Social anxiety and social norms in individualistic and collectivistic countries

Abstract

Background

Social anxiety is assumed to be related to cultural norms across countries. Heinrichs and colleagues [1] compared individualistic and collectivistic countries and found higher social anxiety and more positive attitudes toward socially avoidant behaviors in collectivistic than in individualistic countries. However, the authors failed to include Latin American countries in the collectivistic group.

Methods

To provide support for these earlier results within an extended sample of collectivistic countries, 478 undergraduate students from individualistic countries were compared with 388 undergraduate students from collectivistic countries (including East Asian and Latin American) via self report of social anxiety and social vignettes assessing social norms.

Results

As expected, the results of Heinrichs and colleagues [1] were replicated for the individualistic and Asian countries but not for Latin American countries. Latin American countries displayed the lowest social anxiety levels, whereas the collectivistic East Asian group displayed the highest.

Conclusions

These findings indicate that while culture-mediated social norms affect social anxiety and might help to shed light on the etiology of social anxiety disorder, the dimension of individualism-collectivism may not fully capture the relevant norms.

Although cultural values are hypothesized to have an impact on the level of social anxiety and anxiety-related impairment experienced by individuals within a culture [1], cultural research is still in its infancy [2]. The majority of studies that have addressed cultural influences on psychopathology have focused on the dimension of individualism versus collectivism, a factor shown to vary across countries [3]. Individualistic cultures are said to value the expression and assertion of individual desires whereas collectivist cultures are believed to give greater priority to the maintenance of group harmony [4, 5]. Extant findings consistently indicate that people living in East Asian countries [1, 6] as well as those that migrated from East Asian to western countries [7, 8] report higher levels of social anxiety than those in individualistic cultures.

Heinrichs and colleagues [1] postulated that cultural norms regarding social behavior may explain the relationship between cultural values and social anxiety. To address this issue, these researchers developed an innovative measure that used social vignettes to evaluate cultural norms for the acceptability of socially reticent and withdrawn versus attention-seeking and outgoing behaviors. They conducted a large cross-cultural study that compared individualistic and collectivistic countries and found that collectivistic countries (Japan, South Korea, and Spain) showed higher levels of social anxiety and greater acceptance towards socially reticent and withdrawn behaviors on a cultural level than did individualistic countries (Australia, Canada, Germany, the Netherlands, and the USA). These results may suggest that within collectivist countries, strict social norms designed to ensure group harmony may evoke social anxiety due to feared negative consequences if those norms are violated (stringent norm hypothesis).

Despite those findings, the relationship between individualistic/collectivist values and social anxiety is far from clear. A cross-cultural study by Arrindell and colleagues [9], which examined the association between mean self-reported phobic symptoms and Hofsteade’s dimensions in eleven countries, failed to find a significant association between country individualism scores and social fears. Those findings appear to be inconsistent with the conclusions of Heinrichs et al. In addition, some researchers argue that individuals in collectivist societies, particularly Asian cultures, are more willing to engage in self-criticism and hence are more willing to endorse negative attributes than individuals in individualistic societies [7, 8, 10]. According to this view, higher self-reported levels of social anxiety may reflect a response style related to collectivistic values and low independent self-construal rather than a genuine difference in social anxiety between cultures (response style hypothesis). Furthermore, some researchers found a lower prevalence of social anxiety disorder in East Asians countries [11 - 13], which has led to speculations that individuals from collectivistic countries may report more social anxiety but suffer less impairment due to cultural acceptance of socially reticent behaviors [2]. In sum, while social norms may indeed affect social anxiety, the specific relationship between collectivist values and social anxiety has yet to be identified.

Another shortcoming of the existing literature is that most cross-cultural studies of social anxiety included only East Asian countries as examples of collectivist cultures and failed to include countries from Latin America or Africa [1, 6, 7], both of which are also characterized by collectivist values. Critical analysis of the concept of individualism [14] also supports the idea that the higher levels of social anxiety and lower prevalence of social anxiety disorder found in East Asian countries may be limited to these countries and are not a general characteristic of collectivistic cultures. For example, Allik and McCrae [15] found that people in Latin American countries displayed similar patterns of personality traits to those found in European and North American countries while people in East Asian and African countries displayed different patterns. Furthermore, the study by Arrindell et al. [9] revealed that whereas Guatemala and Venezuela had the lowest individualism scores of eleven countries studied, they did not display significantly higher levels of social fears. Consequently, replication of the Heinrichs et al. [1] results in collectivistic countries other than East Asia is needed to verify whether collectivistic countries do indeed show higher mean levels of social anxiety and greater cultural acceptance of socially reticent behaviors. Therefore, the current study sought to extend previous research findings by examining two geographically distinct collectivist cultures, namely East Asian and Latin American cultures.

