USC Marshall

Educational institutions have increasingly sought to create diverse student bodies. Given these efforts, the present research asks two broad questions. First, to what extent do students’ interactions reflect the diversity of their educational settings? Second, when intergroup interactions occur, how do they impact students’ academic experiences and outcomes?Leveraging a longitudinal design and daily diary methods, we conducted the first large study (N_interactions = 11,460) which tracks the frequency, quality, and consequences of meaningful cross-race and cross-social class interactions in university settings. We found that students reported far fewer cross-race and cross-class interactions than would be expected at chance given the racial and social class diversity of their student bodies. Furthermore, underrepresented racial minority students and students from working-class backgrounds experienced cross-race and cross-class interactions as lower quality than same-race and same-class interactions. Nevertheless, these cross-group interactions predicted better academic performance for racial minority students and those from working-class backgrounds. They did so through different processes: cross-race interactions improved racial minorities’ academic performance by reducing social identity threat, whereas cross-class interactions improved the academic performance of students from working-class backgrounds by increasing sense of belonging. Together, these findings suggest that diversifying educational settings is not enough to encourage meaningful intergroup interactions. Furthermore, fostering intergroup interactions may be one important pathway toward reducing racial and social-class disparities in students’ experiences and academic outcomes in college.

A rich body of research throughout the social sciences demonstrates that bias--people's tendency to display group-based preferences--is a major obstacle in the way of promoting diversity, equity, and inclusion in the workplace. The current article moves beyond the single-level focus of prior theories of workplace bias to propose a novel theoretical model that conceptualizes workplace bias as a multilevel cycle. First, we discuss the theoretical foundations of our bias cycle theory and describe why understanding the nature of workplace bias--and effectively reducing it--requires considering the reciprocal influences of both individual and organizational levels of the cycle. Specifically, we describe how workplace bias operates as a cycle and then propose that successfully reducing workplace bias requires intervening across both the individual and organizational levels of the cycle. Second, because workplace bias is reproduced through both of these levels, we review and bring together literatures that are often considered separately: psychology research on reducing bias at the individual level and sociology and management research on reducing bias at the organizational level. Third, we use our bias cycle theory to formulate general principles for determining how to begin and how to pair interventions across levels. Finally, we conclude by discussing our theoretical contributions and outlining directions for future research.