
Stuart Ritchie: bad science, good science and behavior genetics
Razib Khan's Unsupervised Learning
00:00
What's the Goodheart's Rule?
The file drawer effect, where the studies you see published are not representative of the underlying distribution of results. Goodheart's law states that when the measure becomes the criterion, then then it ceases to be a good measure. And what you actually want is high quality science. The problem is, it's a lot easier to measure just numbers of citations and just tat them up in the way the h in x does, and divide the number of papers. So we need to watch out for things like pre-registration on websites or even people who say they're doing open science but don't really mean it.
Transcript
Play full episode