| dc.description.abstract | Various studies have developed models to predict the number of crash fatalities based on the statistically most significant parameters in police-reported crash data during the Great Recession period (December 2007- June 2009 as per the National Bureau of Economic Research). However, no proper research has been conducted to study the spatial patterns for the fatal crash data during the same period. This study serves as an extension of the study conducted by Project 17-67 funded by NCHRP and aims to understand how the economic downturn affected the spatial patterns of crash fatalities between 2003 and 2017 in the state of Texas. The study divides the fatal crashes dataset into three time periods 2003-2007 (pre-recession era), 2008-2012 (recession era), and 2013-2017 (post-recession era). The study uses the optimized hotspot analysis tool, which first finds the critical distance (fixed distance band) at which spatial correlation between the crash locations is most significant for each dataset. It then finds the hotspots for all the datasets based on the critical distance. The study conducted a hotspot analysis based on two approaches: death counts per crash and death counts/AADT per crash. The use of AADT was considered necessary to remove the factor of traffic flow, which is one of the reasons for the higher number and severity of fatal crashes. The results obtained from the second approach are much more consistent in terms of the number of statistically significant points and their locations across the three time periods than the first approach and is, therefore, used for further analysis. The analysis showed that the number of hotspots and coldspots roughly doubled during the recession period compared to the pre-recession period, despite a reduction in the number of fatalities. Similarly, although the number of fatalities in the post-recession era increased, the number of hotspots and coldspots remained similar in numbers during the same period. | en |
