Apply DEA to the Efficiency Evaluation of R&D projects--A Case study in "Leading Product Development Program

• Main Authors: Chung-ming, Tung, 董鍾明
• Other Authors: LIU, DAY-YANG
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/11189309855595857808

碩士 === 國立臺灣科技大學 === 企業管理系 === 89 === The purpose of this paper is to study the over-efficiency, technical efficiency, input/output slacks, virtual weights and the correlation coefficients between efficiency and government subsidies on R & D of the “Leading Product Development Program” at the Bureau of Industrial Development of the Ministry of Economic Affairs. The research methodology applied to this paper mainly includes the Data Envelopment Analysis to measure the R & D efficiency, model from Charnes, Cooper and Rhodes(1978) to measure the over-efficiency, model from the Banker, Charnes and Cooper (1984) to measure the technical efficiency, input/output slacks method to analyze the extent for which non-efficiency unit could be improved by decreasing the amount of inputs or increasing the amount of outputs, Virtual weights analysis to detect which output contributes the over-efficiency the most, at last, Spearman correlation coefficient to measure the correlation coefficients between efficiency and government subsidies on R & D. Data sources are a sample survey of the 79 research projects in the year 2000 taken from the “Leading Product Development Program.” The main conclusions from this paper include: (1) surprisingly, 62 % of the R & D projects in the year 2000 at the “leading Product Development Program” are classified into non-efficiency unit, (2) through input/output slacks analysis, the result shows that in order to improve their over-efficiencies more than 60 % of the non-efficiency unit of the R & D projects have to reduce more than 50 % of their inputs, (3) via virtual weights analysis, more than 42 % of the R & D projects view derived products output is the most important category, followed by creative and non-creative technology output category with 38 % approval rate, (4) through Spearman correlation coefficients analysis, overall the correlation coefficient turns out an mild inverse relationship between government subsidies on the R & D projects and the over-efficiency.