This is a reflection on a specific MICER17 conference session; for an overview of the conference, start reading here.
Keith S Taber (editor of CERP) gave a fantastic double session on research ethics, and the importance of having a widely-known middle initial. The pre-reading for this session inspired thought, once more, around what really constitutes educational research. Keith has a number of editorials on this, with the opinion that studying a local implementation of a generally effective pedagogical technique is not really research. In this case, to be research it should have control data, and unless the control data is from previous years, splitting into cohorts and running a control in a way known to be disengaging is potentially unethical unless the technique is legitimately novel; in which case, it should be studied alongside best practice, rather than placebo (The reference escapes me but it puts me in mind of a flaw in medicinal chemistry statistics where a new intervention is significant against placebo, but not significant against Existing Best Practice (which itself is not significant against placebo), leading to inappropriate conclusions)
What are some of the reasons these studies happen anyway? Perhaps institutional resistance (Does it work here? Prove it before you change something properly), and perhaps personal doubt (I know it works, but will it work in my hands?). Do I, as a physical scientist, simply trust educational research findings less? Does the increased variation of human research scare me? I would suggest framing both of these issues the same way: We have to put the onus on the person resisting change, whether ourselves or our institution, to prove that the literature supporting change is flawed beyond simply saying “It might not work in our context”.
My takeaway from Keith’s talk was his walk through notable failures of ethics in the history of medicine and psychology: Although the Stanford prison experiment wasn’t on the agenda, we looked at Milgram and Tuskegee, and discussed of the factors that can lead researchers into a situation that is grossly unethical when observed externally. Milgram tells us that people will follow the suggestions of authority into deeply uncomfortable places – deferring our moral judgement in the process. Do we as experimenters (or interviewers) risk accidentally expressing our authority in inappropriate ways? Or can we collectively deceive ourselves that the course of action we are on is justified by the tenets of utilitarianism, as in the extreme example of the Tuskegee incident?
My table had a particularly insightful discussion around the purpose of the debrief – voluntary consent that only becomes informed at the conclusion of the experiment, lest the information affect the outcome. In the Stanford Prison Experiment, trauma is inflicted that goes beyond a simple debrief or disclaimer – and has left people deeply affected, even @decades later. Is it ethical to traumatise someone if it’s all explained later as a fakery? We thought probably not.
All this might seem like a far cry from educational ethics, but badly-implemented research could see students subject to inappropriately difficult tests, potentially harming their self-efficacy and even challenging self-belief. Poorly designed studies can also waste valuable donated time. We also risk a lack of oversight if we are the gatekeepers of our own students – departmental or faculty ethics boards are meant to provide this oversight, but it often amounts to nothing more than a rubber stamp. If we run an experiment with students that view us as a lecturer or leader, can we be sure they feel no implicit coercion? No link between participation and good grades?
We then had an extensive discussion of ethics in publication. Pointing out the limitations of your findings. Not mis-citing sources. When and when not to reveal personally-identifying information. Keith identified a number of “cargo cult behaviours” (my own words), which were seen as making research ethical. Destroying research data and anonymising participants were two given examples and I would add university ethics boards to this under certain circumstances – in that it is possible for a group of people used to assessing medical interventions to rubber-stamp an educational ethics application, but that does not prevent the possibility of straying into subtly coercive behaviour as an interviewer/experimenter. I have no oversight just because my forms are in order!
For a far more elegant summary of the talk, Dr Kristy Turner was also at the conference and sketched several of the talks; her tweet is embedded below with permission, gratefully received!