Information Exchange in Decision-Making Groups Using Group Support Systems

One important factor that can give decision making groups an edge over individual decision makers is the amount and diversity of decision-relevant information held by group members. Prior to group discussion, some of such information may be held uniquely by one or more members (unshared information) while other may be common among all members (shared information). How efficiently groups exchange shared as well as unshared information can influence decision quality. Past research suggests that face-to-face group discussions are frequently biased in favor of shared information: groups discuss more of their shared than unshared information, and unshared information is often introduced late in discussion. In this paper we examine these information sharing biases in the context of computer-mediated group communication. We describe a model of information sampling and use the model to predict sampling behavior. Model predictions are then compared with data obtained through a pilot study that used a Group Support System (GSS) for group communication. Findings of the pilot study are reported.