Title: Assessing the reliability of road accident severity models

Authors: F.F. Saccomanno, S.A. Nassar, J.H. Shortreed

Addresses: Department of Civil Engineering, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada. ' Department of Civil Engineering, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada. ' Department of Civil Engineering, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada

Abstract: The reliability of different statistical road accident severity models is assessed using criteria, such as, goodness-of-fit, robustness of risk factor coefficients, and intuitive acceptability and consistency of output. The results suggest that model reliability is not sensitive to the number of injury classes specified in the model or to the level of model aggregation. All models consistently identified the same risk factors as explaining most of the variation in injury experience in the data: (1) dynamics of the accident, (2) seating position of occupant, (3) use of seat belts, and (4) age of occupant involved. There is no indication that a significant transfer of error takes place from one severity level to another in a sequential model structure. Reliability of the accident severity models was found to depend primarily on the accuracy of information contained in the reported accident data.

Keywords: model reliability assessment; road accident severity modelling; road accident statistics; road accidents.

DOI: 10.1504/IJHVS.1997.054591

International Journal of Heavy Vehicle Systems, 1997 Vol.4 No.2/3/4, pp.266 - 282

Published online: 18 Jun 2013 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article