Differences between countries on acceptance of social behaviors Mean scores and standard deviations of the personal and cultural norms may be found in Table 1. In line with our hypothesis, an overall multivariate analysis of variance with the two norm scales as dependent variables, country group as independent variable and age as a covariate indicated a significant multivariate main effect of country group, F ð2; 905Þ ¼ 100:1, po:001, with a large effect size partial eta2 ¼ .18. Furthermore, age yielded an overall significant main effect, F ð2; 905Þ ¼ 4:2, po:02, with a small effect size partial eta2 ¼ .01. As expected, follow-up univariate F-tests showed that there were differences across country groups in scores on the cultural norm scale, F ð1; 906Þ ¼ 193:6, po:001, partial eta2 ¼ :176 Collectivistic countries scored significantly more accepting toward socially avoidant behavior relative to individualistic countries. The effect size was quite large with more than 17% of the variance between country groups explained by this variable. Unexpectedly, however, country groups did not differ significantly in their personal norms regarding the social acceptance of attention-seeking social behavior, F ð1; 906Þ ¼ 2:3, p4:12, partial eta2 ¼ :003. Individuals from collectivistic and individualistic countries were similarly accepting toward socially avoidant behavior in 2 The vignettes may be requested in English or German via the first author ([email protected]). For all other languages contact the co-author responsible for the country/language of interest. 3 This way the Personal Norm score is based on two items, i.e., evaluation as well as personal acceptability, whereas the cultural norm score is based on one question. We therefore decided to also analyze the data computing both scores in the same way, i.e., just using the personal/cultural acceptability questions. A very similar pattern of results occurred than what is described subsequently.
ARTICLE IN PRESS 1192
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
Table 1 Mean scores (and standard deviations) of the personal and cultural norm scales, social anxiety (SIAS), and blushing (BQ; N ¼ 909) Dimension
Individualistic (N ¼ 515)
Collectivistic (N ¼ 394)
Personal norm (0–80) Cultural norm (0–80) Social anxiety Blushing
36.3 42.0 23.2 25.0
37.0 35.0 27.8 37.6
(5.5) (7.1) (12.8) (19.1)
(4.8) (8.3) (14.1) (22.0)
Note: Personal and Cultural Norm Scale: higher scores indicate more acceptance toward attention-seeking social behavior, lower scores indicate a preference for attention-avoidance. In the group of individualistic countries, there was one questionnaire score missing for the SIAS and the BQ.
their personal view. Similarly, the significant main effect of age only maintained for the cultural but not for the personal norm scale in the follow-up tests with F ð1; 906Þ ¼ 8:2, po:01, partial eta2 ¼ :01, and F ð1; 906Þ ¼ 0:46, p4:49, partial eta2 ¼ :001, respectively. To ensure that differences in gender distribution did not impact these results, we repeated the analyses separately for each gender. The same results were found in males and females on the main dependent variables (personal and cultural norms), although for males age occurred to not play a significant role in contrast to females4. Furthermore, we calculated a discrepancy score (personal norm minus cultural norm) for each individual to explore whether those in collectivistic countries tend to show less discrepancy than those in individualist. In fact, the mean discrepancy score was significantly lower in collectivistic (M ¼ 2:0, SD ¼ 9:1) than in individualistic countries (M ¼ 5:6, SD ¼ 8:8), tð907Þ ¼ 12:8, po:001. Finally, we analyzed whether countries within country type (individualist vs. collectivist) varied on the personal or cultural norm scales. For individualistic countries, the overall univariate effects for country were F ð4; 509Þ ¼ 7:8, po:001 (personal norm scale) and F ð4; 509Þ ¼ 3:6, po:01 (cultural norm scale); and for collectivistic countries, the overall univariate effect for country were F ð2; 390Þ ¼ 9:4, po:001 (personal norm scale) and F ð2; 309Þ ¼ 163:5, po:001 (cultural norm scale). Within the individualist country type, Germany led the rank of countries in terms of acceptance of attention-seeking behavior, followed by the Netherlands and Australia, Canada and the USA with only slight differences in ranking between personal and cultural norms. Within the collectivist country type, on the personal norm scale Korea scored highest in ranking toward acceptance of attention-seeking behavior, followed by Spain and Japan (almost identical). On the cultural norm scale, Spain led the ranking toward acceptance of attention-seeking behaviors, followed by Korea and Japan with fairly large mean differences between each position. Differences between country groups in social anxiety, and blushing As done with the personal and cultural norms, a multivariate analysis of variance was conducted with country group as the independent variable, social anxiety, and blushing as dependent variables and age as covariate. The norm scales were also entered into the equation to control for the number of comparisons although means and significant country group differences were already reported separately. Table 1 demonstrates means and standard deviations for social anxiety, and blushing. As predicted, again a multivariate main effect of country group and age occurred, F ð4; 902Þ ¼ 70:3, po:001, partial eta2 ¼ :24 and F ð4; 902Þ ¼ 3:3, po:02, partial eta2 ¼ :01, respectively. Follow-up univariate F-tests indicated significant country group effects on both anxiety scales, F ð1; 905Þ ¼ 32:5, po:001, partial eta2 ¼ :04 for the SIAS and F ð1; 905Þ ¼ 32:5, po:001, partial eta2 ¼ :04 for the BQ. Age only significantly impacted social anxiety but not fear of blushing, F ð1; 905Þ ¼ 6:0, po:02, partial eta2 ¼ :007 and F ð1; 905Þ ¼ 1:6, p4:20, partial eta2 ¼ :002, respectively, indicating that individuals from collectivistic countries reported more social anxiety and more 4
We also re-analyzed the data with an age and gender-matched subsample of the total sample. These analyses yielded the same pattern of results.
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1193
fear of blushing than individuals from individualistic countries. Age only mattered for social anxiety but did not neutralize the differences between country groups. To ensure that differences in gender distribution did not impact these results, we repeated the analyses separately for each gender. The same results were found in males and females on the main dependent variables (social anxiety and fear of blushing), although for males age occurred to not play a significant role for social anxiety in contrast to females4. Relationships between variables The correlations between the norm scales and social anxiety, and blushing are shown in Table 2. Both types of norms were significantly correlated with social anxiety and fear of blushing. The more accepting the perceived culture and the individual person were toward attention-seeking social behavior, the lower the social anxiety and fear of blushing. In addition, we calculated the correlation between social anxiety and blushing. Blushing was positively associated with social anxiety (r ¼ .40, po.001). Discussion The first aim of the present study was to examine the extent to which certain social behaviors reflecting reticence and modesty at one end versus self promotion and attention seeking at the other are perceived positively by individuals in different countries. When examining participants’ reports of the norms reflective of their own culture, strong differences were found across groups of countries. Specifically, collectivistic countries were said to support behaviors reflecting withdrawal and modesty whereas individualistic countries were said to support behaviors reflecting greater attention seeking. In contrast, there was no difference between groups of countries on participants’ personal views of these behaviors. Participants from both groups of countries appeared to hold equally negative views of individuals who acted in more attention seeking ways relative to those who acted in more attention avoiding ways. Thus participants’ perceptions of the cultural norms held in their countries support the view that collectivistic societies are more positive toward socially withdrawn behaviors while individualistic countries are more positive toward socially extraverted behaviors. However, these differences are not reflected in participants’ own personal views of these behaviors. This pattern was found in both genders and independent of age. The relevance of the statistical significant effect of age for the cultural scale occurred to be small with a maximum of 1% of variance accounted for in the present age range. It was surprising that the two measures used in the study, participants’ reports of their countries’ cultural norms regarding social behavior and participants’ personal views of social behavior showed so little agreement. In fact, the correlation between these two measures approached zero. Clearly participants in the current study saw their own personal views as quite distinct from the norms reflected in their greater societies. Unfortunately, there is little empirical literature on how social norms emerge and transmit (Cialdini & Trost, 1998). It is possible that this effect reflects the age and educational levels of the study’s participants. Participants in the current study were a relatively educated and privileged group, being recruited from university populations. Individuals with greater education will often view themselves as holding different views
Table 2 Associations between perceived norms, social anxiety (SIAS), and blushing (BQ; N ¼ 908)
Social anxiety Blushing
Personal norm for social behavior
Cultural norm for social behavior
.15** .12**
.16** .17**
Note: Personal Norm Scale: Higher scores indicate more personal approval of attention-seeking social behavior. Cultural Norm Scale: Higher scores indicate more believed cultural approval of attention-seeking social behavior. One individual did not fully complete the Social Interaction Anxiety Scale, or the Blushing Scale, respectively. **Po.001.
