A contextual approach to social anxiety in adolescence: The effects of social role discrepancy and self-worth
Girls tend to rate themselves as more socially anxious than boys, and girls who are socially anxious tend to be lonelier and are less accepted by peers than are boys who are socially anxious. Proposed etiologies of social anxiety are diverse, including behavioral, cognitive, physiological, and genetic formulations. However, these approaches leave unanswered questions regarding the sex differences in the prevalence of social anxiety. The purpose of this study was to evaluate a new model, the Social Role Discrepancy Model of Social Anxiety, which could provide an explanation for these differences. Specifically, the investigator hypothesized that social role discrepancy (SRD), as measured by a modified version of the Adolescent Sex Role Inventory (Thomas & Robinson, 1981), would be positively correlated with social anxiety (SA), as measured by the Social Anxiety Scale for Adolescents (La Greca & Lopez, 1998). Furthermore, girls were expected to possess significantly greater levels of SRD than boys. Finally, global self-worth (GSW) as measured by the Self-Perception Profile for Adolescents (Harter, 1988), was expected to function as a moderating variable between SRD and SA for boys, and as a mediating variable between these constructs for girls. The sample consisted of 74 boys and 92 girls in 10th through 12th grades at public schools in the southeastern United States. As hypothesized, girls reported significantly higher levels of social role discrepancy than did boys. Unlike findings from previous research, boys and girls reported similar levels of SA and GSW. Zero-order intercorrelation matrices were computed for the total sample, for girls, and for boys. For girls scoring at least one standard deviation above the mean on SRD scores, GSW scores accounted for 46% of the variance in Fear of Negative Evaluation (FNE) scores, and for 32% of the variance in total Social Anxiety (SAS-T) scores. For girls scoring below this cutoff, GSW scores accounted for only 21% of the variance in FNE scores and 12% of SAS-T scores. Statistical analysis did not support the proposed mediator/moderator models. Implications of the current research and suggestions for future studies are discussed.