Conversation logic effects in the minimal group paradigm: existent but weak

According to a conversation logic (Grice, 1975) analysis of the minimal group paradigm, participants discriminate along group boundaries because they feel obliged to incorporate the provided group membership information into their resource allocation decisions. Conversely, intergroup bias might disappear if the relevance of this information is explicitly attributed to a different task, as first demonstrated by Blank (1997). Two experiments addressing possible alternative interpretations of my earlier results, however, failed to support this expectation. In retrospect, the manipulation of group membership relevance might have been overridden by a counteracting salience effect. In contrast, a third experiment provided support for the conversation logic-based prediction that under some conditions outgroup bias should occur. Overall, however, conversation logic effects seem to be weak, compared to other influences in the minimal group paradigm. The general discussion focuses on the inherent uncertainty of the experimental setting and the heterogeneity of behavioral strategies it induces.