Norms, Defensiveness, and College Heavy Drinking
Principal Investigator: Gina Agostinelli, Ph.D.

Return to Video Presentation

I study normative interventions that are employed to control heavy drinking among college students. These interventions provide students with normative statistics on the prevalence of heavy drinking. Importantly, research indicates that these interventions will only work if they avoid triggering defensive reactions. Yet, to date, controlled research is rare that examines the effects of such interventions on defensive reactions.

Accordingly, I set out to experimentally study how college students at all levels of drinking would respond to an intervention employing normative information on drinking rates, and just how deeply and openly they would process that information, in terms of accepting its various implications. What kind of judgments would it affect?

Here's what I did.
205 undergraduates participated in the "College Life Study" which was administered on a computer. They were randomly assigned to receive or not receive experimental feedback that presented normative information on the low rate of college heavy drinking.

The participants were recruited across all drinking levels. This pie chart shows you the normative feedback that women received about heavy drinking rates on their campus, with heavy drinking defined as having four or more drinks in a row at least once in the last two weeks.

Men received a similar graphic, but heavy drinking was defined as having five or more drinks. As you can see, the provided norms indicate to heavy drinkers that their heavy drinking rate is relatively uncommon.

What did I find?
First, and most importantly, the manipulation worked, eliminating the usual overestimation bias. Those who saw the pie chart norms in the experimental condition were significantly more accurate in reporting the percent of heavy drinkers on their campus, erasing the overestimation bias as can be seen in the control condition. National estimates of college heavy drinking (for which students were never provided specific statistics) also were significantly reduced in the experimental condition, relative to the control.

Second, I wondered whether students who received the norms that indicated that heavy drinking was less common than they initially realized, would then go a step further and modify their standards for how many days they thought most students would think it was okay for a student to engage in heavy drinking. Indeed, students who were provided with the norms indicated that most students would judge significantly less days per year as okay for heavy drinking than those who did not receive the norms.

Third, students rated how socially acceptable it was for a college student to be a heavy drinker along different dimensions such as smart/foolish, good/bad. Interestingly, only abstainers in the experimental condition who just found out that heavy drinking was uncommon, took that information and inferred from it that heavy drinking was also less socially acceptable. Heavy and regular drinkers did not.

Now, let's look at how students judged their own current drinking.

...

Unexpectedly, experimental participants across all drinking levels had a significantly more positive attitude of their drinking habits than did those controls who didn't receive the norms. Other measures also indicated that their attitudes were less ambivalent and more polarized than were the controls'.

....

Finally, for just the drinkers, we measured the extent to which they thought their current drinking habits were problematic.

....

Here, learning the normative information had no effect, not even on the heavy drinkers who learned that their heavy drinking rate was atypical.

So, what does all of this mean?

A computer program that presents normative information to students can be a relatively inexpensive and easy-to-administer brief intervention for correcting overestimation biases that occur in judging the commonness of college heavy drinking.

Yet, heavy drinkers will only modify their perceptions of other students' standards for the number of acceptable heavy drinking days in response to the normative feedback, and not their beliefs concerning the social acceptability of heavy drinking. In addition, heavy drinkers will not infer from learning the norms that their drinking rate is problematic. Instead, their attitudes about their drinking defensively became more positive and less ambivalent.

In contrast, abstainers accept all of the various inferences from learning the norms in judging both themselves and others. Their attitudes concerning their own non-drinking also become more positive and less ambivalent.

A follow-up study is examining the longevity of the obtained effects, and whether the initially observed defensive reactions in heavy drinkers will go on to affect their drinking rates. Fortunately, preliminary results indicate that, five weeks later, the normative information can curtail heavy drinking, although this finding is qualified by other factors. This latter finding suggests that defensive reactions induced by receiving self-implicating normative information may be only short-lived


[Return to Top]