• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.34 no.4
• /
• pp.1279-1287
• /
• 2014

The purpose of this study is to develop the U-turn accident model at signalized intersections in urban areas. The characteristics of the accidents which are associated with U-turn operation at 3 and 4-legged signalized intersections was analyzed and the U-turn accident model was developed by regression analysis in Changwon city. First, in order to analyze the effectiveness on traffic accidents by U-turn installation, the difference of mean of traffic accident number are measured between two groups which are composed by whether or not U-turn installation the groups by Mann-Whitney U test. The result of significance test showed that intergroup comparison on mean by accident types made difference except rear-end accident type and by accident locations exit section only showed difference in significance level at 4-legged intersections, so the accident number have more where the U-turn is permitted than not. Response measures about the number of accidents were classified by whether accidents occurred and accident model were constructed using binomial logistic regression analysis method. The developed models show that the variables of conflict traffic, number of opposing lane are adopted as independent variable for both intersections. The variables of longitudinal grade for 3-legged signalized intersection and number of crosswalk for 4-legged signalized intersection at which the U-turn is permitted is adopted as independent variable only. These study results suggest that U-turn would be permitted at the intersection where the number of opposing lane is more than 3.5 each, the longitudinal grade of opposing road is upward flow and there is need to establish the U-turn traffic sign at signalized intersections.

• The Korean Journal of Applied Statistics
• /
• v.30 no.4
• /
• pp.513-528
• /
• 2017

Statistics are often used to test two samples if they have been drawn from the same underlying distribution. In this paper, we introduce several well-known distribution-free tests to compare distributions and conduct an extensive Monte-Carlo simulation to specify their behaviors. We consider various circumstances of when two distributions vary in (1) location, (2) scale, (3) symmetry, (4) kurtosis, (5) tail weight. A practical guideline for two-sample goodness-of-fit test is presented based on the simulation result.