Evaluating Programs for Impaired Driving

In the years spanning 1997 to 2006 (the latter being the most recent year for which data are available), there were more alcohol-related fatal crashes in 2006 (17,749) than in any of the other years.1 Also in 2006, 4,434 pedestrians (14 or older) were killed in drunk driving incidents, and there are many more injuries each year related to drinking and driving. Clearly, drinking and driving remains a devastating problem in the United States. However, with a range of programs used for universal prevention in schools and recidivism prevention (or intervention) through the courts, it is often difficult to choose the best strategy to combat the problem.

Two major types of strategies for decreasing impaired driving are law-based and attitude-based strategies. Law-based strategies typically focus on imposing harsher punishments or increasing the perceived likelihood of getting caught (such as through use of sobriety checkpoints). Attitude-based strategies attempt to change the person's perceived acceptance of impaired driving by emphasizing the potential of harm to self (such as through shock videos) or others (such as through victim impact panels or VIPs).

Ultimately, law- and attitude-based strategies can be used together to decrease overall rates of impaired driving. Of course, the selection of most law-based strategies is done through the legislative process, while attitude-based strategies tend to be chosen by individuals within a system, such as educators, counselors, and interventionists. Choosing an effective attitude-based strategy is particularly difficult because the majority of programs aimed at impaired driving have very little well-controlled research to inform those working in the field. One reason for this lack of research is that it is difficult to assess programs' effectiveness.

Universal prevention in schools

Shock videos are commonly used in driver's education programs. However, few studies have examined their effectiveness. With the wide use of prevention attempts, it would be valuable for educators and counselors to evaluate how successful shock videos are at changing attitudes.

One of the few studies of its kind examined the effectiveness of a shock video that was already being used in a driver's education class in a high school.2 In the study, 106 students were randomly divided into two groups that watched either an impaired driving prevention video (“Sentenced for Life,” produced in 1988) or a control video unrelated to drinking and driving. Only the shock video group had a significant decrease in rationalizations for drinking and driving, lenient attitudes toward drinking and driving, and perceived likelihood of future impaired driving. However, this study looked only at a short-term attitude change (same day) and did not examine actual behavior change.

Clearly, more research needs to be done with shock videos, and this study was fairly easy to do for a few reasons. First, it measured the effectiveness of a program already in place. Second, it used the Behaviors and Attitudes Drinking and Driving Scale (BADDS; SASSI Research Institute)3, which is a brief self-report questionnaire with good reliability and validity.4 Finally, the control group was able to watch the shock video later that same week.

As another example, Fatal Vision goggles alter one's perceptual field to mimic some of the effects of being intoxicated (such as loss of coordination), and they have been used in countless high schools and universities as a tool to prevent drunk driving. Students wear these goggles while completing sobriety tests (including walking on a straight line), and these exercises are often paired with a video involving one or more stories recounting fatal impaired driving accidents. When we scoured the research in 2001, however, we could not find a single study that investigated the devices' effectiveness.

We decided to put them to the test in the first study on Fatal Vision goggles.5 The study results showed that a group of students who wore the goggles during the program had a greater decrease in drinking and driving rationalizations and reported likelihood of future impaired driving than did those in a control group. However, we measured only the attitude change on the same day of the study.

In a second study on Fatal Vision goggles, we found that the attitude change on the BADDS was not sustained one month following the program.6 Additionally, there was not a significant decrease in self-reported impaired driving at one-month follow-up. Thus, while our research to date is somewhat discouraging, more research needs to be conducted with Fatal Vision goggles to determine if there is added value in including them as part of a comprehensive prevention approach.

Court efforts to prevent recidivism

Prevention and intervention targeting attitude change with adults charged with DWI also deserves examination. Victim impact panels have been used frequently, with the panels typically consisting of several people who have experienced devastating effects from impaired driving (such as losing a family member).

In one of the few well-designed studies of VIPs, 422 first-time DWI offenders were tracked for two years following their conviction.7 All participants received a mandated educational “DWI school.” After completing the school, they were randomly divided into two groups, with one group experiencing a VIP. The overall two-year recidivism rate in the study was about 18%. Unfortunately, members of the VIP group were just as likely as those in the control group to reoffend. On the other hand, a few previous studies have found some positive effects from VIPs, suggesting that more research is needed. Professionals currently using VIPs could add to the research literature by focusing on measuring the panels' effectiveness in terms of attitude change as well as decreasing recidivism rates.

Victim-offender mediation (VOM) is similar to VIPs; however, in VOM the offender participates in a more intimate meeting with the actual victim of his/her crime (or a surrogate). The victim and the offender have a conversation mediated by a neutral facilitator, and the offender's family and friends also provide feedback. One study of this process tracked 620 DWI offenders in Australia in a two-year follow-up.8 They were randomly divided into a VOM group and a control group. At follow-up, both groups were equally likely to reoffend; however, the VOM group reported greater efforts to drink and drive less, believed the overall process was more legitimate, and reported knowing that reoffending would cause them more problems. Once again, this type of program has shown mixed results and would benefit from continued assessment of effectiveness.

