Daniel Hawes

Daniel R. Hawes Ph.D.

Quilted Science

Will You Read This Post?

How Asking Questions Changes Behavior

Posted Oct 28, 2009

Asking questions changes behavior. In the psychology literature, this phenomenon is often referred to as "mere measurement effect". Vicki Morwitz, a professor of marketing at NYU Stern, demonstrated the effect in 1993, when she showed that simply asking questions about people's intentions to buy a car, or a personal computer led to increased purchasing rates for these items. Even earlier than Morwitz study, University of Washington's Anthony Greenwald - also a professor of marketing - demonstrated that students who were asked about their intentions to vote in upcoming elections became more likely to vote than students who were not asked. The theoretical challenge that mere measurement effects could have on a science that relies heavily on questionnaires for its data is an interesting topic of course, but from a more applied perspective the effect might also carry opportunities:

If querying people about intended behavior manipulates them to perform the behavior more frequently, this might be a very convenient way to get them to do what you would like them to do. After all, I did try this manipulative strategy with the title for this post...and you are reading this, right?

The fact that, both Morwitz and Greenwald are marketing experts might be an indication that the private sector has been privy of this opportunity for a while now, but gladly there are also socially conscientious applications of mere measurement effects that can be and are being investigated. For example, in 2008 a study conducted among Canadian blood donors, investigated whether the mere measurement effect could be exploited as a cheap and efficient intervention to increase blood donations. The study recruited a sample of 4672 registered blood donors, aged 18 to 70, and asked 2900 of these donors to anonymously respond to a questionnaire regarding (amongst a number of additional control measures) their intentions of giving blood within the next 6 months. Six months after the survey assignment, and another six months after that, the researchers assessed whether the survey had an effect on the groups' blood donation behavior. It did.

Comparing both groups, the cohort that had answered the questionnaire showed a higher registration rate as well as an increased share of successful donations. After 6 months the surveyed group showed 8.6% higher registration, and after 12 months registration was still 6.4% increased. Given that yearly blood donations in the USA figure around 15 million units of whole blood and red cells, and saves around 4.5 million lives each year, even the 6% increase would be rather impactful.

So does this mean that simply sending out questionnaires will become a part of social policy packages in the future? A little bit of skepticism naturally remains: Most importantly, as the blood study shows, receiving a questionnaire is not enough to induce the mere measurement effect. Rather, people need to actually answer the questions for the effect to occur. In this sense, it is not really asking questions that changes behavior, but answering questions that is behaviorally influential. Additionally, the study contained only registered blood donors, whose attitude towards blood donations is presumably positive. It is not clear whether the effect would carry over to people with different initial attitudes. This criticism becomes especially important when we consider that attitude activation, or heightened attitude accessibility, are the prominant explanations for what causes the mere measurement effect. Also - since we are dealing with measurement effects here - who is to say, really, that the increased donation frequency was not caused by any of the other control items on the questionnaire (e.g. questions about satisfaction with previous donation experiences, or questions associated with "regret"). In listing these types of caveats, the study's authors readily point out the need for further

"research to delineate what factors determinate the magnitude of the impact of questionnaire completion on subsequent behavior".

But for now, the idea of employing mere measurement effects for social change certainly remains intriguing, and of course this also raises a number of questions regarding possible applications (or mis-use?) of the effect in politics and advertisement. At this point, however, I plan to be mindful of the questions I answer...

Main Reference:

Godin, G., Sheeran, P., Conner, M., & Germain, M. (2008). Asking questions changes behavior: Mere measurement effects on frequency of blood donation. Health Psychology, 27 (2), 179-184 DOI: 10.1037/0278-6133.27.2.179