| Summary: | 博士 === 國立交通大學 === 交通運輸研究所 === 92 === Conventional data envelopment analysis (DEA) models do not consider the effects of environmental factors or data noise; nor do these models adjust the “slacks” while making comparison of the relative efficiency for the decision making units (DMUs). As a consequence, factors causing the inefficiency are hardly to be recognized and such performance evaluation is likely biased. To remove these shortcomings, this research proposes a four-stage DEA model in which the effects of environmental factors and data noise are accounted for and the input excesses or output slacks are adjusted. The proposed model is then applied to the evaluation of railway performance, with distinctive measurements between efficiency and effectiveness and between productivity and sales force due to the non-storable nature of the railway transport services. A panel data composed of 44 worldwide railway systems over recent seven years (1995-2001) are used as our tested DMUs. The results of the proposed four-stage DEA have concluded that conventional DEA models understate the efficiency and effectiveness scores but overstate the productivity and sales force measurements. Finally, the implications of the four-stage DEA results are drawn to propose strategies for ameliorating the railway performance. Some possible extensions in the future research are presented.
|
|---|
