Statistical Test of Difference between Expected Accident Frequencies

Abstract

A primer on the testing of some common statistical hypotheses in road safety is presented. The basic notions of statistical hypothesis testing are reviewed and applied to the specific circumstance when one wishes to test a statistical hypothesis about a change in the expected accident frequency beyond what is the result of a change in traffic and similar influences. The hope is that this exposition will illumine the meaning and intricacy of such tests, inform the decision about when such tests are called for, help users choose the right significance or power, and, because software is now available, improve practice in this matter. A companion paper on the detection of deterioration in safety makes use of the foundation provided.

Get full access to this article

View all access options for this article.

References

Hauer

Statistical Tables for the Testing of the Difference Between the Expected Values of two Poisson Distributed Random Variables with Applications in Transport Safety. Publication 78-06. Department of Civil Engineering, University of Toronto, Canada, July 1978.

Crow

E. L.

, Davis

F. A.

, and Maxfield

M. W.

Statistic Manual. Dover Publications, New York, N. Y., 1960.

Hacking

Logic of Statistical Inference. Cambridge University Press, 1965.

Przyborowski

, and Wilenski

Homogeneity of Results in Testing Samples from a Poisson Series. Biometrika, Vol. 31, 1940, pp. 313–323.