The trusted source for
healthcare information and
Incentives for surveys — Can they be coercive?
Sociologist says that IRBs should focus on risk
When researchers are attempting to persuade subjects to answer questions about themselves, whether on the phone or by filling out a form, sometimes the altruism of participating in research isn't enough.
So they resort to incentives — cash, or a non-cash gift, as a reward for participation. IRBs can become concerned about these incentives, and whether they manipulate people into participating in surveys they wouldn't have answered otherwise.
But a sociologist at the University of Michigan argues that such IRB concerns can be overblown — that incentives don't generally induce subjects to work against their own interests, particularly when the study involved simply requires that they answer questions.
Even when the questions involved are difficult ones about violence, or high-risk or stigmatizing behaviors, studies show that subjects don't tend to take on higher actual risks in exchange for money, says Eleanor Singer, PhD, adjunct professor of sociology at the Population Studies Center at the University of Michigan, Ann Arbor, MI.
"Honestly I don't think that incentives are usually an ethical issue," Singer says. "I don't know why IRBs need to concern themselves with them as much as they do. It is a benefit that respondents can be given. I don't see anything wrong with giving it to them, frankly."
Singer recently wrote about the issue of incentives for survey participation for the American Journal of Preventative Medicine. Her paper was part of a larger look at survey work, particularly in the area of violence and violence prevention.
Prepaid vs. refusal-conversion
Singer says there isn't a lot of information about how much researchers, in general, are paying to convince respondents to participate in surveys, but there are a few general ways in which cash incentives are offered:
Singer says it's hard to know how often refusal-conversion payments are used, since they're often used simultaneously with other forms of payments. She says some ethicists disagree as to the appropriateness of refusal-conversion payments, which, in effect, reward subjects more for being less cooperative. But they can be helpful in raising survey participation rates.
In her own work, Singer says she deals with that issue by offering a small prepaid sum to everyone contacted for a survey, with a refusal-conversion incentive offered to those who initially decline.
But the larger issue for IRBs tends to be the effect of the incentive on a person's autonomous decision to participate. Does the prospect of payment coerce a subject into responding?
'Influence,' not 'coerce'
Singer argues that it doesn't, using the Belmont Report's definition of "coercion" as an "overt threat of harm. . .intentionally presented by one person to another in order to obtain compliance."
Another issue raised by the Belmont Report is that of "undue influence," or "an offer of excessive, unwarranted, inappropriate or improper reward or other overture in order to obtain compliance."
"You should never think of incentives as coercive but, obviously, you can think of them as influencing people," she says. "The question is, are there situations in which the size of the incentives raises some questions about the amount of influence that may be exerted on the subjects? Is that ever bad?"
Singer says that several studies looking at the cost-benefit decision-making process of respondents show that individuals do not exchange risk for money — they don't accept greater risk as the compensation offered increases.
There are some concerns that people wishing to participate in a survey might give false information to meet eligibility requirements. But she notes that this can be dealt with through rigorous screening of applicants.
Singer says that if IRBs are concerned about incentives and their effects, they should attack the problem on two fronts:
• Ask subjects' opinions about incentives. Singer says IRBs should do research on what size incentives are considered excessive by the people who are offered them.
"The IRB tends to substitute its own very subjective judgment in these matters and every IRB comes up with a different amount," Singer says. "I think that's wrong, because it subjects research being done by different institutions to very different standards."
She says it would be useful to get the public's input into these issues: Given a certain type of situation, what kind of incentive is excessive? What is considered insufficient?
"You can do it locally, you can do it nationally," Singer says. "You can have discussions about what would be the most appropriate kind of sample to get this kind of opinion from. But there should be at least some input from the populations who are actually affected by the incentives."
Some researchers argue that IRBs, or a federal agency, should monitor levels of incentives being used in survey research to get a better idea of what is being offered to whom. This type of empirical information could help better identify real potential ethical problems.
• Focus on risk, not incentives. Singer argues that the IRB should focus on reducing risks, such as the nature of the questions asked or the possibility that personal information might be disclosed, rather than worrying how much people are being paid.
"The issue isn't whether you can offer incentives in exchange for risks, the issue is can you protect people against risks," Singer says. "If you can't, there are some questions as to whether you should be doing the research at all.
"And if you are protecting them against risk in every reasonable way, then the offer of an incentive is not, to me, unethical."
Even in situations that may give some IRBs pause (for example, surveys dealing with vulnerable populations such as children), incentives are not as important as the risks being asked of the respondents, Singer says.
"What are you influencing these kids to do?" she asks. "Is this against their own interests? If it isn't, if there's no risk to them or if they're protected against the risks such as confidentiality issues, then it's really not the IRB's business to decide what incentives may be offered."
Singer E, Bossarte RM. Incentives for Survey Participation: When are They "Coercive?" Am J Prev Med 2006 Nov; 31(5):411-418.