An Exploratory Investigation of Self-efficacy to Drive While Intoxicated




Rossheim, Matthew E.


Journal Title

Journal ISSN

Volume Title



This exploratory study was designed to address significant gaps in the literature on driving self-efficacy among intoxicated individuals, and as a result, enhance understanding of the situational factors that influence decision-making about alcohol-impaired driving at night in the on-premise setting. Research designed to identify impaired driving risk factors in their social context is needed to better guide alcohol-impaired driving prevention efforts. The research questions addressed in this study set the foundation for future studies aimed at identifying individual- and event-level risk factors for alcohol-impaired driving, which has important implications for preventing impaired driving. Several bodies of research literature on alcohol use were reviewed. First, an overview of effective policies to prevent alcohol-related harm is presented. Then, research on college student alcohol use and related risk behaviors is reviewed, as well as information on the college drinking environment. This is followed by a discussion of alcohol field research which examines nighttime, event-level risk behavior in college bar patrons. Given that college students represent a high-risk group for drinking and driving, this review shows that there is a need for research that assesses event-level risk factors for alcohol-impaired driving episodes in this population. Beliefs that facilitate decisions to engage in alcohol-impaired driving should be assessed when exiting on-premise drinking establishments because from a traffic safety perspective, this is a critical point in time for making decisions about whether to drive a vehicle. The research methods used in this secondary analysis of data are described. After reviewing the study’s purpose, detailed information is provided about the sample and setting, procedures used to collect data, instruments, hypotheses, and data analysis methods. Discussion of study design strengths and weaknesses is provided. Results of the present study are presented. First, information on participation rate and sample representativeness is presented. Overall, the participation rate was good compared to studies with similar methods, and the sample of interviewed bar goers appears to be comparable to the systematically selected bar goers with regard to sex, race, and time of night leaving the bar district. Next, descriptive statistics are displayed for each variable analyzed in the study. For measurement scales, inter-item reliability was assessed using Cronbach’s alpha. With regard to the first study hypothesis, the novel self-efficacy to drive while intoxicated (SEDWI) scale had excellent internal consistency. Bivariate associations between important variables were also reviewed. Finally, to test Hypotheses Two and Three, the results of a series of multivariable linear regression models are presented and interpreted. These models suggest that self-efficacy to drive while intoxicated is more strongly associated with situational variables, i.e., perceived drunkenness and self-estimated blood alcohol concentration (BAC) than enduring personality traits of bar goers, i.e., habitual drinking practices, risk proneness, and sex. Data supported two of three tested hypotheses. Findings regarding each hypothesis are discussed in the context of relevant research literature. Strengths and weaknesses of the study are discussed as well. This dissertation closes with discussion of the implications for influencing bar patron decision-making in the on-premise environment and recommendations for future research.