Watchdogs And Guinea Pigs

Until Timothy Leary came along, psychology researchers at Harvard could do pretty much anything they wanted with their human subjects and still be left in peace. Even with Leary and his colleague Richard Alpert hosting psilocybin parties in private homes, with state and national agencies investigating the legality of the drug tests, and with reports of at least one former subject hospitalized at McLean's with a nervous breakdown-- even then, Harvard didn't do much. The Crimson reported in the spring of 1962 that University officials had known about the research on hallucinogens for at least two years, but had not interfered "because to do so would be an invasion of academic freedom."

Leary did set the University to thinking, for the first time, about overseeing researchers in order to safeguard the rights of subjects. Nothing substantial was done, however, until 1966, when the federal government required the establishment of a standing institutional committee to review all research funded by the Public Health Service. That committee was the ancestor of the Committee on the Use of Human Subjects (CUHS), now charged with the review of experiments performed within the Faculty of Arts and Sciences.

Most of the 14 members of the CUHS are from the Psychology and Social Relations Department, just as most of the projects it reviews are in psychology. The committee is supposed to ensure that no humans are subjected to risk of any kind; or that if an experiment must involve risk, its benefits outweigh its dangers and its subjects fully understand those dangers and freely agree to participate.

The word "risk" often brings to mind medical or biological dangers-- chancy new surgical techniques or the testing of possibly toxic new drugs. But although psychological research sometimes entails physical risk--Leary was not alone in testing drugs, and electric shock, loud sudden noise and alcohol have all been used recently in William James-- more often the dangers are to the mind. Experimental psychology can invade subjects' privacy, stressfully manipulate their minds, or coerce them into unpleasant situations by intentionally deceiving them.

The task of the CUHS is to ensure that research go on without harming its subjects. But Harvard psychologists disagree vehemently on just how much protection the CUHS should provide. Just as the University was reluctant to violate "academic freedom" by restraining Leary, so there remain professors who regard the CUHS as an intrusion and a potential violator of researchers' rights. Others in the department, however, feel that the Committee has not provided enough protection for subjects.

The dispute is clouded further by disagreement among Harvard researchers about what is or is not "ethical." A brief description of three experiments, all recently or currently performed at Harvard, show the complexities involved:

Heart attack. You are walking down a not-too-busy Cambridge street. Suddenly you notice a young man, "a typical college-age student in appearance," collapse to the ground, "clutching his heart" and "in obvious pain." Alarmed, you rush forward to assist the man.

As you do so, a cameraman, hidden, perhaps, behind a nearby tree, records your reactions. Unwittingly you have become part of a psychology study. If you become suspicious when the student rises unhurt, or if you happen to notice the cameraman, you may resent having been tricked. But if you don't catch on, you may never realize the contribution you have made to science.

Moreover, had you-- out of fear, say-- ignored the victim's plight and walked quickly away, your guilt feelings might never have been dispelled. You would never know that the "typical college student" was in fact a "stooge," helping a Harvard graduate student in psychology complete his dissertation.

The scenario and the quotes are from a research proposal approved by the CUHS several years ago. The document, after describing the methods of the study--designed to measure helping behavior-- explains why "debriefing (of subjects) will provide no useful purpose."

"Any subject who helps," the graduate student writes, "will feel good because he has helped and may have negative feelings if he is debriefed and realizes he has been tricked." The proposal does not discuss the feelings of those subjects who decide not to help or of those who discover the trick.

Herbert Kelman, Richard Clarke Cabot Professor of Social Ethics, criticized such studies in a 1972 article in American Psychologist. "Such a procedure," Kelman wrote, "on the basis of a rather massive deception, places the subject in a situation...that may (whether or not he decides to help the victim) be very disturbing to him-- without giving him any choice in the matter at all."

This Course May Be Dangerous to Your Health. One course in the Harvard catalogue requires its student to become research subjects. Psychology and Social Relations 1330, Group Psychology, is taught by Professor Robert Freed Bales. The course, Bales says, is "primarily education" for undergraduates, but he adds that both he and graduate students have used observations of student interaction for their research. To enroll, students must accept the "ethical responsibilities involved in the professional role of scientific investigator." What they investigate is each other.

