A Naturalistic Driving Study Across the Lifespan

In high-risk populations, the ability to drive safely requires striking a balance between maintaining independent mobility and the avoidance of unsafe driving. In the United States, where alternative transportation is often limited, motor vehicle crashes (MVCs) are one of the leading causes of death for individuals across the lifespan. The Senior and Adolescent Naturalistic Driving Study (SANDS) investigated the complexities surrounding driving in two high risk age groups (teens and older adults). We employed naturalistic driving technologies to measure unbiased real-world driving mobility (amount traveled throughout environment), driving safety (crashes/risky driving behavior), and driving behavior (how/when travel occurred). SANDS participants were recruited to participate in the following stages: (1) a comprehensive baseline assessment of demographics, cognitive, sensory, and physical/health functioning; (2) installation of a Naturalistic Data Acquisition Device (N-DAD) into participants’ vehicles for two weeks, providing objective driving data via photographs, high g-force events (i.e., MVCs, near MVCs, and critical incidents), and Geographical Positioning System (GPS); and (3) a post-test assessment during which participants provided self-reported information about driving safety, mobility and behavior and removal of devices. Overall, due to limitations with the N-DAD, the results should be interpreted with great caution. Results indicated a general lack of association between self-reported and objective data, with one exception: there was a significant relationship between self-reported and objectively measured interaction with cell phones. Several possible demographic, cognitive, sensory, and physical predictors of driving emerged for the sample, with different predictors for older adults and teens. Qualitative data suggested that participants were satisfied with the N-DAD overall and see its utility for other’s to review their driving behavior, particularly if incentives were involved (e.g., discount on car insurance). Implications of findings and possible future directions are discussed.