A developmental focus to admissions testing: admissions and placement standards development

Abstract

With the introduction of the Admissions and Placement Assessment Programme (APAP) at the University of Port Elizabeth (UPE) in 1998, it was felt that many learners entering higher education needed some form of academic development. It was also felt that testing at admission should provide information about the level of learners' proficiency on core academic competencies, to assist with decision-making about academic development. This article describes how the research team went about defining performance standards on the battery of measures used at UPE using a multidimensional cluster analysis approach. K-means cluster analysis was used as a method to classify 380 examinees in the Faculties of Science, and Economic and Building Sciences into groups similar in levels of proficiency on the Admissions and Placement Battery (APAB). The top and bottom borderline clusters of the final cluster solution were used to set cut scores on the measures. The clusters and cut scores were validated on a hold-out sample of 725 learners. The cut scores successfully identified learners who were not yet proficient enough to be successful without support, but who would benefit from development, as opposed to those learners who probably would not benefit, or were not in need of development. The importance of using performance standards in a multidimensional manner was emphasised.



(South African Journal of Higher Education: 2003 17 (3): 192-208)