ARTICLE IN PRESS 1194
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
to those of the broader society (Wagner & Zick, 1995) and as being the ones expected to shape attitudes within a society rather than reflect them. Similarly, the late adolescent/ young adult developmental level reflected in the current study’s participants is typically a time of rebellion and formation of individualized attitudes. Despite the mean difference across scales and the overall near zero correlation between personal and cultural norms, it is interesting that when analyzed separately, in fact a positive relation was established between the personal and the cultural perspective within collectivistic countries. This suggests that individuals from collectivistic countries report more consistent responses on the two measures than those from individualistic countries. In addition, the difference between personal and cultural perceived norms is less in collectivistic than in individualistic countries. Individuals in collectivistic countries ‘‘fit in’’ to the perceived cultural norm more. The correlation showed that there is greater agreement between personal and cultural norms in collectivistic countries while the mean discrepancy showed that there is also less difference between these norm types in collectivistic countries. Interestingly, in individualistic countries participants’ personal norms were less positive toward attentionseeking behavior than their perceived cultural norms meaning that students in those countries perceived themselves to be less extreme along the dimension than other people in the same culture. The inconsistency between self-reported personal and cultural norms might at least partially reflect the perception of one’s position relative to other people in the same culture. For example, when a student in the USA reports that she is ‘‘accepting of attention-avoiding social behavior’’, she might mean that she is so in comparison to other people in the USA and she might in fact be not so accepting toward these behaviors in absolute terms. Clearly, replication of these results with more representative population samples would be of value. Data from the personal norm items also needs to be interpreted more cautiously than data from the cultural norm scale due to limits to the psychometrics of the items. Cronbach’s alpha was satisfactory for the cultural norm scale but not for the personal norm scale. Alphas between .60 and .70 are not unusual for instruments employed in cross-cultural research (e.g., Singelis, 1994). Thus the internal consistency of .71 for the scale reflecting cultural norms for social behavior indicated reasonable homogeneity of items. In contrast, the internal consistency of .40 for the personal view scale indicated that individuals’ personal perceptions of the acceptance of the various social behaviors differed considerably within themselves. Another explanation of the unexpected cross-cultural pattern in the personal views of social behaviors may therefore be the low internal consistency of the scale. Thus, it is difficult to confidently interpret the differences between the personal and the cultural view. It is possible that individuals’ views of the acceptance of others’ social behaviors are based more on specific contextual information or on personal experiences and this may be a less coherent construct to assess. It is also possible that the social desirability that affects much of the research into attitudes is more strongly influential for items relating to personal views than for items about general cultural perspectives (Kim, Sarason, & Sarason, in press; Smith & Mackie, 2000). Individuals are motivated to make an impression of being likable as a person (Leary, 1995; Tedeschi, 1981) but there is no need to also put this effort into conveying a ‘‘likable’’ culture. Interestingly, self-enhancement is more common among Americans than among Japanese or other Asian individuals (Kitayama, Markus, Matsumoto, & Norasakkunkit; 1997). Another interpretation of our data may therefore focus on the discrepancy between the personal and cultural view: Asian individuals are more congruent in their personal and cultural view whereas individuals from individualistic countries are more prone to view themselves as less negative than they perceive their culture. Further work is needed to develop both scales, but especially the personal view scale and it is possible that alternate methods to standard surveys need to be developed to assess personal views of others’ social behaviors. Interestingly, both scales also vary across individual countries but within country type. The individualistic country group yielded more consistent variations across both scales with Germany being ‘‘most’’ accepting toward attention-seeking behavior independent of the individual or perceived cultural norm, followed by the Netherlands, and then Australia, Canada and the USA on the personal norm and Australia, the Netherlands, USA and Canada on the cultural norm scale. In contrast, within the collectivistic country type, the perceived cultural norm scale varied widely across the three countries with Spain reporting most acceptance toward attention-seeking behavior and Japan the least. On the personal norm scale, only slight differences occurred. This may indicate at least two things (1) that collectivistic countries are more
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1195
heterogeneous in their cultural norms than individualistic countries or (2) that Spain is in fact moving toward greater individualism, at least in their cultural norm. The second aim of our research was to compare levels of symptoms of social anxiety across groups of countries. The data showed that individuals from collectivistic countries were significantly more socially anxious and fearful of blushing than individuals from individualistic countries. While our data demonstrated significant differences in reports of social anxiety between countries, whereas previous research (e.g., Kleinknecht et al., 1997) appears not to have done so, these findings are in fact in close agreement. In the present data, the difference between collectivistic and individualistic countries in reported social anxiety provided an effect size of d ¼ :34, which is similar to the effect size reported by Kleinknecht et al. 1997; d ¼ :43). Thus, earlier non-significant findings may have been due to a reduced power relative to the present study. In collectivistic countries strict social rules are provided about what behavior is appropriate in certain social situations (e.g., Argyle, 1986) and if an individual deviates from this social norm, anxiety-provoking sanctions may apply. Suh et al. (1998) have argued that it is extremely important for individuals in such countries that their social behavior is evaluated as appropriate and positive. This may explain the higher levels of social anxiety in collectivistic countries. Our data clearly contradicts the idea that obvious and consistent social rules are associated with less social anxiety. These results may also have implication for another related concept: Triandis (2004) suggested differentiating cultures along the dimension ‘‘tightness versus looseness’’. A tight culture has explicit norms and severe sanctions for violating them. No empirically based ranking of cultures on this axis exists but it may be suggested that the collectivistic countries investigated lie closer to the tighter end and the individualistic countries closer to the looser end of this dimension. Thus, similar results using this dimension might be expected in terms of relationships with social anxiety and we cannot be certain which construct is more strongly responsible for the results. Our final aim was to investigate the relationships between preferences for socially reticent behavior and levels of symptomatology. Correlation analyses showed that acceptance of attention-avoiding behaviors on both a personal and a perceived cultural level were associated with greater social anxiety and more fear of blushing. In contrast to the cross-cultural inconsistency between the personal and cultural norm scales, this finding offers a convincing pattern across scales: individuals seem to be more socially anxious if they personally are more accepting toward socially withdrawn behavior and if they believe their culture is. Obviously, the knowledge or at least the belief that a socially reticent norm exists in a country does not reduce or minimize social anxiety. This may be explained by assuming that social slips will be more obvious if social withdrawal is a norm. Adherence to social norms is more important in collectivistic than in individualistic countries (Suh et al., 1998) because sanctions from the group may not be as severe in individualistic countries than in collectivistic countries. What might these findings imply for the clinical condition of social phobia? The relationship between social anxiety as a symptom and social phobia as a disorder may warrant slightly different consideration. Despite high levels of social anxiety, a clinical diagnosis of social phobia requires the additional consideration of life interference (Rapee & Spence, 2004). Causally, it is likely that somewhat different factors influence these two aspects of the clinical diagnosis. Once again, the role of cultural influence on clinical levels of social phobia may work in either of two very different ways. First, it is possible that the prevalence of diagnosable social phobia is higher in more individualistic countries where assertiveness, individuality, and ‘‘standing out from the crowd’’ are highly valued. In such countries, social withdrawal and shyness are more likely to interfere with quality of life and be viewed as a handicap (Rapee & Spence, 2004). In contrast, in countries where it is more valued to not stand out and to fit in with the crowd, social phobia as a diagnosis may be less prevalent because some social withdrawal and shyness may be viewed positively. An alternate hypothesis is that in countries where socially withdrawn behavior is viewed positively, parents and social pressure may work to support introverted behavior during development, and as a result, shy children in the culture will be less likely to learn to overcome their shyness. In turn, this may lead to higher levels of diagnosable social phobia in collectivistic countries. Consistent with both suggestions, some research has indicated that parents in Thailand view externalizing behaviors in their children as more problematic than internalizing, whereas parents in the United States do not differentiate strongly (Weisz, Suwanlert, Chaiyasit, & Walter, 1987; Weisz, Suwanlert, Chaiyasit, Weiss, Walter, & Anderson, 1988).
