Tackling measurement problems with Item Response Theory: Principles, characteristics, and assessment, with an illustrative example

摘要

Interest in measurement issues remains unabated, as evidenced by published research in the area of reliability, validity, and, in particular, scale development. At the same time, psychometricians have continued to generate alternative measurement approaches and models at an explosive pace. Surprisingly, these alternative measurement approaches have been slow to diffuse into the marketing literature despite marketers' inherent interest in measurement issues. This paper discusses one such measurement approach, item response theory (IRT), which can potentially address critical measurement concerns. My focus is on identifying basic principles and key characteristics and on providing an assessment for applied researchers. Toward this end, an empirical example of role conflict (RC) and role ambiguity (RA) concepts is included to illustrate IRT principles and amplify the theory's relevance to resolving measurement dilemmas. In addition, I provide a comparison with the current paradigm of measurement―classical test theory (CTT)―to afford a balanced appreciation of the payoffs of adopting the IRT approach.