Ethical Issues in Psychological Research with Sexual Minorities

Part I

One of the issues raised by these cases is the role of risk/benefit analysis in the Belmont Report and the Code of Federal Regulations human subject research.  It is worth pointing out that European research guideline to not accept that consequentialist emphasis and place much more weight on the protection of human subjects. (Elgesem)  So the issues raised in this case are to some extent created by the consequentialist considerations built into the U.S. guidelines.  It is also worth remembering that federal guidelines are a minimalist ethical approach and nothing precludes using a more stringent moral standard. In this case I shall assume a more stringent standard and start from the assumption that serious harm to the subjects in this case cannot be justified by some generalized good to others. 

In this case it is noted that gay men experience greater degrees of anxiety and depression than heterosexuals, in part because of stress resulting from experiences of discrimination. What we do not know from the case is the severity of such stress and its resultant anxiety and depression.  If that level is already quite severe, then an experiment could increase that stress and the resultant anxiety and depression even more.  If that is a fact, then it  would suggest that the study in Part I ought not be done at all.

We also do not know what the range of risk is within the pool of subjects. If there is a range of levels of reaction to stress within the population to be studied and some are at a very high level of danger to exposure to such stress but others are not, then perhaps the very high risk persons should be removed from the potential pool of subjects to begin with.  Something could be learned from the study, even if the study does not include very high risk subjects.  A claim to be able to learn even more from the study if high risk subjects are included does not override the potential harm to high risk subjects. 

One of the risks identified in the study is that the study will make the participants more aware of discrimination aimed at them and that hyper-vigilance about discrimination in turn may increase their level of depression/anxiety. Additionally, hyper-vigilance about the stress may itself compound problems of depression/anxiety. Making them sensitive to the actual level of discrimination they receive could be a mixed blessing. On the one hand if they are naively unaware of their environment, it may be important for them to have a realistic awareness of the discrimination to which they are subjected. On the other hand, there can be the harm of inflicted insight. By engaging in this experience, they are forced to see things that they really do not want to see. 

There may be another risk of this experiment. It is possible that making subjects hyper-vigilant about discrimination may not only sensitize them to actual discrimination but also to condition them or habituate them to see discrimination where there is none, to distort their perception of reality.  In some ways this may be a more serious harm. There is some value in having an accurate awareness of the challenges of being gay in a discriminatory environment, however painful that may be. But to acquire a distorted perception of the situation cannot be good.

Clearly, full disclosure of these risks will be important as the researcher solicits subjects for this research.  It may be important to have trained counselors to help subjects fully understand the risks during the informed consent process and to work though their experience at the end of the study and, if any show signs of psychological problems worsened by the experience, the researcher has an obligation to provide the appropriate treatment.  If there are insufficient funds in the study to provide such help, if needed, for problems caused by the study, then perhaps the experiment should not be done.

Part 2

If we already know that gay youth are at increased risk of suicide and if the study in Part I shows that that drawing more attention to the stress can increase the stress, then this study will likely increase the risk of suicide for young gay men in the study, particularly if they are not out to their parents, but perhaps otherwise as well. That suggests that the study should build in therapy support for all young subjects who may need it as a result of the study. 

It is serious business to do research on young subjects without the knowledge or consent of parents. Imagine the harm if, as result of the study, a young man who is out to his family, committed suicide without the parents being aware that he was in the study.  Imagine, what is worse, if he were not out to his family. It would be difficult to make out the case that the study was worth the cost to this family. 

Perhaps the study should start by screening out any young gay men who show evidence of suicidal tendencies for whatever reasons and then screen out of the remaining population, all who are not out to their parents. The study population would be those out to their parents and at lower risk of suicide. Thus it would be possible to approach the parents of young gays in the study for permission without the possible repercussions noted for those not out and parents would be alerted to watch for any signs of suicidal tendencies as result of the study. If the study establishes there is still a greater tendency to suicide among this population, do we really need to study the population of those who are not out to parents?

Part 3

Apparently, there was no discussion regarding the right of the center to review the findings and approve the manuscript prior to publication, before its IRB board gave its approval for the study.  It would have been good to have cleared that up before beginning the study. That does not solve the issue, however; it just pushes the issue back a step.

There are parallels here with collaborative research with Native American groups who also have their own IRB’s.  In either case, it raises the issue of the possibility of inappropriate influence on the study findings. Imagine if an ultraconservative anti-gay group offered to completely fund the study but required that they give the manuscript approval before it could be published. We would think that an inappropriate conflict of interest and the same applies to the center.

Is it unethical for Brian to hide scientific findings if they do not turn out the way he wishes the results to turn out?  Brian can choose to become a scientist or not; he can choose to do research in this area or not. But having chosen to be a research scientist and having chosen to do research in this area, then he needs to do the best science he can do and let the chips fall where they may. The very point of science is a search for truth about our physical world. To falsify results or to suppress results is to undermine the very fabric of science.  If Brian is not interested in the pursuit of science, what would be the point of Brian being a scientist?  Sissela Bok has written two classical works on the core value of truth telling and its value in the sciences as well as and the ethical issue of keeping secrets that are both worth considering in this general context. (Bok, 1999, 1989)

If it turns out that it is a healthy option for gays to stay in the closet, wouldn’t they like to know that?  What sort of respect for persons is exhibited by denying them that knowledge?  It seems to me that the subjects who may have been put at risk as subjects in the experiment deserve to have their efforts honored by publishing the truth of the findings. Finally, if any part of this study were financed by anyone besides Brian, it is an abuse of the use of those funds to suppress the results. 