ARTICLE IN PRESS 1196
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
Epidemiological data appear more consistent with the first of these suggestions. Wittchen and Fehm (2003) conclude that there exist marked differences in prevalence of social phobia across countries with the USA reporting the highest levels of clinical disorder and Korea and Taiwan reporting the lowest. Several other Western countries appear to have prevalence rates closer to those of the USA while there is some research demonstrating more modest rates in a Latin American country, Brazil. Thus, there are intriguing suggestions across epidemiological surveys that social phobia as a diagnosis is rarer in collectivistic than in individualistic countries. This seems paradoxical considering that in the present study we found higher levels of social anxiety symptoms in collectivistic than individualistic countries. However, this may indicate that despite higher levels of symptomatology, the clinical diagnosis is assigned less often in collectivistic countries because socially reticent behavior does not make a bad impression or strike somebody as very unusual. The life interference may therefore be weaker in collectivistic countries than in individualistic countries in the presence of similar levels of social anxiety. A limitation of the present study is the restricted sample of students which does not allow a generalization to other populations. Students are a highly advantaged group in comparison to other individuals in the respective country, they are younger and better educated in general. It is likely that the acceptance or rejection of norms is dependent on these variables. In fact, age occurred to be a significant factor to control for in the present despite the limited age range included. It is also important to note the relevance of the Hofstede classification system for the conclusions drawn from the present study. His approach to classifying countries is somewhat controversial. The position of Spain, for example, on the individualism–collectivism dimension is above Japan but below Germany in rank and score (Hofstede, 2001) and it is questionable on what bases to assign it to either group. In the present study, it was assigned to the collectivist group because it falls below the mean score on Hofstede’s scale while Germany was above the mean (and therefore classified into the individualistic group). Further, the data from Hofstede are unfortunately somewhat dated by now and therefore, major shifts may have occurred over time in the position of countries on this dimension. Unfortunately, we are not aware of any sources pertinent to this issue. Nevertheless, we re-analyzed the data by grouping Spain with the individualist group. The results remained generally the same with the exception of the personal norm scale which was now also significantly different between groups (with collectivistic countries being slightly more acceptable toward attention-seeking behavior in the light of overall low scores on this scale). The variance explained by this effect is small (less than 1% of variance). Furthermore, the use of back translation of measures only in Germany is a limitation of the study. It is possible that the meanings of items are not equivalent across the languages tested. However, the high internal consistencies for the SIAS and BQ specifically demonstrate very good homogeneity across countries for social anxiety indicators. Relative to the total alpha for the personal and cultural norm scales, the alpha per country is also not indicative of a specific country which is considerably worse than the others in the vignettes although the alphas vary more than with the SIAS or BQ. It is reasonable to assume that this is associated rather with the concept than with the translations because the same variation occurred between English-speaking countries (e.g., for the cultural scale alpha ranged from .46 to .62) than between non-English-speaking countries (e.g., for the cultural scale alpha ranged from .38 to .65). In sum, our study supports the idea that countries differ in cultural norms perceived by individuals for socially acceptable behavior. Simultaneously, it showed that social anxiety levels differed between these countries. We did not assess for the presence of social phobia, however, but differences in prevalence rates from epidemiological research are consistent with the notion that different cultural norms impact upon the assignment of a clinical diagnosis of social phobia. In this model, the level of social anxiety predicts the clinical diagnosis of social phobia; however, the association between these two concepts is moderated at least in part by cultural norms. It is obvious that we have to be cautious about making this connection but we believe that it may be worthwhile for future research to focus on this relation between the clinical diagnosis of social phobia and perceived cultural norms. Acknowledgement We are grateful to Christina Saalfrank for her help in collecting the data in Spain and partially in Germany under the responsibility of the first author. We also thank Dr. Manuel Mun˜oz for facilitating the coordination and participation of students in Spain.
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1197
References American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders (4th ed.). Washington: APA. Argyle, M. (1986). Rules for social relationships in four cultures. Australian Journal of Psychology, 38, 309–318. Arrindell, W. A., Eisemann, M., Oei, T. P. S., Caballo, V. E., Sanavio, E., Sica, C., et al. (2004). Phobic anxiety in 11 nations: Part II. Hofstede’s dimensions of national cultures predict national-level variations. Personality and Individual Differences,, 37, 627–643. Bo¨gels, S., & Reith, W. (1999). Validity of two measures to assess social fears: The Dutch social phobia and anxiety inventory and the fear of blushing, trembling, and sweating questionnaire. Journal of Psychopathology and Behavioral Assessment, 21, 51–66. Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity, and compliance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 151–192). New York: Oxford University Press. Draguns, J. G., & Tanaka-Matsumi, J. (2003). Assessment of psychopathology across and within cultures: Issues and findings. Behaviour Research and Therapy, 41, 755–776. Hofstede, G. (1984). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage. Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations (2nd ed.). Beverly Hills, CA: Sage. Kim, D-Y., Sarason, B.R., & Sarason, I.G. (in press). Implicit social cognition and culture: Parent–child relations and cultural values, explicit and implicit psychological acculturation, and distress of Korean-American young adults. Journal of Social and Clinical Psychology. Kitayama, S., Markus, H. R., Matsumoto, H., & Norasakkunkit, V. (1997). Individual and collective processes in the construction of the self: Self-enhancement in the United States and self-criticism in Japan. Journal of Personality and Social Psychology, 72, 1245–1267. Kleinknecht, R. A., Dinnel, D. L., Kleinknecht, E. E., Hiruma, N., & Harada, N. (1997). Cultural factors in social anxiety: A comparison of social phobia symptoms and taijin kyofusho. Behaviour Research and Therapy, 11, 157–177. Leary, M. R. (1995). Self-presentation: Impression management and interpersonal behaviour. Madison, WI: Brown & Benchmark. Lucas, R. E., Diener, E., Grob, A., Suh, E. M., & Shao, L. (2000). Cross-cultural evidence for the fundamental features of extraversion. Journal of Personality and Social Psychology, 79, 452–468. Mattick, R. P., & Clarke, J. C. (1998). Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Behaviour Research and Therapy, 36, 455–470. Rapee, R. M., & Spence, S. H. (2004). The etiology of social phobia: Empirical evidence and an initial model. Clinical Psychology Review, 24, 737–767. Scherer, K. R., Wallbott, H. G., & Summerfield, A. R. (1986). Experiencing emotion: A cross-cultural study. Cambridge: Cambridge University Press. Singelis, M. T. (1994). The measurement of independent and interdependent self. Personality and Social Psychology Bulletin, 20, 580–591. Singelis, M. T., & Sharkey, W. F. (1995). Culture, self-construal, and embarrassability. The role of self-construal. Journal of Cross-Cultural Psychology, 26, 636–644. Smith, E. R., & Mackie, D. M. (2000). Social psychology. Philadelphia: Psychology Press. Suh, E., Diener, E., Oishi, S., & Triandis, H. C. (1998). The shifting basis if life satisfaction judgement across cultures: Emotions versus norms. Journal of Personality and Social Psychology, 74, 482–493. Tedeschi, J. T. (Ed.). (1981). Impression management theory and social psychology research. New York: Academic Press. Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview. Triandis, H. C. (2004). Culture and social behavior. Boston: McGraw-Hill. Wagner, U., & Zick, A. (1995). The relation of formal education to ethnic prejudice: Its reliability, validity and explanation. European Journal of Social Psychology, 25, 41–56. Wallbott, H. G., & Scherer, K. R. (1995). Cultural determinants in experiencing shame and guilt. In J. P. Tangeney, & K. W. Fischer (Eds.), Self-conscious emotions. The psychology of shame, guilt, embarrassement, and pride (pp. 465–487). New York: Guilford Press. Weisz, J. R., Suwanlert, S., Chaiyasit, W., Walter, B. R., & Anderson (1988). Thai and American perspectives on over- and undercontrolled child behavior problems: Exploring the threshold model among parents, teachers, and psychologists. Journal of Consulting & Clinical Psychology, 56, 601–609. Weisz, J. R., Suwanlert, S., Chaiyasit, W., & Walter, B. R. (1987). Over- and undercontrolled referral problems among children and adolescents from Thailand and the United States: The wat and wai of cultural differences. Journal of Consulting & Clinical Psychology, 55, 719–726. Wittchen, H.-U., & Fehm, L. (2003). Epidemiology and natural course of social fears and social phobia. Acta Psychiatrica Scandinavica, 108(Suppl. 417), 4–18.
