A Latent Class Modeling Approach for Identifying Injury Severity Factors and Individuals at High Risk of Death at Highway-Railway Crossings Naveen ELURU 1 , Morteza BAGHERI 2 , Luis F. MIRANDA-MORENO 3 and Liping FU 4 1 Assistant Professor, McGill University, Department of Civil Engineering & Applied Mechanics, 817 Sherbrooke Street W., Montreal, Quebec, H3A 2K6 ; PH +1-514-398- 6823; FAX +1-514-398-7361; email: naveen.eluru@mcgill.ca 2 Assistant Professor, Iran University of Science and Technology, School of Railway Engineering, Tehran, Iran, 16846-13114; email: morteza.bagheri@gmail.com 3 Assistant Professor, McGill University, Department of Civil Engineering & Applied Mechanics, 817 Sherbrooke Street W., Montreal, Quebec, H3A 2K6 ; PH +1- 514- 398-6589; FAX +1-514-398-7361; email: luis.miranda-moreno@mcgill.ca 4 Professor, University of Waterloo, Department of Civil Engineering & Environmental Engineering, 200 University Avenue West, Waterloo, Ontario, N2L 3G1; PH +1-519- 888-4567 ext.33984; FAX +1-519-888-4349; email: lfu@uwaterloo.ca ABSTRACT The growing focus on improving railway freight transportation in North America has resulted in increased attention to safety at highway-railway crossings (HRC). Recently, federal government agencies such as US Federal Railroad Administration (FRA) and Canadian Transportation Safety Board (TSB) have alluded to safety concerns associated with HRC. Safety at HRCs are of considerable importance to the government as well as the public due to the significant economic and emotional damages associated with accidents at HRC. To address these safety concerns, transportation researchers are focusing on developing countermeasures that enhance safety at HRC. Earlier research on HRC safety has employed a risk based approach considering both frequency and consequence; however, there has been very little research examining the consequence of the collision. In this paper, we aim to identify the different factors that influence injury severity of highway vehicle occupants, in particular drivers, involved in a vehicle-train collision. The commonly used approach to modeling vehicle occupant injury severity is the traditional ordered response model. However, the ordered response model restricts the effect of various factors on injury severity to be constant across all accidents. It is possible that accidents might be grouped (clustered) into different segments to differentiate the effects of various factors at the segment level. The current research effort proposes an innovative latent segmentation-based ordered response model to study injury severity. In this case, individuals (drivers) are assigned probabilistically to different segments with probability of getting injured specific to each segment. The validity and strength of the formulated collision consequence model is tested using the United States Federal Railroad Administration database which includes inventory data of all the railroad crossings in the US and collision data at these HRC crossings from 1584 ICTIS 2011 © ASCE 2011 Downloaded 10 Feb 2012 to 129.97.120.253. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org