Cultural norms, values, and beliefs are thought to be powerful forces that shape differences between cultures (Triandis, 1989) and may influence the development of psychopathology (Eshun & Gurung, 2009). Such factors have been of particular interest regarding problematic social anxiety (Caldwell-Harris and Ayçiçegi, 2006, Heinrichs et al., 2006 and Xie et al., 2008). Specifically, self-construal has emerged as a potentially important construct in explaining cultural differences (Markus & Kitayama, 1991). Self-construal is conceptualized as a constellation of thoughts, feelings, and actions concerning one’s relationship to others as well as one’s self-identity in relation to others (Singelis, 1994). Markus and Kitayama (1991) summarize theory and research regarding two dimensions of self-construal: Independent and interdependent. We base our discussion of these constructs on Markus and Kitayama’s (1991) summary of theory and research.
1.1. Independent and interdependent self construal: Cultural and individual differences
According to Markus and Kitayama (1991), independent self-construal is a view of the self that focuses on internal attributes and uniqueness of the self, whereas interdependent self-construal involves a focus on the social connectedness of the self. Markus and Kitayama note that independent self-construal has been defined as “a bounded, unitary, and stable” (p. 226) self whereas an interdependent self-construal is a “flexible and variable” (p. 226) self. People who are higher in independent self-construal are expected to emphasize being unique, promoting one’s own goals, and being direct in communication. In contrast, a person with higher interdependent self-construal is expected to see the self as intertwined with others and emphasize public status, social roles, and relationships. Interdependent self-construal has been found to be more prevalent in collectivist cultures such as those of East Asia as compared to the West, in which more emphasis is on independence (Markus & Kitayama, 1991).
Our discussion thus far might be taken to imply that the constructs of independence and interdependence comprise one bipolar dimension, such that being high in one implies being low in the other. However, it has been argued, and shown, that high (or low) levels of both interdependent and independent self-construal can coexist within an individual (Bhawuk and Brislin, 1992, Cross and Markus, 1991 and Singelis, 1994). Thus, it seems that self-construal may be best conceptualized not just as a cultural construct, but as an important individual difference that varies both between and within cultures.
1.2. Self construal and social anxiety
Most authors who have previously tested the relationship between social anxiety and self-construal report that social anxiety has a negative relationship with independent self-construal and a positive relationship with interdependent self-construal (Dinnel et al., 2002, Moscovitch et al., 2005, Norasakkunkit and Kalick, 2009, Okazaki, 1997, Okazaki, 2000 and Singelis and Sharkey, 1995). Further, Hong and Woody (2007) found that cultural differences in social anxiety between Canadian East Asians and Canadian Caucasians were fully mediated by independent self-construal and partially mediated by interdependent self-construal. Norasakkunkit and Kalick (2009) manipulated independent self-construal using a priming task and found that priming independence led to decreased scores on a social anxiety measure. These results suggest a causal link between independent self-construal and social anxiety.
1.3. A model of personality and self-construal
Given the cultural focus of much research regarding self-construal, it would be natural to consider how these variables might help explain cultural differences regarding social anxiety. However, with a growing literature suggesting that self-construal may serve as a crucial individual difference within cultures (e.g., Paukert, Pettit, & Amacker, 2008), it seems important to examine how self-construal interacts with other salient individual differences to produce social anxiety.1 We believe that personality traits, as captured by the five-factor model, are crucial individual differences that are likely to influence self-construal, as well as the relationship between self construal and social anxiety.
Theories of the big five factors of personality have found that personality traits generalize and are expressed across cultures (e.g., Katigbak, Church, Guanzon-Lapena, Carlota, & del Pilar, 2002). Thus, we start with the assumption that the constructs of personality and self-construal have an influence on individuals from all cultures (e.g., developmentally) and further, personality is more likely to affect the development of self-construal than self-construal is the development of personality.2 We make this assumption partially because it seems plausible to us that personality influences the way that cultural messages (such as self-construal) are internalized and expressed, as well as the type of message directed at an individual. For example, within the five-factor model, individuals higher in extraversion are described, in part, as sociable (e.g., Costa & McCrae, 2001). The sociability aspect of extraversion makes it plausible that extraverts would have more exposure to cultural messages conveyed through interpersonal interactions because of their increased likelihood of engaging in social interactions. Further, Mooradian and Swan (2006) found that people higher in extraversion were more likely to rely on information relayed via word of mouth. Thus, we would expect that in Western cultures the predominance of cultural messages promoting independence would lead to higher individualism in people with higher extraversion, who should have more frequent exposure to such messages.
If extraversion does influence self-construal, it is possible that factors related to extraversion might actually be a consequence of self-construal rather than extraversion alone. It is already well established that people with problematic social anxiety tend to have lower extraversion (e.g., Bienvenu et al., 2001), such that (higher) extraversion might be a protective factor for social anxiety disorder. If our analysis is correct, then independence might partially mediate a protective relationship between extraversion and social anxiety. Level of independent self-construal might even be more amenable to change than personality per se, potentially opening up a new avenue for treatment. A clear understanding of how risk and protective factors for social anxiety relate to each other should assist in identifying likely areas for intervention and individuals who are likely to benefit from such interventions.
The available literature provides very few links between self-construal and other individual differences (other than culture of origin). However, it is well established that social anxiety is positively related to neuroticism and negatively related to extraversion (Bienvenu et al., 2001, Kotov et al., 2007, Trull and Sher, 1994 and Watson et al., 2005). Thus, we expect that independent self-construal would have the opposite relationships with those personality traits in Western cultures. However, we could only locate one paper that included any big five personality traits (and only agreeableness and extraversion) and self-construal (only interdependent self-construal): In a study that did not focus directly on the relationship of personality to self-construal, Tams (2008) found that interdependent self-construal was correlated with agreeableness. The literature is clearly in need of basic work identifying how these constructs relate to one another.
Of particular interest are potential interactions between self-construal and personality traits related to social anxiety. Specifically, given the consistently reported relationship between social anxiety and neuroticism (typically a medium-sized effect, e.g., Bienvenu et al., 2001), we expect neuroticism and interdependent self-construal to interact to predict social anxiety, such that individuals who are higher in both neuroticism and interdependent self-construal will exhibit particularly high levels of social anxiety. We expect that these individuals’ tendencies to be anxious in general will be magnified by their tendency to focus on social relationships, making them particularly vulnerable to problematic social anxiety.
1.4. The current study
In the current study we examined the relationship between self-construal, the big five personality traits, and social anxiety. We hypothesized that (a) the relation of extraversion and neuroticism to social anxiety (i.e., because these personality traits consistently exhibit a relationship with social anxiety, Bienvenu et al., 2001), would be mediated by independent self-construal, and (b) interdependent self-construal and neuroticism would interact to predict social anxiety.
