Investigation of key factors for accident severity at railroad grade crossings by using a logit model

Shou Ren Hu, Chin-Shang Li, Chi Kang Lee

Research output: Contribution to journalArticle

42 Citations (Scopus)

Abstract

Although several studies have used logit or probit models and their variants to fit data of accident severity on roadway segments, few have investigated accident severity at a railroad grade crossing (RGC). Compared to accident risk analysis in terms of accident frequency and severity of a highway system, investigation of the factors contributing to traffic accidents at an RGC may be more complicated because of additional highway-railway interactions. Because the proportional odds assumption was violated while fitting cumulative logit modeled by the proportional odds models with stepwise variable selection to ordinal accident severity data collected at 592 RGCs in Taiwan as suggested by Strokes et al. [Strokes, M.E., Davis, C.S., Koch, G.G., 2000. Categorical Data Analysis Using the SAS System, second ed. SAS Institute, Inc., Cary, NC, p. 249], a generalized logit model with stepwise variable selection was used instead to identify explanatory variables (factors or covariates) that were significantly associated with the severity of collisions. Hence, the fitted model was used to predict the level of accident severity, given a set of values in the explanatory variables. Number of daily trains, highway separation, number of daily trucks, obstacle detection device, and approaching crossing markings significantly affected levels of accident severity at an RGC (p-value = 0.0009, 0.0008, 0.0112, 0.0017, and 0.0003, respectively). Finally, marginal effect analysis on the number of daily trains and law enforcement camera was conducted to evaluate the effect of the number of daily trains and presence of a law enforcement camera on the potential accident severity.

Original languageEnglish (US)
Pages (from-to)186-194
Number of pages9
JournalSafety Science
Volume48
Issue number2
DOIs
StatePublished - Feb 2010

Fingerprint

Crossings (pipe and cable)
Railroads
railroad
Accidents
accident
Logistic Models
law enforcement
Law Enforcement
Law enforcement
accident frequency
accident risk
Cameras
traffic accident
stroke
Highway accidents
Traffic Accidents
German Federal Railways
Taiwan
Highway systems
data analysis

Keywords

  • Accident severity
  • Generalized logit
  • Marginal effect
  • Proportional odds
  • Railroad grade crossing

ASJC Scopus subject areas

  • Safety Research
  • Public Health, Environmental and Occupational Health
  • Safety, Risk, Reliability and Quality

Cite this

Investigation of key factors for accident severity at railroad grade crossings by using a logit model. / Hu, Shou Ren; Li, Chin-Shang; Lee, Chi Kang.

In: Safety Science, Vol. 48, No. 2, 02.2010, p. 186-194.

Research output: Contribution to journalArticle

@article{ac5bd3eec86046a8ae1878882c0105d9,
title = "Investigation of key factors for accident severity at railroad grade crossings by using a logit model",
abstract = "Although several studies have used logit or probit models and their variants to fit data of accident severity on roadway segments, few have investigated accident severity at a railroad grade crossing (RGC). Compared to accident risk analysis in terms of accident frequency and severity of a highway system, investigation of the factors contributing to traffic accidents at an RGC may be more complicated because of additional highway-railway interactions. Because the proportional odds assumption was violated while fitting cumulative logit modeled by the proportional odds models with stepwise variable selection to ordinal accident severity data collected at 592 RGCs in Taiwan as suggested by Strokes et al. [Strokes, M.E., Davis, C.S., Koch, G.G., 2000. Categorical Data Analysis Using the SAS System, second ed. SAS Institute, Inc., Cary, NC, p. 249], a generalized logit model with stepwise variable selection was used instead to identify explanatory variables (factors or covariates) that were significantly associated with the severity of collisions. Hence, the fitted model was used to predict the level of accident severity, given a set of values in the explanatory variables. Number of daily trains, highway separation, number of daily trucks, obstacle detection device, and approaching crossing markings significantly affected levels of accident severity at an RGC (p-value = 0.0009, 0.0008, 0.0112, 0.0017, and 0.0003, respectively). Finally, marginal effect analysis on the number of daily trains and law enforcement camera was conducted to evaluate the effect of the number of daily trains and presence of a law enforcement camera on the potential accident severity.",
keywords = "Accident severity, Generalized logit, Marginal effect, Proportional odds, Railroad grade crossing",
author = "Hu, {Shou Ren} and Chin-Shang Li and Lee, {Chi Kang}",
year = "2010",
month = "2",
doi = "10.1016/j.ssci.2009.07.010",
language = "English (US)",
volume = "48",
pages = "186--194",
journal = "Safety Science",
issn = "0925-7535",
publisher = "Elsevier",
number = "2",

