Reanalysis of experimental halo effects.

Data on ratings of individuals, obtained under 2 conditions of judgment and published in 1956, were reanalyzed by a more complete analysis of variance. The usual interaction between raters and individuals, called a halo effect, was found but it was not influenced by judgment conditions intended to maximize it. Hence, the evidence for halo effect due to judging operations remains questionable. (PsycINFO Database Record (c) 2010 APA, all rights reserved)