SPEAKER:
Welcome to the New South Wales Department of Education's Edspresso Series. These short podcasts are part of the work of the Education for a Changing World Initiative, and explore the thinking and ethical literacy skills students need in an AI future. Join us as we speak to a range of experts about how emerging technologies such as artificial intelligence are likely to change the world around us, and what this might mean for education.

How young is too young to develop metacognitive skills? And how can we foster environments in the classroom to enable metacognitive skill development? To find out, we asked Dr Shirley Larkin, a senior lecturer at the Graduate School of Education, University of Exeter, about her research in Educational Psychology, which focuses on how primary-aged students develop metacognition to enable them to learn, in their own way, for their own purposes throughout life. Shirley, what is metacognition? Why will strong metacognitive skills be important in an AI world?

SHIRLEY LARKIN:
Metacognition is a word coined by John Flavell and Ann Brown in the 1970s. And it refers to our ability to reflect on our thinking processes. And there are different models of metacognition but they all really focus on two main elements.

Firstly, what we call metacognitive knowledge. And that's knowledge about ourselves that we build up over time about our own thinking processes. So, just as we build up knowledge about the world around us, we can come to know things about the way that we think. And the second part is what is called metacognitive regulation. This is going on all the time. And this is the idea that a part of our brain is monitoring our thinking and most of the time we are not aware of that at all. But if we hit an obstacle, this monitor becomes apparent and it stops us and tells us to do something or asks us to make a decision.

So, for instance, once we know our fluent readers, we can be reading along quite happily engrossed in the content of something. And that's fine, we are not even aware that we are reading, sometimes. And then we hit a word that we're unfamiliar with, and this monitor will stop us and go, oh, you don't know that word. And then you have to make a decision. What are you going to do? You're going to look the word up? You're going to skip the word? Are you going to write it down for the future?

Those are the two main aspects of metacognition which will appear in all the models. And then we come to metacognitive skills, which are a part of that ongoing regulation, really. And these are things like task analysis, goal setting, predicting, planning, monitoring, evaluating, correcting errors and reflecting. And all of that, I would say, is going to be important in the future. I mean, I think it's really important now and always has been important.

For instance, we know that across all subject domains, experts and skill practitioners use more metacognitive skills than novices. So, what are often called gifted and talented children tend to be highly metacognitive. And you often find that people who succeed at something are also metacognitive. So, metacognitive skills enable us to know what we know and to gather more knowledge to monitor our progress towards a goal and to evaluate our learning and to make plans for future learning. And in an AI world, I think there are three areas where developing metacognitive skills are going to be needed. If we imagine a world where machines take over a lot of the jobs we currently do, humans will need to monitor and understand how these machines are performing and be able to make complex evaluations of where this technology should go next, including the implications for human cognition as well as assuring that the machines respond appropriately to our own needs.

The second area, I think, is thinking about what will humans do if machines take over the jobs? And one area, I think, is likely to be around understanding and empathising with other humans. So, those complicated relationship thoughts such as, ?I know that you know that I know what you are thinking." Those kinds of things we get involved with in relationships very often. And so, developing those social metacognitive skills, I think might become even more vital in a world that includes humans and AI, to enable humans to adapt to this changing world.

And the third area, I think, is that metacognition is vital to making good decisions in very complex scenarios. And machines tend to work logically, but humans are not always logical. And metacognitive skills help in making decisions in what we might call emergent and messy conditions. Where the outcome can't be predicted and where the factors involved include both emotions and intuitions. So, that's what I would say in terms of how they are going to be even more important in the future. But I stress that I think they're vitally important now.

SPEAKER:
Hmm, that?s really interesting. How do young children develop metacognitive skills? And what does metacognition in young children look like in comparison to older children and adults?

SHIRLEY LARKIN:
Children develop metacognitive skills through experience, through practice and through being taught them, really. So, some of the earlier signs of metacognition in infants relate to social relationships. So, there's lots of evidence that infants are aware of how other people feel. They show embarrassment when they know that someone has seen them do something they shouldn't. They're aware of representations of themselves in mirrors and photographs. And these are all the beginnings of metacognition and they are often referred to as theory of mind. And so this involves, firstly, understanding that I think, then understanding that other people think and understanding that these other people can think differently from how I think.

And there are lots of tests called false belief tests, which have been done with children. This shows that by the age of four the majority of children can use their understanding of what another person knows to get a correct answer on one of these kinds of tests. So they can think about what another person is thinking. So, giving children lots of opportunities to practice metacognitive skills and to develop the language of thinking is important. And what does metacognition in young children look like? Well, it depends quite a bit on their age and their language level. So, children talking about their own thinking or the thinking of other people will be metacognition.

