W. Cheng, X. Jia
Hotspot identification (HSID) is standard practice in departments of transportations throughout the US. Most agencies use the methods based on either crash count or crash rate to screen roadway networks for hazardous locations. Even though theoretical analyses of crash count and rate and their relationships exist in previous research studies, a comprehensive comparison of the two crash statistics in terms of their hotspot identification performances is lacking and needed. To address this issue, the authors present a study designed to quantify the performance differences of crash count and crash rate in various situations. Both empirical road section crash data and experimentally derived simulated data are used in this study. Three previously proposed evaluation tests are selected and employed to evaluate different aspects of identification performance. Additionally, various levels of confidence are also explored. The results illustrate that crash count significantly outperforms crash rate in identifying the hotspots under each of the evaluation tests. The findings are consistent with those of most of previous research studies.
Keywords: hotspot identification; crash count and rate; empirical and simulated crash data