Using Rank Statistics for Determining Programmatic Efficiency Differences in Data Envelopment Analysis

Abstract

This paper presents an analytical approach, based on rank statistics, to the issue of comparing programs within Data Envelopment Analysis (DEA) efficiency evaluation framework. The program evaluation procedure distinguishes between managerial and programmatic inefficiency and uses the Mann-Whitney rank statistic to evaluate the statistical significance of the differences observed between a treatment program and its control group program after adjusting for differences in managerial efficiency between the programs. A numerical example, based on the data used to evaluate the educational enhancement of the Program Follow Through, is used to illustrate the proposed statistical procedures.