For example, "I think I know the answer," or, "I understand what to do," or, "I forgot to think about it," or, "It's hard to concentrate when people are talking." So, the focus is on their own thinking. If they're playing together in a group you might hear, ?We don't know how to think about it?, or, "We're thinking about it all wrong," or, "I think this is like something we did before." If they're solving problems together you might hear, "I changed my mind, I think he's got the best idea", or, "I don't understand it," or, I'm good at math, so I should be able to do this." Or in evaluation they're thinking, so, something like, "This isn't working, we need to think about it differently" and so on. So, the focus of the thinking should be another thinking process.

So, a focus on memory, understanding, knowing, believing, wondering, imagining, concentrating. And also, those metacognitive skills, planning, predicting, evaluating, monitoring, that I mentioned earlier. In very young children, of course, the focus might be non-verbal. So, watching children correct their own errors will be evidence of metacognitive regulation. Because to correct an error you'd have to go through the thought process of firstly, monitoring your thinking, stopping, realising the error, making a decision and correcting the error.

First of all, children often do think aloud, actually, they do verbalise pretty much everything they're thinking. So, sometimes you can pick it up from what they say as well as looking at what they're doing. Older children are likely to have more sophisticated language and so they can express metacognition more easily. That doesn't necessarily mean that they are more metacognitive. So, many older students and adults do not develop higher levels or stronger metacognition because no one has taught them or provided opportunities for them to develop metacognitively.

So, some people seem to get it intuitively and they are the ones who tend to do well academically. And adults sometimes, so they are not aware of how to be thinking, about how they process information or how their brain works or their memory. Until someone like a researcher asks them or gets them to do a test, which reveals something to them. And I found this with some research I did with some teachers when one said to me, ?You know, before I did this I had no idea I was using a way of thinking I was taught as a child, but which is no longer really useful to me. And that was a revelation to me." So, it's important, really, to kind of provide opportunities for the people at all levels and all ages.

SPEAKER:
Shirley thanks for that. In your work you talk about metacognitive environments. Could you tell us a little bit more about how teachers can create good metacognitive environments in their classrooms?

SHIRLEY LARKIN:
It's a term used to describe, really, any type of learning environment where the focus is on developing metacognition. So, in a recent project I did on religious education in primary schools, we focused on developing a metacognitive environment under three headings. The physical space, the task and the relationships.

So in terms of the physical space, it was about being flexible. Ensuring that children could work in groups, could work individually and could move around the classroom to do different tasks. Now, we have three zones where we focused on metacognitive questions and prompts for reflection in one zone, understanding your own world view. The second part of our metacognitive environment focused on the actual tasks and activities that we created. And it's really important that tasks and activities actually require metacognition. Because not all of them do. If they're too simple or they have fixed answers, they're less likely to develop metacognition because you don't need metacognition to do that.

So the tasks that we developed we tried to make as authentic as possible but still linked to the religious RE curriculum that we have here in the UK. We tried to make them complex and open ended. And we tried to challenge children to go further and take time to think. Not just by getting an answer but about how they were thinking about the task and whether they could think about it differently. The third area was a focus on relationships, and here we also included language.

So, getting children to think about what are called mental state words such as believing, knowing, wondering, guessing, imagining. We had teachers run a session on what is belief. Because these were all RE teachers in a primary school, we were working with seven and eight-year-olds. And they often talked about the faith beliefs of different faiths. You know, Hindus believe this, Christians believe this, and so on. But they've never taught about what belief is or where beliefs come from or whether we can have different beliefs from each other. Or whether our own beliefs can change over time and that kind of thing.

So, we focused a session around that and that was quite a revelation, I think, for both the teachers and the children. So, the other area of relationships was around trying to develop authentic learning relationships where teachers could be really clear and upfront about what they were doing. So, there's nothing wrong with asking questions to test knowledge. So, questions where the teacher has the right answer already in their head. And they're just looking for a child to say that answer.

But what we suggested was, ?well, why not tell the children that's what you do? Why not tell the children, I'm going to ask you some questions now to test your knowledge about this particular area." And then you can contrast that to asking questions which are much more open ended, which are genuinely open. Where the teachers are thinking along with the children about an issue. And of course, teachers can model good metacognition by thinking aloud. They're making their own thinking visible and challenging students on helpful metacognition. ?Cause we can build up this unhelpful metacognition about ourselves, for instance, that we're not good at certain types of things. Which we could also challenge and say, well, how do you know you're not good at it? Maybe it's just that you don't like it, and so on.

SPEAKER:
That?s really interesting, thank you. Is the role of the educator to directly teach, or to facilitate the development of metacognitive skills, and what could this look like in action?

SHIRLEY LARKIN:
I think the first educators, of course, are parents and caregivers. So, it's important that they help children develop those theory of mind skills I was talking about earlier. So, understanding their own and other people's thinking. And you can do that through lots of play, and puppets are really useful for doing that. Understanding emotions by reading pictures of faces with lots of different emotions on them. Lots of language around developing those mental state words. So, using them a lot.

