Research into the validity of the traffic conflicts technique.

The use of near-misses and conflicts as an indirect measure of safety in widespread. The Traffic Conflicts Technique is one procedure for the indirect measurement of safety which has achieved a degree of formalization and attained a measure of popularity in many countries. However, doubts about the validity of the Traffic Conflicts Technique persist. In this paper we examine the conceptual foundations of indirect safety measurement when as data serve observation of events such as near-misses or conflicts. The key problem addressed is that of "validity." A definition of what validity is to mean is suggested. The statistical machinery for the measurement of validity on the basis of empirical evidence is formulated and tested. With this, it may now be possible to move towards rational concensus about the potential of the Traffic Conflicts Technique and similar procedures for the indirect measurement of safety. As a by-product of the analytical results, a yardstick has been created by which to judge the relative performance of the many extant variants of the Traffic Conflicts Technique.