Mitigating the Common Information Sampling Bias Inherent in Small-Group Discussion (Retracted)

Abstract

Gathering problem-relevant information through small-group discussion is one method for initial decision making. A collective information sampling (CIS) model offered by Stasser and Titus (1987) suggests that two inherent biases, the common information sampling bias and a related recency effect, act in concert to suboptimize the efficacy of group discussion. The objectives of this exploratory study are to refine the predictive validity the CIS model and test a group intervention technique designed to mitigate the common information sampling and recency biases. Analysis of the evidence from a laboratory and field experiment suggests that the refined CIS model accurately predicted the distribution of information items entered into group discussion. While the control groups exhibited both the common information sampling bias and primacy effect, these biases were mitigated in the treatment groups after receiving an intervention technique called Shared Cognition Awareness Training (SCAT). Results obtained from this exploratory research indicate that the refined CIS model can offer researchers a useful prediction tool and the SCAT intervention technique presents a relatively straightforward means of invoking a more complete exchange of potentially valuable information during small-group discussion.