Social dilemmas : group discussion, group decision, and demonstrability

A social dilemma is defined as a scenario in which a person must decide between an individually rational choice and a collectively rational choice. These two choices are referred to as the non-cooperative choice and the cooperative choice. These dilemmas are ubiquitous in society, for example, the decision to use public transport or not, recycle, or conserve energy. A robust finding in social dilemma research is that a period of group discussion prior to choice making increases cooperation, (e.g. van de Kragt, Orbell, Dawes, Braver, & Wilson 1986). Subsequent research offered two explanations for this effect. Specifically that group discussion either, elicits higher levels of group identity (Dawes, McTavish, & Shaklee, 1977), or enables a consensus process to occur (Boaus & Komorita, 1996).

This PhD research investigated the group discussion effect in more detail. Evidence from Hopthrow and Hulbert (2005) suggests that groups treat a social dilemma as a problem to be solved and, as such, groups are influenced by the demonstrability of the cooperative choice. Demonstrability refers to the level to which the correct answer to a problem can be shown to a doubting but capable group member (Laughlin, 1980). The doctoral research presented is unique and original as it uses a group problem solving methodology to gain fresh insights into factors that affect cooperation in social dilemmas. This thesis presents four experiments exploring group level phenomena, measuring cooperative choice by means of six-person face-to-face groups playing a prisoner’s dilemma game and one experiment measuring the impressions and understanding of individuals (meta-cognitions) when asked about group level phenomena (and associated pilot work). The main aim of the thesis is to outline the previously neglected link between the well-defined processes of group problem solving and the processes occurring when groups discuss a social dilemma. Moreover, the thesis examines in detail the concept of demonstrability and the impact of a cooperative group decision upon cooperation in social dilemmas.

Throughout the thesis participants were given a social dilemma and asked for their preference for cooperation or non-cooperation before they were exposed to a group context. This enabled the measurement of the effect of group processes on individual choice making. Strong support was found for the consensus model of the group discussion effect. Groups that did not make a group decision did not show the group discussion effect unless exposed to decision imitating voting. Moreover, the impact of demonstrability on this process is moderately robust. High demonstrability was shown to protect the group from a composition of entirely non-cooperative group members, in that groups became cooperative. These important findings specify more closely the mechanisms driving the group discussion effect. They also impact on the way work-groups should be organised for maximum cooperation. A consistent theme throughout the thesis was that demonstrability had no effect upon individual pre-discussion choices even when participants were asked to imagine the way they would behave in the context of a group discussion. Evidence from the final experiment of the thesis suggests that this process is moderated by participants’ level of meta-cognition of group process. Moreover, people may be more likely to be able to predict potential group process in a dilemma that is less abstract. The meta-cognition findings suggest demonstrability does not merely change the attractiveness of the cooperative payoff as participants indicate no awareness of the differences between the two types of dilemma. Continued exploration of meta-cognitive processes on real-life social dilemmas would allow the development of solutions to large scale dilemmas.