The Dunning-Kruger Effect

I have had several conversations lately that have left me in mind of a very famous graph. The conversations tended to be about things like anthropogenic climate change or evolution or how the universe began and the person I was talking to had ideas that were, shall we say, contrary to the scientific consensus. Apparently humans cannot change the climate, humans ‘didn’t evolve from monkeys’ and there never was a big bang – the universe has always been the same and unchanging.

The graph that I was left thinking of was one from a very famous psychological study written by David Dunning and Justin Kruger of Cornell University in 1999. You can see a version of it below.

dunning

As the rather cutting title makes clear, Dunning and Kruger wanted to see how incompetence has an effect on confidence. The concept was by no means a new one. As the authors themselves point out it had been remarked upon as long ago as the 5th century by Confucius and also by Darwin in 1871 whom they quote:

“Ignorance more frequently begets confidence than does knowledge.”

The study was inspired by the escapades of a man who knew too little. His name was McArthur Wheeler and in 1995 he robbed two banks in broad daylight without any apparent attempt to conceal his identity; unsurprisingly he was caught later the same day. When he was shown recordings from the security cameras he was heard to mutter, “But I wore the juice.” He mistakenly believed that rubbing lemon juice over his face would render it invisible to video recordings in a similar way to its interaction with invisible ink. So whilst Mr Wheeler wasn’t the smartest bank robber of all time he may well have been the freshest smelling.

In their own words, the paper studied how people who are incompetent “…suffer a dual burden: Not only do they reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the ability to realise it.” They proposed that incompetent people will:

  • fail to recognise their own lack of skill.
  • fail to recognise the extent of their inadequacy.
  • fail to recognise genuine skill in others.
  • but that they would recognise and acknowledge their own lack of skill, after they are exposed to training for that skill.

Over the course of four experiments the pair looked at people’s grammatical ability, their sense of humour and their logical reasoning. Having sat a test they then asked people to rank where they think they might have come in the group. The people who did worst were the ones that consistently overestimated how well they had done. On average, those people who actually scored in the 12th percentile ranked themselves in the 62nd percentile.

They also found a second effect, though, whereby those who scored highest would underestimate how well they did. They put this down to them overestimating the skill of others.

Dunning and Kruger came up with a plausible explanation for the effect that has gone on to bare their name: “The same knowledge that underlies the ability to produce correct judgment is also the knowledge that underlies the ability to recognise correct judgment. To lack the former is to be deficient in the latter.”

The graph I showed above demonstrates how when people first learn a small amount about a topic they have massive overconfidence in their abilities – they don’t yet have enough knowledge to recognise how little they know. As you learn more you start to get an understanding in just how complex pretty much anything worth knowing can be and you know how much more you have to learn to become a true expert in that field. After years of study and experience you may well become an expert at which point your confidence is probably well placed although the data shows that you may well underestimate your abilities.

The person I was talking to had watched a documentary on humans and our ancestors and was highlighting the fact that neanderthals had larger skulls, and therefore brains, than modern humans. For them, this was smoking gun evidence that we can’t have evolved from them (no scientist claims that we did) and that either neanderthals were more intelligent than us or that brain size doesn’t matter.

None of those hypotheses are true but my interlocutor lacked the competence to make a valid judgment on them. They had watched a documentary and therefore they knew all there was to know. They failed to recognise that not all documentaries are created equal, some of them are utter crap. There are plenty of documentaries about the paranormal and cryptozoology, but none of them mean that ghosts and Bigfoot are actually real.

I should be clear that incompetent people aren’y necessarily stupid. We are all incompetent in most things. I can speak reasonably well about genetics and a few other scientific topics as well as Star Trek and Lord of the Rings. But I know absolutely nothing about how to take an engine apart, or the music of Brahms, or the Chang Dynasty or, indeed, every other subject on earth. If I were to start talking about them I would very quickly sound like an idiot. The point of this post is to highlight the Dunning-Kruger Effect. If we are aware of it then we can try to factor it in to our thinking in our day to day lives. Let’s try not to hold forth on things that we actually know very little about. I know for a fact that this is something I used to do quite a lot; I hope I don’t do it anymore. I shall remain vigilant.

Advertisements

2 thoughts on “The Dunning-Kruger Effect

  1. Thank you for this insight. I knew that some phenomenon like this existed. I knew neither it’s name or that such a study had been conducted. The icing on the cake was the Darwin quotation.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s