Validation of Crash Modification Factors Derived from Cross-Sectional Studies with Regression Models

Abstract

Crash modification factors (CMFs) can be used to capture the safety effects of countermeasures and play a significant role in traffic safety management. As an alternative to the before-and-after study, the regression model method has been widely used for estimating CMFs. Although before-and-after studies are considered to be superior, the use of regression models for estimating CMFs has never been fully investigated. Consequently, the conditions in which regression models could be used for such a purpose were examined. CMFs for three variables—lane width, curve density, and pavement friction—were assumed and used for generating random crash counts. Then CMFs were derived from regression models by using the simulated crash data for three different scenarios. The results were then compared with the assumed true values. The study results showed that (a) when all factors affecting traffic safety are identical in all segments except those of interest, CMFs derived from regression models should be unbiased; (b) if some factors having minor safety effects are omitted from the models, the accuracy of estimated CMFs can still be acceptable; and (c) if some factors already known to have significant effects on crash risk are omitted, CMFs derived from the regression models are generally unreliable. Thus, depending on missing variables not included in the model, the transportation safety analyst can decide whether CMFs developed from regression models should be used for highway safety applications.