R&D 프로젝트 효율성 상관분석 및 사업포지셔닝 조사를 위한 2단계 DEA/AR-I 성과평가모형

This study aims to propose a two-staged DEA/AR-I performance evaluation model for R&D projects efficiency correlation analysis as well as R&D programs positioning investigation according to the stepwise performance creation process relating to government-funded R&D projects. First, it is examined whether a statistically significant correlation exists between the first stage ‘technical output’ and the second stage ‘business outcome’ DEA/AR-I efficiency score. Especially, in order to enhance the reliability of DEA optimal solutions, a DEA/AR-I revised model is developed by adding Acceptance Region Type I(Ⅰ.e. AR-I) constraints into the DEA basic model. Second, R&D programs(Ⅰ.e. R&D project groups classified by technology and organization types) are tested so as to verify significant differences in their DEA/AR-I efficiency scores in terms of Kruskal-Wallis test of hypothesis that is a nonparametric single-factor analysis of variance method. Additionally, two R&D programs positioning matrices are generated by the test statistic. The empirical dataset is composed of 1,864 R&D projects which is 77.89% of the total of 2,393 R&D projects deployed in ‘Industry Technology Development Program' supervised by Ministry of Knowledge Economy(MKE) and administered by Korea Evaluation Institute of Industrial Technology(KEIT) in the year of 2009. Results show that; 1)the sample correlation coefficient ≒0.05 that is statistically significant but its absolute magnitude is very small between the first stage ‘technical output’ DEA efficiency score DEA/AR-I(1) and the second stage ‘business outcome’ DEA efficiency score DEA/AR-I(2); 2) R&D programs positioning matrices locate, ①‘biological․medical’ in superior, ‘mechanical․material’ in inferior position regarding the technology types classification, ②‘venture business’, ‘small and medium-sized business’ in superior, ‘large enterprise’ in inferior position regarding the organization types classification.