Guidelines for Combining the Judgments of Individual Members in Decision Conferences

Abstract

This paper is concerned with the problem of obtaining judgments from individual group members and combining these individual judgments to arrive at a group decision according to a process which minimizes error. Two preconditions for such a process are established first, based on prior research: (1) that individual judgments should be obtained by means of a nominal group format, and (2) that expression of the individual judgments should be recorded on some type of rating scale. Based on these premises, the research reported here examined the following variables to determine the effect of each on the error contained in the group judgments: (a) the number of intervals on the rating scale used by individual group members; (b) the aggregation rule used to transform individual judgments into a single group representative judgment; (c) the number of group members; and (d) the accuracy of the judgments of the individual group members. Some guidelines for the practicing administrator wishing to structure decision-making processes in accordance with the Implications of this study are then suggested.