Tests of three causal models of halo rater error

Results of a study which compares the explanatory power of halo error models in terms of their ability to account for observed halo rating error show that different causal models (general impression, salient dimension and inadequate discrimination halo error models) can be distinguished either theoretically or empirically. It was further proven that the general impression model of halo error error was the most defensible. However, it is suggested that additional rating methods and levels of analysis need further investigation to determine validity of this assumption.