The debiasing role of group support systems: an experimental investigation of the representativeness bias

Abstract

Past research has demonstrated that individual and group judgments are subject to systematic biases. Although much effort has been devoted to the debiasing of individual judgments, no corresponding work has been found on the debiasing of group judgments. This study examines the usefulness of a group support system (GSS) in addressing an important judgment bias, namely, the representativeness bias, which refers to the bias incurred in posterior-probability estimations by not properly utilizing all information sources, such as the base rates.The formation of a judgment is seen from the perspective of an information integration process. Two orthogonal dimensions of information integration—interpersonalandintrapersonal—are involved in group judgments. Interpersonal information integration concerns the aspect of information sharing among group members, and can be supported with the computer-mediated communication channel of GSS. Intrapersonal information integration deals with the information processing capacities and capabilities of individuals, and can be supported using a problem representation tool, as part of GSS.A laboratory experiment with a 2×2 factorial design was conducted. One hundred and twenty subjects, randomly allocated to 40 three-member groups, took part in the experiment. Data pertaining to both processes and outcomes were collected and analysed. Representativeness bias was reduced by the use of the problem representation tool. Increased use of the tool led to greater awareness about the base rate and, consequently, to better judgments in this problem context. On the other hand, computer-mediated communication did not reduce the representativeness bias. Although computer-mediated communication is capable of improving the interpersonal aspect of information integration, the representativeness bias is primarily a result of cognitive limitations, and benefits little from improved communication among group members. It is also possible that benefits of computer-mediated communication can be more readily derived by larger groups than by smaller groups of size three used in this study.