We know that surveying teenagers about alcohol, drugs and sex makes them more likely to report that they have drunk, taken drugs and had sex in future surveys.

We know that surveying teenagers about alcohol, drugs and sex makes them more likely to report that they have drunk, taken drugs and had sex in future surveys. The question is we're they lying the first time, the 2nd time or did their behaviour actually change.