Exploring a Factor for Domestic Violence Beliefs: A Demonstration of Measurement Ambiguity

Significant effort and money has been spent to change social and legal responses to domestic violence (DV) but, current statistics indicate relatively little decline in incidence rates and the literature contains conflicting findings, resulting in several controversies and heated political debates. Most of the rhetoric surrounding the controversial findings focus on theories and contextual issues related to survey instruments. However, this article seeks to raise another important issue that is common in criminological literature: poor measurement development and practices. Specifically, inconsistencies in scale development, psychometrics, and reporting practices for measurement lead to poor scale development and analytical confusion. We argue that poorly developed measures used repeatedly and slightly altered across varying studies contribute to the controversies in the literature. This article serves as an example of scale development to stimulate conversation regarding measurement issues amongst social scientists by focusing on one example regarding the measurement of DV beliefs. Exploratory and confirmatory factor analyses are used to explore the dimensionality of beliefs to develop the best measure for the data. Results highlight the need for more rigorous statistical estimation models and empirically supported psychometric techniques, and the need for further replication when developing composite measures of latent constructs.