Abstract: | Block (1963) suggested that in estimating the equivalence of tests E should correct for unreliability to give a more appropriate estimate of the "conceptual equivalence" of the tests. In the present paper it is pointed out that: a distinction should be made between conceptual equivalence and functional equivalence; functional equivalence is of prior importance in research seeking relationships of other variables to the variable represented by the tests; in estimating functional equivalence it is not apropriate to correct for unreliability; and, in fact, correction for attenuation may often lead to over-generalization of results. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |