Professor and Chair, Research and Evaluation Methodology
University of Colorado Boulder
In the context of an educational assessment, it is considered good advice to consider practical significance when evaluating an observed difference in means. But this is a lot harder than it appears because such evaluations require inferences about magnitudes, and inferences about magnitudes require well-defined measurement units. It has become conventional for researchers to express practical significance in terms of variability (i.e., effect size units) or some implicit function of time (i.e., weeks of learning). In this presentation Briggs will explore the historical and conceptual precedent for such moves, which can be traced back to the work of Gustav Fechner and Francis Galton between 1860 and 1890, Alfred Binet between 1904 and 1908, and L. L. Thurstone between 1925 and 1932. Briggs will point out some problems with conventional approaches to the evaluation of practical significance, and suggest some ideas for how these might be addressed by taking what he views as a more theory-driven approach to the measurement of psychological and educational attributes.