167: Diluted effect sizes
Everything Hertz
00:00
How to Assess Bias in Research
This is a great idea. By definition, an effect is a true effect and some shit. This was really interesting because the summary, I think the median or the main effect size was 0.36. To actually see what sort of effects that you get, it's a really cool control for how to actually assess bias in research. The closer you get to anything that is dealing with intact people, the further away you get from the fidelity of the measurements that you use. There's a point in time where you are almost forced to assume operationalize or as otherwise make a leap of logic that you can't necessarily prove. It's always difficult. And a comparison with magic is therefore very instruct
Play episode from 05:55
Transcript
