Consider if every person thought he/she was in the 60%ile for some skill. Then, the best people would underestimate their skill and the worst would overestimate.
This would produce the statistical results of the Dunning Kruger test, but I don't think that's nearly as exciting as what is claimed.
I was thinking this as I read it, and the author seemed to be shocked that both extremes of the bell curve thought of themselves as being more average. That just seems logical, to assume you are average.
Social proof. A majority of people who believe X in their social circle. Unfortunately thats kind of a chicken and egg problem :P
Or, get the majority to believe/trust one person on other grounds. Then he/she can convince them together and social proof stays intact if most of them accept most of the message.
Sometimes the authority is an institution instead of a person like science, religion, nationalism.
I have seen D-K written about in the context of beginner-experts; people that don't have more knowledgeable peers to help them see the limitations of what they know (think one-man IT team in a small company, or angry drunk Uncle rambling on about politics in the recliner).
My observation is that D-K comes up in areas that people think are easy to understand (but are actually complex). Especially if the person comes from a domain with high perceived complexity. Example would be computer scientist discussing education or psychology... That seems to be ripe territory for D-K.
> people that don't have more knowledgeable peers to help them see the limitations of what they know
I've definitely met people like that; people who are used to being the smartest people in the room, and so when they make pronouncements without thinking long about a problem, everyone around them is used to assuming what they say is right, because they're smart.
And once they're around other smart people, it's like talking to a brick wall—they're so used to everyone else accepting their bullshit, that they don't know how to handle being called on it, and they don't know how to defend their ideas because they've never had to before.
It probably wasn't quantitative but qualitative, and of few choices, "rate your driving skills as one of the following: above average, below average, or average". In which case, yes, a statistical impossibility. But agreed, they're missing key context.
Lay people confuse average with median. When you ask someone if they're "above average" they often assume you mean "in the top half of the population in ability" which is "above median." In which case 80% is impossible.
What you say might be true, but the article is written by "the editor in chief of Knowing Neurons and a PhD candidate in neuroscience at the University of California". It seems that not only lay people confuse statistics terms, but also the very people that make research and present the results. And this is sad ... and in line with the article in some sense.
> Are you implying that the author [a PhD candidate in neuroscience] doesn't understand the difference between different measures of central tendency?
They probably are. There's this tendency in geek circles to assume that someone—especially an expert in another field—doesn't know what they're talking about unless they spell it out in such excruciating detail that those geeks (who actually don't know what they're talking about) can follow along as well.
I don't know why it is, and it's frustrating as hell. It seems like it's related to the Principle of Charity, but I don't know if it's the same thing, or just a similar thing.
Hmm. Maybe? I don't think that's quite it, either—or at least, that aspect of it doesn't bother me as much. If this type of geek thinks that they're really good in some area that they have no knowledge of, that's silly, but what bothers me more is the assumption that the people who are experts don't know anything. They're related, definitely. But I object more to the tearing-other-people-down, rather than the building-themselves-up.
Like the users above who assume that the author, a PhD candidate in neuroscience, is unaware of the difference between mean and median. Did they explicitly mention which they meant? No, of course not—it's clear from the context which was meant, and to explicitly spell it out would be insulting to the readers. Yet we have readers here who take that omission to mean that the author doesn't even understand the difference.
I see it all the time when a person writes about a problem in their field, and then armchair quarterbacks here confidently state their 5 minutes of thought about it as if it were a novel idea. So sure, there's some Dunning-Kruger in there, but there's also a corresponding assumption that author didn't think of the solutions you get from 5 minutes of thought about the problem. "Why didn't they just consider X!" Well duh—of course they did, because that's obvious.
Glad to read this. I thought I might be the only one picking up on the irony here. I mean, come on. He JUST read about D-K, then demonstrates it in action.
I agree, although as I said to another commenter, I'm more bothered by assuming others to be incompetent even though they are, than I am by assuming yourself to be competent when you're not. And they're related, but not quite the same?
Ever since I learned about Dunning-Krueger, I've systematically questioned many things to which I would normally feel competence or expertise.
It's basically redoubled my Imposter Syndrome. It's not crippling, but I often wonder: "Do I _actually_ know this? Do I know enough to know it if I don't?"
22 comments
[ 0.24 ms ] story [ 59.4 ms ] threadThis would produce the statistical results of the Dunning Kruger test, but I don't think that's nearly as exciting as what is claimed.
Or, get the majority to believe/trust one person on other grounds. Then he/she can convince them together and social proof stays intact if most of them accept most of the message.
Sometimes the authority is an institution instead of a person like science, religion, nationalism.
My observation is that D-K comes up in areas that people think are easy to understand (but are actually complex). Especially if the person comes from a domain with high perceived complexity. Example would be computer scientist discussing education or psychology... That seems to be ripe territory for D-K.
I've definitely met people like that; people who are used to being the smartest people in the room, and so when they make pronouncements without thinking long about a problem, everyone around them is used to assuming what they say is right, because they're smart.
And once they're around other smart people, it's like talking to a brick wall—they're so used to everyone else accepting their bullshit, that they don't know how to handle being called on it, and they don't know how to defend their ideas because they've never had to before.
Oy. Frustrating people.
In a driving test, five drivers score 100, 100, 100, 100, 90. The average is 98, and so 80% of them are above average.
They probably are. There's this tendency in geek circles to assume that someone—especially an expert in another field—doesn't know what they're talking about unless they spell it out in such excruciating detail that those geeks (who actually don't know what they're talking about) can follow along as well.
I don't know why it is, and it's frustrating as hell. It seems like it's related to the Principle of Charity, but I don't know if it's the same thing, or just a similar thing.
Like the users above who assume that the author, a PhD candidate in neuroscience, is unaware of the difference between mean and median. Did they explicitly mention which they meant? No, of course not—it's clear from the context which was meant, and to explicitly spell it out would be insulting to the readers. Yet we have readers here who take that omission to mean that the author doesn't even understand the difference.
I see it all the time when a person writes about a problem in their field, and then armchair quarterbacks here confidently state their 5 minutes of thought about it as if it were a novel idea. So sure, there's some Dunning-Kruger in there, but there's also a corresponding assumption that author didn't think of the solutions you get from 5 minutes of thought about the problem. "Why didn't they just consider X!" Well duh—of course they did, because that's obvious.
It's basically redoubled my Imposter Syndrome. It's not crippling, but I often wonder: "Do I _actually_ know this? Do I know enough to know it if I don't?"