Abstract
Work zone crashes have been a major concern of safety for many government agencies and travelling public. According to Federal Highway Administration(FHWA) Fact & Statistics, there is need for maintenance, rehabilitation and advancement of existing road networks, which in turn lead to setup of large number of work zones throughout US which eventually impacts the regular traffic flow and safety. The main objective of this project was to find work zone crash characteristics and risk factors that affect the crash severity. For that purpose data relating to work zone crashes that occurred on all roadways in the state of California for the years 2011 to 2016 were obtained from Statewide Integrated Traffic Records System (SWITRS). A total of 12 variables with 45136 crashes were used to develop a crash severity model. The first step of the study was gathering literature on previous studies relating to work zone crash analysis. In the second step, the work zone crash data for each variable was analyzed based on crash severity and crash patterns were determined. In the third step, each variable was statistically analyzed to determine significant differences between the variable and the dependent variable using Pearson chi-square statistical test. A total of 10 variables out of 12 variables were found to have a significant relationship between the dependent variable. In the fourth step, a binary logistic regression model was used to determine significant predictors of fatal crashes. In this study, the independent variables such as location of crash, violation category, crash type, motor vehicle involved with, lighting condition, alcohol involved, and driver’s gender were found to be significant predictors of fatal crashes at work zones in the state of California for years 2011-2016.