This project included a background literature synthesis and observational field study. The research goals were to document driving problems and errors at intersections, for older drivers using their own cars to travel familiar and unfamiliar routes, and to measure how well they could be predicted by prior tests given in an office setting. This volume, Volume I, presents the field study methodology and results; Volume II (See C 24145) presents the background synthesis. Field observations of intersection negotiation were conducted using 82 subjects, age 61 and older, referred to the California Department of Motor Vehicles (DMV) for special testing. The subjects first completed a functional test battery measuring vision, attention capabilities, and head/neck flexibility. They then underwent on-road testing administered by DMV examiners. Subjects were administered the test on both an unfamiliar and familiar route, unless the testing was prematurely terminated for safety reasons. During the on-road tests, a miniature, multiple-camera apparatus in the driver's own vehicle recorded visual search behaviours, brake and accelerator use, and traffic events in the forward scene. Analysis of the videotaped data revealed a high incidence of visual search errors. Many common behaviours were included (such as failure to look to the sides when travelling through an intersection on a green light) that were technical errors, but which rarely require an emergency response. The highest error rate for an actual manoeuvre, as captured by the cameras, was making a lane change with an unsafe gap. This problem was exaggerated on the low familiarity test route, where drivers had no expectation of where the next turn would occur. Analysis of errors recorded by the DMV examiners followed the same general pattern as the video-based error classification, where scanning errors predominated across both familiar and unfamiliar test routes, and manoeuvre errors occurred less frequently. Those driving errors observed most often by the examiners included failure to stop completely at a stop sign, stopping over a stop bar, improper turning path, and stopping for no reason. Regression analyses examined the relationships between functional test results and weighted examiners' error scores. Speed of response on visual discrimination tasks was the best predictor, but no single measure accounted for more than 18% of variance on the criterion. (A)
Abstract