Building a consistent system for faculty appraisal using Data Envelopment Analysis

TL;DR

This paper applies Data Envelopment Analysis to evaluate faculty performance across different academic ranks, demonstrating its potential for fairer, democratic decision-making in academic institutions.

Contribution

It introduces a novel application of DEA for faculty appraisal, comparing separate and combined group evaluations to highlight fairness and efficiency implications.

Findings

Efficiency scores ranged from 0.85 to 0.93 across categories.

Treating all faculty as a single group led to efficiency decline.

DEA supports democratic and fair faculty evaluation processes.

Abstract

Data Envelopment Analysis (DEA) appears more than just an instrument of measurement. DEA models can be seen as a mathematical structure for democratic voicing within decisional contexts. Such an important aspect of DEA is enhanced through the performance evaluation of a group of professors in a virtual Business college. We show that the outcomes of the analysis can be very useful to support decision processes at many levels. There are three categories of professors: Assistant professors, Associate professors, and Full professors. The evaluation process of these professors is investigated through two different cases. The first case handles each category of professors as a separate sample representing an independent population. The results show that the mean efficiency scores fall between 0.85 and 0.93 for all professors no matters their category. In spite of enabling more fairness, such an approach suffers from its exclusive character, which is contrary to the democratic spirit of DEA. The second case tries to cope with this deficiency through the assessment of the faculty members as a single sample drawn from the same population, i.e., Assistant professors, Associate professors, and Full professors are treated equally, only on the ground of their respective inputs and outputs, no matters their academic rank. A clear efficiency decline is reported, basically due to the very nature of DEA as a procedure that is more efficiency than output focused.