Nationality Heterogeneity and Interpersonal Relationships at Work Defended on Friday, 13 May 2011
In this dissertation I test three new approaches to extend the ‘classical’ model of workplace diversity. The ‘classical’ model of workplace diversity assumes that diversity affects work outcomes via the mediating effects of social networks. I hypothesize that this model fruitfully can be extended by 1) considering that diversity forms a context in which employees act, 2) testing alternative predictors of network formation and employee behavior (i.e., employee voice), and 3) integrating diversity and social network perspectives in a contingency model. Three empirical studies support these hypotheses. In the first study, I show that the association between leadership and employee voice is stronger for nationality dissimilar employees. The second study finds that employee voice affects the strength of friendship relations but that this effect is contingent on employees’ past position in the social network. Finally, the third study demonstrates that group performance is maximized at moderate levels of task network centralization but lowest at high and low levels of centralization but that this relation is moderated by nationality diversity. Nationality diverse teams required more centralization to achieve high performance than homogeneous teams. Finally, I discuss the implications of these findings for research on diversity and social networks.
diversity, social networks, leadership