Font Size: a A A

Causal analysis of highway crashes: A systematic analysis approach with subjective and statistical methods

Posted on:2002-04-19Degree:Ph.DType:Dissertation
University:Georgia Institute of TechnologyCandidate:Wu, Chi-Hung EvelynFull Text:PDF
GTID:1462390011997999Subject:Engineering
Abstract/Summary:
Due to high fatality rates, traffic safety issues on rural two-lane highways have drawn the attention of the nation. Improving traffic safety in rural areas becomes an important matter to government agencies. Since rural two-lane highways cover the majority of the highway system, the decision of safety improvements is influential and needs extensive evaluation.; This study focused on three categories: crash database, contributing factors, and countermeasure development. The research identified deficiencies of the current crash database, identified the crash contributing factors on rural two-lane highways in Georgia, and proposed prioritized countermeasures to either prevent crashes or reduce severity.; This study investigated the causal factors of fatal crashes utilizing a sample crash database created by a Georgia Tech research team. The causal factor analysis procedure examined two major phases: crash occurrence and crash severity, by means of two analysis methods: subjective and statistical methods.; For the subjective analysis method, a safety expert in the research team reviewed and investigated each crash. For each case, this safety expert developed causal chain relationships for crash occurrence and crash severity. For the statistical analysis method, the author applied Loglinear models to examine the likelihood of involvement in fatal crashes and multinomial Logit models to characterize statistical relationships between contributing factors and crash severity.; The analysis results for both methods were similar. This study found that human related factors were the initial contributing factor in over 90-percent of crashes. Essentially, the fatal crashes were attributed to driving under the influence (DUI), driver errors, horizontal curve, speeding, narrow lane width, and the absence of paved shoulder . Crash severity was associated with safety restraint usage, age, roadside obstacles, roadside slope, and overturn.; Finally, this study developed a list of prospective engineering-based countermeasures and prioritized these countermeasures based on the significance of their remedying factors and the frequency of their application. Research findings suggested that government agencies should allocate budget to improve horizontal curves, pave shoulders, widen travel lane, and eliminate roadside hazards.
Keywords/Search Tags:Crash, Rural two-lane highways, Statistical, Causal, Safety, Subjective, Methods
Related items