4 ms·
Political polarization is partly a sample bias illusion
- chalst 5y agoI'm not convinced by the conclusion, but the idea behind what to test and the data analysis is thought-provoking. I am not convinced that the phenomenon seen is not an artefact of the choices made of the seven questions narrowed in on. To channel Andrew Gelman, this kind of study starts in the middle of the garden of forking paths.
- dynm 5y agoI have to admit that the conclusion is something I "want" to believe. So while I tried to do the analysis fairly, in a Bayesian sense, I can't fault you for some skepticism!