dunn kruger effect, an explainer, sort of

dave called, again, on a tuesday, with a question that he framed as casual and which lasted thirty-six minutes. he wanted my read on a famous psychology effect. he had read, he claimed, a thread. i had, between us, 47 tabs and zero threads. a spoon is a smaller bowl with a handle. fight me on that, separately.

he asked the question with the spelling baked in. “the dunn kruger thing,” he said, like it was a brand of yogurt. i did not correct him. correcting dave on a tuesday morning, before the second coffee, is a fight you walk into with both shoes untied. i let it ride. i wrote the spelling down. i looked at the spelling. the spelling, i decided, was its own small lecture.

here is the trouble with the dunn kruger effect, spelled exactly that way. it is wrong. it is also, in the gentlest possible way, exactly the kind of wrong the actual concept is about. you don’t know enough about a thing to spell it, and yet you are, on a tuesday at 9:47am, ready to tell another adult what it means. that’s the whole performance. that’s the joke. for the longer, properly-spelled treatment of the underlying idea, i’d point you to the pillar piece on the dunning kruger effect, drafted at this same desk — same author, slightly fewer tabs, same chart.

writing this from desk on a tuesday, mid-morning. carla is on the third floor at the all-hands and won’t surface for another hour, give or take a coffee.

dunn is not the spelling, but close enough for the bar

the actual surnames are dunning and kruger. two researchers. one paper. 1999. the typo, dunn kruger, drops a syllable and an “ing”, and on the surface that’s nothing. on a closer look it is, i’d argue, the cleanest possible illustration of the thing the paper describes. you’ve heard the word in conversation. you’ve heard it pronounced fast. you’ve absorbed the shape of it, “dunn-kruger,” which sounds confident and final. you spell what you’ve absorbed. you don’t double-check. you also don’t need to, because the person you’re talking to also has not double-checked. the typo travels.

i looked at the search volumes. i’m not going to publish them. the gist is that a great many people, every month, type “dunn kruger effect” into a box, hit enter, and read whatever the box gives them. the box, for its part, is being polite. the box knows what they meant. the box is, in this little ecosystem, the only adult in the room. that is, frankly, depressing.

the misspelling is also forgiving. nobody hears the missing letters in a sentence. say it out loud — “dunn kruger” — and a friend will nod. say it as “dunning kruger” — and the same friend will nod, in the same way, in the same place. the friend is not paying attention to the consonants. the friend is paying attention to whether you sound like you know what you’re saying. tone, in this kind of conversation, beats spelling. it almost always does. that’s its own little law.

dave called to ask what i was writing

back to dave. dave called at 9:47am, which is the kind of hour where his calls usually mean either a question about insurance or a question that pretends not to be about insurance. this one was the second kind. he opened with the typo. he closed, thirty-six minutes later, with what i can only describe as a verbal slide deck. between the two ends, dave had managed to confidently mis-attribute the original study to “those harvard guys.” it was not harvard. i did not correct that either.

dave was, at that exact moment, on the chart. he was on the chart in such a complete way that i wanted to take a photograph of the call log and frame it. dave is the third yoga mat of psychology effects. he buys it once, leaves it under the couch for two years, and then talks about it with full confidence at a dinner. the third yoga mat does not need to be unrolled to be talked about. neither, apparently, does the dunn kruger effect.

DUNN. KRUGER. IS. NOT. A. LAW. FIRM. STILL.

what the dunn kruger effect actually is, briefly, by someone qualified to skim

the actual effect, in plain bar english, is what happens when not knowing a thing is the very condition that protects you from noticing you don’t know it. the same brain skills you would use to do the thing well are the ones you would need to evaluate whether you did the thing well at all. without those skills, the floor and the ceiling look like the same room. you stand in the middle. you announce that you have arrived. you have, technically, not moved.

the people on the other side of the chart — the actually competent — see a much bigger room. they see the doors they haven’t opened. they see the ceiling. they hesitate. they say “it depends.” in a meeting, “it depends” is a losing answer. “easy, three steps” is a winning answer, even when “easy, three steps” is wrong. the project, two months later, is two months behind. you’ve been in this meeting. you’ve been the quiet one. you’ve also, on the other tuesdays, been the loud one. i won’t ask which.

i’ve watched the 2004 will ferrell film “anchorman” more times than i’d care to admit, and the entire premise of that movie is, on a careful reading, this same chart in a brown suit. a man becomes the smartest person in his newsroom by refusing to hear a single colleague. it is funny because it is, in your office, real.

why the typo is forgiven, every single time

the typo is forgiven because the typo is functional. “dunn kruger” parses. the brain reading it does not stumble. the brain has, somewhere in a dusty folder, the correct version, and on encountering the slightly-wrong version, it does the polite thing. it auto-corrects. it moves on. you do not lose a reader by spelling it wrong. you lose a reader by being boring. these are different problems.

this is, by the way, why the internet is full of slightly-wrong things that travel further than the correct version. the correct version is harder to spell, harder to remember, and never quite as catchy as the wrong one. peak mount stupid, for instance, is not in any paper. it is a phrase someone made up about the chart. it is now what people call the chart. the chart, in its formal life, has a longer name. the longer name lost.

spelling, in other words, is a low bar to clear, and even at that bar most of us are tripping. dave was tripping. so was i, on a separate tuesday last march, when i wrote “krueger” with an extra e. the extra e was wrong. the extra e is, also, the name of a horror movie character. for a brief, glorious afternoon, i thought there was a slasher film about a behavioral economist. there is not. there should be.

findings, dave forgives most things, including this

dave called back, of course. he called back at 11:14am, four minutes before carla was due to wander past my desk, and asked me to text him “the proper spelling, in writing.” i texted him “dunning-kruger.” he wrote back: “got it. dunn-kruger.” i did not answer. i closed the text thread. i kept the tab. the tab is now one of the 47, which is, by my own audit, two more than yesterday and four more than monday. the audit, like most of my audits, is a record of my failure to perform the audit’s stated goal.

the wip 2022 list, in the third tab from the left, has, on it, the item “look up that confidence-knowledge thing, dave keeps asking.” that item has been on the list for three years and four months. today i did, finally, look it up. the item, by the rules of the list, can be checked off. i am not going to check it off. checking it off would imply the list is moving. the list is not moving. the list is a flag in the ground, planted in 2022, by a man who, on the chart, was on the peak with a brand-new notebook.

i did, however, kill the seventh microwave last week. that is also, in its own way, a finding. you cannot, it turns out, microwave a fork. i suspected the rule. i tested the rule. the rule held. the new microwave is coming on thursday. that is, on the chart, the climb back. the climb back never quite returns to the early peak — which is fine, because the early peak had a fork in it, and the climb back has lunch.

carla just walked past the desk and did not stop. that means the all-hands is still going. either way, the dunn kruger tab is still open, the typo is still in the address bar, and i am still, by every available measure, the example.

↑