THE SYSTEMATIC DISTORTION HYPOTHESIS - A CONFIRMATORY TEST OF THE IMPLICIT COVARIANCE AND GENERAL IMPRESSION MODELS

Supervisory ratings of subordinate performance were used to compare tw o variations of the systematic distortion hypothesis: an implicit cova riance model and a general impression (GI) model. Our purpose was to a ssess whether a model specifying a set of correlated performance dimen sions or a model specifying a single, GI factor provided the best repr esentation of actual supervisory performance ratings (N = 491). Confir matory factor analysis was used to test a first-order correlated facto r (i.e., implicit covariance) model against a second-order hierarchica l (i.e., GI) model. The results generally supported the latter perspec tive whereby a GI is thought to serve as the basis for subsequent dime nsional judgments, which in turn serve as the basis for individual ite m ratings. The implications of these results with respect to rater tra ining and rating accuracy are discussed.