Performance evaluation in education under uncertainty: A robust optimization approach

Analysing and enhancing education system performance is of importance to local authorities and policy makers because education improves the human capital, which, in turn, leads to economic growth. This study draws on the secondary data to assess the education performance using data envelopment analysis (DEA). However, the reliability of the efficiency measures calculated by DEA is jeopardized if the inputs and outputs are erroneous, which is likely to occur through secondary data collected by government agencies. This study attends to uncertainty through the lens of robust optimization, which fits into a DEA application. We propose a robust enhanced Russell measure model to consider the extent of inherent uncertainty in the light of uncertain characteristics. We also present a case study in education to demonstrate the applicability and efficacy of the proposed models in practice.