DRIVER JUDGMENT AND ERROR DURING THE AMBER PERIOD AT TRAFFIC LIGHTS With an Appendix on the Calculation of Thresholds and Subjective Scales

Abstract

This paper describes tho behaviour of a small group of drivers who took part in an experiment on stopping at traffic lights. The data of their judgments and subsequent performance aro given and compared. Generally, the judgments of when to try to stop whon the amber signal appeared wore found to be consistent and fairly accurate. However, tho judgment thresholds at each speed—the distances from the traffic lights whon thoy showed amber at which drivers decided to stop on 50 per cent of the occasions—were found to require almost the peak vehicle braking performance, and a third of these decisions to stop lod to errors, hesitations or changes of mind. Equations havo been derived from tho data which link tho proportion of occasions the drivers decided to stop, tho proportion of successful stops, tho overall stopping distance, the initial speod, the drivers' rosponso times, and the average decelerations during the hraking period. Examination of tho errors, hesitations anil changes of mind suggost that the distances in which 95 per cent of the drivers stopped successfully might bo used as design parameter in the calculations of tho minimum amber poriod for a traffic light. In the Appendix tho theory of the subjective scale of difficulty ia devoloped as it applies to this experiment.