Belief bias in individual and collective reasoning

Abstract

In this paper, we investigate whether collaborative group performance is better than individual performance in solving a syllogism evaluation task. We hypothesise that collaborative group settings will outperform individual settings and that the belief bias effect will be mitigated in a group setting. Two empirical studies were conducted with Argentinian undergraduate students. Study 1 (N = 239) used a between-subjects design with two conditions: individual resolution and interactive group resolution. Overall, the group condition performed better than the individual condition, but there were no significant differences in evaluating invalid syllogisms. Study 2 (N = 115) used a within-subjects design with three conditions: individual resolution, interactive group resolution, and individual after-interactive group resolution. Overall, the group condition performed better than the individual condition, and the individual after-interactive group condition showed an increase in accurate answers compared to individual resolution. However, as observed in Study 1, the collaborative group setting did not improve the evaluation of invalid syllogisms. We propose an explanation for the group resolution of invalid believable syllogisms within the framework of the selective processing model of the belief bias. This research provides new data on the effects of collaborative settings in deductive reasoning beyond the Western Educated Industrialised Rich Democratic (WEIRD) cultures.