Abstract: | "The purpose of this paper has been to examine the consequences of the Dingman-Guilford approach to the formation of a composite rating in the case of unit rank. The essence of the Dingman-Guilford approach is the identification of the single common factor with the trait being rated. A first consequence of this identification is that the Dingman-Guilford composite x' is superior to the customary unit-weight composite from both a validity and a reliability point of view. Extending the argument, however, assigning the raters those weights which maximize the correlation with the single common factor yields a composite which is superior to x' on the same counts and for the same reasons that x' is superior to the customary composite." (PsycINFO Database Record (c) 2010 APA, all rights reserved) |