We chose to re-examine the results of Heinrichs et al. [1] in the Latin American culture because research points to unique Latin American cultural scripts, or social norms, that promote sociability and might therefore be relevant to social anxiety. For example, people from Mexico have been found to be more sociable and extraverted than Americans when considering behavioral data [16]: These authors found that Mexicans talked more and spent more time socializing compared to Americans in their daily lives, as measured with an Electronically Activated Recorder. These results fit the cultural script of simpatía, first described by Triandis, Marin, Lisansky, and Betancourt, [17] which is assumed to be a specific characteristic of Latin American culture. According to this cultural script, people are expected to be likeable, fun, easygoing, polite, and respectful. Avoiding conflicts and refraining from criticizing others by emphasizing positive and de-emphasizing negative behaviors are highly valued behaviors in Latin America [18]. According to these findings, Latin American cultures may provide higher rates of modeling sociable, non-anxious behaviors coupled with lower prevalence of negative learning experiences (being criticized and rejected by important others) and therefore may evoke less social anxiety than is the case in individualistic or East Asian countries. These findings would challenge the conclusions of Heinrichs et al. [1], who assumed all collectivistic countries to be characterized by higher levels of social anxiety. To the best of our knowledge, no study has investigated this issue. The results of the Arrindell et al. [9] study is consistent with the possibility that Latin American and East Asian countries may differ significantly despite their shared valuing of group harmony. However, Arrindell et al. did not directly test group differences to determine whether different collectivistic countries do indeed differ on levels of social anxiety. Research that compares Latin American collectivist countries with individualistic European and North American countries on the one hand, and with East Asian collectivist countries on the other, would provide a good test of the link between collectivist values per se and social anxiety.

Extant research also indicates that it is important to consider both gender and age when examining cultural differences. A study conducted on undergraduate students from the United States indicates that gender affected the association between social anxiety and self-construal. Specifically, independent self-construal showed negative associations with social anxiety only in men, while a reversed pattern was found in women [19]. In addition, Heinrichs et al. found a significant main effect of age when comparing collectivistic and individualistic countries on social norms. Furthermore, community studies have consistently found a higher prevalence of social anxiety disorders in women and younger populations [11]. These results highlight the need to investigate genders separately and to control for age when conducting cross cultural research on social anxiety and social norms.

In this study, we collected data from students in those countries included by Heinrichs et al. [1] to replicate results within this new sample. To address our primary question, we also collected data from two Latin American collectivistic countries, Costa Rica and Ecuador, to compare with the individualistic and East Asian countries. We predicted that the findings would replicate the results of Heinrichs et al. [1] in the individualistic and East Asian countries, i.e., we expected higher levels of social anxiety and greater cultural acceptance toward socially reticent behavior in the collectivistic East Asian countries. Based on the simpatía script specific to Latin America and the reasoning described above, we predicted that there would be less acceptance of socially reticent behaviors and lower levels of social anxiety in Latin American countries in comparison to individualistic and East Asian countries.

Method

Participants

A total of 866 undergraduates in psychology (680 women) from nine countries participated (see Table 1 for sample size for each country), including Australia (Macquarie University), Canada (University of British Columbia), Costa Rica (Universidad Latina, Universidad Costa Rica, Universidad Autonoma de Centroamerica), Ecuador (Universidad de Ecuador), Germany (Technical University of Braunschweig, University of Bielefeld, University of Magdeburg), Japan (Tokai Gakuin University, Nagoya Institute of Technology), South Korea (Yonsei University), the Netherlands (University of Maastricht), and the USA (Boston University).

Table 1.

