March 10, 2012

Can questions lead to change? An experiment

Can Questions Lead to Change? An Analogue Experiment
Sara Healing, Janet Beavin Bavelas, Department of Psychology, University of Victoria, Canada

It is commonly assumed that questions can affect behavior and that different sorts of questions affect behavior differently. But little research has been done which provides evidence this assumption. Healing and Bavelas did an experiment in which they tested the differential impact of two different types of questions a) their task performance on a difficult task, and b) their own attributions of their task performance (their interpretation of why they performed as they did).

The researchers chose to examine two types of questions borrowed from Alan Jenkins’s “Invitations to Responsibility” (1990):  a) questions that focused on personal agency (for example: "Did you manage your time effectively?", and "What could you have done differently?") and b) questions that focused on external causes (such as: "What effect did time constraints have on you?", and: "What kept you from succeeding?"). These two contrasting types of questions illustrate the general principle that questions  always contain implicit presuppositions which are unstated but logically implied assumptions. The researchers wanted to test the assumption that questions differing in their presuppositions can affect the person answering them.

The experiment: participants were interviewed about a difficult task that they had just done. The interview questions focused either on the difficulties of the task itself (external causes) or on what the interviewee had done or could have done (personal agency). After the interview, interviewees wrote down, in their own words, the factors responsible for the task score. The following week, interviewees did the task again and got a new score.p

Results: As expected, the interviewees’ answers usually went along with the presupposition in the questions. Attributions by interviewees were overwhelmingly congruent with the types of questions asked to them. These differences were still present, to a lesser degree, one week after the interview. The task performance of the participants what had been asked the personal agency questions improved significantly one week later; the performance of participants who had been asked the external causes questions did not.

Reflection:  Questions contain implicit presuppositions which constrain and orient people to a particular aspect of their experience. Interviewee tend to go along and accept these implicit presuppositions and are affected by them in their behavior (/performance) and their own attributions of their performance. What were initially presuppositions generated by interviewers become mutual presuppositions and the behavior and attributions by interviewees which are based on these presuppositions are eventually owned by the interviewees. As Healing and Bavelas conclude: "All questions are “loaded questions”; the practitioner’s choice is how to “load” them with presuppositions that will be useful to the client."


  1. Hi Coert

    Has this study been published yet?

  2. yes, if you click the link you can see where and when

  3. Very interesting experiment. I found the results validated my own experiences. When I ask "personal agency" questions people tend to look at their own contribution to whatever needed to be accomplished.
    By taking ownership they become aware of their part in bringing about change.



Enter your email address:

Delivered by FeedBurner