Development of a General Legal Confidence Scale: A First Implementation of the Rasch Measurement Model in Empirical Legal Studies

Abstract

Legal capability has long been of evident importance in our understanding of legal problem resolution behavior. Although legal capability remains a contested concept, there is much commonality between specifications. Some aspects are generic, while others—such as legal confidence—are particular to law. Such law‐specific measures as have been developed to date have been developed in an ad‐hoc fashion; with no attempts made to test psychometric properties, using either classical test theory or modern psychometric methods. This has been a shortcoming in the empirical legal field, weakening theoretical development and precluding reliable estimation of changes in levels of legal capability over time. In this article, we set out details of a study aimed at introducing new methods to scale development in the field of empirical legal studies based on the approaches that have evolved in other fields and the latest developments in psychometric modeling. Specifically, we set out details of the use of a specially designed item pool—based on an increasingly demanding legal scenario—and Rasch analysis to develop a general legal confidence scale. Once the 12 item pool items were reduced to a final set of six, this yielded a scale with good psychometric properties: a General Legal Confidence (GLC) Scale. The scale showed good overall fit, item fit, person fit, targeting, and internal consistency. All items had ordered thresholds, there was no response dependence, items were unidimensional, and there was no evidence of differential item functioning. The GLC scale constitutes an effective measure of general legal confidence, and demonstrates it is possible to arrive at robust and coherent law‐specific measures of legal capability through the careful design of questions and application of the latest psychometric modeling techniques.