Demographic Variables and Mean Levels of Social Norms and Social Anxiety

	N	Age M (SD)	% Female	Individualism Score1	Cultural Norm (0 – 85) M (SD)	Social Anxiety (0 – 80) M (SD)
Individualistic Country Group	478	19.60 (1.36)	75.5%	81.6	48.15 (7.88)	21.64 (12.85)
Australia	78	18.95 (1.04)	61.5%	90	47.67 (7.13)	21.95 (13.95)
Canada	40	19.73 (1.32)	70.0%	80	50.48 (8.21)	19.83 (12.73)
Germany	160	20.57 (1.09)	91.3%	67	47.83 (8.63)	20.63 (10.58)
Netherlands	58	19.97 (1.43)	67.2%	80	50.66 (7.90)	16.66 (9.46)
USA	142	18.68 (0.90)	70.4%	91	47.10 (7.00)	25.15 (14.84)
Collectivistic Country Group	388	20.02 (1.26)	69.3%	21.75	44.10 (2.92)	24.48 (14.19)
East Asian Country Group	215	19.94 (1.29)	61.9%	32	43.22 (7.89)	30.27 (15.17)
South Korea	91	20.23 (1.21)	72.5%	18	46.11 (8.46)	26.90 (13.46)
Japan	124	19.73 (1.32)	54.0%	46	41.09 (6.72)	32.73 (15.92)
Latin American Country Group	173	20.12 (1.22)	78.6%	11.5	45.20 (10.19)	17.28 (8.51)
Costa Rica	77	20.18 (1.21)	81.8%	15	48.60 (8.67)	16.11 (9.74)
Ecuador	96	20.07 (1.23)	76.0%	8	42.48 (10.53)	18.22 (7.31)
All countries	866	19.79 (1.33)	72.7%	41.7	46.34 (8.65)	22.91 (13.53)

Note: Cultural Norm: higher scores indicate more acceptance toward attention-seeking social behavior, lower scores indicate a preference for attention-avoidance. Social Anxiety: higher scores indicate higher level of self-reported social anxiety. 1 = after Hofstede [3].

As in our previous study, groups differed on gender, χ²(8, N = 866) = 65.53, p <.001, and age, F(2, 863) = 11.83, p <.001, partial eta² =.03. Post hoc tests revealed significant age differences between participants from individualistic countries and East Asian countries (p <.05, d = 0.26), as well as between participants from individualistic countries and Latin American countries (p <.001, d = 0.4). East Asian and Latin American countries did not differ significantly with respect to age. To control for those differences, we conducted analyses with age as a covariate. We also conducted analyses separately for each gender in addition to the main analysis, which included both genders.

To ensure that variation of ethnicity within a country did not impact differences between country groups, we assessed participants’ ethnicity. As intended, we found very little variance of ethnicity within the three country groups. 99.7% of participants from individualistic countries identifying themselves as either “Caucasian” or “European”, 98.3% of the participants from Latin American countries identified themselves as “Hispanic” and 100% of participants from East Asian countries identified themselves as “Asian”. Furthermore, 99.7 % had always lived in their country and had never lived somewhere else, and all participants identified themselves as native speakers.

Procedure

Participants were recruited via announcements in classes. In some countries students were tested in class groups (Costa Rica, Ecuador, Japan, South Korea, Australia and the United States) and returned the questionnaires anonymously after class. In Canada, Germany, The Netherlands and the United States, students were asked to fill out questionnaires at home and return questionnaires anonymously to the university. All answers were completely anonymous, as students did not write their name onto the questionnaire. In some countries, students received course credit points for their voluntary participation (Australia, Canada, Germany, Japan, South Korea, and the United States). Response rate was not systematically assessed.

Data analysis

Inclusion criteria across countries were: (1) the participant and both parents were born in the respective country and (2) participants’ age had to be between 18 and 22 years. The same criteria were applied in Heinrichs et al. [1]. For consistency with the previous study, we established groups based on Hofstede’s [3] rating of country individualism. Additionally, we grouped collectivistic countries geographically [15]. This approach resulted in three groups: 1) individualistic countries (Australia, Canada, Germany, the Netherlands, and the USA), which included all of the individualistic countries in Heinrichs et al. [1]; 2) collectivistic East Asian countries, which included two (Japan and South Korea) of the three countries that were considered by Heinrichs et al. (Spain was excluded based upon geographical location); and 3) collectivistic Latin American countries (Costa Rica and Ecuador).

To compare country group levels of social anxiety and social norms we conducted an ANCOVA comparing the three country groups (individualistic countries, East Asian countries, and Latin American countries). Because differences between country groups on age and gender could influence the results, age was entered as a covariate for all analyses. We further ran separate analyses for males and females to check for gender differences. All significance tests were two-tailed with p =.05 as the significance threshold. Missing values were replaced by mean scores.

