Negative results

There’s a noted positive bias in scientific publication. When someone tries something and it works, it gets published, but when someone tries something and it doesn’t work — which, by all accounts, is the more common case — it doesn’t get published. Which makes sense, it’s more interesting to read about the one thing that did work than about the hundreds of failures. But in hiding all those failed attempts behind the curtain, the scientist dooms others to waste their efforts trying things that have already been tried, and which are known not to work.

Positive bias goes way beyond academia. Photographers take thousands of photographs — mostly bad ones — and publish the single one in which the monkey was looking at the camera. Writers delete as many words as they type, but you don’t see those in-between constructions, just the final product, which seems to have flowed effortlessly from their fingers. Mathematicians make false starts, programmers write bugs, philosophers think themselves in circles, designers make terrible prototypes. It’s easy to see how impostor syndrome develops — when they do it, it looks perfect and easy and God-given, but when I do it, it’s ugly and hard and God, how will I ever be as good as they are?

So talk about your failures. Publish your half-done things. Make it less lonely for the rest of us.

Show Comments