STATISTICAL EVALUATION OF THE EFFECTS OF HIGHWAY GEOMETRIC DESIGN ON TRUCK ACCIDENT INVOLVEMENTS

This paper illustrates how the Poisson regression model can be used (1) to evaluate the effects of highway geometric design on truck accident involvement rates, and (2) to estimate and quantify the uncertainties of the expected reductions in truck accident involvements from various improvements in highway geometric design. The data source used in this study was the Highway Safety Information System (HSIS), a highway safety data base administered by the Federal Highway Administration (FHWA). Five years of highway geometric, traffic, and truck accident data for rural Interstate highways in one of the HSIS State from 1985 to 1989 were used for illustrations.