Scale Evaluation with Exploratory Structural Equation Modeling and Simultaneous Adjustment for Acquiescence Bias

There is continued debate about the appropriateness of Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) for scale evaluation and latent variable modeling more generally. This study seeks to combine the strengths of EFA, namely flexible evaluation of item-factor structures, while using one of the unique strengths of CFA, the simultaneous adjustment for common method bias in self-report survey data. Exploratory Structural Equation Modeling (ESEM) (Asparouhov & Muthén, 2009) is used to combine the EFA logic with CFA adjustment for individual acquiescence bias in semantically balanced scales. Two applications corroborate the usefulness of this model, using the short personality scale of Rammstedt and John (2007) as an example. The first application shows how the model can be used for general scale evaluation. The second application incorporates the model into testing measurement invariance of instruments while controlling for acquiescence across respondents. Finally, further applications of this model will be discussed.