Determining subgroups of teens for targeted driving injury prevention strategies: a latent class analysis approach.

Objective: To utilize teen traffic safety belief profiles to improve targeting of resources and messages to young novice drivers.

Methods: The National Young Driver Survey is a nationally representative sample 5665 of 9th to 11th graders in the United States. Using latent class analysis, we grouped beliefs about 25 safety-relevant behaviors into a summary set of belief profiles and related these profiles to demographics and driver/passenger experience.

Results: We determined 5 safety belief profiles of teens: "everything" (rated most of the 25 behaviors as important to safety); "drivers/personal responsibility" (rated driver-related behaviors but few others as important); "driver drinking" (rated only driver alcohol use as always important); "distractions/external forces" (rated predominantly passenger rather than driver issues as important); and "nothing" (rated no issues as important). Three key groups emerged who were more likely than their counterparts to belong to the distractions/external forces than the everything class and for whom targeted messaging might be effective: males, non-white adolescents, and teens who had experienced an injury crash as a driver.

Conclusions: The classes appear to organize around the locus of control (LOC) social learning framework, with some teens perceiving crashes resulting primarily from their own behavior and others believing that forces in their environment determine the events that result in a crash. Designing interventions that help young drivers understand their role in crashes may help improve the safety behavior of young drivers. In particular, for those involved in crashes, interventions designed to help them understand that the crash was a result of their actions, rather than a random or externally driven event, may influence them to take control with safety-oriented behaviors.