An early well-known hypothesis of Mourant & Rockwell (1970, 1972), which is based on their eye-movement measurements, states that drivers learn to use peripheral vision in lane-keeping while beginners need foveal vision for it. This hypothesis has not been confirmed in real-life experimental settings, however. We recently showed that when forced to do a foveal in-car task, more experienced drivers are better able to keep the car in the lane than novices when the task eccentricity increases from 7 degrees to 23 degrees, thus supporting the hypothesis. This paper reviews two further experiments using the same forced peripheral vision driving paradigm. The first, using a similar representative sample of young male conscripts in similar conditions (lane width of 3m and speed of 30km/h), confirmed the result. The other, using psychology students in conditions closer to normal highway driving (lane width of 3.75m, speed of 60km/h), also included blindfold driving in order to check the use of kinaesthetic and tactual information in each experience group. The results were in the expected direction but far from significant, presumably due to somewhat different conditions and to the fact that subjects were from highly selected population who were able to develop ad hoc strategies in the task. No experience effect was found in blindfold performance. The results also showed that the foveal task load does not influence peripheral lane keeping performance, by contrast with the concept that attention within the visual field is a function of foveal load.
Abstract