A mathematical ranking model in learning analytics

Abstract

The introduction of new educationa l trends triggers the development of innovative methods to collect, analyse and report the data subsequently generated. This paper discusses the implementation of a multi-stage mathematical class ranking model as a method in learning analytics, applied to a Computer Science module. The model applies the principle of Pareto optimality in an outputs-only data envelopment model to categorise st udents into classes of similar efficiency which are ranked according to dominance. The model is then adapted to calculate improvem ent targets for each studen