THE GENERALITY OF BEHAVIORAL DATA II: REPLICATION OF AN EXPERIMENT ON GENERALIZATION FROM OBSERVATIONS ON ONE OCCASION'

In two experiments previously reported the effect was studied of variation of task and group composition on ratings of behavior based on observation of individuals in group situations. The experiment reported here was planned as a replication of one of these studies. The main results of the two previous experiments were completely confirmed: When the situation variation encompassed both task and group com- position simultaneously the relationship was random between ratings made independently by two teams of judges, after one observation instance for each team.