TO WHAT extent are young people experimenting with sex and drugs? For kids in the US, we’ll know this week. Every two years, some 14,000 teens fill in a questionnaire on various risky behaviours, and the results from 2007 are now ready for public scrutiny.
This snapshot of the vices of America’s teens comes from the . Recent trends are mostly encouraging, with kids reporting falling use of alcohol, tobacco and illegal drugs, and that they are becoming less likely to have sex. Yet according to Gavan Fitzsimons, a psychologist at Duke University in Durham, North Carolina, those who take part in such surveys may later pay a price for their participation. Merely responding to questions about risky behaviours can increase the likelihood that they will indulge in future, he claims in the Journal of Consumer Psychology (vol 18, p 82).
The YRBSS is one of several similar surveys worldwide. , for instance, canvasses some 100,000 16-year-olds every four years. If they are affected as Fitzsimons claims, it would be a serious concern.
Advertisement
Asking people questions about their intentions is known to make them more likely to make purchases or indulge in desirable behaviours such as voting. Fitzsimons bases his controversial warning about similar effects with risky behaviours on studies from his own lab. In 2006, his team claimed that asking about future drug use increases subsequent drug-taking. Students asked how many times they would use illegal drugs in the coming two months subsequently said they did so an average of 2.8 times, compared with an average of once among a control group (Social Influence, vol 1, p 117).
Fitzsimons has obtained similar results after asking questions about drinking, and watching TV rather than studying. These studies all relied on students’ own reports of what they had done – which might not be accurate. So Fitzsimons also asked students about missing classes, and then monitored their actual attendance; this was also less than in a control group ().
Given these results, Fitzsimons suggests that those running surveys should take steps to limit the dangers. As a minimum, he argues, they should warn participants that survey questions can influence later behaviour – which he has shown may reduce the effect.
But does the evidence justify altering such large and important surveys? No, says Laura Kann of the US Centers for Disease Control and Prevention, who directs the YRBSS. “We don’t find anything that we immediately need to act upon,” she says.
Jon Krosnick, an expert in survey methods at Stanford University in California, argues that Fitzsimons’s drug-use study was flawed. When Krosnick reanalysed the data using different tests, and excluded two “outliers” who reported very frequent drug use, the results were not statistically significant (Social Influence, vol 2, p 178).
It is certainly hard to understand why a simple question would exert such a powerful influence, given the myriad mentions of sex and drugs elsewhere. More fundamentally, Fitzsimons’s studies all examine the effect of asking about what people will do in the future, whereas surveys like the YRBSS ask about past behaviours. These are two very different things, argues Kann.
Maybe so, but there is little evidence on the effects of asking about past risky behaviour. The most pertinent study dates from 1994, when researchers at the University of North Carolina, Chapel Hill, compared four groups of boys who had been asked about their sexual behaviour different numbers of times. They found a small effect from completing one questionnaire two years before, with the boys reporting more masturbation, wet dreams and contact with a partner that stopped short of intercourse. But there was little evidence that multiple surveys increased the effect ().
This study used data collected for other projects, so there may have been differences between the groups of boys involved. What is needed to properly test the effects of asking about past sexual behaviour and drug use is a study in which volunteers are randomly assigned to experimental and control groups.
Scientists cannot react to every scare story by immediately launching a study, but in this case, inaction could be dangerous. Survey researchers have a duty of care to those who volunteer to take part. Just imagine the anger from parents – not to mention lawsuits – if it turns out that Fitzsimons is right.
What’s more, research into drug use and sexual behaviour is notoriously susceptible to political interference. Most recently, of distorting and suppressing science to support its abstinence-centred approach to reproductive health. To leave untested the suggestion that surveys about sex and drugs can be harmful could give ammunition to those who want to base policies on moral agendas, rather than scientific evidence, and ultimately undermine public health.
“Research into drug use and sexual behaviour is notoriously susceptible to interference”
Teenagers – Learn more about a uniquely human phenomenon in our comprehensive special report.
Focus on America – Delve into the science and technology questions facing the USA in our special report.
