Expertise Gaps in Value-Added Modeling

Abstract

This study introduces the notion of an "expertise gap," a mismatch between researcher expertise and research content. I show that these gaps can exist and warrant our attention in research endeavors that involve multidisciplinarity embedded in complex research designs. Writing as an education researcher concerned with educator quality, I show that several expertise gaps are embedded in a particular statistical approach to the measurement of teacher effectiveness: the use of value-added models. My analysis lays out the basic argument structure behind these models, locates eight consequential assumptions that researchers are obligated to make, and considers the kind of disciplinary expertise that these assumptions call for. I argue for the importance of illuminating expertise gaps and studying how they shape public understanding of science and technology.