Structural Equation Analysis of Three Methods for Evaluating Conjoint Measurement Results

Abstract

This paper applies a structural equation methodology (LISREL) to the evaluation of the results obtained from conjoint measurement. Contrary to previous work, the results indicated that stress, sign, and prediction give consistent evaluations of the results of conjoint measurement. The solution given by the LISREL analysis was criticized; however, following a Spearman factor analysis, the same conclusion was drawn: stress, sign, and prediction appear to be measuring the same construct.