View article

The misuse of simple correlational analysis for assessing the relationship between a multiplicative composite and a simple varaible (as is found, for example, in a value expectancy formulation) is addressed. Despite the warnings of Schmidt (1973) and the alternatives proposed (eg, Cohen, 1978), such misuse is pervasive. Examples of this misuse are provided from the several areas in social and organizational psychology. Finally, researchers, authors, reviewers, and journal editors are urged to be very sensitive to these problems.