大學教學績效之標竿管理路徑分析──圓錐比例層級資料包絡分析法之應用

Benchmarking of University Teaching Performance: An Application of the Cone-Ratio Tiered Data Envelopment Analysis Method

臺灣因少子化現象加劇大學間的競爭，各校莫不積極提升教學、研究與服務表現來提高學生的就讀意願。透過標竿管理（Benchmarking）有系統地尋找合適的參考對象，並據以規劃績效改善的路徑，確實是一種可行的目標管理方式。過往有許多文獻利用資料包絡分析法（Data Envelopment Analysis, DEA）來進行標竿管理研究，然而，傳統資料包絡分析法存在兩個缺點：一、權數分配不合理，導致績效鑑別度不足；二、標竿選取不當的兩個缺點。過去雖有研究曾利用AR-DEA解決權數分配不合理的問題，並採用TDEA進行大學教學表現的效率分層。但是其AR-TDEA卻無法明確地找出該無效率DMU之逐層標竿大學，據以規劃教學效能的改善路徑。因此本研究結合圓錐比例法（Cone-Ratio, CR）與層級資料包絡分析法（Tiered Data Envelopment Analysis, TDEA），提出CR-TDEA模型改善上述缺點。並利用此模型針對臺灣的大學教學表現進行績效分層並據以提出改善路徑規劃方法。實證結果顯示，CR-TDEA能有效改善教學效能的鑑別度及標竿選擇的適用性，為教學效能不佳的大學提供短、中、長期的績效改善方案。本研究展示了CR-TDEA在教學效能提升中的實用性，並為臺灣高等教育標竿管理提供實用工具。

The declining birthrate in Taiwan has intensified competition among universities, compelling institutions to actively enhance their performance in teaching, research, and service. Benchmarking offers a systematic approach to identifying suitable references and planning performance improvement pathways, making it a viable goal management strategy. Previous studies have frequently applied Data Envelopment Analysis (DEA) in benchmarking research. However, traditional DEA has two notable drawbacks: (1) unreasonable weight allocation, leading to poor performance discrimination, and (2) inappropriate benchmark selection. While prior research has employed AR-DEA to address the first issue and utilized TDEA for efficiency stratification of universities' teaching performance, the AR-TDEA approach fails to explicitly identify stepwise benchmark universities for inefficient Decision-Making Units (DMUs) to plan teaching performance improvement pathways.
To address these limitations, this study integrates the Cone-Ratio (CR) method and Tiered Data Envelopment Analysis (TDEA) to propose the CR-TDEA model. This model is applied to stratify the teaching performance of universities in Taiwan and propose methods for planning improvement pathways. Empirical results indicate that CR-TDEA significantly enhances performance discrimination and improves the applicability of benchmark selection, providing underperforming universities with short-, medium-, and long-term performance improvement strategies. This study highlights the practicality of CR-TDEA in enhancing teaching performance and offers a valuable tool for benchmarking in Taiwan's higher education sector.