Whistling in the Wind

This study examined the relationship between perceived workgroup sexual orientation dissimilarity and participant perceptions of group process and withdrawal. Based on the theory of relational demography within groups (Riordan, 2000) and recent research on moderators of the dissimilarity-outcome relationship (e.g., Stewart and Garcia-Prieto, 2008), the study argued that: (1) perceived sexual orientation dissimilarity would be associated with negative group process effects and increased withdrawal for all study participants, (2) that the relationship between perceived sexual orientation dissimilarity and outcomes would be stronger for heterosexual individuals than for those who identified as lesbian, gay or bisexual (LBG), and (3) that participants' level of sexual orientation identity development would moderate their responses to increased sexual orientation dissimilarity in their workgroups. Three hundred and ninety-eight graduate students at Columbia University were asked to respond to an online questionnaire designed to assess their perceptions of workgroup dissimilarity, communication, conflict and peer relations as well as their individual levels of withdrawal. Hypotheses were tested using hierarchical multiple regression analysis. Results indicated that perceived dissimilarity was positively related to increased relationship conflict, task conflict and withdrawal and negatively related to peer relations among all study participants. Additionally, the effects of perceived dissimilarity on task conflict and withdrawal were moderated by participant sexual orientation and participant sexual orientation identity development, consistent with study hypotheses. Slightly different patterns of findings emerged when the results were examined for LGB and heterosexual individuals separately. Though not hypothesized, values dissimilarity was found to mediate the relationship between perceived sexual orientation dissimilarity and several of the group process outcomes, particularly for heterosexual individuals. The contributions and implications of these findings for relational demography and sexual orientation workplace research are also discussed.