Behaviour Research and Therapy 44 (2006) 1187–1197 www.elsevier.com/locate/brat
Cultural differences in perceived social norms and social anxiety Nina Heinrichsa,, Ronald M. Rapeeb, Lynn A. Aldenc, Susan Bo¨gelsd, Stefan G. Hofmanne, Kyung Ja Ohf, Yuji Sakanog a
Institute of Psychology, Technical University of Braunschweig, Spielmannstr. 12a, 38106 Braunschweig, Germany b Macquarie University, Australia c University of British Columbia, Canada d University of Maastricht, The Netherlands e Boston University, USA f Yonsei University, Korea g Waseda University, Japan Received 12 April 2005; received in revised form 10 August 2005; accepted 6 September 2005
Abstract Cultural considerations in social anxiety are a rarely investigated topic although it seems likely that differences between countries in social norms may relate to the extent of social anxiety. The present study investigated individuals’ personal and perceived cultural norms and their relation to social anxiety and fear of blushing. A total of 909 participants from eight countries completed vignettes describing social situations and evaluated the social acceptability of the behavior of the main actor both from their own, personal perspective as well as from a cultural viewpoint. Personal and cultural norms showed somewhat different patterns in comparison between types of countries (individualistic/collectivistic). According to reported cultural norms, collectivistic countries were more accepting toward socially reticent and withdrawn behaviors than was the case in individualistic countries. In contrast, there was no difference between individualistic and collectivistic countries on individuals’ personal perspectives regarding socially withdrawn behavior. Collectivistic countries also reported greater levels of social anxiety and more fear of blushing than individualistic countries. Significant positive relations occurred between the extent to which attention-avoiding behaviors are accepted in a culture and the level of social anxiety or fear of blushing symptoms. These results provide initial evidence that social anxiety may be related to different cultural norms across countries. r 2005 Elsevier Ltd. All rights reserved. Keywords: Culture; Anxiety; Norms; Behavior; Blushing
Introduction The past decade has seen a growing recognition of the intricate interplay between culture and psychopathology. Some research has focused on psychopathologic manifestations of a specific disorder across cultures (Draguns & Tanaka-Matsumi, 2003) while in other studies, the focus has been on comparing Corresponding author. Tel.: +49 531 391 2808; fax: +49 531 391 8105.
E-mail address: [email protected] (N. Heinrichs). 0005-7967/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.brat.2005.09.006
ARTICLE IN PRESS 1188
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
disorder-typical symptoms across cultures (e.g., Kleinknecht, Dinnel, Kleinknecht, Hiruma, & Harada, 1997). In addition, Draguns and Tanaka-Matsumi (2003) concluded that, although culture has a considerable impact upon psychopathology, there is a lack of knowledge about ‘‘kinds of features or dimensions of culture’’ that ‘‘are implicated in generating the distinctive manifestations of disturbance of a given time and place’’ (p. 767). It is the purpose of the present study to compare social anxiety across cultures and to examine the association between social anxiety and one such cultural feature, social behavioral norms (could also say ‘‘norms for social behavior’’). In considering the relationship between culture and social anxiety, there are two competing hypotheses. First, it might be expected that countries1 that hold clear and stringent norms about socially appropriate behavior will be characterized by lower levels of social anxiety than countries in which social behavior is less norm governed. In cultures where unequivocal social norms are held by the group, individuals know precisely what is expected from them in a social situation, and this in turn may help to minimize social distress. On the other hand, the opposite may be true. Specifically, levels of social anxiety may be higher in countries in which social norms are clear and extensive than in countries with more relaxed social norms because the consequences of breaking these norms are considerably greater. Hofstede (1984, 2001) introduced the dimension of ‘‘Uncertainty avoidance’’. This is the extent to which members of a country can tolerate ambiguous situations. These are usually situations which are novel or unknown and may include some uncontrollability. This intolerance of ambiguity at the cultural level has been measured with the Uncertainty Avoidance Index and it is strongly related to anxiety. Thus, there is at least some evidence that a cultural phenomenon is related to anxiety and well-being on an individual level. Little attention has been focused on the question of which countries hold more stringent social norms and expectations. However, another related concept that has been given a considerable degree of attention in cross-cultural research is the notion of individualism/collectivism (Hofstede, 1984, 2001; Singelis, 1994). In collectivistic cultures, harmony within the group is the highest priority and individual gain is considered to be less important than improvement of the broader social group. Thus, in collectivistic countries it is likely that more overt social norms will exist to maintain social harmony. In contrast, in individualistic societies, individual achievements and success receive the greatest reward and social admiration. Hence, social norms may be less frequent and less overt since fitting into society is a less valued concept. Lucas, Diener, Grob, Suh and Shao (2000) demonstrated that social contacts serve different purposes in individualistic versus collectivistic cultures. In individualistic cultures, individual feelings and thoughts determine behavior more directly. In collectivistic cultures, harmony within the group is the highest priority, and norms and role expectations impact on behavior considerably. Thus, in collectivistic cultures more rules and guidelines for social behavior should exist that make social slips more obvious than in individualistic cultures. In continents and countries like Asia, South America, or Pacific Islands, strict social rules are supposed to be provided about what behavior is appropriate in certain social situations (e.g., Argyle, 1986). If an individual deviates from these social rules, they are threatened by sanctions, such as exclusion from the group. It is therefore highly relevant for individuals in such countries that their social behavior is evaluated as appropriate and positive (Suh, Diener, Oishi & Triandis, 1998). Further, norms are strong predictors of life satisfaction in collectivistic but not individualistic countries (Suh et al., 1998). A concept that is related to social anxiety is embarrassment (APA, 1994). Again, it is possible that embarrassment and, more specifically, the fear of embarrassment is related to social norms in a culture. Singelis and Sharkey (1995) have suggested that it is easier to embarrass individuals from South-East Asia because more rules for social behavior exist there. Asian individuals should therefore be more concerned about their social behavior because social deviations are easier to detect. Other authors have also suggested that embarrassment is more common in collectivistic cultures because it is induced by external sanctions whereas
1
Although often done, considering countries as the unit of analysis is certainly too global because cultures are not equal to countries (Triandis, 1995). However, since nationality is easier to assess than culture and prevalence rates usually have been reported by countries and not by cultures, we use these terms interchangeably for the purpose of present study. The reader who is interested in the differences between these two concepts and the implication for genuine cross-cultural research is referred to Triandis (1995).
