« Bayesian model averaging paper | Main | Twitter roundup »

April 20, 2010


No doubt Ms. Aschwanden is correct that there will be resistance to evidence-based conclusions about best practices. At least some of this resistance might result from skepticism about the validity of the research.

A brief look at Ms. Aschwanden's leading example is perhaps illustrative. She discusses a study of the prophylactic use of ibuprofen by marathon runners that concluded that the ibuprofen was of no benefit. When presented with the findings by the researcher, the runners persisted in their belief that the ibuprofen was helping them.

Although the study Ms. Aschwanden discusses is behind a firewall, the abstract indicates that the study involved 29 runners who used ibuprofen (in two different doses) and a control group of 25, in a single race event. It seems to me that so small a study can at best provide only tentative support for the proposition that prophylactic use of ibuprofen has no benefit. Might the runners' skepticism about the findings have been not entirely irrational?

Ms. Aschwanden does not label as a "myth" the idea that prophylactic use of ibuprofen is helpful, though that's clearly the context of her discussion of the runners' resistance to changing their behavior. Perhaps in deciding that the research about ibuprofen was conclusive, Ms. Aschwanden was guilty of the very resistance to information outside the narrative that she laments. Perhaps anyone who characterizes a proposition as a myth needs to be alert to that danger.

Perhaps you should stop using weasel words like "perhaps", Rob.

The comments to this entry are closed.