01333532

Component

https://doi.org/10.3141/2236-01

http://scholar.google.com...hita&publication_year=2011

Structural equation modeling (SEM) is a confirmatory, multivariate technique used to examine causal relationships between variables. Related to path analysis with a goal of selecting a model that best explains underlying relationships between variables, SEM is a useful tool for traffic safety research. This study examined the severity of crashes in terms of factors commonly attributed to accidents. These factors included human, vehicle, and roadway factors, along with accessibility measures that were considered relevant in previous studies. In this study, SEM was used to test an a priori model of crash severity. The analysis was carried out in a two-step process. The measurement model was first tested with a confirmatory factor analysis. After the validity of the measurement model was established, a four-latent-factor structural model was run. With an acceptable model fit, the magnitude of standardized path coefficients from the exogenous latent variables provided a means to assess the relative importance of the latent factors on crash severity. The results showed that the human latent factor was the most influential. Although a positive statistical relationship existed between roadway factors and crash severity, accessibility factors had the opposite effect on crash severity, that is, increased accessibility was shown to reduce crash severity.

01360407

11-2663

English

pp 1-10

2011

Transportation Research Record: Journal of the Transportation Research Board

9780309167567

Print

Figures (4) ; References (33) ; Tables (6)

Accessibility; Crash causes; Crash severity; Highway factors in crashes; Highway safety; Human factors in crashes; Structural equation modeling; Validity; Vehicle factors in crashes

Data and Information Technology; Highways; Safety and Human Factors; I80: Accident Studies

TRIS, TRB, ATRI

Feb 17 2011 6:14PM

• Analysis of Naturalistic Driving Event Data: Omitted-Variable Bias and Multilevel Modeling Approaches
• Application of Latent Class Growth Model to Longitudinal Analysis of Traffic Crashes
• Conflict Models for Single-Lane Roundabout Slip Lanes from Microsimulation: Development and Validation
• Disaggregate Exposure Measures and Injury Frequency Models of Cyclist Safety at Signalized Intersections
• Effects of Mobile Phone Use on Accident Risk: Problems of Meta-Analysis When Studies Are Few and Bad
• Estimation of Traffic Conflict Risk for Merging Vehicles on Highway Merge Section
• Model of Relationship Between Interstate Crash Occurrence and Geometrics: Exploratory Insights from Random Parameter Negative Binomial Approach
• Relation of Flow, Speed, and Density of Urban Freeways to Functional Form of a Safety Performance Function
• Safety Performance Functions for Bicycle Crashes in New Zealand and Australia
• Separation of Safety Effects of Multiple Improvements by Alternate Empirical Bayes Methods
• Speed Profile Variation as a Road Network Screening Tool
• Understanding Causality of Intersection Crashes: Case Study in Virginia