一直以來大學辦學的績效排名即是產官學各界所關注的焦點， 雖然已有不少的文獻在探討大學績效，但是在衡量方法、衡量指標、影響績效因素上還是有一些議題值得再深入研究，本論文主要的目的在於針對大學績效的衡量方式以及影響績效之因素進行探討，包括三個議題，分列於三個章節討論。內容摘要如下： 第二章「大學評鑑之分群與標竿選擇」，目的再利用資料包絡分析法所延伸之交叉效率法搭配叢集分析對大學進行分類，並且找出各類別中最值得仿效的標竿大學。根據本研究之結果，本章所使用之客觀方式與教育部之主觀方式，兩種分群方式之結果顯然不同。根據實證結果，利用研究型或非研究型大學為依據進行分類更為適合。 第三章「以學生為本位之大學商學院之績效評估：保證區域資料包絡分析法之應用」，本章主要目的，以台灣地區大學商學院為對象，建構一個以學生為本位的大學商學院表現評估模式。使用商學院畢業生的就業市場成就與多元能力培養程度，來做為績效衡量指標。本研究結合模糊層級分析法的專家調查結果與保證區域資料包絡分析法，進行實證分析。結果發現，平均而言國立與私立大學績效表現無顯著差異，但就個別學校來看，某些私立大學的績效優於國立大學，此結果與國立皆優於私立的傳統印象相當不同。 第四章「社會資本與研究產出之關係---以台灣地區經濟學者為例」，社會資本是由社會學家所提出的概念，近幾年來受到其他領域的重視，不管是企管、經濟、政治學等等，都開始探討社會資本在各領域中的相關議題。本章的主要目的在於建構一個包含社會資本存量的學術生產函數，並以線性參數模型與半參數平滑參數模型作為實證模式，探討社會資本與研究產出間的相關性。實證結果發現，社會資本變動導致學術生產函數產生非中立性的技術變動。同時，平均而言，台灣地區經濟學者的社會資本投入，尚處於規模報酬遞增的階段，他們應多從事交流活動，以增加社會資本的投入來提高研究產出。 College and University rankings have been the foci of attention when the performance is considered. Although there were many studies had discussed the measurement methodology, measurement index, and the factors affecting performance. But still have many questions that need to investigate. The main goals in this dissertation are constructing the new measurement methodology of college evaluation and investigating the factors affecting college performance. There are three chapters in this dissertation: In the first chapter, we try to classify the universities and find benchmark of the universities when government assesses the universities’ performance. In their recent assessment of university performance in Taiwan, the educational authorities classified the universities in a subjective way. They accorded the year and specialty of university to classify. We want know if we classify them with objective quantitative methodology. Is the result of classification the same? For this propose, we employed the cross-efficiency analysis and cluster analysis to classify the universities and try to find the benchmark of universities. According the empirical result, we found that classify the universities by their effort in academic research is better for assessing universities’ performance. In the second chapter, we measured the relative performance of college of business in Taiwan using AR/DEA model. The purpose of this chapter is to construct a students-based performance evaluation system for business school. Depend on the empirical results, we find that public schools’ performance is similarly to the private schools on average. However, there are still some private schools perform better than the public ones when performances are evaluated from students’ wage and performance. In the third chapter, we try to investigate the relationship between social capital and research output in higher education. We proposes a semiparametric smooth coefficient production model to study the relationship between social capital and research output where coefficients are unknown function of the social capital. Empirical results form 380 economists in Taiwan show that, social capital is the factor of non-neutral technical change, and the marginal product of human capital decreasing as the social capital increasing. We also found economist is subject to increase return to scale of social capital. Thus, we suggest the economist in Taiwan should collaborate with others more in research.
