His research program is focused on four mutually supportive areas. The development of a framework for statistical modeling (Explanatory Item Response Modeling) that allows for the extension and adaptation of psychometric models in ways that make them more responsive to problems that arise in education and other areas of application. The application of measurement framework—the BEAR Assessment System (BAS) —to a range of assessment situations across the social sciences, but concentrating on the development of a body of work that can support use of sound assessment approaches by teachers and other professionals. The exploration of philosophical and historical perspectives on the area of psychometrics. The development and dissemination of policy positions in educational testing and assessment, based on the research mentioned above. Throughout all four strands of this broad research agenda, there is a continuing theme of the application of newer conceptual and statistical methods to pressing problems of educational and psychological research. The philosophy underlying his approach to these areas is that there are critical and distinctive problems involved in the measurement, statistical analysis, reporting and usage of psychological and educational constructs and that research methods used to investigate these problems must be sensitive to the best available substantive theory.