The logic of empirical testing of accident prediction models

Rune Elvik

doi:10.55329/lpdp1271

Authors

Rune Elvik Institute of Transport Economics (TØI), Norway https://orcid.org/0000-0003-1460-350X

DOI:

https://doi.org/10.55329/lpdp1271

Keywords:

accident prediction model, empirical testing, falsification, research synthesis

Abstract

This paper explains the logic of empirical testing of accident prediction models. The key element of empirical testing is to make out-of-sample predictions of the number of accidents. This means that a model developed in sample A is applied, without modification, to predict the number of accidents in sample B. The procedure is illustrated in two samples formed by randomisation. A model fitted to the first sample was applied to predict the number of accidents in the second sample. The model was only partly supported. In general, any accident prediction model is likely to be merely a local statistical description of a particular data set. If tested by means of out-of-sample predictions, the model is very likely to be falsified. This does not mean that accident prediction models do not show general tendencies, but these tendencies are likely to be empirically supported only at a qualitative level, or at best an ordinal level of numerical measurement. In this sense accident prediction models are similar to many models developed in economics. The models predict the direction, and in some cases the relative strength of statistical relationships, but not their precise numerical values.

Downloads

Download data is not yet available.

Author Biography

Rune Elvik, Institute of Transport Economics (TØI), Norway

Rune Elvik is a senior road safety researcher at the Institute of Transport Economics, Oslo, Norway. He started at the institute in 1980. He was associate editor of Accident Analysis and Prevention from 1997 to 2004, and Editor-in-chief (with Karl Kim) of the journal from 2005 to 2013. He has currently published 190 papers in scientific journals.

CRediT contribution: Conceptualization, Formal analysis, Writing—original draft, Writing—review & editing.

References

AASHTO (2010). Highway Safety Manual, Washington D.C.: American Association of State Highway and Transportation Officials.

Avelar, R. E., Dixon, K., Kutela, B., Klump, S., Wemple, B., Storm, R., & Morgan, M. (2021). Simple index to assess the calibration quality of safety performance functions based on multiple goodness-of-fit metrics. Transportation Research Record, 2675, 1598-1610. DOI: https://doi.org/10.1177/03611981211008896

Elvik, R. (2010). An exploratory study of mechanisms by which exposure influences accident occurrence. Transportation Research Record, 2148, 76-82. DOI: https://doi.org/10.3141/2148-09

Elvik, R. (2015). Some implications of an event-based definition of exposure to the risk of road accident. Accident Analysis and Prevention, 76, 15-24. DOI: https://doi.org/10.1016/j.aap.2014.12.011

Elvik, R., Erke, A., & Christensen, P. (2009). Elementary units of exposure. Transportation Research Record, 2103, 25-31. DOI: https://doi.org/10.3141/2103-04

Elvik, R., & Høye, A. K. (2023). The use of low-level theory to guide the interpretation of road safety evaluation studies. Safety Science, 158, 105963. DOI: https://doi.org/10.1016/j.ssci.2022.105963

Elvik, R., & Høye, A. K. (2023). Changes over time in the relationship between road accidents and factors influencing them: the case of Norway. Accident Analysis and Prevention, 183, 106989. DOI: https://doi.org/10.1016/j.aap.2023.106989

Farid, A., Abdel-Aty, M., Lee, J., Eluru, N., & Wang, J-H. (2016). Exploring the transferability of safety performance functions. Accident Analysis and Prevention, 94, 143-152. DOI: https://doi.org/10.1016/j.aap.2016.04.031

Fridstrøm, L., Ifver, J., Ingebrigtsen, S., Kulmala, R., & Thomsen, L. K. (1995). Measuring the contribution of randomness, exposure, weather, and daylight to the variation in road accident counts. Accident Analysis and Prevention, 27, 1-20. DOI: https://doi.org/10.1016/0001-4575(94)E0023-E

Friedman, M. (1953). The methodology of positive economics. In: Essays in Positive Economics, Chicago: The University of Chicago Press.

Hauer, E. (2025). HSM models: Warts, blemishes and unexamined assumptions. Unpublished manuscript

Hauer, E., & Bamfo, J. (1997). Two tools for finding what function links the dependent variable to the explanatory variables. Proceedings of 10th ICTCT Workshop, Lund, Sweden:

Kvisberg, J. (2003). Analyse av kryssulykker på hovedvegnettet i Region Øst. Hovedoppgave for faggruppe veg og samferdsel. Trondheim: NTNU, Fakultet for ingeniørvitenskap og teknologi, Institutt for bygg, anlegg og transport.

La Torre, F., Domenichini, L., Branzi, V., Meocci, M., Paliotto, A., & Tanzi, N. (2022). Transferability of the highway safety manual freeway model to EU countries. Accident Analysis and Prevention, 178, 106852. DOI: https://doi.org/10.1016/j.aap.2022.106852

Lord, D., & Mannering, F. (2010). The statistical analysis of crash-frequency data: A review and assessment of methodological alternatives. Transportation Research Part A, 44, 291-305. DOI: https://doi.org/10.1016/j.tra.2010.02.001

Mannering, F. L., Shankar, V., & Bhat, C. R. (2016). Unobserved heterogeneity and the statistical analysis of highway accident data. Analytic Methods in Accident Research, 11, 1-16. DOI: https://doi.org/10.1016/j.amar.2016.04.001

Sawalha, Z., & Sayed, T. (2006). Transferability of accident prediction models. Safety Science, 44, 209-219. DOI: https://doi.org/10.1016/j.ssci.2005.09.001

Srinivasan, R., Carter, D., & Bauer, K. (2013). Safety performance function decision guide: SPF calibration vs SPF development. Final Report. Report FHWA-SA-14-004, US Department of Transportation, Federal Highway Administration.

Srinivasan, R., Colety, M., Bahar, G., Crowther, B., & Farmen, M. (2016). Estimation of calibration functions for predicting crashes on rural two-lane roads on Arizona. Transportation Research Record, 2583, 17-24. DOI: https://doi.org/10.3141/2583-03

Tang, D., Yang, X., & Wang, X. (2020). Improving the transferability of the crash prediction model using the TrAdaBoost.R2 algorithm. Accident Analysis and Prevention, 141, 105551. DOI: https://doi.org/10.1016/j.aap.2020.105551