Sex and communication effects in a mixed-motive game

Abstract

This study investigates the effects of restricted communication opportunity and the sex of the subject on co-operative responses made by British dyads in a 180-trial matrix version of Prisoner's Dilemma. The first experiment employed 20 male, 20 female and 20 mixed-sex dyads who were assigned to one of four verbal communication conditions: none allowed; allowed before trial l; allowed after trial 31; allowed throughout. No differences due to the sex of the dyad were demonstrated but communication allowed throughout elicited significantly higher levels of co-operation. A financial incentive was introduced into the second experiment. Two communication conditions were retained: none allowed; allowed after trial 31. 10 male and 10 female dyads took part. Communication opportunity elicited higher levels of co-operation, especially for male pairs. Subjects were more than twice as co-operative when offered a financial incentive. Both of these experiments were conducted by a female E, the author. When experiment 2 was replicated by a male E communicaction effects, but no sex effects, were observed. However, there were no effects due to the sex of the subject but there was a main communication effect when Expts. 2 and 3 were replicated in a 8aianced sex of E.design. However, the female experimenter elicited higher levels of cooperation from all subjects regardless of their sex. The data from Expts. 2-4 were analysed in a multiple E design. Communication opportunity elicited significantly higher levels of cooperation and female Ss were less co-operative in the presence of male experimenters, but only as the duration of the inter-action increased. The experimenters themselves were found to elicit different levels of co-operation from subjects regardless of the sex of the player or of the sex of E. Experiment 2 was replicated by the author using 32 American students. American males were more co-operative than British men and both American and British women. No communication effects were observed. Experiment 2 was then repeated with E absent from the room. No significant main or inter-action effects were observed. It seems that the presence of the experimenter, whatever their sex, helps to define the psychological environment in which the subjects' strategic inter-action occurs.