3 ms·

This may be a great article. As an aside, all Dunning and Krueger showed is that everybody thinks they're in the top 1/3 to 1/4. (At least everybody in undergr

by nyeah 2y ago

This may be a great article. As an aside, all Dunning and Krueger showed is that everybody thinks they're in the top 1/3 to 1/4. (At least everybody in undergrad school at Cornell.)

dahart 2y ago
Oh it’s much worse than that, and it wasn’t everybody in Cornell undergrad, it was a grand total of like 45 undergrads at Cornell who signed up for extra credit (which I imagine is a pretty big confounding factor - the A students don’t need it, and the F students don’t bother, on top of the massive confounding factor of only measuring undergrads from Cornell - they didn’t measure any people who are truly incompetent or unskilled). The DK experiment depends on people ranking themselves against the cohort, other people who they don’t know. The DK effect probably doesn’t exist, it has been argued compellingly that the paper does not demonstrate what it claims to. “To establish the Dunning-Kruger effect is an artifact of research design, not human thinking, my colleagues and I showed it can be produced using randomly generated data.”(!!!) https://www.scientificamerican.com/article/the-dunning-kruger-effect-isnt-what-you-think-it-is/ https://www.scientificamerican.com/article/the-dunning-kruge... “the asymmetry reported by Kruger and Dunning actually goes away, and even reverses, when the ability tests given to participants are very difficult.” https://talyarkoni.org/blog/2010/07/07/what-the-dunning-kruger-effect-is-and-isnt/ https://talyarkoni.org/blog/2010/07/07/what-the-dunning-krug... “The Dunning Kruger Effect is probably not real” https://www.mcgill.ca/oss/article/critical-thinking/dunning-kruger-effect-probably-not-real https://www.mcgill.ca/oss/article/critical-thinking/dunning-...
- nyeah 2y ago
  If you are downvoting, I hope you've read at least one of D&K's papers. Or at least looked at the plots. Not meta-stuff pro or con, but what Dunning and Krueger wrote.
  - dahart 2y ago
    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/ https://digitalcommons.usf.edu/numeracy/vol9/iss1/art4/
    - nyeah 2y ago
      Oh, no, I'm asking people who downvoted your comment to please get their heads in the game. I don't understand their point of view.
      - dahart 2y ago
        Oh haha, thanks!

by nyeah 2y ago

This may be a great article. As an aside, all Dunning and Krueger showed is that everybody thinks they're in the top 1/3 to 1/4. (At least everybody in undergrad school at Cornell.)

dahart 2y ago

Oh it’s much worse than that, and it wasn’t everybody in Cornell undergrad, it was a grand total of like 45 undergrads at Cornell who signed up for extra credit (which I imagine is a pretty big confounding factor - the A students don’t need it, and the F students don’t bother, on top of the massive confounding factor of only measuring undergrads from Cornell - they didn’t measure any people who are truly incompetent or unskilled). The DK experiment depends on people ranking themselves against the cohort, other people who they don’t know. The DK effect probably doesn’t exist, it has been argued compellingly that the paper does not demonstrate what it claims to. “To establish the Dunning-Kruger effect is an artifact of research design, not human thinking, my colleagues and I showed it can be produced using randomly generated data.”(!!!) https://www.scientificamerican.com/article/the-dunning-kruger-effect-isnt-what-you-think-it-is/ https://www.scientificamerican.com/article/the-dunning-kruge... “the asymmetry reported by Kruger and Dunning actually goes away, and even reverses, when the ability tests given to participants are very difficult.” https://talyarkoni.org/blog/2010/07/07/what-the-dunning-kruger-effect-is-and-isnt/ https://talyarkoni.org/blog/2010/07/07/what-the-dunning-krug... “The Dunning Kruger Effect is probably not real” https://www.mcgill.ca/oss/article/critical-thinking/dunning-kruger-effect-probably-not-real https://www.mcgill.ca/oss/article/critical-thinking/dunning-...

nyeah 2y ago
If you are downvoting, I hope you've read at least one of D&K's papers. Or at least looked at the plots. Not meta-stuff pro or con, but what Dunning and Krueger wrote.
- dahart 2y ago
  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/ https://digitalcommons.usf.edu/numeracy/vol9/iss1/art4/
  - nyeah 2y ago
    Oh, no, I'm asking people who downvoted your comment to please get their heads in the game. I don't understand their point of view.
    - dahart 2y ago
      Oh haha, thanks!