Dunning-Kruger and Curse of Knowledge

Recently I have been looking into the idea of Cognitive Biases (https://yourbias.is/), both from a personal level and an educational level. As I read more research, I am becoming more aware of the fact that I have certain "gut reactions" to what I read, and the discussions I have with people, and these are explained by our own biases. In a future post I am going to look at some more of these, and link them to my own experiences, but in this post I want to focus on two that I think play a huge role in teaching.

The Dunning-Kruger Effect

The Dunning-Kruger Effect (https://youtu.be/pOLmD_WVY-E and https://youtu.be/kTFUrXiYWl0) states that novices in a particular domain are more likely to overestimate their abilities and knowledge in that domain, and conversely, that those with more knowledge are likely to underestimate their abilities.

For example, a situation many students find themselves in:

• Student A hardly studies for a test, and afterwards thinks they have done well. They are surprised when receiving a low mark for the test

• Student B works really hard, but feels the test did not go well. They are pleasantly surprised when they receive a good mark for the test.

I know I have been in both these situations, as a student myself and talking to students I teach.

The reasons for this are that when you do not know much about a particular topic you are unaware of all the things there are to know. Our knowledge is built up over time in a complex network of connecting ideas, and with little knowledge we cannot make these connections. It also means we are less likely to spot a mistake we have made, and even less likely to be able to fix it. If we cannot see the mistakes, we feel confident in our answers.

On the other hand, with more knowledge you have more connections, and can often see something is not right. This makes you feel less confident in your response, even though it might only be a small part of your answer.

When you become an "expert" in a field then you will also be able to fix the mistakes you spot, once again leading to an increase in confidence, though this is unlikely to be the case for many school aged students, as true expertise requires a great deal of time to develop.

What does this mean for the classroom?

When teaching new content students will often go through a phase of false confidence in their new knowledge/understanding. This is completely normal, and we are all prone to this particular cognitive bias when we encounter new material. This can be compounded when looking at performance in the short term instead of long term learning (see post here: http://markhamtl.wixsite.com/teaching-learning/single-post/2017/06/15/Learning-Versus-Performance).

As a teacher we need to help students understand this process of learning, but also ensure we are not assuming students are able to do something just because they say they can.

Does this mean we should scrap ideas like traffic lights and thumbs up/thumbs down?

Possibly. These strategies give us a flavour of student confidence, but in many ways, the Dunning-Kruger effect suggests that if the whole class says they are super confident on something that is new to them, then it is likely they actually have a very superficial understanding.

Whatever we do, when students express their confidence about a particular topic, we should always be wary of their own analysis, and use some objective measure (through formative assessment techniques) to ascertain their actual competence.

The Curse of Knowledge

The Curse of Knowledge is a bias that experts in a field are prone to. When we have reached a level of expertise, our risk of falling for the Dunning-Kruger effect diminishes for that field, but we open ourselves up to a new cognitive bias: when you know something, it is very hard to imagine not knowing it.

In the ground breaking study by Elizabeth Newton (https://hbr.org/2006/12/the-curse-of-knowledge) that led to this idea, Newton assigned people to one of two groups: "tappers" or "listeners". The tappers were instructed to tap the rhythm to a well known song, such as "Happy Birthday", and the listeners had to guess the song. The success rate was only 2.5%. This in itself is not a particularly interesting result, however, Newton also asked the tappers to predict if their listener would guess the song. They predicted that the listeners would guess 50% of the songs.

This is a perfect example of the curse of knowledge. The tappers had the knowledge of the song in their head, and so for them it was impossible to not know what the song was. But for the listeners, it just sounded like a bunch of taps. Without the knowledge it was hard to guess. But more importantly, with the knowledge, the tappers assumed everybody else would have the same knowledge.

This is something that we as teachers must always be on the lookout for. But not in our students, this is something in our own behaviour we must try to avoid.

We are experts in our subjects, and find ourselves explaining things, that we probably don’t remember learning ourselves, to students. We have a vast network of connections already built in our brains, and new knowledge can attach itself to this network much more easily than to a smaller network (i.e. A students). Something which is obvious to us, is not obvious to a student, even if they have studied it before.

A common example in Maths is solving equations. There are multiple steps to solving an equation, and as an experienced mathematician, I can usually jump a couple of steps which seem obvious. But students do not find this obvious because they have not had the time to automate the required skills. This takes many years of practice, in a variety of contexts.

So when solving an equation as an example, I have to write out every step, no matter how small or "obvious" I might think it is. The same is true of every other area of my subject!

It amazes me how those working in Early Years are able to teach such fundamental concepts, such as addition of single digit numbers, or sounding out letters. These are things we all take for granted, and because of that they can be very difficult to teach. This is also why being an expert in your subject does not necessarily make you an expert subject teacher, as being able to avoid the curse of knowledge is probably one of the most important skills of being a teacher.

What does this mean for the classroom?

Whenever we are talking with students we must ensure that we break work down into as small steps as possible. Try it now. Choose an example of a concept and try breaking it into steps. Now look more closely, and I bet you can find some more intermediary steps you didn’t think of at first. This exercise can be an excellent way to spend a department meeting!

It also means that when we fall prey to this particular cognitive bias, we should own up to it with our students. If they don’t understand something you have said, don’t blame them, but rather explain that you have not explained it well because of the curse of knowledge. It is good for our students to see us making mistakes, and owning up to them, and trying to fix them.