基于理想窗寬的DEA視窗分析模型的我國高�？蒲性u價
發(fā)布時間：2018-08-26 10:21
【摘要】：高等學校是我國科研活動的重要力量，尤其在基礎(chǔ)研究活動中占有重要地位，我國政府對高校的科研創(chuàng)新一直都很重視。作為一個發(fā)展中國家，在國內(nèi)紛紛提出建設(shè)世界一流大學的時代背景下，尤其在當前國家急需創(chuàng)建科技創(chuàng)新體系的形勢下，如何充分利用我國高校有限的科研資源，提高我國高校在全球的學術(shù)競爭力，建設(shè)世界一流的大學，對高校的科研效率進行科學合理的評價顯得尤為重要和迫切。 科研評價方法有很多種，考慮到高�？蒲谢顒泳哂卸嗤度攵喈a(chǎn)出的特征，以及數(shù)據(jù)包絡(luò)分析方法（Data Envelopment Analysis, DEA）在解決多投入多產(chǎn)出的綜合評價問題所具有的優(yōu)點，國內(nèi)外許多學者都選擇了DEA方法進行研究，取得了許多研究成果。但是，已有的研究成果都忽視了一個問題——大部分對高校的科研效率評價都是基于傳統(tǒng)DEA模型，而傳統(tǒng)DEA模型則隱含假設(shè)“投入產(chǎn)出發(fā)生在同一個時間段”，即高校當年投入當年得到產(chǎn)出，這顯然與“高�？蒲型度氘a(chǎn)出分布在多個時間段”的現(xiàn)實情況不相符，即高校的科研經(jīng)費是分多個年份撥入高校，并且高校的科研成果在連續(xù)多年里按照科研進度分階段發(fā)表。此外，在現(xiàn)實的許多研究中，需要經(jīng)常對決策單元（Decision MakingUnit, DMU）進行多個時期的考察，在此基礎(chǔ)上進行面板數(shù)據(jù)分析來研究DMU的效率的波動和變化情況，一些學者選擇應用DEA視窗分析模型（WindowAnalysis, WA）來進行動態(tài)分析。然而，在應用DEA視窗分析模型進行動態(tài)分析的研究過程中，仍然存在一些問題：學者基于主觀選擇的不同窗寬所得出的結(jié)果存在比較大的偏差；不同窗寬的結(jié)果存在偏差，如何確定理想的窗寬使得偏差最��；投入產(chǎn)出之間存在一定的滯后期，而滯后期對于視窗分析結(jié)果存在比較大的影響，這一點在已有的研究成果中都沒有引起重視。針對上述這些問題，引出了本文所研究的問題。 由于已有研究成果的科研評價指標比較零散，并且主觀選擇的隨意性比較大，不能夠較全面的涵蓋高�？蒲谢顒�，具有一定的片面性。所以，在第一章介紹國內(nèi)外DEA模型研究現(xiàn)狀的基礎(chǔ)上，本文在第二章基于科研評價指標體系的建立標準和構(gòu)建步驟，建立了我國高校的科研評價指標體系；其次，在第三章，基于我國高校的科研評價指標體系，應用DEA視窗分析模型，對我國高校在2003-2007年、2003-2008年和2003-2009年的科研效率進行動態(tài)評價，并對不同窗寬的分析結(jié)果進行了對比分析；在第四章，通過對比不同窗寬的分析結(jié)果，揭示了“窗寬的確定”對DEA視窗分析模型結(jié)果有重要影響，先從理論上進行分析，，研究確定理想窗寬，建立理想窗寬的DEA視窗分析模型，對我國高校多年的科研效率進行動態(tài)分析；最后在第五章，增加考慮高�？蒲型度氘a(chǎn)出的滯后期，基于理想窗寬的DEA視窗分析模型，定量分析滯后期對我國高�？蒲行实挠绊�。 研究結(jié)果表明：隨著窗寬的增加，我國各省市高校的技術(shù)效率（TechnicalEfficiency, TE）和純技術(shù)效率（Pure Technical Efficiency, PE）呈現(xiàn)遞減的變化趨勢，而規(guī)模效率（Scale Efficiency, SE）呈不規(guī)律的變化；科研效率（包括TE，PE和SE）在不同年份對窗寬變化的敏感程度不同，即在某些年份，科研效率值對窗寬的變化比較敏感，表現(xiàn)為不同窗寬下的效率值相差比較大，在某些年份，科研效率值對窗寬的變化不敏感，表現(xiàn)為不同窗寬下的效率值相差不大；選擇不同的窗寬會使得研究者對科研效率的長期變化趨勢得出大相徑庭的結(jié)論。通過對比理想窗寬的DEA視窗分析模型與傳統(tǒng)DEA模型的分析結(jié)果，發(fā)現(xiàn)兩者之間存在很大的偏差，這進一步說明了建立新的DEA視窗分析模型，對我國高�？蒲行蔬M行動態(tài)分析是很有必要的，此DEA視窗分析模型的分析結(jié)果更符合科研工作的實際情況�？紤]滯后期與不考慮滯后期的分析結(jié)果存在很大的偏差，說明滯后期對我國高�？蒲行试u價具有很大的影響，在高�？蒲性u價中必須引起重視。本文基于理想窗寬的DEA視窗分析模型的分析結(jié)果，對我國高校的科研評價提出了一些對策建議。
[Abstract]:Colleges and universities are an important force in our country's scientific research activities, especially in basic research activities. Our government has always attached great importance to the scientific innovation of colleges and universities. Under the situation, how to make full use of the limited scientific research resources of our universities, improve the academic competitiveness of our universities in the world, and build a world-class university, it is particularly important and urgent to evaluate the scientific research efficiency of universities scientifically and rationally.
