The effect of vehicle mass in two-car crashes has been studied intensively in the past. The most well-known results are given by Evans (1) who has shown that in a crash between two cars of different masses, the fatality risk ratio (R) of the lighter to heavier car increases as a power function of mass ratio (µ) of the heavier to the lighter car. This paper explicitlyreviews the methodology that uses this relationship to estimate the effect of µ on R, introduces a new methodology based on generalized linear modeling techniques and disaggregate analysis of crash data, and compares the two methodologies in detail using US crash fatality data. Comparison of results between the two methodologies showed that for the data used, the point estimates of R from Evans methodology were not significantly differentfrom those estimated based on the introduced methodology confirming that Evans relationship can be used as a simple method to reliably estimate R for a given range of µ in a specific set of two-car crashes. However, it was shown that the introduced methodology has some advantages which make ita more appropriate approach for a detailed analysis of two-car crashes. Some of these advantages include its underlying assumption on the relationship between absolute driver fatality risk and vehicle change in speed, itsflexibility in controlling for or estimating the effects of other factorscontributing to fatality risk, and a more reliable estimate of confidenceintervals for R.
Abstract