Developmental theorists describe friendship development in terms of stages that mature well into adolescence. The purpose of this prospective study is to investigate the possible differences between high school students, with and without social phobia, and their levels of friendship quality and friendship expectation. It is hypothesized that students with social phobia will have lower quality friendships as well as less mature expectations of friendship compared to students without social phobia. A MANOVA analysis will be utilized to identify interactions between these variables. Gender and grade level differences will also be examined. Implications for treatment supplements are discussed. / text

<p>Three studies focus on an intriguing paradox in the associations between gender, friendship quality, and loneliness, and examine whether gender differences in friendship expectations help explain why the paradox occurs. Study 1 (n = 1761 college undergraduates) documents the three elements of this paradox: (1) females reported higher levels of various positive features in their friendships than did males; (2) higher levels of positive friendship features were associated with lower levels of loneliness; and (3) males and females reported similar levels of loneliness. Consistent with this paradox, when friendship features were statistically controlled, a statistical suppression effect was found such that females reported higher levels of loneliness than did males. </p><p>Study 2 (n = 1008 young adults aged 18 to 29) replicated each of the findings from Study 1 using a revised and expanded measure that reliably assessed a broader set of distinct friendship features. In addition to measuring friendship features and loneliness, Study 2 also examined friendship satisfaction, and here too a striking suppression effect emerged. Specifically, although females reported slightly higher levels of friendship satisfaction than did males, females reported lower levels of friendship satisfaction than did males when friendship features were statistically controlled. Another noteworthy finding was that several friendship features were more strongly related to friendship satisfaction for females than they were for males, suggesting that females may be more "sensitive" to subtle variations in friendship features than are males. </p><p>Study 3 (n = 419 young adults aged 18 to 29) further replicated the suppression effects observed in Studies 1 and 2, and was designed to learn whether gender differences in friendship expectations would help explain the paradox and suppression effects. Two different facets of friendship expectations were hypothesized and assessed with newly developed, highly reliable measures of each facet. The first facet, referred to as "feature-specific friendship expectations," focused on the degree to which individuals expect a best friendship to be characterized by each of the friendship features that were assessed in Study 2. The second facet, referred to as "feature-specific friendship standards," focused on identifying where individuals "set the bar" in deciding whether or not a friend's actions have fulfilled expectations in various friendship feature domains. </p><p>Gender differences were found for both facets of friendship expectations with females generally having higher expectations for their friends than did males. The two facets were only moderately correlated, and related in distinct ways to other variables of interest. Findings indicated that higher levels of feature-specific friendship expectations were generally associated with more positive functioning in the social domain (i.e., higher levels of positive friendship features and friendship satisfaction), whereas higher levels of feature-specific friendship standards were associated with potentially more problematic functioning (i.e., more negative responses to ambiguous violations of friendship expectations). </p><p>Study 3 also tested the hypothesis that discrepancies between feature-specific friendship expectations and the quality of a person's best friendship on each of the same features are associated with loneliness and also with friendship satisfaction. Polynomial regression analysis, rather than the traditional difference score approach, was used to test this hypothesis. The discrepancy hypothesis was not supported with regard to either loneliness or friendship satisfaction; possible explanations for this finding are discussed. </p><p>Together, findings from the three studies provide evidence of the replicability of the observed paradox, identify friendship quality as a suppressor variable on gender differences in loneliness and friendship satisfaction, and provide evidence for the existence of two distinct facets of friendship expectations. Results from this dissertation suggest important directions for future research designed to better understand the linkages among gender, social cognition, and social experience in contributing to emotional well-being for young adults.</p> / Dissertation