TO: Panel on Research Ethics

TO: Panel on Research Ethics

FROM: Matthew Cooper, Professor & Chair, Dept of Anthropology, McMasterUniversity; Member, McMaster Research Ethics Board

RE: Needed changes to the TCPS

DATE: 30 September 2018

Researchers experience the Tri-Council Policy Statement through their REBs. While this statement may seem obvious, it is crucial for understanding some of the changes needed in the system of ethics review. Indeed, it appears to be the variability in the interpretation and implementation of the TCPS by different REBs that accounts for much of the dissatisfaction and concern expressed by researchers in the social sciences and humanities. Thus, I believe that changes are necessary to the TCPS and also to the ways in which many REBs operate.

1. Problem: Lack of considerationof non-experimental forms of research. In the TCPS there are only four paragraphs that might be applicable, under the heading “Naturalistic Observation.” The choice of that rubric itself as well as the brief discussion expresses the unexamined assumption that experimental research is the norm, while other kinds of research are deviant.

1. Recommendations:
2. A separate section of the TCPS devoted to non-experimental forms of research particularly focusing on the social sciences and humanities.

2.2.This section must begin by recognizing and emphasizing the legitimacy of these kinds of research.

2.3.The section must discuss the diversity and complexity of such research. . For example, the assumption should not be made that all such research necessarily is “qualitative.”Ethnographic research by anthropologists, for example, often has quantitative as well as qualitative components. It may include participant observation, participatory research, census taking, mapping, survey research, various forms of interviewing from the most structured to the most informal, etc. Some research methods in the humanities, e.g., content analysis, may also be quantitative. The section, therefore, should discuss several major kinds of research in separate sub-sections, laying out some of the characteristics of the particular kind of research.

2.4.The section must clearly discuss the epistemological differences between the major research traditions. It must make it clear that positivism is not the “default” perspective in the social sciences and humanities, although there certainly is a great deal of research done from that perspective in certain fields. It must make it clear, to take one example, that research in the interpretive tradition rests on legitimate philosophical underpinnings.

2.5.The section must recognize that in much ethnographic and participatory research questions to be studied generally emerge in the course of the research. Often the focus of the research shifts. Indeed, it has been argued that if such shifts do not occur in ethnographic research then the research was superficial, at best. Thus, REBs should be enjoined to recognize that it may not be possible or appropriate for all researchers to specify their research questions in their REB applications beyond indicating more general topical concerns. As well, researchers whose focus shifts during the course of the project should not be required to clear this change with the REB unless the new research focus would endanger either the researcher or the research participants.

2.6.The section should recognize clearly that most ethnographic and qualitative research does not aim to test hypotheses. Therefore, for example, in descriptive case studies researchers should not be expected by REBs to specify control groups.

2.7.The section should recognize that much ethnographic research is long term and does not necessarily have clear start and end dates. Further, much anthropological field research may involve repeated visits over a period of many years. While each visit may be construed as a project, they are not clearly separable nor should they be.

2.8.Destruction of Data: Data collected in such projects as those mentioned in 2.7 should not be subject to the requirement of many REBs that data, e.g., tape recordings, videos, interview transcripts, must be destroyed after the project has ended. Similarly, other data that have continuing documentary or archival value must be exempt from this requirement.

2.9.Informed Consent: The section must recognize the greater complexity of the issue of informed consent in ethnographic research than in clinical or other experimental research.

2.9.1.The TCPS appears to assume that research participants are passive; they lack agency. This is certainly not the case in ethnographic research, at least in part because the research takes place in the participants’ own communities. If we assume that participants are agents, not mere subjects, then it follows that the power differential in the field between researcher and participant may be variable and not always favour the researcher.

2.9.2.The TCPS assumes that consent occurs at a single point in time, i.e., before the interview begins or the subject is enrolled in a clinical trial. However, in long term ethnographic research consent occurs (or does not) at many times. The same individual may agree to talk or be observed at one point but refuse at another, even if the subject of the interview is the same.

2.9.3.While the TCPS allows for the possibility of non-written forms of consent, many REBs appear strictly to enforce the requirement that consent must be obtained in written form. Two points must be made here. First, in some social situations written consent may be impossible to obtain, for example, because many of the participants are illiterate or because it goes against cultural norms of trust in relationships. Attempting to obtain written consent in such situations may make it impossible to obtain people’s permission to be interviewed. Second, in some situations written consent may endanger the research participants. Thus, the possibility of alternate, non-written forms of consent must be further emphasized in this section.

2.9.4.Furthermore, particularly in participant observation research, in which much interaction between researcher and participants is informal, it may not be reasonable or even possible to require the researcher to obtain consent for many encounters. This is particularly salient given that most such research is long term. That is, if the researcher has obtained consent previously from the participant and/or community and has notified all potential participants (for example, at a meeting or through a community newspaper) that he or she is a researcher, then obtaining consent for each encounter should not be required.

2.9.5.Clearer, more nuanced guidelines must be provided for purely observational research (i.e. “naturalistic observation”). Strictly interpreted, the current guidelines in the TCPS would make most such research impossible. If, for example, a student doing observational research in a coffee shop must obtain consent from every patron, the project quickly would become impossible. Similarly, a simple observational study of people in a university cafeteria would present insuperable problems. Yet such projects, in which subjects are neither identified nor recorded (e.g., on videotape), are truly minimal risk and are important for the training of students.

C:ethics/ipre-tcps/tcps-revisions-recommendations mcooper.doc