Discussion and consensus in rater groups: Implications for behavioral and rating accuracy

Very little research has explored the importance of discussion and consensus among raters in determining both behavioral and rating accuracy; yet groups of raters are routinely used in personnel practices such as assessment centers and panel interviews. Results, based on 225 raters in 75 three-person teams, show that not only does the anticipation of group discussion result in initial superior behavioral accuracy, but also requiring raters to reach consensus results in even greater improvements in behavioral accuracy. On the other hand, even though behavioral accuracy improved in anticipation of group discussion, rating accuracy simultaneously decreased. Rating accuracy, however, did significantly increase during the process of reaching consensus to levels comparable to those groups not anticipating discussion. Overall, group discussion and consensus appear to have larger effects on behavioral accuracy than rating accuracy. Implications, especially for both assessment centers and panel interviews, are discussed.