TY - JOUR TI - Estimation of bicycle crash modification factors (CMFs) on urban facilities using zero inflated negative binomial models AU - Raihan, Md Asif AU - Alluri, Priyanka AU - Wu, Wensong AU - Gan, Albert T2 - Accident; Analysis and Prevention AB - The objective of this study was to develop crash modification factors (CMFs) for bicycle crashes for different roadway segment and intersection facility types in urban areas. The study used four years (2011-2014) of crash data from Florida to quantify the safety impacts of roadway and traffic characteristics, bicycle infrastructure, and bicycle activity data on bicycle crashes. A cross-sectional analysis using Generalized Linear Model (GLM) approach with Zero Inflated Negative Binomial (ZINB) distribution was adopted to develop the relevant regression models in this study. Lane width, speed limit, and grass in the median were observed to have positive impacts on reducing bicycle crashes. On the contrary, presence of sidewalk and sidewalk barrier were found to increase the bicycle crash probabilities. Increased bicycle activity was found to reduce the bicycle crash probabilities on segments, while increased bicycle activity resulted in higher bicycle crash probabilities at intersections. Bus stops were found to increase the bicycle crash probabilities at intersections, whereas, protected signal control had a positive impact on bicycle safety. This research provides a greater insight into how various characteristics affect bicycle safety, a topic that is seldom considered by researchers and practitioners. DA - 2019/02// PY - 2019 DO - 10.1016/j.aap.2018.12.009 DP - PubMed VL - 123 SP - 303 EP - 313 J2 - Accid Anal Prev LA - eng SN - 1879-2057 KW - Accidents, Traffic KW - Bicycling KW - Built Environment KW - Crash modification factor (CMF) KW - Cross-Sectional Studies KW - Cross-sectional analysis KW - Crowdsourced bicycle activity data KW - Environment Design KW - Florida KW - Humans KW - Models, Statistical KW - Roadway characteristics KW - Safety KW - Zero inflated negative binomial ER -