In recent years researchers have become increasingly involved in the analysis of accident data in an attempt to make inferences on the processes of accident “causation.” As a result of this interest, many statistical hypotheses have been formulated and examined using such data, and many elaborate projects have been undertaken to develop information collection and retrieval schemes based upon these data. In many cases this increased activity has proceeded in an ad hoc manner. In part this is because of the complexity of this type of research. In addition, however, it is due to the fact that a general framework has not been utilised to organise and guide this research. The purpose of this paper is to introduce such a general framework by applying the concepts of probability theory and statistics. A probabilistic approach is desirable because of the residual uncertainties which always remain when drawing conclusions from accident data analyses. This is due to the nature of the accident process, which is stochastic and not deterministic. In the first section of this paper concepts of probability theory are utilised to develop the basic analysis. Next, techniques for making inferences on these probability distributions are reviewed. Finally, various problems which arise when implementing this approach are briefly examined.
Samenvatting