Article révisé par les pairs
Résumé : This study tested a relational demography model of workgroup identification. We theorized that early in workgroup formation, (a) racial identification would moderate the influence of racial dissimilarity on member communication behavior (frequency and length of verbal participation) during a group task, (b) member communication behaviors would mediate the effect of racial dissimilarity on workgroup identification, and that (c) these effects would be affected non-symmetrically by racial group membership (White, Black). This model was tested among a sample of 211 undergraduates (113 White and 98 Black), working in 43 temporary workgroups. Whites' racial identification moderated the relationship between racial dissimilarity and member communication behaviors, which in turn partially mediated the effect of racial dissimilarity on workgroup identification, generally consistent with our model. Among Black workgroup members support for our model was limited. Copyright © 2008 John Wiley & Sons, Ltd.