Red Turn Arrow: An Information Theoretic Evaluation

Abstract

An information‐theoretic statistical procedure for the analysis of categorical or qualitative variables or count data not necessarily arrayed in a multi‐way cross‐classification or contingency table is presented. This procedure can be used to provide a uniform approach to both statistical testing and estimation in various kinds of traffic studies. In particular, a statistical procedure is presented for analyzing counts of violations observed at various intersections during cycles of red left turn arrow or red ball signals. Data was collected at fifteen intersections of different types at different times of the day and located in different geographical locations.