메타분석을 활용한 중앙분리대의 사고감소효과 분석 = The Effects of Installing Median-Barrier on Crash Reduction using Meta-Analysis

다국어 초록 (Multilingual Abstract)

In this study, we tried to suggest a new method of effect analysis and necessity of median barrier installation as analyzing crash reduction effect of median barrier using meta-analysis. The meta-analysis is a statistical analysis than combine the results of multiple scientific studies. We used the data of ‘Effective Analysis of Improvement Project at Black Spot’ while the meta-analysis is used the results of the scientific studies. First we analyzed the Homogeneity test to assure a heterogeneity of the data. The analysis of Homogeneity showed that the probability I² to have a heterogeneity of data is 77.89%. It means that the data is suitable a Random effect model than fixed effect model. The Random effect model analyzed that the crash reduction effect of median barrier installation is 32%. Therefore, the data were analyzed using a random effect model. In case of using a random effect model, the crash reduction effect of median barrier installation were analyzed 32% while a fixed effect model were analyzed 28%. We performed the sensitive analysis to verify the result. Its result showed that the aggregated estimation have a significant representative. Therefore, we conclude that the crash reduction effect of median barrier is 32%.