Performance measure selection problem with both orientations of envelopment forms in DEA

Performance evaluation of some independent and homogeneous organizations, based on relative efficiency, has economic production theory as its foundation. Data envelopment analysis (DEA) is a nonparametric approach to evaluate the performance of some homogeneous decision making units (DMUs) with multiple inputs and multiple outputs. If the number of performance measures is high in comparison with the number of DMUs, then a large percentage of DMUs will be identified as efficient, which practically is not favorable. We extend a new approach which combines both input-and output-oriented envelopment DEA models to select a subset of performance measures under variable returns to scale assumption. In this approach, a mixed binary linear programming (MBLP) model is formulated which employs the rough rule of thumb in DEA to select the most influential performance measures. Practically, we utilize a real dataset as an example to illustrate the potential application of the proposed approach.