References

• Elgesem, Dag. “What is special about the ethical issues in online research?” Ethics and Information Technology, Volume 4, Number 3, 2002, pp. 195-203.
• Bok, Sissela. Lying: Moral Choice in Public and Private Life. Vintage, 1999.
• Bok, Sissela. Secrets: On the Ethics of Concealment and Revelation. Vintage, 1989.

Although this case raises a variety of ethical issues, all with their own subtleties and complexities, two particular issues will be discussed in this commentary. 

Assessing Risk/Benefit Ratio

The first issue, which is raised in Part 1 of the case, concerns risk/benefit ratios when working with human participants.  One of the central tenets of the Belmont Report (1979) is that researchers must (1) do no harm and (2) maximize possible benefits and minimize possible harms.  This case demonstrates several of the complexities inherent in the principle of beneficence.  

First, how should risk be defined and determined?  In Brian’s case, it is possible but not definite that focusing on stress experiences will have negative effects for the participants.  How should Brian determine the probability of risk?  If there is no realistic way to do this, how should Brian proceed?

Many scientists conducting research similar to this have relied on the fact that, in most cases, the negative impact of stress exposure is short lived.  In other words, while people may experience negative effects in the moment, they quickly return to baseline and are fine in the long-term. Should calculations of risk differentially weight short-term versus long-term consequences?

Additionally, researchers conducting this type of work have stood behind their informed consent procedures.  These researchers reason that as long as they are up-front from the very beginning about potential repercussions of study participation then they are within the guidelines of ethical practice.  Responsibility for participant safety is ascribed to the consenting participant as opposed to the researcher.  This practice may be fine in cases where the consent document is completely clear, where the potential participant is fully engaged in understanding the document and fully capable of consenting, and where the researcher is capable of answering questions about procedures and potential threats.  Unfortunately, these circumstances are rare, especially in psychological research where the majority of participants are undergraduates fulfilling a course requirement.   Who is ultimately responsible for determining the risk of participation? Who is ultimately responsible for the safety of research participants?  The researcher? The consenting participant? The IRB?

Second, who should be considered when calculating risk/benefit ratios? Only the research participants? Society-at-large?  Future generations?  Brian is aware that he is possibly putting participants at risk by asking them to focus on the stress they experience in their daily lives.  However, he is also aware that basic information about gay-related stress must be collected in order to create effective interventions and to help myriad gay men and lesbians cope with their minority status. If Brian were to consider only his research participants in his calculations, he would likely conclude that the benefits do not outweigh the potential harms.  However, if Brian were to consider his research participants, the gay community and future generations of gay individuals, he would likely arrive at a very different conclusion.  Is there an appropriate way to calculate risk/benefit ratios? Is it ever ethical to sacrifice a few individuals for a larger goal?

Bias in Social Science Research

The second issue, raised in Part 3, concerns personal or political bias in research.  Many social scientists emphasize the importance of objectivity in the pursuit of knowledge.  These individuals assert that science must be free of bias, and that the researcher must be neutral in relation to the topic and communities being investigated.  Elias (1987) summarized this position nicely, stating that those who study human groups must learn to “keep their two roles as participant and enquirer clearly and consistently apart and . . . to establish in their work the undisputed dominance of the latter” (cited in Perry, Thurston & Green, 2004; p. 135).  Others, however, argue that objectivity is impossible to attain and that better science is derived from active involvement on the researcher’s part.  In their discussion of qualitative research, Perry and colleagues (2004) argued that a critical piece of the research process involves interpretation and that the researcher necessarily plays a central role in this analysis.  These researchers concluded that instead of ignoring “emotional involvement” in research, we should recognize “the inevitability of involvement and the potentially significant part it can play in developing a more reality-congruent picture of complex aspects of the social world. . . .” (p. 139). Is there such thing as objective social science?  If not, should scientists be responsible for revealing their biases?  Is science valid if there is bias linked with data?

In his research on gay-related topics, Brian wears two hats: one as a scientist and one as a gay man.  While these two identities do not have to conflict, they nevertheless can conflict.  In the case, for example, Brian discovers that those who are “out” experience more gay-related stress than those who are not “out.”  “Scientist Brian” finds this discovery interesting, while “Gay Brian” finds it problematic. Brian fears that revealing such a finding could encourage people to lead closeted lives and could subsequently set the gay rights movement back several decades.  In deciding whether to publicize this finding or not, which one of Brian’s hats should have more weight? Is it possible to wear both hats at the same time? How?

In discussing this piece of the case it is important to realize that there are both pros and cons to Brian’s involvement in the research.  On the pro side, Brian’s status as a gay man gives him credibility with the people he is studying.  The gay men that Brian is reaching out to are more likely to trust him and to involve themselves in the research given that he is “one of them.”  This becomes crucial when studying a group such as gay men, a group that has long been manipulated by researchers and ostracized by the psychological community.  Many gay men are suspect of the research enterprise and want to know that they are not being used for research that will come to haunt them and their community later on.  If Brian were heterosexual, many potential participants may opt out of the research, skeptical of the ultimate aims.  On the con side, Brian’s insider status interferes with his ability to be objective. What are other pros and cons of Brian’s involvement in research on a community to which he belongs?

References

• Elias, N. (1987). Involvement and detachment.  Oxford, UK: Basil Blackwell.
• Perry, C., Thurston, M., & Green, K. (2004).  Involvement and detachment in researching sexuality: reflections on the process of semistructured interviewing.  Qualitative Health Research, 14:135-148.
• The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (1979). The Belmont Report: Ethical principles and guidelines for the protection of human subjects of research. Retrieved June 30, 2005 from http://www.hhs.gov/ohrp/humansubjects/guidance/belmont.htm.