A recent analysis spanning 40 years of surveys including more than 100 papers on the 'question-behavior effect,' a phenomenon in which asking people about performing a certain behavior influences whether they do it in the future, offers insight to marketers, policy makers and others seeking to impact human behavior.

The authors conclude that asking about performing a future behavior changes the likelihood of that behavior happening. That means parents asking their children, 'Will you drink and drive?' should be more effective than saying, 'Don't drink and drive.' For people making New Year's resolutions, a question like, 'Will I exercise -- yes or no?' may be more effective than declaring, 'I will exercise.'

The belief by the marketing scholars is that when people are asked 'Will you recycle?' it causes a psychological response that can influence their behavior when they get a chance to recycle. The question reminds them that recycling is good for the environment but may also make them feel uncomfortable if they are not recycling. Thus, they become motivated to recycle to alleviate their feelings of discomfort. That is a long daisy chain of effects to be drawn from surveys but the authors do it nonetheless and conclude questioning is a simple yet effective technique to produce consistent, significant changes across a wide domain of behaviors. They claim the technique can sway people toward cheating less in college, exercising more, recycling, or reducing gender stereotyping. So if you get asked if you'll stop discriminating against men in psychology programs, you might be less likely to discriminate against men.

Benefits of using the technique

"We found the effect is strongest when questions are used to encourage behavior with personal and socially accepted norms, such as eating healthy foods or volunteering," said Eric R. Spangenberg, first author and dean of the Paul Merage School of Business, University of California, Irvine. "But it can be used effectively to even influence consumer purchases, such as a new computer."

Problems with the results are obvious: the question-behavior effect to be strongest when questions are administered via a computer or paper-and-pencil survey, and when questions are answered with a response of 'yes' or 'no.' They also found that those using the technique are better off not providing a specific time frame for the target behavior.

Those are severe confounding factors.

They suggest that the technique will be less impactful on habits or behaviors that consumers have done a lot. The researchers also advise using caution asking about vices like skipping class or drinking alcohol. In their review, they found a study showing that people asked about vices later did them more than a control group.