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1189
guilt and self-blame are more common in individualistic cultures because they are induced by internal sanctions (Scherer, Wallbott & Summerfield, 1986; Wallbott & Scherer, 1995). Thus, there is some evidence and considerable conjecture regarding different social norms between collectivistic societies including South-East Asian, and South American, and individualistic societies such as most Western countries. To our knowledge, no studies have yet been undertaken specifically comparing perceived individual and social norms between such societies. However, at least one study has assessed levels of social anxiety between countries (Kleinknecht et al., 1997). Although comparisons between societies are difficult to conduct due to possible differences in interpretation of questionnaire items, these authors reported means on social interaction anxiety of 24.6 (11.6) for participants from the USA and 29.8 (12.6) for Japanese participants. Close to one half standard deviation difference is actually quite large across a population. Therefore, it is possible that these data are consistent with the suggestion that stronger social rules in a society lead to greater levels of social anxiety. But these results clearly need to be replicated in a larger sample. In sum, it is not yet clear whether cultural factors may shape levels of social anxiety. There is little clear evidence relating to levels of symptoms of social anxiety or embarrassment across cultures but at least some evidence has suggested possible higher levels of social anxiety and a higher social significance of embarrassment in collectivistic relative to individualistic cultures. In addition, work is needed to determine whether social behavioral norms contribute to the hypothesized differences in social anxiety across countries. The first aim of the present study was to examine the extent to which social reticence vs. expressions of individuality are accepted or rejected by individuals in a given country. The second aim was to compare levels of social anxiety symptomatology and fears of blushing as an indicator of fear of visible embarrassment between countries. Finally, we put these measures together and aimed to investigate how the level of acceptance of social behavior relates to levels of social anxiety, and blushing. We expected that individuals from different countries would vary in the extent to which they accept social reticence or expression of individuality. We predicted that individuals from collectivistic countries would be more accepting toward the expression of social reticence and less accepting toward the expression of individuality on a personal as well as on a perceived cultural level. We also predicted that individuals from collectivistic countries would report greater social anxiety and fear of embarrassment than individuals from individualistic countries. And finally, we expected social anxiety and fear of blushing to be positively related with acceptance of social reticence on a personal as well as on a perceived cultural level. Method Participants Participants were in total 909 undergraduates in psychology at eight different universities with each university representing one country, including Australia (Macquarie University, Sydney; n ¼ 64), Canada (University of British Columbia, n ¼ 141), Germany (Technical University of Braunschweig, n ¼ 106), Japan (Waseda University, n ¼ 126), Korea (Yonsei University, n ¼ 159), the Netherlands (University of Maastricht, n ¼ 104), Spain (Universidad Complutense Madrid, n ¼ 109) and USA (Boston University, n ¼ 100). Inclusion criteria for the study were: (1) the participant and both parents were born in the respective country, and (2) aged between 18 and 22 years. Attempts were made across countries to collect relatively similar proportions of each age to minimize any differences in age. Similarly, researchers in each country attempted to collect about 75% female participants because more psychology students are female than male. Organization of participant countries Selection of countries for participation in the study was based on convenience. Each researcher in the study contributed data from one country and volunteered their participation based on interest. We classified the countries into two groups (collectivistic and individualistic) based upon a continuum of individualism/collectivism described by Hofstede (1984, 2001; see also Arrindell et al., 2004). In a work-goal related study, Hofstede (1984) identified four dimensions of culture including the dimension
ARTICLE IN PRESS 1190
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
individualism–collectivism. In order to simplify analyses, we conceptualized the countries in the present study in to two groups based on Hofstede’s data placing countries along this dimension. The individualistic group included USA, Australia, Canada, the Netherlands, and Germany. The collectivistic group included Japan, Spain and Korea. Age and gender differed significantly across country group, tð907Þ ¼ 11:0, po:001, and w2 ð1Þ ¼ 4:2, po:05, respectively. Students in individualistic countries (M ¼ 19:5, SD ¼ 1:3) were significantly younger than in collectivistic countries (M ¼ 20:3, SD ¼ 1:0). The sample from individualistic countries also had a slightly higher proportion of females (75%) than the one from collectivistic countries (69%). Measures Social behavior vignettes We developed specifically for the current study 19 hypothetical vignettes describing the social behavior of individuals in a variety of social situations. Some of the vignettes were developed to tap social assertive, attention-seeking social behavior (e.g., ‘‘You are sitting in a crowded bus. The man in front of you suddenly gets up and shouts loudly across the bus to a second person that they both will get off at the stop after the next.’’). Others were written to tap socially withdrawn, attention-avoiding social behavior (e.g., ‘‘You are sitting in a math class. The lecturer writes a problem on the board and asks if anybody can solve the problem. You can see that the woman sitting next to you has already worked out the problem but she does not step forward.’’). Vignettes were contributed from the whole research group. The lead authors (NH and RR) began to collect vignettes based upon their initial conception of the construct of interest (i.e., societal norms that may ultimately have an influence on levels of social anxiety). The remaining authors later added vignettes and comments. The construct under investigation developed gradually across this process of collecting vignettes and discussing them among the authors. Each author was required to consider the intended themes of the vignettes with respect to relevance to his/her country (e.g., in terms of activities involved) and to add more vignettes to the initial core. Participants were asked to read each vignette and provide a response to each of three different questions: the respondent’s personal judgment about how appropriate the displayed behavior was (individual norms for social adequacy), the respondent’s degree of positivity/negativity toward the vignette’s main actor (social evaluation), and how appropriate the main actor’s behavior would be in the respondent’s culture (cultural norms for social adequacy). Each question was answered on a 6-point Likert-type scale (0–5). Social anxiety (Social Interaction Anxiety Scale, SIAS, Mattick & Clarke, 1998) The SIAS assesses fears of general social interaction. Social interaction anxiety as measured with the SIAS refers to distress when meeting and interacting with other people. The central concern surrounds fears of being inarticulate, boring, sounding stupid, not knowing what to say or being ignored. The scale has high levels of internal consistency (Cronbach’s alpha in a sample of undergraduates .88 and in patients with social phobia .93, and test-retest-reliability r ¼ :92 for 4 and 12 weeks, respectively) and established validity (e.g., high convergent and satisfactory divergent validity with other anxiety and mood measures). In the present study, internal consistencies varied across countries from .85 (Germany, Spain) to .93 (Japan). Fear of blushing (Blushing Questionnaire, B-Q; Bo¨gels & Reith, 1999) The B-Q measures a fear of blushing. Six questions surrounding the topic of blushing are scored by means of a visual analogue scale (0-100). A blushing phobia is usually assessed if the mean score of the six items exceeds 50. Cronbach’s alpha for this questionnaire is .95. In the present study, internal consistencies varied across countries from .83 (USA) to .91 (Korea). Translation of measures For the five non-English speaking countries in the present study, the principal researcher for each site translated the vignettes and the blushing questionnaire into the native language of their respective country. Back-translation was only conducted in Germany due to the lack of financial support for this study.
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1191
Procedure Participants were recruited via a notice placed on university information boards or announcements in classes. In most countries (Australia, USA, Canada, Korea, Germany in part), undergraduates received course credit for their voluntary participation. Students were tested in class groups (in Australia, USA, Korea, Spain, Japan, Germany in part, The Netherlands in part) or individually (Canada, Germany in part, The Netherlands in part). Upon arrival, they were asked to read an information sheet about the study. Subsequently, they completed the package of questionnaires. Participation was anonymous and each respective university ethics committee approved the methodology where required. Results Development of vignettes to assess perceived social norms2 We originally developed 19 vignettes describing different behaviors displayed in social situations. Three vignettes were dropped based on consensus due to lack of convergence with the others, leaving 16 vignettes. Analyses of responses to the three questions relating to each vignette indicated high correlations between the first two questions (individual norms for social adequacy and social evaluation, r ¼ :80). Responses to these two items for each vignette were averaged because they were not sufficiently independent for separate dimensions. We were therefore left with two dimensions across the 16 vignettes: the respondent’s personal view of the appropriateness of the behavior (which we labeled Personal Norm) and the extent to which the respondent believed their culture would view the behavior as appropriate (which we labeled Cultural Norm)3. Both scales consisted of 16 items with each scale yielding a range from 0 to 80. Lower scores indicate a preference for attention-avoiding social behaviors while higher scores indicate a preference for attention-seeking behavior in the personal and the cultural norm. Cronbach’s alpha for the Cultural Norm Scale was good for a cross-cultural instrument (a ¼ :71; range across countries: .38–.65), the Personal Norm Scale, however, was less than satisfactory (a ¼ :40, range across countries: .30–.53). The Personal and Cultural Norm scales were not significantly correlated in the entire sample, rð909Þ ¼ :05, ns. When running separate analyses for country groups, individualistic countries demonstrate the same zero correlation (r ¼ :04, ns), whereas a significant association in collectivistic countries was found (r ¼ :13, po.02). Differences between countries on acceptance of social behaviors Mean scores and standard deviations of the personal and cultural norms may be found in Table 1. In line with our hypothesis, an overall multivariate analysis of variance with the two norm scales as dependent variables, country group as independent variable and age as a covariate indicated a significant multivariate main effect of country group, F ð2; 905Þ ¼ 100:1, po:001, with a large effect size partial eta2 ¼ .18. Furthermore, age yielded an overall significant main effect, F ð2; 905Þ ¼ 4:2, po:02, with a small effect size partial eta2 ¼ .01. As expected, follow-up univariate F-tests showed that there were differences across country groups in scores on the cultural norm scale, F ð1; 906Þ ¼ 193:6, po:001, partial eta2 ¼ :176 Collectivistic countries scored significantly more accepting toward socially avoidant behavior relative to individualistic countries. The effect size was quite large with more than 17% of the variance between country groups explained by this variable. Unexpectedly, however, country groups did not differ significantly in their personal norms regarding the social acceptance of attention-seeking social behavior, F ð1; 906Þ ¼ 2:3, p4:12, partial eta2 ¼ :003. Individuals from collectivistic and individualistic countries were similarly accepting toward socially avoidant behavior in 2 The vignettes may be requested in English or German via the first author ([email protected]). For all other languages contact the co-author responsible for the country/language of interest. 3 This way the Personal Norm score is based on two items, i.e., evaluation as well as personal acceptability, whereas the cultural norm score is based on one question. We therefore decided to also analyze the data computing both scores in the same way, i.e., just using the personal/cultural acceptability questions. A very similar pattern of results occurred than what is described subsequently.
