Asking someone how likely they are to take illegal drugs in the future can actually increase the likelihood that they will indeed take drugs – a finding with worrying implications for health research.
Patti Williams and colleagues recruited 167 undergrads and asked some of them about their intentions to take drugs, and the others about their intentions to exercise. Two months later, the students were contacted again, and those who had been asked about drugs reported taking drugs an average of 2.8 times in the intervening period, compared with an average of 1.1 times among the students previously asked about exercise.
The effect was even more dramatic when those students who said they hadn’t taken any drugs at all were omitted from the analysis. Among the remaining students, those asked about their drug-taking intentions said they’d used drugs an average of 10.3 times over the past two months, compared with an average of 4 times among the students previously asked about their exercise intentions.
This observation, together with further analysis, suggested it wasn’t that new drug users had been created, but rather that the questioning had led to increased use among current users who presumably had a positive attitude towards drugs in the first place.
“The results of the current study may well be troubling for researchers trying to survey respondents in at-risk populations”, the researchers said. “By virtue of surveying the at-risk population in an attempt to help them, serious harm may actually be done to the sampled group”.
It wasn’t all bad news – those students asked about their intentions to exercise subsequently reported having exercised more than the students who were earlier asked about their drug-taking intentions. But the message remains – asking someone a question about their intentions can alter their future behaviour, sometimes in negative ways.
Williams, P., Block, L.G. & Fitzsimons, G.J. (2006). Simply asking questions about health behaviours increases both healthy and unhealthy behaviours. Social Influence, 1, 117-127.
UPDATE: Click 'comments' below - one of the authors of this research answers a reader's question and kindly provides a link to some forthcoming research that expands on the current findings.