TY - JOUR
T1 - Analysis on crash reduction factors for road segment safety
AU - Oh, Jutaek
AU - Park, Dongjoo
N1 - Publisher Copyright:
© 2014, © 2014 The Institute of Urban Sciences.
PY - 2014/9/15
Y1 - 2014/9/15
N2 - The CRF (crash reduction factor), which has credibility that can quantitatively explain effects on traffic accidents in the field of traffic safety, is an important area which can evaluate the safety of roads. In this research, the influence on traffic safety using accident data was investigated, focusing on road factors and surrounding environmental factors from road engineers’ point of view. In this study, the predictive levels of CRFs, which were used to evaluate the safety of roads by quantitatively expressing effects on traffic accidents, were developed. Accident models were developed by using data from roadways. It was shown that rather than Poisson or negative binomial, zero-inflated Poisson was more appropriate when the frequency in which 0 (zero) case occurs is high. It is also shown that the horizontal curve and radius, whether a concave section exists, the number of lighting systems, whether level terrain or mountainous terrain exists, and the number of crosswalks have influence on accident in certain sections of road.
AB - The CRF (crash reduction factor), which has credibility that can quantitatively explain effects on traffic accidents in the field of traffic safety, is an important area which can evaluate the safety of roads. In this research, the influence on traffic safety using accident data was investigated, focusing on road factors and surrounding environmental factors from road engineers’ point of view. In this study, the predictive levels of CRFs, which were used to evaluate the safety of roads by quantitatively expressing effects on traffic accidents, were developed. Accident models were developed by using data from roadways. It was shown that rather than Poisson or negative binomial, zero-inflated Poisson was more appropriate when the frequency in which 0 (zero) case occurs is high. It is also shown that the horizontal curve and radius, whether a concave section exists, the number of lighting systems, whether level terrain or mountainous terrain exists, and the number of crosswalks have influence on accident in certain sections of road.
KW - Poisson regression model
KW - arterial roads
KW - negative binomial regression model
KW - old road
UR - http://www.scopus.com/inward/record.url?scp=84912054505&partnerID=8YFLogxK
U2 - 10.1080/12265934.2014.955124
DO - 10.1080/12265934.2014.955124
M3 - Article
AN - SCOPUS:84912054505
SN - 1226-5934
VL - 18
SP - 396
EP - 403
JO - International Journal of Urban Sciences
JF - International Journal of Urban Sciences
IS - 3
ER -