Students wanting to join PSR 1330 must agree to accept "any possible risks." The course calls for participation in small groups with subsequent self-analysis, and Stephen Williamson, a pre-doctoral fellow associated with the course, says that it can indeed be an "intense experience."

"It raises question," he says, "and it can be painful."

Bales insists, however, that the risks are "very minor" and that students are not psychologically harmed by his course. "I've had no problem of breakage," he says. Williamson agrees with Bales that PSR 1330 is not dangerous. "If it is harmful," he says, "the person was heading for danger anyway."

Williamson himself is doing research based on data collected from observing the course in years past. There is no indication in the agreement students sign that written observational data about them will "routinely" be maintained, as Bales says they are. Bales called the lack of information in the agreement an "omission by error. Now that I note it isn't there," he says, "I'll put it in."

Love and sex and all that. Zick Rubin, associate professor of Psychology, is an authority on "couples research." Edward L. Pattullo, director of the Center for the Behavioral Sciences and chairman of the CUHS, says that such research is one of the most ethically difficult fields of inquiry--last spring the Committee discouraged a graduate student from conducting a study in the area because, according to Committee member David Gartrell, the student lacked "the requisite counseling skills." But the CUHS has not hindered Rubin's research-- although according to Gartrell, Rubin himself found his work was "very intrusive and was creating problems."

Through questionnaires, interviews and laboratory experiments, Rubin studies the relationships between men and women in student dating. "This tends to be a problem, first, in terms of recruiting," Pattullo says. "Each member of the couple should feel equally free to say no." But one member may ask the other, why don't you participate? What are you unwilling to share?

Once underway, the study may continue to influence its subjects and their relationships. "In some cases," Rubin writes in an upcoming article in the American Psychologist coauthored by his former research assistant, Cynthia Mitchell, "our study served to strengthen a relationship; in other cases, to facilitate its dissolution." Rubin believes that the study affected, one way or the other, "considerably more than half of the couples."

Personal, probing questions were the most frequent cause of problems. Questionnaires raised issues that many couples had never discussed and that in some cases, subjects had not even though about themselves. Specific questions included:

Who would you say is more involved in your relationship, -- or you?

How likely would you say it is that you and -- will eventually marry each other?

What are the worst things about your relationship with --?

In at least some cases, Rubin says, participation "led to conflicts in which one partner was more ready to disclose his or her feelings..."

Thus, Rubin not only observed but also influenced and, in some cases, helped bring about the demise of the relationships he sought to study. Rubin believes that his project did not cause "many" breakups which "would not have occurred anyway." The risks exist, however, and lead to questions of how or whether subjects can be apprised of the dangers of participation before they become involved.

Pattullo says that some subjects "may regret" participating in couples research. Those subjects, he believes, should be offered "some help" afterwards by the researcher.

It is rarely possible to point to experiments and say, this one is ethical, this one is unethical-- and for this reason, the Committee has avoided sweeping regulation. On the issue of deceit, for example, it has formulated no general policy. No consensus exists within the Psychology and Social Relations Department on the morality of deception. Kelman believes, for example, that deception "may create self-doubts, lower self-esteem, or create temporary conflict, frustration or anxiety." Even when it may seem benign, Kelman writes, deception "violates the respect to which all fellow humans are entitled."

But the CUHS is not overly offended by deception. "Researchers tend to think of deception as a white lie," Gartell explains. And, he says, many CUHS members have depended on it in their own research work. "As practitioners who use deception, we're not going to say that deception per se necessarily involves risks."

Although the CUHS has never lapsed into self-righteousness, it continues to inspire resentment among some Harvard researchers. Bales is one professor who worries about the potential consequences of committee review. "I think there are ethical problems in research," he said, "and I like to see them discussed. But I don't think people should feel intimidated by committees prejudiced against them." Bales does not think the CUHS is an intimidating force right now-- "but whenever you have a regulatory mechanism, the possibility exists."

Even some Committee members have doubts. Reid Hastie, assistant professor of Psychology and a member of the CUHS, says, "The Committee does affect the kind of research that's done, and I think often those effects are bad. I don't think people should be cowed."

Such doubts certainly account at least in part for Pattullo's relaxed attitude toward review. In its nine-year history, Pattullo's committee has not rejected a single proposal.

The CUHS has enormous power to affect psychology research, but Pattullo prefers to emphasize deference and compromise. "We've always been able to work out changes in the procedure," he says, "or the investigator has been able to be persuaded to do something else." Here Pattullo emphasizes "persuaded" -- not, he says, "intimidated."

