With a probability sample of 3000 suburban drivers, driving infractions in the past year (number of accidents of any severity and number of violations plus warnings) were obtained from both personal interviews and official files and examined in relation to age for men and women aged 16-24 and 35-44. Two non-linear methods were used to control for exposure - separate analyses within mileage brackets and an adjustment of infraction scores by the multiple classification (mca) program to remove effects associated with mileage and driving conditions. Under both procedures, significant age effects were found; young men aged 18 and 19 had more violations than men either younger or older. With road experience controlled by age when driving was learned the same peak at 18 or 19 appeared except that among men who learned at 15, the "recommended" age for driver education, the crash and violation peaks were delayed one or two years. (Author/publisher)
Abstract