Attention Factors Compared to Other Predictors of Simulated Driving Performance Across Age Groups

Groups of young, middle-aged, and older adults performed a battery of computer-based attention tasks, the UFOV® and neuropsychological tests, and simulated low-speed driving in a suburban scenario. Results from the attention tasks were submitted to Maximum Likelihood factor analysis and 6 factors were extracted that explained more than 57% of the task variance. The factors were labeled speed, switching, visual search, executive, sustained, and divided attention in descending order of amount of task variance explained. The factor scores were used to predict simulated driving performance. Step-wise regressions were computed with driving performance as the criterion, and age, sex and the factor scores, the UFOV® scores, or the neuropsychological test scores as predictors. Results showed that the perceptual-motor speed and divided attention measures from the UFOV® and attention battery were more likely to explain driving performance variance than the neuropsychological tests.