I didn’t downvote anyone, in fact I upvoted your comment. Why do you ask? I am getting downvoted here, but I don’t know why. Yes I have read D&K’s paper and studied the plots and methodology multiple times. I thought I was mostly agreeing with you here. What part of their plots or paper are you thinking of? In my opinion, the paper dramatically overstates the claims of what it actually demonstrates experimentally. The title alone primes the reader to jump to a conclusion that isn’t supporter by their data; they did not show people who are unskilled or “incompetent” on an absolute scale, they did not test all skills in general (they only tested 4 relatively easy “skills” and one of those was subjective), they did not test the general population. Looking at the plots again won’t address the confounding issues I mentioned, even if I don’t dispute their data at all.

Every time I read the DK paper, even now, I’m struck by how wild it is that so many people refer to this as a human-wide cognitive bias, when there was so little evidence to begin with, the study was so small and biased, so many people have called out problematic issues with their data and experiment design. If randomly generated data shows the same effect, then what, exactly, did D&K demonstrate?? https://digitalcommons.usf.edu/numeracy/vol9/iss1/art4/