But in its tolerance, the CUHS opens itself to criticism from the other side. Some feel that the Committee is not vigilant enough, that it fails to take its task as seriously as it should and that researchers in the department respond by treating the CUHS and its responsibility with less than total respect. In fact, the Committee is less strict, both in policy and practice, than may similar groups at other institutions.

Last summer, a graduate student in psychology, Joanne Martin, wanted to conduct some pilot experiments for her thesis on relative deprivation. Martin planned to recruit subjects from a summer school psychology class: in return for their participation, she says, she promised to help subjects prepare for their examinations.

But Martin wanted to conduct her experiment without CUHS approval. She had "unexpectedly got funding," she says, "and the Committee wasn't meeting--and I really wanted to run just a couple of pilots." So Martin went directly to Pattullo to request permission. Pattullo told her to go ahead with her research.

"There's really nothing about the study which would be morally debatable," Martin says. But it does involve deception, and Martin says that, in general, she sees "lots of problems" with deceiving subjects. "I was very leery of doing it in this study, and I'm still considering other options," she says. "I guess I feel very uncomfortable lying to anybody."

Because of the deception involved, Pattullo told me, he should have waited for the next CUHS meeting. "I simply sort of overstepped the bounds a bit, acting beyond my authority." But it is "fairly common" for Pattullo to approve proposals for researchers who cannot wait until the CUHS returns to session in the fall.

But it is its official policy that the CUHS permissiveness becomes more evident. In some ways, the Committee sets looser standards for non-government research than the government-- that is, the Department of Health, Education and Welfare--sets for its own. For example, HEW tends to view informed consent--the concept that, before participating, subjects should clearly understand both the risks of a study and their right to withdraw from it at any time-- as the keystone of regulation. But at Harvard, according to Pattullo, informed consent forms are "seldom required."

Most striking, perhaps, is that the CUHS does not require that all research projects be reviewed. The Committee must, by law, review any HEW-funded study at least once a year. But for other research, much is left to the discretion of the investigator.

"We don't automatically require that they come to us," Pattullo explains. "We rely more heavily on the good judgment of the investigator than HEW does." Pattullo believes the Committee reviews about three-fourths of the research actually performed at William James.

But the CUHS has no way of knowing exactly how much or what kind of research is going on. Elizabeth Hepner, a recent addition to the Committee, seems puzzled that the Committee does not review all research in the same way: "If you're looking at the rights of subjects, it shouldn't matter who funds it."

Some confusion has resulted. "It's been hard to find out exactly what kinds of research have to go through the Committee," says assistant professor of Psychology Peter A. de Villiers. Many department members routinely bring any project involving human subjects to the Committee's attention. But others are either not sure of the Committee's rules or not eager to cooperate.

"There's doubtless a few people who just never bother to ask the Committee when they should," says Hastie. "It's quite possible that there are stressful experiments going on that the the committee simply hasn't heard about."

Another committee member, Gartrell, has "known a couple of cases" of potentially risky research never having been reviewed. And Pattullo himself acknowledges the problem: "There's quite a lot of research that goes on which should come to our attention but which never gets reviewed."

For now, however, the CUHS has no intention of changing the regulation which allows researchers to bypass the Committee in certain cases, if they so choose. Many tend to see the problem as one of education, not stricter regulation. "I'd rather handle this through an education process than through a stronger policing effort," Kelman says. And Pattullo believes that he has a "considerable" educational responsibility. "It's a task which is never done," he says.

Certainly, sensitivity to ethical problems in research has grown enormously during the decade since HEW required committee review. If a Leary came to Harvard wanting to experiment with psilocybin or LSD, the CUHS--if it knew of the tests--would certainly interfere, and few would mourn the assault on Leary's academic freedom. "The Committee has raised everyone's consciousness during the last few years," Bales says.

But the consciousness-raising process is far from complete. Whether or not a review committee should exist is no longer disputed, but many complex problems remain--whether to permit the deception involved in a heart attack study, for instance, or the manipulation in couples research. On these, the CUHS has yet to provide strong leadership. If the Committee tries to do so in the furture--if it opts for a more active role in the protection of human subjects--it is sure to encounter resistance from many Harvard researchers