Abstract

Sample sets of crash data are used to examine the similarities in crash-contributing factors among various counties in the state of Arkansas that have similar effects on spatial autocorrelation. Moran's I and Getis–Ord Gi* statistics were used to determine the correlation, and multinomial logistic regression was used to identify the crash-contributing factors. Seventy-five counties were divided into five categories on the basis of the Z-values of the Getis–Ord Gi* statistic. Depending on the sample data size, for each category crash data from a county or a group of counties were used, and crash-contributing factors were identified on the basis of the crash severity index. Results indicated that most of the crash-contributing factors identified for each category were also identified by the sample crash data from a county or a group of counties in that category. Pulaski County, with the highest Z-value from the first category, had the largest cluster of crashes and identified the highest percentage (55%) of factors that contributed to crashes in the category by using the sample crash data. From the sample data used, the multinomial logistic regression indicated the following factors to be positively associated with crash severity: nighttime driving, driving under the influence of alcohol, roadway gradient, alignment on a curve, rural areas, and collision types head-on and sideswipe-same-direction. The results of this research can be used for better allocation of funds by departments of transportation by analyzing smaller sets of data to identify crash-contributing factors associated with higher levels of crash severity.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call