Factors Affecting Deviations of Efficiency in Distance-Based Common Set of Weight -DEA Models

There are different approaches to generate a common set of weights in DEA based on the 𝑝 - distance measure. Deviation of an efficiency score derived from a CSW from target efficiency score may be related to the model and the parameter 𝑝. In this study, we try to clarify points about choosing 𝑝, model, and data set if it is necessary to produce an efficiency score with the least deviation by a CSW. Two improved linear models are developed by analyzing the result of available models. The results of the proposed models have smaller individual and overall efficiency than corresponding prior ones. It has been confirmed with numerical examples and simulation analysis.