Study On Multi-aspects Research Performance Assessment And Optimization For University Colleges And Departments

Colleges and universities always act as the principal parts of scientific research. The levels or degrees of their researches should be at the forefront of science and technology developments, which makes colleges and universities the most important force to enhance the overall national strength as well as the research and technological competitiveness. Therefore, more and more researches have been focused on improving the research abilities of colleges and universities in recent years. I believe that strengthening the management of university research performance evaluation is effective in improving the research ability of university. Early research performance evaluations were most based on qualitative assessment, namely peer-review, which evaluates the academic research work based on the assessment of the invited experts and administrative staff in the same field. In recent years, in order to improve the evaluation of scientific research, a number of research institutions and universities began to introduce the concept of performance evaluation abroad, while trying to adopt new methods to evaluate the level of research performance. However, the theory for the performance evaluation is neither systematic nor scientific. The related researches are too general, lacking pertinence.This thesis first analyses the characteristics of the research performance of departments of engineering universities, then systematically expounds current researches on university research performance evaluation. The connotation and the measurement range of the related indicators of the university research performance are analyzed in detail through extensive surveys and interviews with experts. The weight of each indicator is determined by the use of U-Model method and the combination of subjective and objective entropy. A comprehensive performance evaluation model is constructed based on six indicators, namely the journal articles, government sponsored research projects, patents, research awards, academic posts and achievements appraisal. Then the model is applied for the empirical analysis of the research performance of one secondary college of a university. A research performance assessment and management system has been developed based on Java. Meanwhile, based on data samples from four secondary colleges of five years, the thesis analyzes the correlations between different indicators as well as those between the individual indicators and the comprehensive research performance level. Therefore, effective ways and means to promote the research performance level of universities have been proposed.In addition, the data envelopment analysis (DEA) method is introduced to construct the static and dynamic models of university research performance measure. In the static measurement model, by taking into account of the phenomenon of increasing marginal revenue problem in research activities, we adopt the transfer function to transform the potential production set corresponding to the reference set, in order to meet the convexity requirements of the traditional DEA method on the potential production set, allowing the use of the traditional DEA method to analyze and evaluate the research activity data. In the dynamic measurement model, we adjust the time combination of the input and the output data, considering the hysteresis nature of the input-output process of research activities. Therefore, an input-output system reflecting the hysteresis is constructed, using the Malmquist index to decompose the dynamic evolvements of the indicators of university research activities. The proposed static measurement model that considers the marginal benefits incensement and the dynamic measurement model considering the hysteresis have been adopted to the empirical analysis of the scientific input and output data of ten secondary colleges, which validates the effectiveness of the proposed DEA evaluation model for the university research performance.Finally, we use the Production-theoretical Decomposition Approach (PDA) to establish a university research driving factors decomposition model, which decomposes the factors that affect the research achievements into seven indicators, namely the changes in funding and outcomes, the changes in potential per capita funding, the faculty changes, the changes in technical efficiency of funding, the changes in technical efficiency of faculty, the changes in technical reform of funding, and the changes in technical reform of faculty. The driving factors of changes in research achievements of ten secondary colleges in the period of year 2010 to 2013 are analyzed, where the influence factors of papers, patents and awards are analyzed respectively. The result of the empirical analysis provides a useful reference to plan the excitation mechanism for university research.