Description
Background: In 2010, 3,092 people were killed in crashes involving a distracted driver and an estimated additional 416,000 were injured in motor vehicle crashes involving a distracted driver. There is currently a 17 item scale that predicts the behavior of distracted driving. The purpose of this study is to model distracted driving using the least number of questions, while maintaining a high predictive value. Methods: A cross-sectional study was performed with an anonymous online survey of college-attending drivers, with a 100% probability sample from students attending 18 different educational institutions across San Diego County, California. Regression analysis identified correlations between predictive variables and total distracted driving scale. Factor analysis assisted in the creation of a scale assessing distracted driving behavior, which served as the key outcome. Results: The four strongest predictors of distracted driving were: self-efficacy in driving while multi-tasking (_ = .37), perception of how safe multi-tasking while driving is (_ = .19), social norms (_ = .29), and driving frequency (_ = .12). The final regression equation (F (17, 4945) = 224.31; p < .0001; R2 = .44) explained 44% of the total variance of an individuals' total distracted driving score. The internal consistency correlation, using Cronbach's standardized alpha coefficient, was 0.88. Principal axis factoring was accomplished using varimax rotation. A five factor scale (eigenvalue 1.01-6.78 with 64% variance explained) was derived. Factor loadings ranged from .43 to .87. Criterion validity of the brief distracted driving scale was examined by comparing with total distracted driving score. Twelve items were deleted based on inter-item (less than .30) and item-total (less than .55) correlations and factor loadings (less than .45). Five factors emerged and represented behaviors relating to the frequency of talking on the phone, texting, using plug-in devices, and using non-plug-in devices while driving. Conclusions: Data from this study suggest that a distracted driving scale with five questions can be used to assess distracted driving behaviors among college students. Public Health Implications: An online distracted driving scale, such as this one, may be used to deliver immediate personalized feedback as part of an individualized prevention program or applied to a larger audience to evaluate the success of public awareness campaigns.