ARTICLE IN PRESS 1192
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
Table 1 Mean scores (and standard deviations) of the personal and cultural norm scales, social anxiety (SIAS), and blushing (BQ; N ¼ 909) Dimension
Individualistic (N ¼ 515)
Collectivistic (N ¼ 394)
Personal norm (0–80) Cultural norm (0–80) Social anxiety Blushing
36.3 42.0 23.2 25.0
37.0 35.0 27.8 37.6
(5.5) (7.1) (12.8) (19.1)
(4.8) (8.3) (14.1) (22.0)
Note: Personal and Cultural Norm Scale: higher scores indicate more acceptance toward attention-seeking social behavior, lower scores indicate a preference for attention-avoidance. In the group of individualistic countries, there was one questionnaire score missing for the SIAS and the BQ.
their personal view. Similarly, the significant main effect of age only maintained for the cultural but not for the personal norm scale in the follow-up tests with F ð1; 906Þ ¼ 8:2, po:01, partial eta2 ¼ :01, and F ð1; 906Þ ¼ 0:46, p4:49, partial eta2 ¼ :001, respectively. To ensure that differences in gender distribution did not impact these results, we repeated the analyses separately for each gender. The same results were found in males and females on the main dependent variables (personal and cultural norms), although for males age occurred to not play a significant role in contrast to females4. Furthermore, we calculated a discrepancy score (personal norm minus cultural norm) for each individual to explore whether those in collectivistic countries tend to show less discrepancy than those in individualist. In fact, the mean discrepancy score was significantly lower in collectivistic (M ¼ 2:0, SD ¼ 9:1) than in individualistic countries (M ¼ 5:6, SD ¼ 8:8), tð907Þ ¼ 12:8, po:001. Finally, we analyzed whether countries within country type (individualist vs. collectivist) varied on the personal or cultural norm scales. For individualistic countries, the overall univariate effects for country were F ð4; 509Þ ¼ 7:8, po:001 (personal norm scale) and F ð4; 509Þ ¼ 3:6, po:01 (cultural norm scale); and for collectivistic countries, the overall univariate effect for country were F ð2; 390Þ ¼ 9:4, po:001 (personal norm scale) and F ð2; 309Þ ¼ 163:5, po:001 (cultural norm scale). Within the individualist country type, Germany led the rank of countries in terms of acceptance of attention-seeking behavior, followed by the Netherlands and Australia, Canada and the USA with only slight differences in ranking between personal and cultural norms. Within the collectivist country type, on the personal norm scale Korea scored highest in ranking toward acceptance of attention-seeking behavior, followed by Spain and Japan (almost identical). On the cultural norm scale, Spain led the ranking toward acceptance of attention-seeking behaviors, followed by Korea and Japan with fairly large mean differences between each position. Differences between country groups in social anxiety, and blushing As done with the personal and cultural norms, a multivariate analysis of variance was conducted with country group as the independent variable, social anxiety, and blushing as dependent variables and age as covariate. The norm scales were also entered into the equation to control for the number of comparisons although means and significant country group differences were already reported separately. Table 1 demonstrates means and standard deviations for social anxiety, and blushing. As predicted, again a multivariate main effect of country group and age occurred, F ð4; 902Þ ¼ 70:3, po:001, partial eta2 ¼ :24 and F ð4; 902Þ ¼ 3:3, po:02, partial eta2 ¼ :01, respectively. Follow-up univariate F-tests indicated significant country group effects on both anxiety scales, F ð1; 905Þ ¼ 32:5, po:001, partial eta2 ¼ :04 for the SIAS and F ð1; 905Þ ¼ 32:5, po:001, partial eta2 ¼ :04 for the BQ. Age only significantly impacted social anxiety but not fear of blushing, F ð1; 905Þ ¼ 6:0, po:02, partial eta2 ¼ :007 and F ð1; 905Þ ¼ 1:6, p4:20, partial eta2 ¼ :002, respectively, indicating that individuals from collectivistic countries reported more social anxiety and more 4
We also re-analyzed the data with an age and gender-matched subsample of the total sample. These analyses yielded the same pattern of results.
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1193
fear of blushing than individuals from individualistic countries. Age only mattered for social anxiety but did not neutralize the differences between country groups. To ensure that differences in gender distribution did not impact these results, we repeated the analyses separately for each gender. The same results were found in males and females on the main dependent variables (social anxiety and fear of blushing), although for males age occurred to not play a significant role for social anxiety in contrast to females4. Relationships between variables The correlations between the norm scales and social anxiety, and blushing are shown in Table 2. Both types of norms were significantly correlated with social anxiety and fear of blushing. The more accepting the perceived culture and the individual person were toward attention-seeking social behavior, the lower the social anxiety and fear of blushing. In addition, we calculated the correlation between social anxiety and blushing. Blushing was positively associated with social anxiety (r ¼ .40, po.001). Discussion The first aim of the present study was to examine the extent to which certain social behaviors reflecting reticence and modesty at one end versus self promotion and attention seeking at the other are perceived positively by individuals in different countries. When examining participants’ reports of the norms reflective of their own culture, strong differences were found across groups of countries. Specifically, collectivistic countries were said to support behaviors reflecting withdrawal and modesty whereas individualistic countries were said to support behaviors reflecting greater attention seeking. In contrast, there was no difference between groups of countries on participants’ personal views of these behaviors. Participants from both groups of countries appeared to hold equally negative views of individuals who acted in more attention seeking ways relative to those who acted in more attention avoiding ways. Thus participants’ perceptions of the cultural norms held in their countries support the view that collectivistic societies are more positive toward socially withdrawn behaviors while individualistic countries are more positive toward socially extraverted behaviors. However, these differences are not reflected in participants’ own personal views of these behaviors. This pattern was found in both genders and independent of age. The relevance of the statistical significant effect of age for the cultural scale occurred to be small with a maximum of 1% of variance accounted for in the present age range. It was surprising that the two measures used in the study, participants’ reports of their countries’ cultural norms regarding social behavior and participants’ personal views of social behavior showed so little agreement. In fact, the correlation between these two measures approached zero. Clearly participants in the current study saw their own personal views as quite distinct from the norms reflected in their greater societies. Unfortunately, there is little empirical literature on how social norms emerge and transmit (Cialdini & Trost, 1998). It is possible that this effect reflects the age and educational levels of the study’s participants. Participants in the current study were a relatively educated and privileged group, being recruited from university populations. Individuals with greater education will often view themselves as holding different views
Table 2 Associations between perceived norms, social anxiety (SIAS), and blushing (BQ; N ¼ 908)
Social anxiety Blushing
Personal norm for social behavior
Cultural norm for social behavior
.15** .12**
.16** .17**
Note: Personal Norm Scale: Higher scores indicate more personal approval of attention-seeking social behavior. Cultural Norm Scale: Higher scores indicate more believed cultural approval of attention-seeking social behavior. One individual did not fully complete the Social Interaction Anxiety Scale, or the Blushing Scale, respectively. **Po.001.
