Issue: 2021/Vol.31/No.3, Pages 41-65
PERFORMANCE ANALYSIS IN THE PRESENCE OF BOUNDED, DISCRETE AND FLEXIBLE MEASURES
Sohrab Kordrostami
, Monireh Jahani Sayyad Noveiri
Cite as: S. Kordrostami, M. J. S. Noveiri. Performance analysis in the presence of bounded, discrete and flexible measures. Operations Research and Decisions 2021: 31(3), 41-65. DOI 10.37190/ord210303
Abstract
In conventional data envelopment analysis (DEA) models, the relative efficiency of decision-making units (DMUs) is evaluated while all measures with certain input and/or output status are considered as continuous data without upper and/or lower bounds. However, there are occasions in realworld applications that the efficiency of firms must be assessed while bounded elements, discrete values, and flexible measures are present. For this purpose, the current study proposes DEA-based approaches to estimate the relative efficiency of DMUs where bounded factors, integer values, and flexible measures exist. To illustrate it, radial models based on two aspects, individual and aggregate, are introduced to measure the performance of entities and to handle the status of the flexible measure such that there are bounded components and discrete data. Applications of approaches proposed in the areas of quality management, highway maintenance patrols, and university performance measurement are given to clarify the issue and to show their practicability. It was found that the introduced procedure can determine practical projection points for bounded measures and integer values (from the individual DMU viewpoint) and can classify flexible measures along with evaluation of DMUs relative efficiency.
Keywords: data envelopment analysis, efficiency, bounded data, discrete data, flexible measure
Received: 21 September 2020 Accepted: 9 July 2021