This paper presents a qualitative analysis of data obtained in earlier work, complementing the theoretical and empirical approaches used there. It considers the behaviour of drivers and their observed actions at traffic signals, in terms of driving behaviours and situations which could not be identified from purely numerical data. Red-running and non-red-running vehicles from three signal-controlled junctions in different English sites were matched on a site-by-site basis. There was first a pilot study of 17 matched pairs at the first site. The main sample consisted of 17, 43, and 133 matched pairs from the three sites. It consisted of all vehicles which passed the stopline after amber onset. The vehicles were classified into eleven categories according to their relation to other vehicles, road conditions, and their drivers' behaviour. The qualitative analysis showed the following factors to be associated individually with red-running: (1) obscuration; (2) the presence of other vehicles on the road; (3) the presence of pedestrians; (4) complex road and traffic conditions; (5) dithering by drivers; and (6) being in a platoon but not leading it. The last three factors were significant when considered together. It was clear that dithering was especially strongly associated with red-running, indicating that red-running is related to drivers' difficulties in reaching appropriate decisions. This suggests that a remedy, which simplifies decision making, would help to reduce red-running.
Abstract