Measures

Reactions to Social Behavior Questionnaire-Revised (RSBQ-R)

The Reactions to Social Behavior Questionnaire (RSBQ) was developed by Heinrichs and colleagues [1] to assess cultural and personal acceptance of reticent versus outgoing social behavior. The original RSBQ comprised 16 social behavior vignettes that depicted individuals in a variety of social situations. In the Heinrichs et al. [1] study they assessed personal and cultural norms but the Personal Norm scale displayed low internal consistency (α =.40; range across countries:.30 – .53). The low internal consistencies of this scale were the rationale for developing a revised version of the RSBQ. The modified questionnaire [20] included 17 vignettes (featuring six of the original vignettes of Heinrichs et al.). As in Heinrichs et al. [1], the newly developed vignettes described socially reticent (ten vignettes) and attention-seeking behaviors (seven vignettes) in specific situations (either interaction or performance situations). The vignettes did not mention the emotional state of the actor: we excluded terms such as anxious, shy, or self-confident. For instance, “You are sitting in a math class. The lecturer writes a problem on the board and asks if anybody can solve it. You can see that the women sitting next to you has already worked out the problem but she does not step forward.” In the present study, we focused on perceived cultural norms. As in Heinrichs et al., cultural norms were assessed by asking the participants to rate whether the described behaviors are typical or atypical for their own culture, e.g. “How typical is this behaviour for individuals in your culture in general?” The scale ranged from very typical = 0 to very atypical = 5. The items were summed to yield a total score. The newly developed vignettes were translated in all non-German speaking countries and back-translated in Ecuador, Costa Rica, and Japan. The revised Cultural Norm scale of the RSBQ-R showed satisfactory to good internal consistencies (α = .72; range across countries: .68 – .78).

Social Interaction Anxiety Scale [21]

The SIAS consists of 20 items that reflect anxiety in social interaction situations, e.g., fears of being inarticulate, being boring, or not knowing what to say. Individuals are asked to indicate to what degree the statements describe themselves on 5-point scales (“not at all” to “extremely”). The SIAS has been shown to have good internal consistency and well-established validity in clinical and non-clinical samples [21]. In the present study internal consistency was good within all countries (range across country groups: .79 – .91).

Results

Differences between countries on acceptance of social behaviors

Table 1 shows means and standard deviations of the cultural norms. We conducted univariate analyses of variance (ANCOVAs) with cultural norm as the dependent variable and group (i.e., individualistic, East Asian, Latin American) as the independent variable, controlling for age as a covariate. Analysis revealed a significant main effect of group, F(2, 862) = 28.84, p <.001, with a small effect size (partial eta² = .06). Age did not yield a significant main effect, F(2, 862) = 2.85, p = .09. Post-hoc tests revealed significant differences between individualistic and collectivistic countries. Participants from individualistic countries were less accepting toward socially reticent behavior than were participants from East Asian (p < .001, d = 0.63) and Latin American countries (p < .001, d = 0.33), while participants from East Asian and Latin American countries did not differ on their cultural norms (p = .08, d = 0.22). Because previous research found a significant interaction between gender, self-construal and social anxiety [19] and because social anxiety is known to be higher in women [11], we repeated the analyses separately for each gender to check whether the different proportion of males in the different country groups affected the results. Again, the analysis revealed significant main effects of group, both in women (F(2, 626) = 19.17, p < .001, partial eta² = .06) and men (F(2, 232) = 10.00, p < .001, partial eta² = .08). Age did not yield a significant main effect in either females or males. Follow up post hoc tests showed that female participants from individualistic countries were less accepting toward socially reticent behavior than were those in East Asian (p < .001, d = 0.59) and Latin American countries (p < .001, d = 0.37), while East Asian and Latin American countries did not differ on cultural norms (p = .74, d = 0.15). In male participants, a different pattern emerged: individualistic (p < .001, d = 0.68) and Latin American countries (p < .05, d = 0.45) were less accepting towards socially reticent behavior than East Asian countries, and did not differ from each other (p = .76, d = 0.14).

