There's a handful of comments kind of getting tripped up here. Scott is relying on a 1-2 punch-- First, that openness to experience was decently different in th

There's a handful of comments kind of getting tripped up here. Scott is relying on a 1-2 punch-- First, that openness to experience was decently different in the two groups (73rd percentile vs 69 percentile) and then that most Big5 tests aren't done well. ( Big5 is a valid personality test-- it gets decent test/re-test scores, but it's still weak. Myers-Briggs fails because, despite being very similar to the Big5, it doesn't have good enough test/re-test scores.) Scott cites another study that says when you test openness properly, you get a wider distribution. So if you pair those two together, you get a decent sized effect. Maybe.