Supporting teen leaders: Validation of the I Drive Smart Survey.

摘要

Recent policy change allows states to spend federal dollars directly on teen-led driver safety efforts and requires regular evaluations of effectiveness. There are currently no standardized instruments to measure change in teen driving behavior relevant to teen leaders. This study serves the Tennessee Department of Safety and Homeland Security, Tennessee Highway Safety Office and their network of teen leaders to empirically test and refine the I Drive Smart survey developed by partners and grounded in the Theory of Planned Behavior (TPB). The survey is designed to be administered by teen leaders to their peers and produce data relevant for use in improving planning as well as tracking changes occurring from their work. The survey measures attitudes, perceptions of social norms (peer, family, and law enforcement), perceptions of behavior control, and both driving and passenger behavior intentions. The I Drive Smart web survey was administered by a group of teen leaders to 175 of their peers. Findings were used to inform local planning and in this quasi-confirmatory study aimed at optimizing the survey. An exploratory factor analysis revealed a four factor model aligned with TPB that explained 61.618% of variation. Item reliability analysis demonstrated high internal consistency for the behavior intention scale with a Cronbach's alpha of .884. An ordinary least squares regression test found the predictive validity of the identified components to be strong, explaining 64.5% of variation in the model and identifying perceptions of behavior control as the best predictor of behavior intentions, followed by family and peer norms. The behavior control component retained so much variation that the optimized survey assesses both volitional and non-volitional control concepts. Further, teen leaders were able to successfully administer the survey and found data helpful in supporting their planning. This study demonstrates that teen leaders are capable of directing evaluation activities and that the refined version of the I Drive Smart survey has appropriate psychometric properties for teen leaders in highway safety to use. Standard procedures for using the survey are discussed along with recommendations for analysis that includes triangulation with other local data points.