In this study, the authors use ANOVA (Analysis of Variance) modelling techniques to examine the relationship between injury scores and three independent variables including vehicle class (car, van, pickup), crash type (head-on, rear-end, rollover, etc.), and restraint use in Hawaii. In addition to estimating mean injury levels for various vehicle classes and crash types, the authors also describe the actual mean injury levels for various combinations of vehicle class, crash type, and restraint use. Findings demonstrate the effectiveness of seat belts in reducing injuries across virtually all combinations of vehicle class and crash type. The authors also demonstrate which types of crashes and vehicles have the highest reported injury levels. The analysis is based on comprehensive police crash data for 62,754 motor vehicle occupants involved in collisions in Hawaii during 1990.
Abstract