A: Clinical equipoise is an ethical problem with statistical repercussions. When you're doing research with human subjects, a particular research project is ethical only if the researcher has substantial doubt as to what the answer to the research question is. So if you're doing a new treatment versus a placebo control, and in your heart you feel reasonably sure that this drug or this treatment is better than that placebo, then you are ethically bound not to participate in that project. In recent years, I have forborne to participate in placebo control projects in many cases because I know from talking to the researcher that if the results of the study came out in the opposite direction from the way they want it to come out, they won't accept the results. From what I understand about this particular treatment, and this particular control, I'm pretty sure that the treatment is better than the control. I am ethically bound not to be in this project.

Now the reason this has statistical overtones is that if you have reasonable certainty of knowing the answer, it is very difficult to be dispassionate and objective as you are designing, executing, analyzing, and interpreting results of the study.

A good paper to read more about this was published by Freedman in the New England Journal of Medicine (Freedman, 1987). In his article, Freedman says that the study must be designed in a way that is very likely to disturb clinical equipoise. You do not do a study basically to prove the point that you wanted to prove in the first place. It has to challenge your theory.

And many philosophers have put forward the idea that the null hypothesis, particularly of randomness, is never true. And so putting a lot of time and effort into disproving the null hypothesis of randomness seems false somehow or other. Having clinical equipoise means that you're trying to disturb the credibility of that null hypothesis. And if you can do that, then you can progress from there. So whether or not you can be dispassionate about the results of your study is a major issue in the conceptualization of your research.