}

TY - JOUR

T1 - Investigation of key factors for accident severity at railroad grade crossings by using a logit model

AU - Hu, Shou Ren

AU - Li, Chin-Shang

AU - Lee, Chi Kang

PY - 2010/2

Y1 - 2010/2

N2 - Although several studies have used logit or probit models and their variants to fit data of accident severity on roadway segments, few have investigated accident severity at a railroad grade crossing (RGC). Compared to accident risk analysis in terms of accident frequency and severity of a highway system, investigation of the factors contributing to traffic accidents at an RGC may be more complicated because of additional highway-railway interactions. Because the proportional odds assumption was violated while fitting cumulative logit modeled by the proportional odds models with stepwise variable selection to ordinal accident severity data collected at 592 RGCs in Taiwan as suggested by Strokes et al. [Strokes, M.E., Davis, C.S., Koch, G.G., 2000. Categorical Data Analysis Using the SAS System, second ed. SAS Institute, Inc., Cary, NC, p. 249], a generalized logit model with stepwise variable selection was used instead to identify explanatory variables (factors or covariates) that were significantly associated with the severity of collisions. Hence, the fitted model was used to predict the level of accident severity, given a set of values in the explanatory variables. Number of daily trains, highway separation, number of daily trucks, obstacle detection device, and approaching crossing markings significantly affected levels of accident severity at an RGC (p-value = 0.0009, 0.0008, 0.0112, 0.0017, and 0.0003, respectively). Finally, marginal effect analysis on the number of daily trains and law enforcement camera was conducted to evaluate the effect of the number of daily trains and presence of a law enforcement camera on the potential accident severity.

AB - Although several studies have used logit or probit models and their variants to fit data of accident severity on roadway segments, few have investigated accident severity at a railroad grade crossing (RGC). Compared to accident risk analysis in terms of accident frequency and severity of a highway system, investigation of the factors contributing to traffic accidents at an RGC may be more complicated because of additional highway-railway interactions. Because the proportional odds assumption was violated while fitting cumulative logit modeled by the proportional odds models with stepwise variable selection to ordinal accident severity data collected at 592 RGCs in Taiwan as suggested by Strokes et al. [Strokes, M.E., Davis, C.S., Koch, G.G., 2000. Categorical Data Analysis Using the SAS System, second ed. SAS Institute, Inc., Cary, NC, p. 249], a generalized logit model with stepwise variable selection was used instead to identify explanatory variables (factors or covariates) that were significantly associated with the severity of collisions. Hence, the fitted model was used to predict the level of accident severity, given a set of values in the explanatory variables. Number of daily trains, highway separation, number of daily trucks, obstacle detection device, and approaching crossing markings significantly affected levels of accident severity at an RGC (p-value = 0.0009, 0.0008, 0.0112, 0.0017, and 0.0003, respectively). Finally, marginal effect analysis on the number of daily trains and law enforcement camera was conducted to evaluate the effect of the number of daily trains and presence of a law enforcement camera on the potential accident severity.

KW - Accident severity

KW - Generalized logit

KW - Marginal effect

KW - Proportional odds

KW - Railroad grade crossing

UR - http://www.scopus.com/inward/record.url?scp=70449522968&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70449522968&partnerID=8YFLogxK

U2 - 10.1016/j.ssci.2009.07.010

DO - 10.1016/j.ssci.2009.07.010

M3 - Article

AN - SCOPUS:70449522968

VL - 48

SP - 186

EP - 194

JO - Safety Science

JF - Safety Science

SN - 0925-7535

IS - 2

ER -