Differential Subject Performance and the Problems of Selection

Publisher Website
Google Scholar
Abstract
Methods of test equating and scaling have been studied in the statistically advanced literature of educational and psychological measurement, and applied extensively where mass testing and selection procedures are required; for example, selection for tertiary study. Usually the procedures themselves require large data sets, and they are applied in order to ensure that students are not advantaged or disadvantaged because of the arbitrary features of the origin and units of measurement of the various tests. However, equating may be necessary where the data sets are relatively small. This paper provides an example of selection into a professional faculty within a university, following study from a range of subjects available in various faculties at the first-year level, and shows the need and a procedure for equating the grades in the different subjects. The procedure itself is based on a psychometric model studied extensively in the psychometric literature, and the purpose of this paper is to introduce the rationale for its application from first principles for a broader professional education audience and to illustrate its application with a practical example.