Social value orientation and contextual cues influence cooperation and trustworthiness

The higher than predicted levels of cooperation in social dilemmas have motivated others to develop social preference models of behaviour (Dufwenberg and Kirchsteiger, 2001; Fehr and Schmidt, 1999; Rabin, 1993). However, norm-based models of behaviour in social dilemmas (Bicchieri, 2006) may complement social preference models. Six experiments examined the impact of situational cues on cooperative and trustworthy behaviour of proself and prosocial introductory psychology students within a single-play Prisoner’s dilemma or modified Trust game. Using 195 participants, the first study supported the hypotheses that prosocial participants subjectively transformed the Prisoner’s dilemma into a coordination problem and made choices based on these subjective transformations, while proself participants preferred and made choices that reflected the incentive structure of the Prisoner’s dilemma. Findings suggested that the sequential nature of the Prisoner’s dilemma promoted cooperative behaviour in prosocial participants and non-cooperative behaviour in proself participants. The finding that situational cues moderated the predictive value of social value orientation was explored in later studies. In Study 2, using 222 participants, the stimuli for selfish, cooperative, and altruistic expectations were developed for a modified Trust game in Study 3. Using 228 participants in Study 3, proselfs displayed significantly less trustworthiness, compared to prosocials, in the selfish and altruistic trust conditions. However, the difference between the two social value orientations was attenuated in the cooperative trust expectation. Study 4 removed the possibility that participants were acting differently because they did not believe they were playing against a real person. Using 218 participants, proselfs in the high immediacy condition (face-to-face) compared to the low immediacy condition (anonymous) displayed significantly more trustworthiness in response to selfish expectations. This finding suggested that expectations associated with trust and the immediacy of social influence moderated individual differences based on social value orientation. Study 5 proposed and tested a model by which expectations associated with trust activated or deactivated social norms that mediated trustworthy behaviour. Using 265 participants, this study replicated the findings of the selfish and cooperative trust condition in Study 3, although it found no significant difference in the reaction times that underlie the proposed psychological mechanisms mediating trustworthy behaviour. In the altruistic trust condition, proselfs did not significantly keep more money than prosocials, yet displayed significantly faster reaction times for the compliance construct. In the previous studies of this thesis, situational cues had impacted prosocial participants, compared to proself participants, to a lesser extent. One possible reason for this was that prosocials, when playing with an interactive other, were guided by shared expectations of equality, irrespective of the situational cues within the experiment. In Study 6, prosocial participants played against a computer, in which the computer randomised choices between the decision to cooperate and defect. With shared expectations of behaviour absent in this design, prosocial participants cooperated significantly less in a sequential Prisoner’s dilemma when their interactive partner was a computer, compared to another participant. Overall, these findings suggested that situational cues influenced proselfs and prosocials differently. Proselfs were more likely to resist the temptation to take advantage of trust if cooperative expectations were either made salient or in response to a selfish expectation from an immediately-present partner. Furthermore, the findings proposed a social norm perspective of decision making which complemented social preference models of behaviour in interdependent situations