StATS: Characterizing reasons for refusal (created 2006-01-17).
A participant in the IRB Forum raised a question about a research study where people were asked to participate, and when some of them said "no" they were then asked why they did not want to participate. The reasons were tallied across all of the refusals, and percentages were published as part of the full research study. Since the people had already said "no" to participating in the research study, did that also imply that they did not want to share information about their reasons for not participating?
This is a difficult area, and most of the other comments in the IRB Forum have been very critical. In contrast to those comments, I'd like to defend this practice. The authors are trying to examine the extent to which volunteer bias might influence the results of the research. This is an important, but often neglected, issue in research. Almost every research study requires informed consent. The major exceptions are certain retrospective studies, certain studies in an emergency situation, and studies involving a relatively benign intervention like prayer.
Those people who offer consent to participate almost certainly differ from those who refuse. One example that I recall from my earlier work at the National Institute for Occupational Safety and Health involved workplace studies of reproductive toxicants. When you go into a workplace where there is exposure to a chemical that might make men sterile, you usually don't have trouble recruiting volunteers. Even with those people who are unconcerned about their ability to reproduce, they usually end up volunteering to help out their co-workers.
But the control group is a different story. There is no strong incentive to participate, because the chemical being studied is not a problem at the place where you work. So you have to worry about volunteer bias.
It turns out that one of the benefits for participation is getting a complete reproductive health assessment. Now the sort of person who views this as an incentive might be someone who was wondering "Hmm, my wife and I have been trying to have children for the past six months and it just isn't working." If those are the sorts of people who volunteer in greater numbers in the control group, you will end up with a very bad bias in your research. Not all of the couples who have been trying for six months will have problems, of course, but enough might to seriously skew the control group.
So in these reproductive studies, we would always ask people for the reason that they decided to participate (was it for the money? to help out your co-workers? to get a free exam?) and if possible, we would ask the people who refused as well why they declined to participate. I never did those surveys, but as I recall, if someone refused to participate, we would ask them if they would mind filling out a half page questionnaire about their reasons for not participating. If the participants differ from the refusals on these questions, especially if the pattern of reasons is different between the control group refusals and the exposed group refusals, you have learned something very valuable about volunteer bias and about the limitations on generalizing the results of this research.
It seems pretty obvious to me that some of the people would object to all of the time and trouble required to participate in a full reproductive assessment might not object to answering three questions on a survey.
So if someone refused to participate in a research study, I do not see a problem with asking them "why" as long as the researcher is not rude or pushy. They always have the option of saying "none of your business" but the ones who do agree to offer reasons help us to better assess the scientific validity of the study.
Now, it is unclear if the research situation cited above showed adequate respect for the patients' right to be left alone, but I know that if you are careful, you can query refusals without violating their rights.
This page was written by Steve Simon while working at Children's Mercy Hospital. Although I do not hold the copyright for this material, I am reproducing it here as a service, as it is no longer available on the Children's Mercy Hospital website. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Exclusions in research.