The demands of actual automobile driving and concurrent noise stress on human information-processing capacity for eight licensed, college-aged drivers were estimated from the decrement in performance on the delayed digit recall subsidiary task, using multivariate techniques and a counterbalanced design. Under high load, drivers were much more likely to reduce accuracy than sacrifice speed. However, noise did not result in driving error when presented in the absence ofadditional load. This conclusion parallels the 1973 findings of moscowitz, who investigated the effect of alcohol on driving performance. As expected, the subsidiary task measure was sensitive to the additional information-processing demands imposed by the unpredictable noise stimulus, but contrary to expectation inclusion of the subsidiary task tended to interact slightly with noise in impairing drivingperformance. Perhaps in the low-risk driving environment, maintenance of performance on the subsidiary task may have had sufficiently high subjective utility to demand a disproportionately large share ofinformation-processing capacity.(A)
Abstract