The role of structural equation modeling in scale development

Abstract

Across a variety of disciplines and areas of inquiry, reliable and valid measures are a cornerstone of quality research. This is the case because to have confidence in the findings of our studies, we must first have confidence in the quality of our measures. This article briefly reviews the literature on scale development and provides an empirical demonstration of the scale development process. The example considered is the development and validation of a condom influence strategy questionnaire-short form (CISQ-S), a scale to measureways individuals persuade their partners to use condoms. A special focus is put on the unique contribution that structural equation modeling techniques, particularly confirmatory factor analysis, bring to scale development. Latent variable modeling and its applications to scale development are also considered. Suggestions and implications for scale developers are discussed. © 2003, Lawrence Erlbaum Associates, Inc.