Empirically determines if two approaches, i.e., individual predictor information versus group information, would yield different predictive results, and since the second approach involved the additional expense of grouping whether the prediction of criterion measures by individual data could be enhanced by the addition of group data. (Author/RC)