Data envelopment analysis: An exposition and critique

Abstract

One of the difficulties with the use of Data Envelopment Analysis (DEA) has been the relative obscurity of the mathematical techniques with which it operates. This paper outlines the nature of these techniques and by using a simple example demonstrates their inadequacy for the measurement of school performance. It also argues against the use of aggregate level data in studies of school effectiveness and suggests that the only secure basis for such research is provided by multilevel models which use measurements made on individual students.