ARTICLE IN PRESS 1194
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
to those of the broader society (Wagner & Zick, 1995) and as being the ones expected to shape attitudes within a society rather than reflect them. Similarly, the late adolescent/ young adult developmental level reflected in the current study’s participants is typically a time of rebellion and formation of individualized attitudes. Despite the mean difference across scales and the overall near zero correlation between personal and cultural norms, it is interesting that when analyzed separately, in fact a positive relation was established between the personal and the cultural perspective within collectivistic countries. This suggests that individuals from collectivistic countries report more consistent responses on the two measures than those from individualistic countries. In addition, the difference between personal and cultural perceived norms is less in collectivistic than in individualistic countries. Individuals in collectivistic countries ‘‘fit in’’ to the perceived cultural norm more. The correlation showed that there is greater agreement between personal and cultural norms in collectivistic countries while the mean discrepancy showed that there is also less difference between these norm types in collectivistic countries. Interestingly, in individualistic countries participants’ personal norms were less positive toward attentionseeking behavior than their perceived cultural norms meaning that students in those countries perceived themselves to be less extreme along the dimension than other people in the same culture. The inconsistency between self-reported personal and cultural norms might at least partially reflect the perception of one’s position relative to other people in the same culture. For example, when a student in the USA reports that she is ‘‘accepting of attention-avoiding social behavior’’, she might mean that she is so in comparison to other people in the USA and she might in fact be not so accepting toward these behaviors in absolute terms. Clearly, replication of these results with more representative population samples would be of value. Data from the personal norm items also needs to be interpreted more cautiously than data from the cultural norm scale due to limits to the psychometrics of the items. Cronbach’s alpha was satisfactory for the cultural norm scale but not for the personal norm scale. Alphas between .60 and .70 are not unusual for instruments employed in cross-cultural research (e.g., Singelis, 1994). Thus the internal consistency of .71 for the scale reflecting cultural norms for social behavior indicated reasonable homogeneity of items. In contrast, the internal consistency of .40 for the personal view scale indicated that individuals’ personal perceptions of the acceptance of the various social behaviors differed considerably within themselves. Another explanation of the unexpected cross-cultural pattern in the personal views of social behaviors may therefore be the low internal consistency of the scale. Thus, it is difficult to confidently interpret the differences between the personal and the cultural view. It is possible that individuals’ views of the acceptance of others’ social behaviors are based more on specific contextual information or on personal experiences and this may be a less coherent construct to assess. It is also possible that the social desirability that affects much of the research into attitudes is more strongly influential for items relating to personal views than for items about general cultural perspectives (Kim, Sarason, & Sarason, in press; Smith & Mackie, 2000). Individuals are motivated to make an impression of being likable as a person (Leary, 1995; Tedeschi, 1981) but there is no need to also put this effort into conveying a ‘‘likable’’ culture. Interestingly, self-enhancement is more common among Americans than among Japanese or other Asian individuals (Kitayama, Markus, Matsumoto, & Norasakkunkit; 1997). Another interpretation of our data may therefore focus on the discrepancy between the personal and cultural view: Asian individuals are more congruent in their personal and cultural view whereas individuals from individualistic countries are more prone to view themselves as less negative than they perceive their culture. Further work is needed to develop both scales, but especially the personal view scale and it is possible that alternate methods to standard surveys need to be developed to assess personal views of others’ social behaviors. Interestingly, both scales also vary across individual countries but within country type. The individualistic country group yielded more consistent variations across both scales with Germany being ‘‘most’’ accepting toward attention-seeking behavior independent of the individual or perceived cultural norm, followed by the Netherlands, and then Australia, Canada and the USA on the personal norm and Australia, the Netherlands, USA and Canada on the cultural norm scale. In contrast, within the collectivistic country type, the perceived cultural norm scale varied widely across the three countries with Spain reporting most acceptance toward attention-seeking behavior and Japan the least. On the personal norm scale, only slight differences occurred. This may indicate at least two things (1) that collectivistic countries are more
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1195
heterogeneous in their cultural norms than individualistic countries or (2) that Spain is in fact moving toward greater individualism, at least in their cultural norm. The second aim of our research was to compare levels of symptoms of social anxiety across groups of countries. The data showed that individuals from collectivistic countries were significantly more socially anxious and fearful of blushing than individuals from individualistic countries. While our data demonstrated significant differences in reports of social anxiety between countries, whereas previous research (e.g., Kleinknecht et al., 1997) appears not to have done so, these findings are in fact in close agreement. In the present data, the difference between collectivistic and individualistic countries in reported social anxiety provided an effect size of d ¼ :34, which is similar to the effect size reported by Kleinknecht et al. 1997; d ¼ :43). Thus, earlier non-significant findings may have been due to a reduced power relative to the present study. In collectivistic countries strict social rules are provided about what behavior is appropriate in certain social situations (e.g., Argyle, 1986) and if an individual deviates from this social norm, anxiety-provoking sanctions may apply. Suh et al. (1998) have argued that it is extremely important for individuals in such countries that their social behavior is evaluated as appropriate and positive. This may explain the higher levels of social anxiety in collectivistic countries. Our data clearly contradicts the idea that obvious and consistent social rules are associated with less social anxiety. These results may also have implication for another related concept: Triandis (2004) suggested differentiating cultures along the dimension ‘‘tightness versus looseness’’. A tight culture has explicit norms and severe sanctions for violating them. No empirically based ranking of cultures on this axis exists but it may be suggested that the collectivistic countries investigated lie closer to the tighter end and the individualistic countries closer to the looser end of this dimension. Thus, similar results using this dimension might be expected in terms of relationships with social anxiety and we cannot be certain which construct is more strongly responsible for the results. Our final aim was to investigate the relationships between preferences for socially reticent behavior and levels of symptomatology. Correlation analyses showed that acceptance of attention-avoiding behaviors on both a personal and a perceived cultural level were associated with greater social anxiety and more fear of blushing. In contrast to the cross-cultural inconsistency between the personal and cultural norm scales, this finding offers a convincing pattern across scales: individuals seem to be more socially anxious if they personally are more accepting toward socially withdrawn behavior and if they believe their culture is. Obviously, the knowledge or at least the belief that a socially reticent norm exists in a country does not reduce or minimize social anxiety. This may be explained by assuming that social slips will be more obvious if social withdrawal is a norm. Adherence to social norms is more important in collectivistic than in individualistic countries (Suh et al., 1998) because sanctions from the group may not be as severe in individualistic countries than in collectivistic countries. What might these findings imply for the clinical condition of social phobia? The relationship between social anxiety as a symptom and social phobia as a disorder may warrant slightly different consideration. Despite high levels of social anxiety, a clinical diagnosis of social phobia requires the additional consideration of life interference (Rapee & Spence, 2004). Causally, it is likely that somewhat different factors influence these two aspects of the clinical diagnosis. Once again, the role of cultural influence on clinical levels of social phobia may work in either of two very different ways. First, it is possible that the prevalence of diagnosable social phobia is higher in more individualistic countries where assertiveness, individuality, and ‘‘standing out from the crowd’’ are highly valued. In such countries, social withdrawal and shyness are more likely to interfere with quality of life and be viewed as a handicap (Rapee & Spence, 2004). In contrast, in countries where it is more valued to not stand out and to fit in with the crowd, social phobia as a diagnosis may be less prevalent because some social withdrawal and shyness may be viewed positively. An alternate hypothesis is that in countries where socially withdrawn behavior is viewed positively, parents and social pressure may work to support introverted behavior during development, and as a result, shy children in the culture will be less likely to learn to overcome their shyness. In turn, this may lead to higher levels of diagnosable social phobia in collectivistic countries. Consistent with both suggestions, some research has indicated that parents in Thailand view externalizing behaviors in their children as more problematic than internalizing, whereas parents in the United States do not differentiate strongly (Weisz, Suwanlert, Chaiyasit, & Walter, 1987; Weisz, Suwanlert, Chaiyasit, Weiss, Walter, & Anderson, 1988).
