That's So Random! (But, Seriously: Why Do We Trust Randomized Trials To Guide Our Decisions?)

Randomized trials are supposed to be the gold standard for Science with a capital S. Free of selection bias. Unambiguous. Measurable. Replicable. Except — as it turns out — there are a whole host of problems with all of these assumptions. For example, inherent biases are hard-baked in to many trials; and we should all know by now that many of even the most well-known clinical trials can't actually be replicated by other scientists.

But there are a few other factors that confound the supposed usefulness of random trials. The first is: humans don't seem to pay much attention to the results, and tend to go with what we already believed in the first place. For example, numerous trials have shown no benefits from multivitamin supplements, and yet a large number of people around the world continue to take them, perhaps even on the advice of health professionals. (The same might be true for flossing!) Then there is a second issue, which is: randomized trials might be able show us that certain things do or don't work, but they don't always explain why this is so. (And, as noted above, they don't always work the way they say they do). 

These rather complex discussions form the basis of a rather excellent book review, which doubles up as a very insightful article, looking at a new book from Yale University Press called 'Randomistas: How Radical Researchers Are Changing Our World'. The reviewer warns that reading the book (perhaps even the multiple anecdotes contained in his article) might turn you into a 'terror' on the cocktail party circuit. It might be worth it to finally smack down some of our communally held placebos.