"For some reason, we've all just sort of assumed that we can take students' responses on surveys at face value," said Aaron M. White, assistant research professor in the department of psychiatry at Duke University Medical Center and first author of the study, "that if they say they had three drinks, then they really had three drinks. This study suggests that it's just not that simple. Students tend to have pretty liberal views about what constitutes a single drink. In fact, if a student tells us they had three drinks, there's a good chance it was more like five or six. This is a big difference, particularly if we're trying to figure out how many students qualify as 'binge drinkers' based on their self-reported drinking habits."
White and his colleagues asked 106 undergraduate students (54 males, 52 females) to complete a 12-item survey designed to gather basic information about students' current drinking habits, and three tasks. The tasks involved free-pouring either beer, a shot of hard liquor, or alcohol for a mixed drink into cups of different sizes according to each subject's estimation of a "standard" drink. The student-poured volumes were then compared to volumes of standard drinks used in the Harvard School of Public Health College Alcohol Study survey(s).
For every cup size in each of the three tasks, students overestimated how much fluid they would need for a "standard" drink size. "Regardless of which type of drink we asked students to pour," said White, "they almost always poured too much. When asked to pour a standard size beer into a 32-ounce cup, some students filled the cup to the top! For these students, each of their drinks actually equaled 2.5 standard drinks."
Furthermore, in all three pouring tasks, the magnitude of the discrepancy increased with cup size.
"These findings suggest that students drink more than they think," said White, "which means that survey data probably underestimate actual drinking levels on college campuses. This is obviously not good news, neither for those of us who use surveys in our research, nor for those of us trying to deal with alcohol misuse on college campuses. The scale of the problem could be bigger than we thought."
"The fact that many students probably consume more alcohol than their survey responses suggest could help explain some of our previous findings about the consequences of drinking," added Courtney Kraus, second author for the study. "We've observed that a surprisingly high percentage of college students experience alcohol-induced memory blackouts, more than might be expected based on their self-reported consumption. The high incidence of blackouts makes more sense if students are actually drinking more than they think."
Hingson said the study's results raise important issues that need to be clarified. "We need to repeat this study with a larger random sample of students," he said. "The national surveys ought to ask whether alcohol is self served or in bottled containers. When they're collecting information about drinks in these surveys, they ought to provide more information about what a 'standard' drink really is. In addition, we ought to study if alcohol-related problems are associated with miscalculation of the amount of alcohol that it takes to make a standard drink."
"We somehow need to teach students, health educators, administrators, and anyone else involved in dealing with college-drinking issues how to accurately define a drink," added White. "Until then, we have to be cautious about the conclusions that we draw from survey data, and about the levels of consumption that we promote to college students as 'safe' or 'normal.' Telling a student that his or her peers typically drink three or four drinks when they go out could do some damage if that kid defines a drink as a 10-ounce cup of booze with a splash of Coke." White also suggested a new kind of beverage labeling.
"When someone picks up a box of cookies or a bag of potato chips," he said, "one of the first things they often do is look for information about serving sizes, calories, etc. Doesn't it make sense that these labels, or at least a rudimentary form of them, should be placed on drinks that contain alcohol?"
