|Home | About | Journals | Submit | Contact Us | Français|
Brief individually focused alcohol interventions which provide personalized feedback (e.g., the BASICS; Dimeff, Baer, Kivlahan, & Marlatt, 1999) have consistently been found effective in reducing drinking among problem drinking college students (Larimer & Cronce, 2002). These interventions typically include a number of components including; a review of the student’s drinking behavior and experienced consequences, risk factors such as family history, expectations regarding potential alcohol effects (e.g., social lubrication, tension reduction, and enhanced sexual performance), and moderation tips such as alternating alcoholic drinks with non alcoholic drinks, information regarding blood alcohol concentration, and information designed to correct misperceptions regarding typical college student drinking behavior (Walters & Neighbors, in press).
As a social psychologist, my initial reaction to reviewing multi-component alcohol interventions was to wonder whether the social influence component (i.e., normative feedback) was strong enough to reduce drinking all by itself. My colleagues and I recently completed three randomized controlled trials evaluating the efficacy of computer delivered personalized normative feedback (PNF) as a stand alone intervention (Neighbors, Larimer, & Lewis, 2004; Neighbors, Lewis, Bergstrom, & Larimer, under review; Lewis & Neighbors, in preparation). Personalized normative feedback is individually tailored and tells students: “1) this is how much you drink, 2) this is how much you said you think other students drink, and 3) this is how much other students actually drink.”
We have consistently found medium to large effects of normative feedback on normative misperceptions (i.e., overestimation of how much other students drink), and small to medium effects on subsequent drinking behavior. There is evidence to suggest that the amount by which drinking is reduced is a direct function of the extent to which normative misperceptions are corrected. We have found similar effects at 1 month, 2 month, 3 month, and 6 month follow-up assessments. We have also found that normative feedback works better for students who drink largely for social reasons and for students who are more concerned about the expectations of others. Thus, the answer to the initial question regarding whether PNF as a stand-alone intervention is effective in reducing alcohol consumption among heavy drinking students is clearly yes.
We are currently nearing the end of year 1 of a 5 year NIAAA funded project (R01AA014576) designed to evaluate web-based PNF over longer follow-up periods, and to examine moderators of intervention efficacy. Among the specific questions addressed by this investigation are: How long does the effect of normative feedback last? Does repeated exposure to normative feedback reduce drinking significantly more than single exposure? How does framing effect the impact of norms information on abstaining and light drinking students?
In reviewing the work we have done with personalized normative feedback and in comparing the strength and consistency of personalized normative feedback results with results of social norms marketing campaigns, we have come to several conclusions regarding implications for practice, evaluation, and implementation of norms based approaches.
Changes in perceived norms will almost certainly be accompanied by smaller but significant changes in drinking (Borsari & Carey, 2000; Mattern & Neighbors, 2004; Neighbors et al, 2004). If a norms based intervention fails to change perceived norms, it is unlikely to have any impact on drinking. In social norms marketing campaigns, posters, flyers, and newspaper ads may never be seen by many of the students they are intended to target. Evaluating the efficacy of an intervention among students who didn’t actually receive the intervention is tenuous. Alternative methods of delivery (e.g., e-mail, mail, classroom presentations, small group presentations, etc.) can help ensure that students get the information.
Exposure to accurate norms information may not always be sufficient to reduce drinking. An implicit assumption behind norms based approaches is that it is the realization that one is drinking more than other students which causes them to reduce their drinking. Personalized normative feedback facilitates this realization by providing an explicit comparison. With typical social norms marketing ads it is not clear whether students think about how the norms information presented compares with either their preconception of typical drinking or their own drinking behavior. The inclusion of statements or questions that increase the likelihood students will compare the norms with their preconceived estimates and their own behavior (e.g., “Are you surprised?” or “How does your drinking compare?”) may help ensure that students consider how the information relates the them personally.
A substantial body of basic research in social influence supports the conclusion that people are most influenced by others when they identify with those others and when they care about what those others think of them. In general, the more psychologically proximal a referent group is to an individual, the more likely the individual will be influenced by information about the norms of that group. Thus, students are more likely to be influenced by norms of their immediate peers (students on campus, female students, Greek students, and athletes), than they are by norms of the general population. There are two important caveats to the recommendation of incorporating norms from more specific referent groups. First, all other things being equal, lower group norms are preferable to higher group norms. Telling a student that their best friends drink 25 drinks a week is unlikely to have a good outcome versus telling a student that the average student on their campus drinks 4 drinks per week. Second, it is important to keep in mind that the more immediate and specific the referent group is, the less likely there will be a disconnection between how much an individual thinks members of that group drink and how much members of that group actually drink. For example, students have a better idea of how much their best friends drink than how much students on their campus in general drink. Thus, among very specific referent groups there may be no misperception to correct. In sum, we suggest considering more specific referent groups (e.g., females or athletes) when there is reason to expect normative misperceptions for that group and when the norms of that group are relatively low.
In evaluating the impact of social norms interventions, one can not expect a reduction in drinking among non-drinkers. We thank an anonymous reviewer for pointing this out to us in our evaluation of a social norms marketing campaign among residence hall students (Mattern & Neighbors, 2004). It makes sense to evaluate whether perceived norms have been corrected (i.e., reduced) among all students following a social norms intervention. But any evaluation of drinking reductions which includes abstainers is potentially misleading. In the best hypothetical case (all students who are abstainers before the social norms intervention remain abstainers after the social norms intervention) inclusion of abstainers in the evaluation leads to a greater likelihood of concluding no effect of the intervention. In the more likely scenario (some students who are abstainers before the social norms intervention report drinking after the social norms intervention) inclusion of abstainers increases the likelihood of concluding that drinking goes up after a social norms intervention. It is unclear to what extent inclusion of abstainers in evaluations has been responsible for some of the null findings reported for social norms marketing interventions. At minimum evaluators of social norms interventions should consider evaluating drinking reductions with and without including abstainers.
One of the primary strengths of norms based approaches (whether delivered via social norms marketing, web-based personalized feedback, or other formats), is their ability to reach a large number of students at a relatively low cost. When an intervention results in a large proportion of students on a given campus reducing their drinking, even if by only a few drinks per week, the impact can be substantial in terms of reduction in negative consequences (e.g., DUI’s, damaged property, academic problems, and sexual assaults).
The “Social Norms Approach” to reducing drinking (Perkins & Berkowitz, 1986; Berkowitz, 2004) has become a dominant approach in alcohol prevention for good reasons. It is based on a solid and rich theoretical foundation. The assumptions underlying the approach have been repeatedly validated in empirical studies. Moreover, it’s inexpensive, it’s novel and interesting, and it works when well implemented. Several keys to successful implementation of social norms marketing campaigns can be derived by considering features of other empirically supported interventions which share the same basic theoretical foundations, such as computer delivered personalized normative feedback. This article was designed to elaborate upon some of these.
Preparation of this article was supported in part by National Institute on Alcohol Abuse and Alcoholism Grant R01AA014576.