V. S. Helgeson (1994) offered a promising model to explain sex differences in well-being. Three meta-analyses of gender research relevant to this model are provided along with an empirical investigation of the relations between agency and communion, unmitigated agency and unmitigated communion, and self-and peer reports of well-being and distress in a sample of 201 college students. The results extended Helgeson's model by showing that agency and communion can be assessed at the level of interests and role behaviors and that the association of agentic traits with well-being may be inflated by self-reports. Results also point to problems in distinguishing trait measures of agency from unmitigated communion and communion from unmitigated agency.

This study examined frequency and severity of physical, symbolic, and psychological aggression between college men and women in 85 heterosexual dating relationships and the extent to which gender role constructs predicted reports of aggression. Although there were no differences on self-reports of perpetration, men reported higher victimization levels than women and higher physical and psychological victimization levels than perpetration levels, whereas women reported higher symbolic perpetration levels than victimization levels. As a result, averaging reports from both partners suggested that women in existing college dating relationships are more aggressive than men. For both genders, stereotypically negative masculine (i.e., instrumental) characteristics were the best predictors of aggressive acts. Perpetrators’ positive masculinity and femininity predicted self-reports of decreased aggression that were not confirmed by their partners. Whereas men’s traditional attitudes about the male role predicted greater male aggression, women’s less traditional attitudes predicted increased severity of female physical aggression.

A history of childhood sexual abuse (CSA) is strongly associated with adult depression. The goal of the present study was to explore potential mediators of the CSA-depression link. The potential mediators were variables known to be associated with depression: interpersonal problems, gender role orientation, sociotropy, and self-silencing. The participants were 109 women and 83 men recruited from the community, approximately one third of whom had a history of CSA. The results indicated that gender role orientation, sociotropy, and self-silencing were not associated with a history of CSA. However, both men and women with a history of CSA reported more inter-personal problems than did individuals without this history. Whereas women reported being distant and controlling, men reported lacking assertiveness and taking too much responsibility in their relationships. These interpersonal variables partially mediated the link between CSA and depressive symptoms. Thus, in part, CSA survivors may be at risk for depression because they experience interpersonal problems.

Helgeson proposed a comprehensive model of agency, communion, and well-being that was based on a review of the literature on agentic and communal personality traits, power and intimacy motivation, and self-critical and dependent depressive styles. The present study empirically examined the overlap among these personality dimensions as well as their relations to positive and negative affect, life satisfaction, and social adjustment. A factor analysis of the various personality measures yielded two factors, labeled communion and agency. Power motivation loaded negatively with four communal qualities: communal traits, communal role behaviors, intimacy motivation, and dependency. Self-criticism loaded negatively with two other agentic qualities: agentic traits and agentic role behaviors. Both agency and communion were significantly positively associated with positive affect and social adjustment. Agency also was significantly associated with higher life satisfaction and lower negative affect. The positive adjustment outcomes related to agency were confirmed by peer reports.

The present article reviews some of the central conceptual issues confronted by gender researchers as they have tried to forge a theory of gender identity that can account for the complexity and diversity of gender-related characteristics displayed by women and men. An emerging consensus suggests that gender is incorporated into an individual's self-concept in multiple and loosely connected ways. We review one example of this emerging multiplicity perspective, Spence's (1993) multifactorial gender identity theory, and describe three recent studies testing its usefulness. We also discuss ways in which multiplicity models of gender could benefit from considering parallel developments in the general personality literature regarding the problem of levels or domains. In particular, it is argued that McAdams's (this issue) integrative three-level model of the structure of personality offers a helpful framework for guiding future test construction and theory development in gender research.