There are many evaluation methods for scientific research. Considering the characteristics of multi-input and multi-output of scientific research activities in Colleges and universities, and the advantages of Data Envelopment Analysis (DEA) in solving the problem of multi-input and multi-output comprehensive evaluation, many scholars at home and abroad have chosen DEA method for their research, and have made a lot of research. However, the existing research results have ignored a problem - most of the evaluation of scientific research efficiency in universities are based on the traditional DEA model, while the traditional DEA model implicitly assumes that "input and output occur in the same period of time", that is, the University input in the same year is output, which is obviously in line with "the distribution of university scientific research input and output in the same period of time." The reality of "multiple time periods" is not consistent, that is, the research funds of universities are allocated to universities in different years, and the scientific research achievements of universities are published in stages according to the progress of scientific research for many years in succession. On the basis of panel data analysis to study the fluctuation and change of DMU efficiency, some scholars choose to use the DEA window analysis model (WA) for dynamic analysis. The results obtained from the same window width have larger deviations; the results from different window widths have deviations; how to determine the ideal window width to minimize the deviation; there is a certain lag period between input and output, and lag period has a greater impact on the analysis results of the window, which has not been paid attention to in the existing research results. In view of the above problems, the problems in this paper are introduced.
Because of the scattered evaluation indexes of existing research results and the arbitrariness of subjective selection, it can not cover the scientific research activities of colleges and universities comprehensively, so it has certain one-sidedness. Secondly, in the third chapter, based on the scientific research evaluation index system of Chinese universities, the research efficiency of Chinese universities in 2003-2007, 2003-2008 and 2003-2009 is dynamically evaluated by using DEA window analysis model, and the results of different window widths are analyzed. In the fourth chapter, by comparing the analysis results of different window widths, it is revealed that the determination of window widths has an important influence on the results of DEA window analysis model. Finally, in the fifth chapter, considering the lag period of the input and output of university scientific research, the influence of lag period on the efficiency of university scientific research in China is analyzed quantitatively based on the DEA window analysis model with ideal window width.
The results show that with the increase of window width, the technical efficiency (TE) and pure technical efficiency (PE) of universities in various provinces and cities in China show a decreasing trend, while the scale efficiency (SE) shows an irregular change; the research efficiency (including TE, PE and SE) varies with the window width in different years. In some years, the efficiency of scientific research is more sensitive to the change of window width. In some years, the efficiency of scientific research is not sensitive to the change of window width. In some years, the efficiency of scientific research is not sensitive to the change of window width. By comparing the results of the DEA window analysis model with the traditional DEA model, it is found that there is a great deviation between the two models. This further shows that it is necessary to establish a new DEA window analysis model for the dynamic analysis of scientific research efficiency in Chinese universities. The results of this DEA window analysis model are more in line with the actual situation of scientific research. Considering the lag period and not considering the lag period, there is a great deviation in the analysis results, which shows that the lag period has a great impact on the evaluation of scientific research efficiency in China's universities and must be paid attention to in the evaluation of scientific research in universities. The analysis results of the model provide some countermeasures and suggestions for the scientific research evaluation of universities in China.
【學位授予單位】：哈爾濱工業(yè)大學
【學位級別】：博士
【學位授予年份】：2012
【分類號】：G644
本文編號：2204561
[Abstract]:Colleges and universities are an important force in our country's scientific research activities, especially in basic research activities. Our government has always attached great importance to the scientific innovation of colleges and universities. Under the situation, how to make full use of the limited scientific research resources of our universities, improve the academic competitiveness of our universities in the world, and build a world-class university, it is particularly important and urgent to evaluate the scientific research efficiency of universities scientifically and rationally.
