A bstract
An experiment examined the effects of role ambiguity and relative group status in an intergroup situation where a superordinate goal was salient. One-hundred-andfifty-six subjects in groups of three undertook a cooperative task under conditions where the groups' roles were clearly Different, Similar, or not allocated (Control).
In addition, the groups'perceived competence at the task was either equally High, equally Low, or unequal. Financial rewards were contingent on successful completion of the joint task. Two contrasting hypotheses were derived from Social Identity Theory: one (HI) predicted most favourable intergroup attitudes in the Control condition where lack of any clear differentiation in group roles might facilitate a superordinate iden tif cation. In contrast, (H2) predicted least favourable attitudes in this condition on the grounds that groups lacked a distinctive identity. Support for the second hypothesis was found since friendliness towards the outgroup decreased with increasing role ambiguity. The status variable also had consistent effects. As predicted, mutual evaluations tended to reflect the consensually agreed status differences: least bias being shown towards high status outgroups, most by high status ingroups. Task performance was also affected by role ambiguity. Judges' ratings of the group products were found to be more favourable as role ambiguity increased, in contrast to the friendliness data. The implications of these findings for Social Identity Theory, the Contact Hypothesis, and theories of group performance are discussed.
This research was carried out as part of an E.S.R.C. grant awarded to Rupert Brown and Jennifer Williams (Grant No. HR 7953). The authors would also like to acknowledge the assistance of Audrey Mathews in the preparation of the experimental materials and the coding of the data.