Analyzing the efficiency values of two-stage DEA model

• Main Authors: Trung Nguyen, 阮文中
• Other Authors: Ming-Chi Tsai
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/28401754077947371508

碩士 === 義守大學 === 管理碩博士班 === 103 === Data Envelopment Analysis (DEA) technique, a nonparametric method in operations research and economics for the estimation of production frontiers, has been applied to many fields to evaluate operational efficiency such as banking, insurance, sport, etc. Based on the original idea, in 1999, Seiford and Zhu first developed and utilized a two-stage DEA model in their research. Subsequently, both of these models have been being used widely and studied in many researches. Regardless of numerous relevant researches, the relationship between efficiency values in these two models remains unknown. This study is going to examine the two-stage DEA model to figure out the relationship between efficiency values and the effect of set of weights on them. Theorems about this relationship are developed and a numerical example is employed to illustrate those theorems. The results reveal that two-stage analysis can provide more information about operational performance compared to the traditional DEA model and set up weights may cause different effects on efficiency values. The study is the first of its kind to examine the connection between efficiency values of a two-stage DEA model, hence only the original CCR model is chosen to examine. The contribution can be used as a starting point for future in-depth work and the method can be applied to other DEA models. The finding from the study will support significantly decision makers in making decision and prove that applying two-stage DEA method in evaluating performance efficiency is reasonable.