Deceiving subjects — What’s an IRB to do?
Deceiving subjects — What’s an IRB to do?
Develop guidelines for researchers
Deception has long been a tool of investigators conducting social and behavioral research. Subjects agreeing to a study may not know exactly what it is about, in order to elicit unguarded responses. Key information about their tasks in the study may be withheld from them. In some cases, investigators may flatly lie to a subject to test a hypothesis.
Although IRBs may approve studies that include deceptive elements, such research raises serious concerns regarding informed consent and possible risk to subjects.
As they review proposals, IRBs need to ensure that subjects are protected by carefully thinking through the necessity of the deception, and providing for a "debriefing," or full disclosure to subjects at the end of the experiment, along with a chance for them to withdraw from the study if they choose.
Fred Rhodewalt, PhD, associate dean and professor of psychology at the University of Utah in Salt Lake City, recently worked with the university’s IRB to develop a set of guidelines for researchers who use deception in their work.
He says deceptive research can be carried out with sensitivity and, in fact, often is viewed with satisfaction by participants once they know the truth. He cites surveys of students, who make up the bulk of the subject pool for campus social-behavioral research.
"On the question of educational value and enjoyment, students think they learn a lot more and enjoy much more the deceptive experiments than the plain dry nondeceptive ones," Rhodewalt says.
That’s even true for the most controversial deceptive research — for example, experiments in social conformity conducted at Yale University by Stanley Milgram in the early 1960s, says Bryan Benham, PhD, an assistant professor of philosophy at the University of Utah.
In Milgram’s experiments, subjects were ordered to administer what they believed were severe electric shocks to a victim, who was in fact unharmed and only acting as if he were injured. After the experiment, subjects were told of the deception.
When researchers went back later to interview the subjects, "a large number said that even though it was disturbing to find out they could do things like harm people, they found it to be an extremely valuable and life-changing experience," Benham says.
Benham, who has studied the use of deception in research, says the notoriety of those experiments led to a reevaluation of deceptive research. He says many of the experiments conducted during the 1950s and 1960s would never be approved by an IRB today.
However, lesser deceptions are fairly common, particularly in social psychology. Benham says that based on various studies of published research, he would estimate that 40% to 50% of articles in social psychology’s leading journals use some form of deception.
The principal defense offered by researchers for deceptive methods is that they are the only way to get truly honest responses from subjects, Benham says.
Rhodewalt says some of his work involves testing subjects and then giving them different types of feedback about how they did. They then are given an opportunity to retake the test and choose the conditions under which they take it. That second offer is deceptive, Rhodewalt says. There never is any intention to re-administer the test; he simply wants to know what conditions the subjects would choose.
But he says many studies employ passive deception, in which the subject is told everything about his participation in a study except the purpose of the study.
"In terms of the procedures, there’s nothing deceptive," he says. "You’re doing exactly the things you would have done otherwise. It’s just the reasons you’re being asked to do it — you’re not told that."
But Benham contends that any degree of deception raises issues that IRBs should be concerned about. How often is it truly necessary? How should informed consent be handled? And what happens when subjects are finally told the truth about their involvement in the study?
The impact on informal consent
Those issues came to the forefront when the University of Utah recently reviewed its practices regarding deception in social-behavioral research, Rhodewalt says. He says some board members suggested that the informed consent documents in those experiments should include boilerplate language warning subjects that they weren’t being told everything about the study.
But he and other researchers who use deception in their work argued that that requirement could bias subjects.
"We had to tell them we were deceiving them and then tell them that only if they were the kind of person who didn’t mind being deceived, should they go ahead," Rhodewalt says.
As a liaison to the IRB, he instead proposed a protocol based on professional guidelines such as those put forward by the American Psychological Association.
The guidelines allow deception, but restrict it in a number of ways:
- The researcher must explain to the IRB exactly what the deception entails.
- Deception may not be used in studies that are expected to cause physical pain or severe emotional distress.
- The researcher must justify the use of the deception, explaining to the IRB why the research is important enough to warrant it, and why using nondeceptive methods isn’t possible.
- The researcher must provide a script for the debriefing session, explaining in detail how subjects will be told about the deception. The researcher also has the obligation of determining whether the subject really understands the deception, regardless of whether he says that he does.
- The subjects all must be given the option of withdrawing from the study and having their data removed.
Rhodewalt says the guidelines comply with federal regulations, which allow informed consent to be waived or altered if the research involved "could not practicably be carried out without the waiver or alteration" (45 CFR 46.116.d), and lays out requirements for protecting subjects.
He adds that research he has seen generally supports the idea that what protects subjects best is a good debriefing phase. Rhodewalt says there’s a fairly low incidence of subjects withdrawing from studies during debriefing, and that’s usually when the information elicited was particularly sensitive to the subject.
"If somebody found they were likely to act aggressively toward somebody else in an experiment, for example, that would be an uncomfortable bit of information," he explains, "but the vast majority do not withdraw."
While the board ultimately decided not to require disclaimers in the informed consent documents, one member, Leslie Francis, PhD, chair of the philosophy department, says she still believes they could be used without prejudicing subjects.
She says researchers could inform students who sign up to be in the pool of research subjects at a university that some of the studies for which they might be recruited may have deceptive elements. That, she says, wouldn’t bias any one particular study.
And because many of the subjects are themselves psychology students, Francis says they need to understand the ethical issues surrounding research and protection of human subjects.
"They’d better well know before they join the psych subject pool that some psych studies sometimes involve not giving full disclosure at the beginning," she says. "That ought to be part of the discussion of ethics in research and psychology, anyway."
Benham and Rhodewalt also suggest other possible ways that IRBs and researchers can take extra steps to protect subjects:
• Recruitment. Benham notes that one of the drawbacks to deceptive studies is that subjects may be recruited who never would have agreed to participate had they known the true purpose of the study.
He says recruitment materials should be designed in such a way that they would not attract a significant number of those subjects. There is a way to test for that possibility: Once subjects are chosen for the study, a group of them are picked at random, told the true details of the study and asked if they still would agree to it.
"If you get a substantial number of people saying yes, then you have good grounds for thinking this is not the type of deception that would change people’s minds," Benham says.
But he says that because those subjects would no longer be eligible for the study, many researchers consider it to be a waste of the effort required to recruit them.
• Vulnerable subjects. Benham says using deceptive studies in children, mentally ill patients, or other vulnerable groups can be especially troubling.
"If we were to put children in a circumstance that was wholly constructed — false —they may not be able to distinguish that from reality, even after a debriefing period," he says. "I think that for people such as children or people who suffer from mental illness, or dementia, deception ought not be used, generally speaking, because we can’t adequately protect them through a debriefing process."
• Using debriefing to re-establish self-esteem. Rhodewalt points out that one of the damaging aspects of a deceptive study is that the subjects may end up feeling duped, or gullible. He says it’s possible to structure the debriefing so that it bolsters the subjects’ self-image at least back to the point where it was when the study began.
At the end of his experiments, Rhodewalt asks subjects to tell him whether they suspected that they were being deceived.
"If they say no, we tell them that most people weren’t, that we’ve designed this so that the majority of people think what we say is happening," Rhodewalt says. "If they say, I knew what you were doing all along,’ we say, Well, you’re pretty perceptive.’
"We added up in my years at Utah; I’ve had between 2,500 and 3,000 research participants, and nobody’s ever complained," he continues. "So you can do this in a sensitive way, and in a way that people feel like they learned something and that it was worthwhile."
Deception has long been a tool of investigators conducting social and behavioral research. Subjects agreeing to a study may not know exactly what it is about, in order to elicit unguarded responses.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.