Issues in fitting theoretical and measurement models in marketing

This article reviews the applications of linear structural equation modeling in business. It further examines the extent to which measurement models are constructed to fit theory models. The consensus appears to be that modelers intuitively use congeneric models without considering the appropriateness of such use. The other measurement models, such as tau equivalent or parallel form models are used very little in the literature. In addition, there is no use of common factor measurement models. The article also examines the extent to which various indices of fit are used in modeling marketing phenomena. Although chi square tests of model fit are used most frequently, the study documents increasing use of yet other indices, which perhaps more accurately reflect the extent to which the model fits.