다 기간 성과평가를 위한 Super Efficiency DEA 모델 및 사례적용

Abstract

This study proposed super efficiency DEA (Data Envelopment Analysis) models for multi-period performance evaluation and presents case studies demonstrating their application. DEA is widely utilized as an analytical tool to evaluate the efficiency of decision-making units (DMUs) that involve multiple inputs and outputs, where causal relationships are difficult to identify. DEA models define a production possibility set based on input and output data, assessing the relative efficiency of DMUs by their position within this set. When evaluating performance over multiple periods, such as in projects involving multi-period inputs and outputs, the scope of input and output data for comparison differs from that of single-period evaluations. This allows for various definitions of production possibility sets, enabling the establishment of multiple DEA models. In this study, several production possibility sets were defined for multi-period performance evaluation, and corresponding super efficiency DEA models were proposed. Super Efficiency DEA models enable the assessment of relative efficiency among efficient DMUs by excluding the input and output data of the DMU being evaluated from the reference set. The proposed models were applied to a multi-period workforce development initiative, specifically the Engineering Education Innovation Project. Data from 10 engineering education innovation centers over a four-year period were used for the analysis. The evaluation results showed that the proposed Super efficiency DEA models, which include multi-period inputs and outputs as comparison targets, provided more differentiated efficiency scores than single period super efficiency DEA models, although there was little difference in the ranking of DMUs.