Large Truck Crashes in Texas: A Predictive Approach for Identifying Those at Higher Risk
Jodi L. Carson, Texas A&M University, August 2007, 82 pp. (473700-00089-1)
The objective of this research is to characterize large truck safety levels in Texas on the basis of driver, vehicle, cargo, and carrier traits while controlling for the effects of crash and operating environment conditions. Data to support this investigation considered a three-year time span from January 1, 2004 to December 31, 2006. During this time, 44,012 total large truck crashes occurred in Texas. Historical large truck crash and carrier profile information was collected from three sources: the Motor Carrier Management Information System (MCMIS) Crash File, the MCMIS Census File, and the Texas Department of Transportation intrastate carrier database. Crash severity served as a surrogate measure for large truck safety and was modeled using ordered probit regression methods. The relationships between the significant (denoted through t-statistics values ³|1.645|) explanatory variables and crash severity were for the most part intuitive and in agreement with previously reported findings. Despite their individual significance for contributing to large truck crash severity levels, these variables in combination achieved a poor overall goodness of fit (r2=0.0029), likely attributable to missing data, timeline inconsistencies between crash and census data, and repeated measures (i.e., if a carrier was involved in more than one crash between 2004 and 2006, the same carrier characteristics were repeated). As indicated from the modeling exercise, roadside and carrier-based, on-site safety enforcement practices should be focused towards: (1) single-unit, three-axle and truck tractor (bobtail) vehicle configurations, (2) tank cargo body types, (3) grain/feed/hay cargo classifications, (4) carriers based in Illinois , and (5) private property carriers. Focusing safety efforts toward these factors, which have shown a significant influence on the severity of a potential crash, provides a more proactive approach to enhancing large truck safety.
Keywords: Large Truck Safety, Commercial Vehicle Safety, Crash Severity, Ordered Probit Regression
ENTIRE REPORT (Adobe Acrobat File – 904 KB)