For example, ?Is that a guess or do you know it and how do you know it?" And so on. There is a role, I think, for direct teaching at metacognitive skills as well. So, how to use knowledge of your own thinking to improve your learning. So, thinking about what parts of your thinking you might need to develop in different ways. Teaching self-questioning strategies. So, getting children to ask themselves, ?How do I know that? How sure am I of that?"

There is a useful checklist for some activities. So, there will be checklists such as, you know, understand the task, set a goal, think about any single task, make a decision about planning, monitor progress, stop, reflect, make corrections, evaluate. All those metacognitive skills. And you can teach children strategies for when they get stuck and have these as reminders around the classroom. ?If I get stuck on something, I don't know what to do. What should I do? I could stop and think about, is this like anything else I've done before? What other things could I do? Is there a resource bank I could use? Is it appropriate to ask somebody else or ask the teacher?" And so on. But to actually think through the options and then make a decision is important.

There's also a role, of course, for modelling good metacognition, as I just said. So, teachers engaging in think aloud when they're demonstrating a task and being explicit about when they change their mind or get something wrong. I think we've got to get away from the idea that getting things wrong is bad. So, one activity, some kind of a subject, might be looking at role models of people who got lots of things wrong before they achieved some great breakthrough. And there are lots of those in different fields.

But teachers, I think, also need to be facilitators. They need to step back, and I think this is hard for teachers, particularly with smaller children. But it's really important that they step back and allow children to try and to fail and then to try again. So they need to provide tasks that require metacognition. Project work is often very useful. And in my experience of working with many primary school teachers, they're endlessly inventive and creative when it comes to creating activities, as long as they know what it is they are trying to achieve. And so I would say that if you are a teacher, just getting to grips with some of the theory of metacognition and what it is you are trying to achieve. And then you can go off and create lots of lovely activities.

SPEAKER:
Does metacognition only develop through strong content knowledge, or can it be developed as a general skill?

SHIRLEY LARKIN:
Sometimes it's easier if there's a strong subject focus, because metacognition looks different depending on the subject. So, metacognition, in reading, might focus on things like comprehension, knowing whether you understand the words or not, as well as reflecting on your knowledge of strategies for decoding unfamiliar words. Knowing why you're reading, what your aim is, making judgements about how much time will be needed, how to read strategically to get the information you need, and so on.

So, conscious monitoring while you're reading and self-questioning. But that might be very different from, for instance, metacognition in certain math situations. If you're focusing on a more problem solving approach in mathematics, where the focus might be on understanding the problem and then reflecting on the rubric of steps that you might take to solve the problem. And maybe, perhaps comparing it much more to other types of problems that are very similar to that problem. So, having said that, the subject domains really do matter, because it looks different in different subjects.

But at some point, for me anyway, you'd want these metacognitive skills to transfer across all subjects. So, for me, developing good metacognitive skills in a math classroom, but not in the English classroom, would really be only half a job. And that's why I've always advocated like a whole school approach to developing metacognition in every subject. But you can develop some general metacognitive skills.

So, you can focus on things like study time and how to allocate resources. Thinking about, "Well, how do I study now? And how could I improve on that?" For instance, "How do you know you know enough to pass the test?" would be a question you might ask more generally, cause that will be common to all subject areas. And what could you do to improve your chances to pass an exam or a test, those kinds of things. So, some of them, those metacognitive skills are generic. Things like planning, predicting and so on. And just have a note of caution, really.

For me, what seemed really important is knowing why you're doing the skill. Why you are planning, why you are predicting, why you are evaluating, and whether or not you need to at any given time. I think that is metacognitive skilfulness. And that's where we want to get to. Not just having children by rote endlessly repeating metacognitive skills without any understanding of why they're doing it.

SPEAKER:
Thanks so much Shirley. And our last question, if you could go back in time and give advice to yourself as a school student, what would you tell yourself to focus on, to help you prepare for what was to come? And would you give different advice to students, today?

SHIRLEY LARKIN:
It's quite a clich�, I suppose. But it would be to focus on something that you love doing and not give up. The world is a really hard place, and it's easier to give up on something when it's not working out than to kind of stick at it for just a bit longer and see and something might happen. So, I would say, "Focus on something you love and don't give up," to myself. And I have given up on things. I've had lots of different careers.

So, to my own students I would say something similar but also to do one thing at once. I teach a lot of doctoral students. They seem to be busy, busy, busy all the time and I don't believe that our brains are great at multitasking. And I think rest is really important. I also think silence is really important. And I try to build silence into all the metacognitive projects that I do. Periods of silence in classrooms. And I think silence is a great way to reflect on your thinking and to develop metacognition. So, I would say, slow down, don't try to do so much. Do one thing and rest.

SPEAKER:
Thank you for listening to this episode of the Edspresso Series. You can find out more about the Education for a Changing World Initiative via the New South Wales Department of Education's website. There you can sign up to our mailing list. Or you can join our conversation on Twitter @education2040.