In part 1 and part 2 of this post, I described how I found myself inadvertently writing a recovery memoir—a type of book that my own research had shown is often harmful to people with an eating disorder—and how I realized once I had a full draft that it was ethically important to find out how readers would respond to it before deciding whether to publish it.

The experimental design involved clear “don’t publish” criteria: If there was a worsening in participants’ scores on the main clinical severity measure we used (the EDE-Q; Fairburn & Beglin, 2008), and if that worsening was greater in the group reading the memoir than the group reading the control text, and if the effect size of the difference was large by a statistical definition (Cohen’s d >0.8). There was no expectation that the book should do good, only that it should not do demonstrable harm.

The book passed the test: It did not do harm beyond the threshold we’d decided was acceptable. Indeed, reading it turned out to do significant good on both the measures we used. Interestingly, this also proved true for the book read by the control group (Ten Zen Questions by fellow Psychology Today contributor [and my mother!] Sue Blackmore, a first-person book about Zen meditation). The paper reporting on our rationale, methods, and results was recently published in the Journal of Eating Disorders, entitled “Ethics-testing an eating disorder recovery memoir: a pre-publication experiment” (Troscianko, Riestra-Camacho, & Carney, 2024).

The unexpected main finding raises all kinds of new questions about what factors were actually doing the good: something shared by the two books or something about the structured reading setup with questions being asked at intervals? As ever, one experiment opens the door to more. For now, though, it’s satisfying to have taken this fundamentally........

© Psychology Today