Differences between country groups in social anxiety

Table 1 shows means and standard deviations of social anxiety in all country groups and countries. Univariate F-tests with age as a covariate indicated a significant country group effect, F(2, 862) = 55.28; p <.001, partial eta² = .11; and a significant effect of age, F(1, 862) = 3.91, p < .05. Follow-up post hoc tests showed significant differences between all country groups with Latin American countries showing the lowest and East Asian countries showing the highest mean level of social anxiety. Specifically, participants from Latin American countries displayed significantly lower levels of social anxiety compared to participants from individualistic (p < .01, d = 0.40) and East Asian countries (p < .001, d = 1.09). Participants from individualistic countries displayed lower levels of social anxiety than did participants from East Asian countries (p < .001, d = 0.62). To test the effect of gender differences between country groups, we repeated the analyses separately for each gender. The same results were found in males and females with the exception of male social anxiety levels in individualistic and Latin American countries, which were not significantly different.

Discussion

The purpose of the present study was to determine whether collectivistic cultures displayed more acceptance of socially reticent behaviors and higher levels of social anxiety compared to individualistic countries. The results of the current study replicated the earlier findings of Heinrichs et al. [1] in new samples from individualistic and East Asian countries. Notably, however, the higher levels of social anxiety were not found in Latin American countries, which are also considered to share the collectivist values found in East Asian cultures.

As expected, collectivistic East Asian countries showed higher levels of social anxiety and greater cultural acceptance of socially reticent behaviors than did individualistic countries. Thus, the earlier findings of Heinrichs et al. [1] were replicated with new samples, using a modified version [20] of the Reactions to Social Behavior Questionnaire. Those findings are consistent with previous research on cross-cultural differences in social anxiety between East Asian and Western individuals [6,7, 22, see 1 for a detailed discussion of these results]. Consistent with our predictions, the Latin American group displayed the lowest social anxiety level within the three country groups while the collectivistic East Asian group displayed the highest social anxiety level. These results suggest that collectivistic values per se do not explain differences in social anxiety across cultures and are in agreement with results of the study by Arrindell et al., [9] which also found social anxiety to be unrelated to country levels of individualism.

One potential interpretation of the country differences on social anxiety is that Latin American cultures promote ways to ensure group harmony (like being sociable, friendly, talkative) that decrease social anxiety while East Asian cultures promote ways to ensure group harmony (like being submissive, quiet) that may increase social anxiety [23]. For example, in East Asian countries such as South Korea and Japan, people might be more conscious of other people’s feelings and try to avoid situations where they might hurt other people’s feelings [6], which reflects an indirect approach to ensuring group harmony. Perhaps in Latin American countries people take a more active approach to fostering interpersonal harmony. Thus, in East Asian countries collectivistic values may increase self-consciousness, which in turn increases social anxiety, while in Latin American countries the emphasis on interpersonal harmony may not necessarily lead to an increase in self-consciousness, and therefore not necessarily to increased social anxiety. The cultural script of simpatía [17] is in line with this explanation. People from Latin America are expected to be likeable, fun, easygoing, polite, and respectful. Avoiding conflicts and criticizing others by emphasizing positive and deemphasizing negative behaviors are highly valued behaviors in Latin America [16]. According to this script and recent findings, Latin American culture may provide greater modeling of sociable, non-anxious behaviors coupled with lower prevalence of negative learning experiences (being criticized and rejected by important others) and therefore evoke less social anxiety than is the case in individualistic or East Asian countries. Further research is needed to test the assumption that the cultural script of simpatía in Latin American can explain cross cultural differences in social anxiety. Furthermore, the inclusion of instruments that assess culturally specific types of social anxiety (like Taijin-Kyofusho [6, 24,25]) would strengthen future studies. However, social anxiety measures such as the SIAS were found to show homogeneity in internal consistency and factor structure when applied to different countries throughout the world indicating that social anxiety is likely to be assessed with similar instruments.

Contrary to our predictions, Latin American countries displayed greater acceptance of socially reticent behaviors than the individualistic countries and similar levels to the East Asian countries. This result raises the possibility that cultural acceptance of socially reticent behavior, while unrelated to levels of social anxiety, may be related to prevalence rates of social anxiety disorder. In line with this possibility, researchers have found a lower prevalence of social anxiety disorder in East Asian and Latin American countries [11 – 13]. In line with the findings of lower prevalence rates in East Asia, a recent study of our research group [26] found that the association between social anxiety and life impairment was greater in individualistic than in collectivistic East Asian countries. These recent results point to the more negative impact of withdrawn and socially reticent behaviors for people from Western countries relative to those from East Asia. Thus, individuals from collectivistic countries may suffer less impairment due to cultural acceptance of socially reticent behaviors [2], and therefore lower levels of social phobia.

