| Summary: | 碩士 === 國立勤益科技大學 === 流通管理系 === 102 === In this paper, the input and output data drawn from 31 chain stores belonging to a hairdressing company in central Taiwan are used to detect the difference of the aggregate efficiency scores obtained under conventional and parallel DEA models. In parallel DEA model, the 31 stores are divided into five parallel groups according to their performance level. Four dividing methods are employed to constitute four experimental types of parallel groups. The experimental results reveal that the aggregate efficiency score of each store obtained by parallel DEA model is less than or equal to that by conventional DEA model, and more importantly, the number of efficient stores under parallel DEA model is less than that under conventional DEA model. These findings exhibit that parallel DEA model has higher discriminability. Among the four experimental parallel groups, complementary performance type and hierarchical performance type perform better than the others.
|
|---|
