Measurement models for marketing constructs

Researchers who seek to understand marketing phenomena frequently need to measure the phenomena studied. Yet, constructing reliable and valid measures of the conceptual entities of interest is a nontrivial task, and before substantive issues can be addressed, the adequacy of the available measures has to be ascertained. In this chapter, we discuss a wide variety of measurement models that researchers can use to evaluate the quality of their measures. It is assumed that, generally, multiple measures are necessary to capture a construct adequately. We first present the congeneric measurement model, in which continuous observed indicators are seen as reflections of an underlying latent variable, each observed variable loads on a single latent variable, and no correlations among the unique factors (measurement errors) are allowed. We contrast the congeneric measurement model with the formative measurement model, in which the observed measures cause the composite variable of interest, and we also consider measurement models that incorporate a mean structure (in addition to a covariance structure) and extend the single-group model to multiple groups. Finally, we address three limitations of the congeneric measurement model (zero loadings of observed measures on non-target constructs, no correlations among the non-substantive components of observed measures, and the assumption of continuous, normally distributed indicators) and present models that relax these limiting assumptions.