Dunning-Kruger and the duty to explain

You have probably heard of the Dunning-Kruger effect: the idea that people who are not expert in a particular field tend to over-estimate their level of competence, while people who are more expert in a particular field tend to under-estimate it. This is often expressed as the claim that non-experts don’t know what they don’t know (‘how hard can it be?’) while experts know just how much they don’t know (‘I’m an imposter!’). The effect was originally described by David Dunning and Justin Kruger in their 1999 paper, Unskilled and Unaware of it: How Difficulties in Recognising One’s Own Incompetence Leads to Inflated Self-Assessments.

Even if you are not familiar with Dunning-Kruger, I expect that you are familiar with the effect it describes, especially if you have ever used the words,’Well, it’s a bit more complicated than that . . .’

This is particularly true if you work in a complex, technical field such as enterprise technology. Those of us who work in such roles frequently get frustrated with the people who ask us to do things (sometimes business leaders, but also often technology leaders who are no longer close to technology) fail to understand what they are asking for, and fail to understand the degree to which they don’t understand.

The designer has drawn the screens for the mobile app. All you have to do is build it. How hard can it be?

This vendor told me yesterday about their great AI solution. All you have to do is give them our data and integrate the models. How hard can it be?

The team’s got ten people and I’ve given you another ten. Go twice as fast. How hard can it be?

There are two ways we can respond to the Dunning-Kruger effect.

The first way, which I must admit can be tempting at times, is to use it to confirm our prejudices and preconceptions.?

We knew that this stakeholder’s demands were unreasonable: we can now blame them for being so ignorant that they don’t recognise their own ignorance.

Of course this person doesn’t get it: they don’t even know enough to know what it is.

There’s no point in trying to explain the solution to them: they haven’t got the basic concepts to grasp it.

This response might feel satisfying for a while. It might give us material to gossip with our fellow technologists, lamenting the technical shortcomings of the people we have to work with, and their lack of appreciation for what we do. It might let us enjoy a few moments of smugness and superiority. But, ultimately, it doesn’t do us any good, and, moreover, it’s not fair to those people we feel smug about.

I think that the second response is more useful and interesting. Rather than use it to confirm our suspicions that those around us are even more ignorant than we thought, we can take it as an insight into universal human cognitive deficits. To me, the interesting thing about the Dunning-Kruger effect is that it applies to everybody. At the same time that those of us who work in technology are lamenting the confident but uninformed pronouncements of non-technologists, we are almost certainly underestimating our own ignorance and the complexity of the problems we are not expert in. Why do you do it that way? is a good question to ask? when designing systems, but only if we listen properly to the answer. I know that I have been guilty of assuming that the business people whose needs I was attempting to serve just hadn’t thought the problem through properly, or were making things more complicated than they needed to be, when what was really needed was for me to try harder to understand their context.

If we think about the Dunning-Kruger effect in this way, then we can act with more empathy and with more impact. When we discover that someone is so unfamiliar with our area of expertise that they don’t know what they don’t know, then it amplifies our duty to explain : to provide them with sufficient foundations to have an informed conversation and make good decisions. When we are dealing with an area with which we are unfamiliar, we also have a duty to examine our own level of ignorance and seek a better level of understanding. Perhaps, when we are working together to design new solutions, our first goal should be to figure out whether we know enough to understand each other.

It's also worth noting that the difference in these two responses (smug superiority vs actionable insight into human cognition) is the difference between cynicism and generosity, between complacency and curiosity. I find that we face this choice almost every day, and that while cynicism and complacency are tempting, generosity and curiosity are always worth the extra work.

And, of course, I have to end this article by declaring that I am not an expert in psychology or behavioural science, and my interpretation of Dunning and Kruger’s research may be flawed: I don’t know what I don’t know!

(Views in this article are my own.)