|Title||Headway Time Errors, Safe Driving Skill, and Experience: An Initial Validation of the Simulated Driving Assessment|
|Publication Type||Conference Paper|
|Year of Publication||2014|
|Authors||Winston FK, McDonald CC, Kandadai V, Loeb H, Tanenbaum JB, Seacrist T, Scarfone SR, Winston Z|
|Conference Name||Transportation Research Board 93rd Annual Meeting|
|Keywords||driver performance, driving experience, driving simulators, evaluation and assessment, headways, high risk drivers, linear regression analysis, recently qualified drivers|
High novice teen crash rates highlight the need for a validated, reproducible method to assess safe driving skill before independent licensure, the aim of the new Simulated Driving Assessment (SDA). In order to validate the SDA in its ability to differentiate drivers according to experience and driving skill, the authors examined headway time errors as one key safety metric. This two group comparison study included an inexperienced group (n=21): 16-17 year olds with 90 days or fewer of provisional licensure and an experienced group (n=17): 25-50 year olds with at least 5 years of Pennsylvania licensure, >100 miles driven per week and no self-reported crashes. The SDA involved a 35-40 minute drive with scenarios representing common crash configurations and was delivered on a fixed based high fidelity simulator. Simulator-derived data were analyzed from the seven potential rear-end scenarios included in the SDA. A second source of data came from professional driving evaluator ratings across eight domains of driving based on video review. For this paper, a cut-off score for the total domain score was created to categorize participants as skilled/less skilled. Skill categorization was compared with police-reported driving records to obtain an indication of validation. A multivariable linear regression model revealed that, independently, both driving skill and experience significantly predicted the simulator-derived total headway time errors across the seven potential rear-end collision scenarios (F(2,35)=13.1; skilled p<0.01; experienced p<0.01). These results provide early support for the SDA as stimulus which can produce driving performance that differentiates drivers according to experience and skill.