Earticle

본 연구는 대한민국 정부의 기술혁신형 국가연구개발사업 중 하나인 “지식경제기술혁신사 업”을 대상으로 R&D 성과창출 성공확률의 예측모형을 추정하고, 추정된 예측모형의 민감도 검토를 위한 실험설계분석을 실시한다. 분석 대상 R&D 주관기관 유형은 “대기업”, “중소기 업”, 두 가지 영리기관이며, “국가R&D사업관리서비스”에 매년 등록 관리되는 여섯 가지 핵 심 R&D 성과 항목 중 특히, 특허등록, 매출액, 두 가지를 분석 대상으로 선정한다. 또한, 타 기관 간 R&D 협력 여부도 영향요인으로 본 연구에 포함한다. 일반적인 R&D 논리모형에 기 초하여 예측모형을 수립하며, 분석 샘플 n = 1,929개 종료 R&D 과제를 대상으로 분석 샘플 특성을 고려한 로짓 (Logit) 회귀분석을 실시한다. 실증분석 결과를 요약하면, 첫째, 정부지원 금액 대비 성과창출 성공확률의 양 (+)의 상관성은 확인되나, 그 영향력은 특허등록과 같은 단기적, 기술적 산출에서 매출액과 같은 중기적, 경제적 결과로 진행될수록 감소하였다. 둘째, 매출액 성과창출 여부는 직전 단계에 위치한 특허등록 창출 여부에 의존적임이 확인되었다. 셋째, “단독 R&D”와 비교할 때 “협력 R&D”에서 특허등록, 매출액 성과창출 성공확률이 더 높고, “대기업”보다 “중소기업”의 매출액 성과창출 성공확률이 더 높았다. 넷째, 민감도 실험 설계분석을 통해, 정부지원금액이 증가할수록 특허등록 성과창출 성공확률은 S자형 비선형 성장곡선 형태의 증가 추세를 명확히 보인 반면, 매출액 성과창출 성공확률은 상대적으로 선 형에 근사한 둔감한 증가 추세를 보였다. 본 연구가 제안하는 정책적 함의로서, 각 R&D 과 제별 순차적인 성과창출 유도, 타 기관 간 R&D 협력 활성화, 중소기업 친화적인 국가연구개 발사업의 기획 전개 관리 등이 바람직하다고 판단된다. 또한, 정부지원금액 증액을 통해 성 취 가능한 R&D 성과 항목은 한정적일 수 있으므로, 각 R&D 과제별 정부지원금액 규모의 적절성을 세밀히 검토하는 절차가 향후 좀 더 강화될 필요가 있다고 판단된다.

This study presents an empirical case where a prediction model is estimated for the performance creation success probability regarding a national technology innovation research and development (R&D) program. Additionally, a framework for a design of experiments is established to conduct sensitivity analyses on the predicted probability. The prediction model is developed based on typical R&D logic models, and it encompasses the total of five distinct R&D characteristics such as R&D Government Subsidy (X1), Patent Registration (Y1), Sales (Y2), Company Type (T1) and Collaboration Type (T2). In the analysis, the sample of n = 1,929 observations (i.e., R&D projects) is analyzed using Logit regression models to cope with the sample’s nonparametric characteristics. In particular, an interesting finding from this study is that the predicted probability for Patent Registration (Y1) performance creation success shows an S-shape nonlinear growth curve trend as R&D Government Subsidy (X1) increases. As practical implications, we suggest the following; (1) promoting the sequential R&D performance creation, (2) enabling the R&D collaboration between other agencies, (3) planning, deployment and management for SME-friendly national R&D programs. In addition, a meticulous review process should be incorporated into the R&D management scheme because it was found that the increased amount of R&D Government Subsidy (X1) might have a limited effect up to the short term and technical outputs only.

Since the deregulation of the aviation market, low-cost carriers (LCCs) have had a major impact on the aviation network and are expected to intensify competition with existing full-service carriers (FSCs). By using a survey of 2,882 passengers, this study determines factors that can influence passengers’ choice to use Korean FSCs and LCCs by employing a logit regression model. As a result of the analysis, there were differences in the main determinants of the selection of Korean FSCs and LCCs. The regression results indicate that the airfare is the main determinant in choosing LCCs, while service quality, safety, and airline image are determinants in choosing FSCs.

지금까지 제안되어 있는 다항 로짓 회귀모형의 적합도 검정 방법들에 대하여 저자들이 제안한 방법들이 타당한지를 확인하고자 본 연구를 진행하였다. 여러 검정 통계량들 중 그룹화 전략을 이용한 통계량들 (Fagerland 등, 2008; Bull, 1994; Pigeon과 Heyse, 1999)을 선정하였고, 이러한 통계량의 기반이 되는 피어슨 ${\chi}^2$ 통계량 또한 같이 비교하였다. 제안된 분포가 모의실험의 상황 하에 얻어지는 귀무분포와 유사한지, 그리고 부적절한 모형의 판별을 적절히 수행하는지에 대하여 확인하였으며, 실제 자료에 세 가지 방법을 적용한 결과를 비교, 평가하였다.

Several goodness-of-fit test statistics have been proposed for a multinomial logit regression model; however, the properties of the proposed tests were not adequately studied. This paper evaluates three different goodness-of-fit tests using grouping strategies, proposed by Fagerland et al. (2008), Bull (1994), and Pigeon and Heyse (1999). In addition, Pearson (1900)'s method is also examined as a reference. Simulation studies were conducted to evaluate the four methods in terms of null distribution and power. A real data example is presented to illustrate the methods.

The source of national wealth creation is humans and no matter how much research and development investment is made without humans, it cannot be represented in the form of knowledge and technology. In other words, to enhance the productivity to raise the competitiveness of either the country or enterprises, it is necessary to deal with employment productivity of human resources. Up to now, employment productivity has not yet been clearly defined, but it is defined from the aspects that increase in income along with employment stability reinforces social stability and social and economic vitalization, and setting the variables related to employment productivity, which is dealt with through the theories, as the independent factors, the analysis was conducted. The results of the analysis are as follows. First, it was confirmed that educational background and age had effects not only on employment status such as finding jobs but on the stability and continuity of actual employment. Second, in the case of gender, unlike general viewpoint that males are relatively higher in employment stability and permanent employment status, the results of this research supported the results of the past researches that females are higher in employment productivity than males. Third, it was confirmed that the vocational training factors (vocational training experience and vocational training intention) had positive effects on employment productivity. Particularly, it was confirmed that these factors had positive effects on the probability of becoming permanent job status and also contributed positively to the annual income level. Since the vocational training factor is clearly an important variable that has effects on employment productivity, it is necessary to supplement and maintain it systematically.

This paper investigates the effect of weather factors(such as winds, rain, snows, temperature, clouds and humidity) on transmission line outages. The result shows that weather variables have significant effects on the transmission line historical outages and the relationship between them is nonlinear. Multiple regression analysis using Logit model is proved to be appropriate in forecasting line failure rate in KEPCO systems. It could also provide system operators with useful informations about system operation and planing.