I just started “Seven Drafts: Self-Edit Like a Pro from Blank Page to Book” by Allison K. Williams. This book was recommended for my future coaching clients.
Two pages in, I already see an important tip:
“If you're worried, that's a good sign. The Dunning-Kruger effect helpfully covers this: the more competent someone is at a particular skill, the more self-critical they're likely to be. The less someone knows about a skill, the better they think they are at doing it.”
@androcat Wouldn't a self-referential assessment be an unavoidable part of the process, given the topic? It seems an unfortunate reality of studying human thought and behavior.
Regardless, I think there's truth and practical value in the Dunning-Kruger effect. Just... far from the level of reliability of, say, germ theory. Very far.
@Alexis_WordsUnbound The problem is the math. If your measure contains a component of one thing on both axes, it will look like there is a correlation even if that correlation is not there.
Like, one axis shows "respondents with high confidence" and the other shows "respondents with high confidence ranked by their skill level in independent assessment", then you'd need to correct for that before concluding.
And from what I heard, this was not done.
But of course, the finding does fit with a suspicion most people have, so there is some bias in favor of accepting it. And once that happens, there is survivorship bias in favor of noticing examples where it fits.
There could just be some set of people so confident (privileged) that they believe themselves to be elite, and that belief makes them accept shallow understanding as complete, leading them to become confident nincompoops.
This would also explain observations, and wouldn't be a general tendency of all unskilled people or of all confident people.
Alexis Miller
Androcat