Sunday, June 19, 2011

What’s the harm in asking?


As the parent of high school kids, I see the number of surveys that go to kids trying to determine their risky behavior.  They are frequently asked about drug use and sex.  On the one hand, it seems like a good idea to have a sense of what high school students are doing as a group.  On the other hand, there can be a real downside to asking questions.  For example, a 2008 issue of the Journal of Consumer Research had a nice research dialogue on the influence of asking questions on a person’s future behavior.

The target article was written by Gavan Fitzsimons and Sarah Moore.  Gavan, along with his colleague Vicky Morwitz have done quite a bit of research over the past few years on the paradoxical effect that asking people questions about future behavior may actually influence the behavior itself.  To take a simple example, there is a classic study by Jim Sherman demonstrating that asking people whether they will volunteer for a good cause leads them to overestimate how likely they will be to volunteer relative to people who are not asked to predict whether they will volunteer.  However, this over-prediction becomes a self-fulfilling prophecy, because this group ends up volunteering more often than a control group that is not asked to predict their future volunteer behavior.

The question of interest in the target article by Fitzsimons and Moore is whether asking teens about risky behaviors like sex and drug use will actually increase the likelihood that these kids will engage in the risky behavior.  There is a growing body of data suggesting that this question-behavior effect does occur.  (There is a commentary on the Fitzsimons and Moore article by Jim Sherman who argues that a lot more research is required to really demonstrate how pervasive this effect is, but there is certainly enough data around to be concerned.)

To be clear, the issue here is that asking kids about whether they plan to use drugs in the near future might make them more likely to use drugs in the near future.  Asking kids whether they plan to have unprotected sex in the near future might make them more likely to have unprotected sex.  Furthermore, there are number of large-scale studies that are being conducted in which these kinds of questions are asked of teens, so this is not an idle concern.

Happily, there are some ways to guard against the question-behavior effect.  Most importantly, there is evidence that if people are told about the question-behavior effect in advance, they don’t seem to be affected by the questions they are asked.  One reason why knowing about the effect may reduce the question-behavior effect is that if you respond to a question about a risky behavior, it will bring to mind both the knowledge that the behavior is risky as well as knowledge about the attractive aspects of the behavior.  The positive feeling about the risky behavior may hang around even after the memory of the survey has faded, leaving you with a positive feeling about a potentially dangerous behavior and no clear source of where that positive feeling came from.  If the opportunity to engage in that risky behavior then arises, this residual positive feeling may lead you to engage in the behavior, because you mistakenly think this positive feeling indicates you want to engage in that behavior.  Knowing about the question-behavior effect in advance gives you an explanation for the positive feelings about the risky behavior, making it less likely that you will believe that these feelings indicate that you want to engage in the behavior.

As a parent, that means that if you find out that your kids are going to participate in a survey or if you find out that they are going to get any kind of sex or drug education in school, you should talk to them beforehand about the fact that being asked a question about a risky behavior can affect future behavior, but primarily when you don’t know that being asked a question can affect that behavior. 

In addition, as a parent, you should talk to your kids about the survey or education program after it is over.  We all hate to talk to our kids about sex and drugs.  It is easier to hope or assume that they are not having sex and taking drugs.  However, just asking about the survey or education program is much easier than having to talk to your kid about why they are taking drugs or having sex (protected or unprotected).  So intervene with your kids before and after a survey to eliminate the impact of questions on future behavior.

And by the way, kids are not the only ones who are susceptible to the question-behavior effect.  In one study, adults who were asked how likely they would be to buy a car in the next six months were significantly more likely to buy a car in that period than a control group that was not asked that question.  So, before you participate in any kind of questionnaire, remind yourself that being asked questions about your future behavior can affect that future behavior.

And finally, even though you now know about the question-behavior effect, if you are given the chance to volunteer your time, do it.