Another potential interpretation of the unexpected result regarding cultural norms may be a response style bias. Consistent with that speculation, Ramirez-Esparza et al. [16] found students from Mexico, in comparison with students from the USA, to be less sociable and extraverted on self-report measures but to be more sociable on behavioral measures. According to this result, higher acceptance of socially reticent behaviors in the RSBQ-R may reflect a response style. However, results pertaining to social anxiety argue against this speculation. One would expect that a response style would be consistent across self-report measures and therefore that there would be similar levels of social anxiety in East Asian and Latin American countries. This was clearly not the case.

There are several limitations of the current study that need to be acknowledged. First, the results are based solely on self-report data from university students with an overrepresentation of females and limited age variability, and require replication in more representative samples. Another direction for future studies would be to examine the interaction between gender, culture, and social anxiety. In this study, we found differences between Latin American and individualistic countries only in women, whereas men in individualistic and Latin American countries showed a similar pattern of results. One possibility is that the relationship between culture and social anxiety is influenced by gender-related differences in self-construal. For example, in a USA sample, Moscovitch, Hofmann and Litz [19] found that independent self-construal was negatively associated with social anxiety only in men, whereas the reverse was found in women. The possible moderating influence of self-construal on gender differences in social anxiety warrants study in other cultures. Another explanation for the current findings is that low power due to small sample sizes may have affected analyses conducted with males.

The current study used questionnaires developed primarily from a Western cultural perspective. Although six of the original vignettes of the RSBQ were developed by all countries involved in the study [1], the 11 new vignettes were developed primarily by German researchers [20]. In a similar vein, the social anxiety measure was developed in a Western country. A further limitation is the use of a single measure to assess each construct. This is especially warranted for the revised version of the RSBQ, which was not validated again. Furthermore, back translation of the RSBQ-R was conducted only in Costa Rica, Ecuador and Japan, and it is possible that the item meanings are not equivalent across languages. On a positive note, the validity of the revised RSBQ-R was supported by the replication of the Heinrichs et al. [1] results in new samples from individualistic and East Asian countries. Furthermore the RSBQ-R showed acceptable to good internal reliability in all countries (country range: α = .68 – .78), which supports the cross-cultural applicability of the instrument. However, future research would benefit from using back-translation for all languages and assessing the construct with more than one measure. For example, the use of behavioral data and on-line responses of cognition, emotion and motivation [16, 27] may provide further information as to whether people living in Latin American and East Asian countries show greater acceptance towards socially reticent behaviors than people living in individualistic countries and further, how this cultural pattern relates to social anxiety.

The current study included only two countries within the East Asian and Latin American groups. Replication in a larger sample of Latin American and East Asian countries would be helpful as well as including collectivistic countries from other geographical regions. Although the lower level of social anxiety in East Asian countries has been found in other studies [1,6,7], to the best of our knowledge no study has tested the country differences between other Latin American and individualistic or East Asian countries [3]. Future research might also examine the influence of Hofstede’s other cultural dimensions on social anxiety. Costa Rica and Ecuador are similar to many Latin American countries in respect to Individualism, Masculinity and Uncertainty Avoidance; however, Costa Rica has the lowest Power Distance score of all Latin American countries, indicating that inequality of power and wealth are less accepted in Costa Rica than in other Latin American countries. On the other hand, Arrindell et al. [9] found the country level of Power Distance to be unrelated to country level of social anxiety within eleven nations.

As intended, the present sample was ethnically homogenous within countries. Future studies may focus on social norms across diverse ethnic groups within one country to determine whether cross cultural differences in social anxiety and social norms can be generalized to the entire population. A further limitation is the limited knowledge of response rate, as it may vary between country groups and may affect results. Also, data collection varied between countries (questionnaires were filled out in class groups or at home). However, differences in social anxiety level cannot be fully explained by different recruitment methods, as data collection was similar in collectivistic East Asian and Latin American countries (in-class questionnaires). Nonetheless, future studies should keep the data collection as similar as possible across countries and carefully assess response rates.

In sum, results indicated that cross-cultural differences in social anxiety cannot be fully explained by individualism-collectivism. To our knowledge this was the first study to investigate social anxiety and social norms in different collectivistic countries, and therefore sheds light on cultural influences on social anxiety.