Teen Drug Use Survey Says…

How worthwhile is the "data" when it comes to kids and drug use? See today's press release from the Marijuana Policy Project, headlined "Teen Drug Surveys Contradict Each Other." An excerpt:

Results from the 2003 Monitoring the Future survey of teen drug use, released today, show a trend directly opposite to that seen in the National PRIDE Survey, released in late August. The two annual youth surveys have been designated by Congress as measures of the success of the White House Office of National Drug Control Policy's (ONDCP's) efforts to reduce teen drug use.

While Monitoring the Future, conducted by University of Michigan researchers and funded by the U.S. Department of Health and Human Services, indicates a decrease in teen use of marijuana and other drugs, the privately-funded PRIDE Survey showed a sharp rise in drug use: Monthly use of marijuana by junior-high students rose 51 percent from 2002 to 2003, and monthly use of heroin rose 60 percent. Despite the differences, both surveys confirm that ONDCP has failed by a large margin to meet goals set for it by Congress. Full PRIDE Survey results are available at http://www.pridesurveys.com.

"Apparently, Drug Czar John Walters only likes to publicize surveys that make federal policies look good, while ignoring everything else," said Rob Kampia, executive director of the Marijuana Policy Project in Washington, D.C. "Walters was strangely silent about this year's PRIDE Survey, though he touted it last year. The taxpayers deserve an honest discussion of the differences between these two surveys, because the bulk of the data suggest our current policies are failing. The only thing these results tell us for sure is that ONDCP has spectacularly failed to meet Congressional goals."

It has never struck me as likely that surveying teenagers about drug use is apt to result in accurate answers, with motivations for lying in every direction, from fear of punishment to desire to mess with people's heads to general disdain for this sort of time-wasting nonsense. So it doesn't surprise me a bit to find two such surveys coming up with such wildly varying results. Such surveying is more a consumption expense for bureaucrats than an investment in human knowledge.

Comments

[amr]

I lied so much on these at my high school. I sniffed glue and took heroin in the past month but had never touched alcohol. Both lies. But I told the truth about tobacco and some other drugs. I can't remember what I answered or what the truth was regarding marijuana. Then, after the survey, I compared crazy combinations with friends. It was a surprise who was wholly truthful and who played around with it.

I was just trying to get my small, rural Catholic high school convinced that it had a heroin problem.

Our survey also had other stuff like sex, self-esteem, self-mutilation, physical abuse, etc.

I was probably just flagged as an outlier and discarded.

[PapayaSF]

I made this comment on Jacob Sullum's thread above, but I see it fits better here:

I'm always suspicious of drug use statistics based on surveys of schoolchildren. Back in the early '60s when I was in elementary school, all the students got a drug survey to take. We were assured it was anonymous and private, but to a wannabe class cutup, it was quite a laugh. There were drugs listed I'd never heard of, and I thought it would be funny to say "Yes" to some of them. No doubt some other pranksters did, too.

To this day I wonder if anyone really believed there was a gang of opium smokers at Green Acres Elementary in Warren, Michigan....

[splash]

You can design surveys to get at information people might not want to disclose. One method is to direct the person being surveyed to flip a coin and answer Yes on heads and answer truthfully on tails. That way, the person being surveyed can give embarassing info without fear.

However, this requires a more expensive survey technique than simply handing out fliers, which is typically done.

The biggest problem with surveys lies not in the accuracy of the replies but in the randomness of the sample. To pick at surveys of teen drug use I would go there, as it seems unlikely that such a survey would correctly sample the desired population.

[J]

I wrote my Master's Thesis on social desirability in Internet survey behavior... this is not news. The adults that took my survey -- I didn't get any teenagers -- lied blithely about privacy policies, Internet cookies, reading instructions, and so on. It only follows that this is even more widespread in teenagers, for the reasons the author listed, and especially when talking about something "illegal" like drug use, and extra-especially on a face-to-face basis.

After all, if I had all those liars in an online survey, imagine what it's like when it's in real life.

[wilda]

i dont think any one really replies truthfully to those surveys.

[gorkjbj]

jxnvhyekgnvidchvdglmv

[davidkulligan]

Doubtful survey, no one is even not ready to give a damn for drugs.
http://www.surveytool.com/web-survey/