[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 247 KB, 1300x1041, VirginDunningKruger.jpg [View same] [iqdb] [saucenao] [google]
10586811 No.10586811 [Reply] [Original]

Dunning Kruger effect feels like it's endlessly cited in online discourse. Is there actual science to back it up?

>> No.10586854

No, it's fake news. Assume you have a group of people with same exact level of skill, say 100. Then you measure their skill and have them report their self-assessed skill level. Because your measurement is noisy, you will see some people measured to have skill of 90, some of 100, and yet others of 110. But because they all have the same skill they will all report skill of 100 and so you will see that people who you think have skill of 90 are overreporting their skill level and people will skill level 110 are underreporting it. So it's just a statistical artifact of noisy measurements. Of course you don't actually need skills to be the same, the same effect is seen if they are normally distributed.

>> No.10586861

Orbulon has observed the term is often misused, both grammatically and in terms of meaning, especially when it comes to general intelligence.

>> No.10586863

>>10586811
Why do the captions make me so angry?

>> No.10586868

>>10586811
To an extent it is real, but people only ever cite it backwards. The standard online Dunning-Kruger shitpost goes like
>you claim to have competence in X
>therefore, by Dunning-Kruger, you actually have no competence in X
which is obviously not what the actual statement of the effect says.

>> No.10586897

>>10586811
Dunning Kruger effect belongs to le redditard science.
Used by said redditfags as a better formulated "no u" argument to shut those down who dare to go against their pc/trans/islam-friendly internet culture

>> No.10586912

>>10586897
>>10586868
>>10586854
D-K in action

>> No.10586917

>>10586863
because you know everything (knows nothing)

>> No.10586918

>>10586811
Nobody uses DK correctly. DK showed the dumber you are, the more wrong you are about your test score. So people who guessed they got 80% actually got ~75%, and people who guessed they got 60% actually got ~48%.

So dumber people overestimate their intelligence more, but still know they're dumb

When you're "so dumb" that you think you're smarter than everyone, you have a psychological disorder, not the DK effect.

>> No.10586938

>>10586811
> Is there actual science to back it up?
Hmmm, i wonder... if only there were a scientific paper one could read about this effect.... maybe perhaps a paper with the first author having the name Dunning and the other author having the name Kruger..... i guess we'll never know

>> No.10586941

>>10586912
this poster is suffering from the dunning-kruger effect.

>> No.10586955

>>10586854
>they will all report skill of 100
You haven't dealt with people much have you? I have worked with several groups with about the same skill level and you get people asking for comformation on something they know is right because they lack confidence and others claiming to be much more skilled than they are.

>> No.10586968

>>10586912
Just cracked the mystery, OP.

People aren't applying the theorem recursively. eg., they assume the know how the Dunning Kruger effects works, and attempt to use it to "win" arguments.

TL;DR: You see it a lot because brainlets misapply it to make themselves feel better.

>> No.10586971

>>10586811
My main problem is that it is heavily used as an ad hominem in discussions. But point is, nobody can really have a full-range knowledge about any topic. There is always a lack of knowledge in some subject, and yes it can happen that you show that. That though doesn't make you an immediate victim to the Dunning-Kruger effect like sophists in the Internet claim and feel cool with once they used that rhetorical trickery on you.

>> No.10586978

Just another clever-sounding term for a euphoric internet expert to throw out there without properly understanding it.

Also on the list are ad hominem (doesn’t just mean ‘an insult’), and non sequitur (doesn’t just mean ‘irrelevant’).

>> No.10587044
File: 30 KB, 375x339, 1554917571536.png [View same] [iqdb] [saucenao] [google]
10587044

>>10586938
but why do none of the charts look like the typical one always cited, like in OP (minus the meme of course)

>> No.10587080

>>10586811
There is actual science, but the memetic usage of the term won't correlate very highly with evidence.

There are a thousand different recursive tangents I could break off into right now, theories of mind, Bayesian induction, why this happens and how we can verify an explanation, but they would mean precisely nothing to the memeplex.

>> No.10587094

>>10587044
Looks pretty similar, just less exagerated, there's even a "mount stupid"

>> No.10587105

>>10586811

Dunning Kruger is culture specific. Americans tend to overestimate their competence whilst Asians tend to underestimate it.

I'm highly skeptical of any effect that doesn't translate across Europeans and Asians.

>> No.10587133

>>10586863
...bcoz you Dun been Kruger'd

>> No.10587167

>>10586854
you worked hard on that one m8, it's almost convincing
if only the actual data from the paper supported it
but in the actual paper you have people overestimating their skill 5 times over

>>10586912
This

>>10587044
>proceeds to post a chart that actually validates all the meme charts

>>10587105
>falling for tatemae

>> No.10587176

>>10586897
It's not that DK is necessarily wrong, but that reddit suffers from it more than anyone.

>> No.10587188

>>10587167

>We do know from other people's work (and one publication in our lab: Balcetis, Dunning, & Miller, 2008) that there are cross-cultural differences in how much people over-rate themselves relative to reality. In North America and Europe, it's rather pervasive. (In fact, a recent study this year found that convicted criminals in the UK rated themselves as more moral than the average Britisher.)
>
>But in other areas of the world, such as Japan and the Far East, one does not find this overrating--and it is quite an active area of research why and when this might be. How it relates to the DKE has not been studied at all. My speculation is that negative feedback when you perform poorly is more prevalent and honest in these other cultures, and that's a hypothesis I would like to test. In the States, poor performance just means you are a little less awesome than you normally are.

See https://www.reddit.com/r/science/comments/2m6d68/science_ama_seriesim_david_dunning_a_social/

>> No.10587229

>>10587105
Addendum: >>10587080
>would mean precisely nothing

Nothing I say will outweigh the memetic propagation rate and impact of the XKCD comic that reference this concept, in an American industrialized context.

>> No.10587301

>>10587105
imagine not understanding the effect

>> No.10587365

>>10587167
>overestimating their skill 5 times over
Those charts have percentiles, not raw scores, genius.

>> No.10587370

>>10586811
>Is there actual science to back it up?
Yeah, Dunning and Krugers science, you fuckwit.

>> No.10587594

>>10586811
Just visit /a/. You'll have a sizable sample of people who don't know shit about anything yet delude themselves into believing that they are knowledgable in various fields.

>> No.10587602

>>10587044
The only difference between this and the memechart is that the smart people never acknowledge their intelligence. Which is very realistic, given how smart people tend to consider themselves unknowledgable due to their firm grasp of the topic and realizing how much they still have to learn.

>> No.10587619

>>10586811
There's a Zen koan. At first mountains were mountains, then mountains were not mountains, finally mountains are mountains.

>> No.10587638

>>10587044
What were the tests? Seems like it was a hard multiple choice and people all guessed what they'd get based on past hard tests. Then randomly some people guessed better than others. How did the test measure intelligence or skill in any meaningful way? I don't know the test, but this is my guess working backwards.

>> No.10587980

>>10587094
Are you joking?
Maybe the difference between the perceived graph and the actual would look like the OP meme graphic, but what the graph actually shows is that people all think that they're slightly above average regardless of actual skill.
OP's graph misleads people into thinking that stupid people are the most confident, and bolsters the Reddit-tier notion that anyone who is confident must be a fool, but they themselves (who are struggling) are actually slightly above average. (Which fittingly follows what the original dunning-kruger observes)