STRONG TESTS OF EXPECTATION-STATES HYPOTHESES

A task-oriented group such as a jury, a committee, or a work team is a network of patterned social relations. Most of the current research on task groups, particularly that guided by Berger, Fisek, Norman, and Zelditch's expectation-states theory, assumes that these patterns are functions of their diffuse and specific status characteristics and of pertinent verbal and nonverbal cues. Exactly how these factors operate (the important question for efforts to reduce gender or racial inequalities) is a subject of much inquiry. Yet previous assessments of such ideas have mainly tested task groups' tendency to become vertically organized, a minimal prediction that does not differentiate competing hypotheses or models. In this paper I present a procedure for assessing the distinctive predictions of a model, which concern group members' locations in such a hierarchy. I illustrate this procedure, an application of the quasi-independence idea, with data from Smith-Lovin and a specific expectation-states formulation set forth by Fisek, Berger, and Norman. This leads to findings not evident from earlier analyses.