Drunk With Good Intentions
by the Times-Picayune
Nobody thinks it's a good idea for teen-agers to go stumbling around plastered, but overstating the severity of the teen drinking problem will do nothing to encourage young people to make responsible choices about alcohol.
And by proclaiming last week that Americans between 12 and 20 account or 25 percent of the nation's alcohol consumption, the National Center on Adiction and Substance Abuse at Columbia University merely damaged its own reputation.
The notion that teen-agers account for a quarter of the nation's alcohol consumption just doesn't square with reality, and no one at the research center should have believed it. That age group makes up about 13 percent of the U.S. population and has to rely on fake ID cards, indifferent sales clerks or complicit older friends to get alcohol. Can they possibly drink a third as much as the 70-odd percent of Americans who are of legal age? Of course not.
Nevertheless, the Columbia researchers extracted the 25-percent figure from a federal government poll called the Household Survey on Drug Abuse and trumpeted it to the media. And perhaps because of the center's academic affiliation and high-profile leadership -- its president, Joseph Califano, who was the nation's health secretary under President Carter -- a number of news organizations reported on the study.
But it quickly became clear that the center's "research" was fishy. Officials of the group soon admitted that they had not used the government survey properly. Teen-agers were significantly over-represented among survey subjects, but the Columbia researchers did not take that into account.
In fact, the error seems to be intentional. Federal statisticians had estimated that teen-agers are responsible for 11.2 percent of the alcohol consumption in the United States, but the Columbia researchers decided that number wasn't high enough.
"We didn't reweight the data," one research center official told The New York Times. "But we thought the 11.4 percent number is way too low, since there's so much under-reporting."
In other words, the researchers simply assumed that teen-age alcohol abuse is more widespread than the federal government had estimated, and they knowingly skewed data to come up with a figure that fit their preconceptions.
Researcher's are free to dispute the government's data, but the only way to prove that those numbers are wrong is to collect more accurate information. Fabricating new numbers won't do the job.
You can't blame the Columbia researchers for trying to combat teen-age alcohol abuse. But their role should be to provide useful information, not to make up numbers that will scare parents and government officials into action.
Meanwhile, academic administrators at Columbia might want to keep a closer eye on the accuracy of the research being conducted and circulated under that university's good name.
*Reprinted with permission of The Times-Picayune. 2002 The Times -Picayune Publishing Co. All rights reserved.