Kruger-Dunning effect "mostly a statistical artifact"

Get your science fix here: research, quackery, activism and all the rest
Post Reply
User avatar
sTeamTraen
Catbabel
Posts: 790
Joined: Mon Nov 11, 2019 4:24 pm

Kruger-Dunning effect "mostly a statistical artifact"

Post by sTeamTraen » Sun Sep 13, 2020 8:29 am

It looks like Skeptics in the Pub attendees are about to experience their other favourite psychological effect, cognitive dissonance, as new analyses show that the Kruger-Dunning* effect is mostly not there. (The canonical empirical article from cognitive dissonance research is under a bit of a cloud, too.)

*Popularly called the Dunning-Kruger effect, but the paper was by Kruger and Dunning, in that order. Apparently they don't speak to each other much any more, as Kruger feels that Dunning has been rather over-egging his claims all these years, eg by swapping the names around.
Careering along in my creosote car

User avatar
Boustrophedon
Dorkwood
Posts: 1353
Joined: Mon Nov 11, 2019 3:58 pm
Location: Lincolnshire Wolds

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by Boustrophedon » Sun Sep 13, 2020 2:42 pm

I know I know nowt about statistics. And I have read before that the Dunning Kruger effect may just be regression toward the mean. But does that mean that the effect is not there but merely the appearance of an effect?

When does the appearance of an effect become an effect?
We're marchin' for freedom today ... hey!
Turn on your headlights and sound your horn
If people get in the way

User avatar
Bird on a Fire
Stummy Beige
Posts: 3533
Joined: Fri Oct 11, 2019 5:05 pm
Location: nadir of brie

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by Bird on a Fire » Sun Sep 13, 2020 3:05 pm

sTeamTraen wrote:
Sun Sep 13, 2020 8:29 am
It looks like Skeptics in the Pub attendees are about to experience their other favourite psychological effect, cognitive dissonance, as new analyses show that the Kruger-Dunning* effect is mostly not there. (The canonical empirical article from cognitive dissonance research is under a bit of a cloud, too.)

*Popularly called the Dunning-Kruger effect, but the paper was by Kruger and Dunning, in that order. Apparently they don't speak to each other much any more, as Kruger feels that Dunning has been rather over-egging his claims all these years, eg by swapping the names around.
I've only read the abstract, so maybe this is dealt with in the paper.

I normally hear the KD effect cited for very specific skills - for instance, somebody tone deaf might not realise that they're singing out of tune. The problem with using "intelligence" as the skill they're assessing, rather than performance in a particular task, is that it's not easy to participate in society without quite quickly (during childhood) getting an externally-derived sense of your own relative intelligence, which is probably accurate enough for people to know where they are at the study's resolution of four quartiles.

If they're claiming it's a statistical artefact, surely the thing to do is simulate random data, apply the analytical methods being called into question, then demonstrate the spurious appearance of said effect.
Ceterum autem censeo humanitatem esse delendam

#ShowYourStripes

User avatar
bob sterman
Clardic Fug
Posts: 212
Joined: Mon Nov 11, 2019 10:25 pm

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by bob sterman » Sun Sep 13, 2020 6:17 pm

Bird on a Fire wrote:
Sun Sep 13, 2020 3:05 pm
I normally hear the KD effect cited for very specific skills - for instance, somebody tone deaf might not realise that they're singing out of tune. The problem with using "intelligence" as the skill they're assessing, rather than performance in a particular task, is that it's not easy to participate in society without quite quickly (during childhood) getting an externally-derived sense of your own relative intelligence, which is probably accurate enough for people to know where they are at the study's resolution of four quartiles.
Errr...with the way promotions work in some organisations it would be quite easy to get an externally-derived overinflated sense of your own relative intelligence!!!

User avatar
JQH
Catbabel
Posts: 898
Joined: Mon Nov 11, 2019 3:30 pm
Location: Sar Flandan

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by JQH » Sun Sep 13, 2020 9:08 pm

bob sterman wrote:
Sun Sep 13, 2020 6:17 pm
Bird on a Fire wrote:
Sun Sep 13, 2020 3:05 pm
I normally hear the KD effect cited for very specific skills - for instance, somebody tone deaf might not realise that they're singing out of tune. The problem with using "intelligence" as the skill they're assessing, rather than performance in a particular task, is that it's not easy to participate in society without quite quickly (during childhood) getting an externally-derived sense of your own relative intelligence, which is probably accurate enough for people to know where they are at the study's resolution of four quartiles.
Errr...with the way promotions work in some organisations it would be quite easy to get an externally-derived overinflated sense of your own relative intelligence!!!
See the current Cabinet for examples.
And remember that if you botch the exit, the carnival of reaction may be coming to a town near you.

Fintan O'Toole

User avatar
Grumble
Dorkwood
Posts: 990
Joined: Mon Nov 11, 2019 1:03 pm

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by Grumble » Sun Sep 13, 2020 9:49 pm

I knew it was all along.
I know this is vitriol, no solution, spleen venting, but I feel better having screamed, don’t you?

User avatar
sTeamTraen
Catbabel
Posts: 790
Joined: Mon Nov 11, 2019 4:24 pm

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by sTeamTraen » Mon Sep 14, 2020 10:35 am

Grumble wrote:
Sun Sep 13, 2020 9:49 pm
I knew it was all along.
The appearance of this new article means that the KD effect is now poised perfectly for meta-jokes down to any depth.
Careering along in my creosote car

User avatar
Little waster
Catbabel
Posts: 768
Joined: Tue Nov 12, 2019 12:35 am
Location: About 1 inch behind my eyes

Re: Kruger-Dunning effect "mostly a statistical artifact"

Post by Little waster » Thu Sep 17, 2020 3:52 pm

sTeamTraen wrote:
Mon Sep 14, 2020 10:35 am
Grumble wrote:
Sun Sep 13, 2020 9:49 pm
I knew it was all along.
The appearance of this new article means that the KD effect is now poised perfectly for meta-jokes down to any depth.
It's Kruger-Dunnings all the way down.
Shamelessly recycling old jokes since 1952.

Post Reply