Bayesian Analysis of Road Accidents: A General Framework for Multinominal Case

Denis Bolduc () and S. Bonin

Working Papers from Laval - Recherche en Politique Economique

Abstract: The detection of dangerous road sites is usually performed using empirical methods which focus on observed accident frequencies and/or proportions of accidents with a given feature. The most widely used detection tools have an empirical Bayes (EB) background. The EB approaches rely on the comparison of frequencies and/or proportions of accidents at a given site with the amounts that would normally occur at similar sites. Currently, analytical techniques for accident proportions describe the number of accidents with a given feature using a binomial distribution. This paper extends to the multinomial case the general EB technique that we recently suggested to analyze road accident proportions. Our proposed approach is a full-information Bayes method that allows for both deterministic and random heterogeneity as well as spatial-correlation among the sites under investigation. The technique can also be used to analyze accident frequencies. An empirical example based on accident data taken from the Québec city database, will serve to demonstrate its usefulness.

Keywords: ROAD ACCIDENTS; ECONOMIC MODELS (search for similar items in EconPapers)
JEL-codes: C50 C52 L91 (search for similar items in EconPapers)
Date: 1998

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: http://EconPapers.repec.org/RePEc:fth:lavape:9802

Access Statistics for this paper

More papers in Working Papers from Laval - Recherche en Politique Economique
Address: UNIVERSITE LAVAL, GREPE DEPARTEMENT D'ECONOMIQUE, QUEBEC G1K 7P4.
Contact information at EDIRC.
Series data maintained by Thomas Krichel ().