ARTICLE IN PRESS 1196
N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
Epidemiological data appear more consistent with the first of these suggestions. Wittchen and Fehm (2003) conclude that there exist marked differences in prevalence of social phobia across countries with the USA reporting the highest levels of clinical disorder and Korea and Taiwan reporting the lowest. Several other Western countries appear to have prevalence rates closer to those of the USA while there is some research demonstrating more modest rates in a Latin American country, Brazil. Thus, there are intriguing suggestions across epidemiological surveys that social phobia as a diagnosis is rarer in collectivistic than in individualistic countries. This seems paradoxical considering that in the present study we found higher levels of social anxiety symptoms in collectivistic than individualistic countries. However, this may indicate that despite higher levels of symptomatology, the clinical diagnosis is assigned less often in collectivistic countries because socially reticent behavior does not make a bad impression or strike somebody as very unusual. The life interference may therefore be weaker in collectivistic countries than in individualistic countries in the presence of similar levels of social anxiety. A limitation of the present study is the restricted sample of students which does not allow a generalization to other populations. Students are a highly advantaged group in comparison to other individuals in the respective country, they are younger and better educated in general. It is likely that the acceptance or rejection of norms is dependent on these variables. In fact, age occurred to be a significant factor to control for in the present despite the limited age range included. It is also important to note the relevance of the Hofstede classification system for the conclusions drawn from the present study. His approach to classifying countries is somewhat controversial. The position of Spain, for example, on the individualism–collectivism dimension is above Japan but below Germany in rank and score (Hofstede, 2001) and it is questionable on what bases to assign it to either group. In the present study, it was assigned to the collectivist group because it falls below the mean score on Hofstede’s scale while Germany was above the mean (and therefore classified into the individualistic group). Further, the data from Hofstede are unfortunately somewhat dated by now and therefore, major shifts may have occurred over time in the position of countries on this dimension. Unfortunately, we are not aware of any sources pertinent to this issue. Nevertheless, we re-analyzed the data by grouping Spain with the individualist group. The results remained generally the same with the exception of the personal norm scale which was now also significantly different between groups (with collectivistic countries being slightly more acceptable toward attention-seeking behavior in the light of overall low scores on this scale). The variance explained by this effect is small (less than 1% of variance). Furthermore, the use of back translation of measures only in Germany is a limitation of the study. It is possible that the meanings of items are not equivalent across the languages tested. However, the high internal consistencies for the SIAS and BQ specifically demonstrate very good homogeneity across countries for social anxiety indicators. Relative to the total alpha for the personal and cultural norm scales, the alpha per country is also not indicative of a specific country which is considerably worse than the others in the vignettes although the alphas vary more than with the SIAS or BQ. It is reasonable to assume that this is associated rather with the concept than with the translations because the same variation occurred between English-speaking countries (e.g., for the cultural scale alpha ranged from .46 to .62) than between non-English-speaking countries (e.g., for the cultural scale alpha ranged from .38 to .65). In sum, our study supports the idea that countries differ in cultural norms perceived by individuals for socially acceptable behavior. Simultaneously, it showed that social anxiety levels differed between these countries. We did not assess for the presence of social phobia, however, but differences in prevalence rates from epidemiological research are consistent with the notion that different cultural norms impact upon the assignment of a clinical diagnosis of social phobia. In this model, the level of social anxiety predicts the clinical diagnosis of social phobia; however, the association between these two concepts is moderated at least in part by cultural norms. It is obvious that we have to be cautious about making this connection but we believe that it may be worthwhile for future research to focus on this relation between the clinical diagnosis of social phobia and perceived cultural norms. Acknowledgement We are grateful to Christina Saalfrank for her help in collecting the data in Spain and partially in Germany under the responsibility of the first author. We also thank Dr. Manuel Mun˜oz for facilitating the coordination and participation of students in Spain.
ARTICLE IN PRESS N. Heinrichs et al. / Behaviour Research and Therapy 44 (2006) 1187–1197
1197
References American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders (4th ed.). Washington: APA. Argyle, M. (1986). Rules for social relationships in four cultures. Australian Journal of Psychology, 38, 309–318. Arrindell, W. A., Eisemann, M., Oei, T. P. S., Caballo, V. E., Sanavio, E., Sica, C., et al. (2004). Phobic anxiety in 11 nations: Part II. Hofstede’s dimensions of national cultures predict national-level variations. Personality and Individual Differences,, 37, 627–643. Bo¨gels, S., & Reith, W. (1999). Validity of two measures to assess social fears: The Dutch social phobia and anxiety inventory and the fear of blushing, trembling, and sweating questionnaire. Journal of Psychopathology and Behavioral Assessment, 21, 51–66. Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity, and compliance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 151–192). New York: Oxford University Press. Draguns, J. G., & Tanaka-Matsumi, J. (2003). Assessment of psychopathology across and within cultures: Issues and findings. Behaviour Research and Therapy, 41, 755–776. Hofstede, G. (1984). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage. Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations (2nd ed.). Beverly Hills, CA: Sage. Kim, D-Y., Sarason, B.R., & Sarason, I.G. (in press). Implicit social cognition and culture: Parent–child relations and cultural values, explicit and implicit psychological acculturation, and distress of Korean-American young adults. Journal of Social and Clinical Psychology. Kitayama, S., Markus, H. R., Matsumoto, H., & Norasakkunkit, V. (1997). Individual and collective processes in the construction of the self: Self-enhancement in the United States and self-criticism in Japan. Journal of Personality and Social Psychology, 72, 1245–1267. Kleinknecht, R. A., Dinnel, D. L., Kleinknecht, E. E., Hiruma, N., & Harada, N. (1997). Cultural factors in social anxiety: A comparison of social phobia symptoms and taijin kyofusho. Behaviour Research and Therapy, 11, 157–177. Leary, M. R. (1995). Self-presentation: Impression management and interpersonal behaviour. Madison, WI: Brown & Benchmark. Lucas, R. E., Diener, E., Grob, A., Suh, E. M., & Shao, L. (2000). Cross-cultural evidence for the fundamental features of extraversion. Journal of Personality and Social Psychology, 79, 452–468. Mattick, R. P., & Clarke, J. C. (1998). Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Behaviour Research and Therapy, 36, 455–470. Rapee, R. M., & Spence, S. H. (2004). The etiology of social phobia: Empirical evidence and an initial model. Clinical Psychology Review, 24, 737–767. Scherer, K. R., Wallbott, H. G., & Summerfield, A. R. (1986). Experiencing emotion: A cross-cultural study. Cambridge: Cambridge University Press. Singelis, M. T. (1994). The measurement of independent and interdependent self. Personality and Social Psychology Bulletin, 20, 580–591. Singelis, M. T., & Sharkey, W. F. (1995). Culture, self-construal, and embarrassability. The role of self-construal. Journal of Cross-Cultural Psychology, 26, 636–644. Smith, E. R., & Mackie, D. M. (2000). Social psychology. Philadelphia: Psychology Press. Suh, E., Diener, E., Oishi, S., & Triandis, H. C. (1998). The shifting basis if life satisfaction judgement across cultures: Emotions versus norms. Journal of Personality and Social Psychology, 74, 482–493. Tedeschi, J. T. (Ed.). (1981). Impression management theory and social psychology research. New York: Academic Press. Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview. Triandis, H. C. (2004). Culture and social behavior. Boston: McGraw-Hill. Wagner, U., & Zick, A. (1995). The relation of formal education to ethnic prejudice: Its reliability, validity and explanation. European Journal of Social Psychology, 25, 41–56. Wallbott, H. G., & Scherer, K. R. (1995). Cultural determinants in experiencing shame and guilt. In J. P. Tangeney, & K. W. Fischer (Eds.), Self-conscious emotions. The psychology of shame, guilt, embarrassement, and pride (pp. 465–487). New York: Guilford Press. Weisz, J. R., Suwanlert, S., Chaiyasit, W., Walter, B. R., & Anderson (1988). Thai and American perspectives on over- and undercontrolled child behavior problems: Exploring the threshold model among parents, teachers, and psychologists. Journal of Consulting & Clinical Psychology, 56, 601–609. Weisz, J. R., Suwanlert, S., Chaiyasit, W., & Walter, B. R. (1987). Over- and undercontrolled referral problems among children and adolescents from Thailand and the United States: The wat and wai of cultural differences. Journal of Consulting & Clinical Psychology, 55, 719–726. Wittchen, H.-U., & Fehm, L. (2003). Epidemiology and natural course of social fears and social phobia. Acta Psychiatrica Scandinavica, 108(Suppl. 417), 4–18.