Modeling Alcohol-Related Crashes for Wisconsin Counties

According to the Wisconsin Dept of Transportation, many people die in alcohol-related traffic crashes every year. In 2002 alone, 290 people were killed in alcohol-related traffic crashes in Wisconsin. A thorough knowledge of the factors leading to alcohol-related traffic crashes is therefore, required in the development of safety improvement plans to minimize the occurrence of these crashes. In addition, knowledge of the timing for any safety improvement requires good assessment of the current and future alcohol-related crashes within a given highway system. Traffic crash prediction involving several independent variables has traditionally used multiple linear regression methods. With several independent variables, the likelihood of multicollinearity problems increases, resulting in inaccurate regression coefficients. This study applied a recursive modeling approach to predict alcohol related traffic crashes for Wisconsin counties based on a set of hypotheses. The hypotheses involved examining the relationships between population density and liquor license availability, liquor license availability and drinking driving, drinking driving and OWI arrests, OWI arrests and OWI convictions, OWI convictions and availability of driver safety plans, driver safety plans and number of offenders completing plans, and ultimately using the number of offenders completing plan as the final predictor of alcohol related crashes. The results of the analysis revealed that alcohol related crashes increased with increasing number of people completing driver safety plans (DSP). Based on the observed anomaly, it is suggested that an investigation into OWI offender attitude towards DSP, as well as effectiveness of DSP categories may be necessary.