Implications of a measurement problem for expectancy theory research

Abstract

Expectancy theories of work motivation specify a multiplicative relationship between valence of job outcomes (V) and expectancy (E) that effort will lead to these outcomes. But both E and V are typically measured using scales lacking a rational zero point and thus at best interval in nature. Multiplication of such scales is theoretically not a meaningful operation. This paper demonstrates that transformations of E and V scales which change only the location of the nonmeaningful zero point can greatly alter obtained correlations between E × V and measures work motivation or effort. Two potential solutions to this problem are presented.