교차로 안전개선대책의 사고보정계수 분석: 도로·교통 특성 조절 효과 중심으로 = Crash Modification Factor Analysis of Intersection Safety Countermeasures: Moderating Effects of Road and Traffic Characteristics

다국어 초록 (Multilingual Abstract)

This study empirically examines the effects of various traffic safety countermeasures implemented at intersections on crash frequency and severity, with consideration of road and traffic characteristics. Using data from improvement projects targeting high-risk intersections over the past decade, crash records before and after intervention were collected for 1,892 sites. A total of 104 moderated regression models were constructed, incorporating four key contextual variables: intersection type, major-road speed limit, traffic volume, and road classification. Of these, 95 models yielded statistically significant results, with those including road classification demonstrating particularly strong explanatory power. The analysis revealed that the effectiveness of countermeasures varied by condition: channelization was most effective at three-leg intersections; skid-resistant pavement on high-speed roads; and lighting and jaywalking-prevention facilities at high-volume intersections. Moreover, national highways exhibited greater crash-reduction effects than metropolitan or municipal roads, suggesting a combined influence of road function and speed environment. To support policy development, the study categorized countermeasures into three types—safety facilities, control and regulation facilities, and road/pedestrian space improvements—and proposed installation strategies tailored to specific road conditions. In addition, Korea-specific Crash Modification Factors (CMFs) were derived and compared with international benchmarks, providing a localized empirical basis for evaluating safety interventions. The findings offer practical guidance for prioritizing improvement projects, designing countermeasure combinations, and allocating resources. They also lay the foundation for establishing a Korean Safety Performance Function (SPF), thereby contributing to data-driven traffic safety planning and evaluation.