There are many evaluation methods for scientific research. Considering the characteristics of multi-input and multi-output of scientific research activities in Colleges and universities, and the advantages of Data Envelopment Analysis (DEA) in solving the problem of multi-input and multi-output comprehensive evaluation, many scholars at home and abroad have chosen DEA method for their research, and have made a lot of research. However, the existing research results have ignored a problem - most of the evaluation of scientific research efficiency in universities are based on the traditional DEA model, while the traditional DEA model implicitly assumes that "input and output occur in the same period of time", that is, the University input in the same year is output, which is obviously in line with "the distribution of university scientific research input and output in the same period of time." The reality of "multiple time periods" is not consistent, that is, the research funds of universities are allocated to universities in different years, and the scientific research achievements of universities are published in stages according to the progress of scientific research for many years in succession. On the basis of panel data analysis to study the fluctuation and change of DMU efficiency, some scholars choose to use the DEA window analysis model (WA) for dynamic analysis. The results obtained from the same window width have larger deviations; the results from different window widths have deviations; how to determine the ideal window width to minimize the deviation; there is a certain lag period between input and output, and lag period has a greater impact on the analysis results of the window, which has not been paid attention to in the existing research results. In view of the above problems, the problems in this paper are introduced.
Because of the scattered evaluation indexes of existing research results and the arbitrariness of subjective selection, it can not cover the scientific research activities of colleges and universities comprehensively, so it has certain one-sidedness. Secondly, in the third chapter, based on the scientific research evaluation index system of Chinese universities, the research efficiency of Chinese universities in 2003-2007, 2003-2008 and 2003-2009 is dynamically evaluated by using DEA window analysis model, and the results of different window widths are analyzed. In the fourth chapter, by comparing the analysis results of different window widths, it is revealed that the determination of window widths has an important influence on the results of DEA window analysis model. Finally, in the fifth chapter, considering the lag period of the input and output of university scientific research, the influence of lag period on the efficiency of university scientific research in China is analyzed quantitatively based on the DEA window analysis model with ideal window width.
The results show that with the increase of window width, the technical efficiency (TE) and pure technical efficiency (PE) of universities in various provinces and cities in China show a decreasing trend, while the scale efficiency (SE) shows an irregular change; the research efficiency (including TE, PE and SE) varies with the window width in different years. In some years, the efficiency of scientific research is more sensitive to the change of window width. In some years, the efficiency of scientific research is not sensitive to the change of window width. In some years, the efficiency of scientific research is not sensitive to the change of window width. By comparing the results of the DEA window analysis model with the traditional DEA model, it is found that there is a great deviation between the two models. This further shows that it is necessary to establish a new DEA window analysis model for the dynamic analysis of scientific research efficiency in Chinese universities. The results of this DEA window analysis model are more in line with the actual situation of scientific research. Considering the lag period and not considering the lag period, there is a great deviation in the analysis results, which shows that the lag period has a great impact on the evaluation of scientific research efficiency in China's universities and must be paid attention to in the evaluation of scientific research in universities. The analysis results of the model provide some countermeasures and suggestions for the scientific research evaluation of universities in China.
【學位授予單位】：哈爾濱工業(yè)大學
【學位級別】：博士
【學位授予年份】：2012
【分類號】：G644
【參考文獻】
相關(guān)期刊論文 前10條
1 童樺,唐慧君,唐暉;研究型大學科研評價指標體系研究[J];湖南大學學報(社會科學版);2005年03期
2 杜偉;高�？蒲性u價現(xiàn)狀與完善途徑探析[J];高等教育研究;2004年04期
3 毛娜;;淺議高�？蒲性u價制度創(chuàng)新[J];黑龍江教育(高教研究與評估);2008年Z2期
4 安寧;鄧開喜;莫雷;;高�？蒲性u價體系構(gòu)建的探討——以華南師范大學科研業(yè)績評價體系構(gòu)建為例[J];華南師范大學學報(自然科學版);2008年03期
5 田東平 ,苗玉鳳;基于DEA的我國高�？蒲行试u價[J];理工高教研究;2005年04期
6 楊印生，李樹根，郝海;數(shù)據(jù)包絡(luò)分析（DEA）的研究進展[J];吉林工業(yè)大學學報;1994年04期
7 周文燕;;我國高�？蒲性u價存在的問題及對策[J];吉首大學學報(自然科學版);2006年04期
8 侯光明,晉琳琳;DEA方法在研究型大學建設(shè)績效評價中的應用[J];高教發(fā)展與評估;2005年05期
9 韓海彬;李全生;;基于AHP/DEA的高校人文社會科學科研效率評價研究[J];高教發(fā)展與評估;2010年02期
10 田東平,苗玉鳳,崔瑞鋒;我國重點高�？蒲行实腄EA分析[J];科技管理研究;2005年08期
相關(guān)博士學位論文 前3條
1 李兆瓊;含有雙性變量的系統(tǒng)的DEA研究[D];中國科學技術(shù)大學;2010年
2 陳敬全;科研評價方法與實證研究[D];武漢大學;2004年
3 張亞杰;高�？萍紕�(chuàng)新能力評價研究[D];武漢理工大學;2010年
本文編號：2204561
本文鏈接：http://sikaile.net/jiaoyulunwen/suzhijiaoyulunwen/2204561.html
最近更新
教材專著
