Could you elaborate? From what I read, Dunning and Kruger did find a real phenomenon where people with limited competence in a domain overestimate their ability, but they did not suggest these individuals thought they were smarter than experts; and one theory holds that it is a statistical truism, which still means it exists.
The Dunning-Kruger effect also emerges from data in which it shouldn’t. For instance, if you carefully craft random data so that it does not contain a Dunning-Kruger effect, you will still find the effect. The reason turns out to be embarrassingly simple: the Dunning-Kruger effect has nothing to do with human psychology.1 It is a statistical artifact — a stunning example of autocorrelation.
I dont understand. The additional experiment data is fairly convincing, but the random data example doesnt seem to disprove the effect in itself.
With random data you are going to get a predicted score of 50 for every group, which is what is shown, but this seems to still indicate that, if this is really what people predicted, that low skill people are overestimating their ability.
Obviously random data would exhibit the effect; why should it not?
Edit: i think i get it. The random data doesnt show that the low performers dont underestimate and the high performers dont overestimate on average, but this is the natural result if everyone has no idea how they performed. Thus my question above is exactly what they are trying to say; if everyone predicts randomly (everyone equally bad at predicting) the effect arises. So there might be no relashionship between performance prediction and performance
Could you elaborate? From what I read, Dunning and Kruger did find a real phenomenon where people with limited competence in a domain overestimate their ability, but they did not suggest these individuals thought they were smarter than experts; and one theory holds that it is a statistical truism, which still means it exists.
What happens is people have the Dunning Kruger effect on the Dunning Kruger effect itself. People call it up far too often and misuse the label
I’ve often heard it’s misunderstood and used in inappropriate situations, but it’s still a real phenomenon.
Like laypeople tossing around “OCD” when they shouldn’t. Absolutely real, but not in the same way that it’s commonly used.
They’re probably talking about this. It’s been too long since I read it so I won’t be discussing it, but I’ll share a paragraph so folks don’t have to click the link to see the gist. https://economicsfromthetopdown.com/2022/04/08/the-dunning-kruger-effect-is-autocorrelation/
I dont understand. The additional experiment data is fairly convincing, but the random data example doesnt seem to disprove the effect in itself. With random data you are going to get a predicted score of 50 for every group, which is what is shown, but this seems to still indicate that, if this is really what people predicted, that low skill people are overestimating their ability. Obviously random data would exhibit the effect; why should it not?
Edit: i think i get it. The random data doesnt show that the low performers dont underestimate and the high performers dont overestimate on average, but this is the natural result if everyone has no idea how they performed. Thus my question above is exactly what they are trying to say; if everyone predicts randomly (everyone equally bad at predicting) the effect arises. So there might be no relashionship between performance prediction and performance
Like I said after my semicolon (
;), that it’s a statistical truism doesn’t mean it doesn’t exist. https://en.wikipedia.org/wiki/Dunning–Kruger_effect#Statistical_and_better-than-average_effect One can prove that this effect will always exist for statistical reasons? That’s good to know, that not-so-competent people are more prone to overestimating.I’m not making an argument, just saying this is likely the thing they’re thinking of.