研發組織績效評量過去多採用主觀評量,未建立多投入與多績效(產出)的總體衡量模型,因此未能針對不同部門提出應如何提升績效之改善建議。Data Envelopment Analysis (DEA)能解決上述問題,因此本研究運用構建國內工業技術研究院(簡稱工研院)各研發單位多投入及多產出效果的評量模式。本研究採用工研院1999年及2000年之投入產出資料進行分析,分別求解CCR效率、A&P效率、交叉效率,及多目標效率。此外,亦將CCR效率進一步區分為純粹技術效率(BCC效率)與規模效率。除了效率值比較和相關管理意涵的討論外,還進行規模報酬分析,最後則進行工研院各研究單位此兩年度之經營效率分析與比較,研究結果顯示工研院各研究單位應擴大研發規模,以發揮研發綜效。經與工研院高階管理團隊認為交叉分析模式結合多目標模式產生之結果可信度高且較以往績效評量客觀公正,故此評量模式亦應適用於一般研發組織之研發績效評量。
Most of the performance assessment of R&D organizations or institutions is based on their self-appraisal or subjective judgments. An overall scheme to measure the performance of R&D organizations involving multi-input and multi-effects (output) has not been established. The information regarding the improvement of such performance is also lacking. The method of Data Envelopment Analysis (DEA) may meet the above needs. This study use DEA method to build a model to evaluate the performance of Industrial Technology Research Institute (ITRI) in Taiwan, using the input and output data for years 1999 and 2000. We will solve for the CCR, A&P, Cross and Multi-Purpose Efficiencies. Both CCR and A&P models are of the self-appraisal type, the Cross model is of the peer-evaluation type, while the Multi-Objective is of the overall-evaluation type. Furthermore, the CCR efficiency is divided into purely technical (BCC) and scale efficiencies. In addition to a comparison of efficiencies and associated discussions, an analysis of the scale-return is offered. Finally, the performance of various organizations of ITRI in these two years is analyzed. The result is brought up in interviews with top executives of ITRI. They concluded that the result based on Cross and Multi-purpose models is more credible and realistic. The performance evaluation method devised accordingly is suitable for adoption.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。