Measuring programs' effects

The programs discussed in this article offer just a sampling of the range of attitude-based strategies for decreasing impaired driving. While impaired driving prevention and intervention programs are well-intentioned, it is clear that more research needs to be done to determine what combination of strategies is most likely to be effective for what type of person. For example, some strategies might work better for high school students, college students, first-time offenders, or repeat offenders. The effectiveness of a specific program for a particular person is also likely to be influenced by personality variables, the person's “stage of change,” and the person's specific rationalizations for or against drinking and driving.

While the gold standard for measuring program effectiveness is tracking future DWI arrests, this type of assessment is flawed because it requires a “wait and see” strategy and also because many people continue to drink and drive without being caught. While good program evaluation techniques often include some form of pretest-posttest measure of attitude change, our own investigation into the literature shows that many programs rely solely on participants' anecdotes as evidence. Also, often the measurement is based on only a single item or just a few items with weak or even nonexistent psychometric data.

In order to assess changes in impaired driving attitudes and behaviors with reliability and validity, we developed the Behaviors and Attitudes Drinking and Driving Scale (BADDS; https://www.thebadds.com) to be a comprehensive self-report measure with five primary subscales (rationalizations, lenient attitudes, likelihood, drinking and driving behaviors, and riding behaviors).

A Guide to Sentencing DWI Offenders, Second Edition9 emphasizes the importance of the assessment of substance-related issues as part of the sentencing process. The BADDS can be used as part of this process, and it should be combined with other tools, such as measures of substance use disorders and clinical interviews. The BADDS also can be used to measure program effectiveness with a single person or groups of people. Considering the mixed results of existing prevention and intervention programs, it is important for each program to evaluate effectiveness with each program participant.

One startling finding from our research with high school students (most of whom were under 16) was that they had nearly as many rationalizations for drinking and driving as adult DUI offenders did.3 Rationalizations for high school students and DUI offenders are substantially higher than those for both a control group of adults (i.e., those attending a traffic court) and the average college student. Thus, universal prevention programs need to start at the high school level, and they should be monitored for effectiveness over time. Additional preventative attempts can be added for students whose attitudes or behaviors do not change.

High recidivism rates for offenders also point to the need for effective evidence-based intervention strategies utilized through the courts. Measures such as the BADDS could be used to assess future DWI risk at pre-sentencing in drug courts, before and after court-ordered program participation, and throughout an individual's probation. Our research shows a relationship between self-reported attitudes on the BADDS and future self-reported impaired driving behavior.3 Information such as this could prove invaluable in attempts to prevent impaired driving as well as to reduce recidivism.

As we have examined the literature, it is clear that there are a variety of programs without enough of a research basis. Anecdotal evidence is a good start, but it needs to be followed by well-designed outcome studies. Much of the needed evidence can and should be generated by professionals in the field who are already using attitude-based prevention programs. With the development of the BADDS, we have a reliable and valid measure of change that can be used by those working in the addiction field. It is our hope that this measure will allow for more evaluation of what we do as professionals.

References

1. National Highway Traffic Safety Administration Traffic Safety Facts 2006, Early Edition Washington, D.C. National Center for Statistics and Analysis, Research and Development 2007
2. Collins A The impact of shock videos and informational videos on adolescent drunk driving attitudes and behaviors. Unpublished master's thesis, Southern IllinoisUniversity Edwardsville, 2007
3. Jewell JD Hupp SDA Lazowski L et al BADDS: User's Guide and Manual Springville, Ind. SASSI Research Institute 2007
4. Jewell JD Hupp SDA Segrist D. Assessing DUI risk: examination of the Behaviors and Attitudes Drinking and Driving Scale (BADDS). Addictive Behav (in press)
5. Jewell JD Hupp SDA Luttrell G The effectiveness of Fatal Vision Goggles: disentangling experiential versus onlooker effects. J Alcohol Drug Educ 2004 48 63–81
6. Jewell JD Hupp SDA Examining the effects of Fatal Vision Goggles on changing attitudes and behaviors related to drinking and driving J Prim Prev 2005 26 553–65
7. Polacsek M Rogers EM Woodall WG et al MADD Victim Impact Panels and stages-of-change in drunk-driving prevention. J Stud Alcohol 2001 62 344–50
8. Tyler TR Sherman L Strang H et al Reintegrative shaming, procedural justice, and recidivism: the engagement of offenders' psychological mechanisms in the Canberra RISE Drinking-and-Driving Experiment. Law Soc 2007 41 553–85
9. National Highway Traffic Safety Administration A Guide to Sentencing DWI Offenders, Second Edition Washington, D.C. 2005