An experiment was devised to apply role theory to the prediction of communication behavior by observing dominance behavior as a function of gender and attitudes toward sex role stereotypes. Task type was used as an exploratory variable. The hypothesized interaction of gender and sex role attitudes did not occur, and the hypothesis that men would be more dominant than women was also rejected. Women with varying sex role attitudes, however, did exhibit dominance behavior in accordance with role theory predictions. Task type had no impact on the subjects' display of dominance or submission. (Author)