We Must Stop Jealously Citing the Dunning-Kruger Effect

What the Dunning-Kruger study actually demonstrated: Among 65 Cornell undergrads (ergo, a pretty selective, smart bunch to start, likely accustomed to comparing well), the “worst” performing thought their performance would be slightly above the average of the group, while the best performing thought their performance would be highest of all. The average performing thought their performance would also be above average.

The participants had no measure to compare against each other, but from a general perspective were likely, to an individual, far above the normal population average. They also had the difficult task of ranking not by actual performance, but by percentile: It was a situation where one could score 95 out of 100 on a difficult assignment and still end up in at the bottom of the percentile ranking. As a group that shared enormous commonalities (same academic background, life situation, all getting into an exclusive school), there is no surprise that self-evaluations compressed towards the center.

What many in this industry endlessly think the Dunning-Kruger study demonstrated1: People who think their performance is above average must actually be below average, and the people who think they are average or below must actually be above average (the speaker almost always slyly promoting their own humility as a demonstration of their superiority, in a bit of an ironic twist. Most rhetoric is self-serving). The shallow meme is that people with confidence in their abilities must actually be incompetent…Dunning-Kruger and all, right?

Cheap rhetoric turns cringe worthy when it’s cited to pull down others. Do a search for Dunning-Kruger cites on developer forums or blogs and you’ll find an endless series of transparent attempts to pitch why the speaker is better than anyone else.

No one has ever gained confidence, position or ranking by projecting some myth that they think undermines the people who have it. It just makes the speaker look worse. The same thing can be seen in competitive online gaming like CS:GO where everyone better than the speaker is a hacker/spends too much time playing the game, and everyone worse is just naturally less skilled and should delete the game they’re so terrible. It’s good for a laugh, at least until it gains truth through repeated assertion.

This is one of those posts that if you’re a blogger isn’t a good way to grow subscribers: Invariable there are some readers who spend their professional life calling everyone “below” them hacks, and everyone “above” them hacks who suffer the Dunning-Kruger effect. It’s common on any developer related forum. Eh. Thankfully I don’t care about reader numbers.

1: At least one author of the study is complicit. This study quite literally made them famous, and scientists and researchers are people too, so often it’s best to just go with the flow and allow people to puff it up to assuage their own insecurities.