S8 E11: Cognitive load theory: Four items at a time, with Greg Ashman

Greg Ashman—author of multiple books including A Little Guide for Teachers: Cognitive Load Theory, deputy principal, and professor—sits down with Susan Lambert on this episode to discuss cognitive load theory and how it applies to how students learn and how to best teach them. Together their conversation covers cognitive load theory, including an exploration of working memory and long-term memory; intrinsic load and extraneous load; biologically primary vs. biologically secondary knowledge; and how to apply these concepts in the classroom. Ashman also provides listeners with helpful advice on ensuring their teaching practices are based on evidence.

Show notes:

• Book:  “A Little Guide for Teachers: Cognitive Load Theory” by Greg Ashman
• Read: Greg Ahsman’s “Quick Insight Series”
• Subscribe: Greg Ashman’s Substack “Filling The Pail”
• Read: Barak Rosenshine’s “Principles of Instruction”

Quotes:
“I now know I shouldn't have felt guilty, but I also know that I could have taught that from the outset in a much more structured way where the students would have left understanding the concepts better without wasting time.” —Greg Ashman

“This idea that kids don't need to know anything anymore, they just need to practice skills is really quite a pernicious and damaging idea.” —Greg Ashman

“Think about the teaching methods that you're being presented with. Ask about the evidence and question whether this is really the optimal way of teaching literacy or whatever it is, or whether it's more based on wishful thinking.” —Greg Ashman

Episode timestamps*
2:00 Introduction: Who is Dr. Gregg Ashman
5:00 Feeling guilty about the way you had been teaching
7:00 Book talk: A Little Guide for Teachers on Cognitive Load Theory
8:00 Defining cognition
11:00 Working memory and long-term memory
13:00 Retrieval of long-term memory
15:00 What is cognitive load?
19:00 Working memory holds 4 items: What is an item?
24:00 Automaticity
26:00 Biologically primary vs biologically secondary knowledge
31:00 Mythbusting: “Long-term memory is like a computer system”
34:00 How can educators use cognitive load theory?
38:00 Explicit teaching 
42:00 Productive struggle and productive failure
49:00 Final advice
*Timestamps are approximate, rounded to nearest minute

Transcript

Greg Ashman: 0:31

This idea that kids don't need to know anything anymore, they just need to practice skills, is really quite a pernicious and damaging idea. They DO need lots of knowledge, in long-term memory, if they're gonna be able to think in a sophisticated way, and develop critical thinking skills. Because knowledge is what you think with.

Susan Lambert: 0:52

This is Susan Lambert. And welcome to Science of Reading: The Podcast, from Amplify, where the Science of Reading lives. "A Little Guide for Teachers: Cognitive Load Theory." That's the title of a recent book authored by today's guest, Dr. Greg Ashman. During this conversation, Dr. Ashman delves into topics from this fascinating book, including what Cognitive Load Theory is, the relationship between working memory and long-term memory, as well as the differences between intrinsic load and extraneous load. All of this will help explain how new knowledge is built. And all of it is very useful for thinking about how to help students most effectively in the classroom. But before we jump in, just a little bit of background about Dr . Greg Ashman. He's deputy principal at Ballarat Clarendon College in Victoria , Australia. He's also an honorary fellow at the Australian Centre for the Advancement of Literacy at Australian Catholic University. And a part-time professor at Academica University of Applied Sciences in Amsterdam. But right now, he's here on Science of Reading: The Podcast. We hope you enjoy. Well, Dr. Greg Ashman, thank you so much for joining us on today's episode.

Greg Ashman: 2:16

It's a pleasure.

Susan Lambert: 2:17

I would love if you could take a little bit of time to introduce yourself to our listeners, and give us a little bit about your background.

Greg Ashman: 2:25

I'm a teacher. I did a degree in natural sciences at Cambridge University and I went off to do a teaching qualification . That was meant to be a backup plan, but I am in the interim, I went to Uganda to do some teaching. Absolutely fell in love with teaching, realized that's what I wanted to do. So I spent about 13 years teaching in London. During that time, I met my wife, who is Australian, who was working in London. We had a couple of girls, and we decided that a better place to bring them up would be Australia. My wife is from a town about just over a hundred thousand people, outside Melbourne called Ballarat. So that's where we are now. I came to Ballarat. And I sent my resume to all the schools in Ballarat. Another local town, called Geelong, didn't get much interest from the government schools, which is where I'd always worked in London. And so I ended up picking up a job in an independent school, Ballarat Clarendon College. And there I fell in love with teaching for the second time. In the sense that I discovered research. It hadn't really figured prominently my career before. And I started reading about teaching and learning methods. And I picked up a book, called "Visible Learning" by John Hattie. I'm not convinced by Hattie's approach anymore, but that was a really influential book. It led me to a paper, called "Why Minimal Guidance During Instruction Does Not Work" by Paul Kirschner , John Sweller, and Richard Clark. And that was really influential. I'd started a blog. Started blogging about it. Made contact with John Sweller. And ended up doing a PhD in this field, called "Cognitive Load Theory," which is the field that John Sweller founded in the eighties. And during that time, I actually got ... maybe angry is too strong a word for it, but I was frustrated that I'd spent the early part of my career not knowing about all these ideas that I knew about now. I always thought that the best way to teach, 'cause this is what they essentially teach you when you're training, is what was known as a constructivist approach. Where kids figure things out for themselves. Teachers facilitate or guide at the side. I could never get that to work very well. So I always ended up having to teach things a bit more explicitly. But, I later learned that there were better, and worse, ways of teaching things explicitly. And I wish I'd known that at the time. I got quite frustrated about that. And that's fired my enthusiasm for blogging, and writing, and things like that. So that's a potted history, and I've probably rambled on too much about that.

Susan Lambert: 5:01

Oh , no, no, no. That's great! And I just wanna go back a little bit, 'cause I think in the pre-call, when we were chatting, you mentioned that when you were teaching this constructivist approach, and then tried to do things a little more explicitly, you felt a little guilty about that. Is that right? It's like, "Well, maybe I'm not doing this right. Why isn't this working?"

Greg Ashman: 5:21

Yeah, I did! I felt guilty. So say you are teaching science; the early part of my career I taught science. And say you want to teach students about how the surface area affects the rate of reaction in chemistry. And you do this classic experiment with marble chips of different sizes. You weigh them out, so they're the same mass, but they're different size chips, and you put them in acid. You collect the gas that's produced in a gas syringe. And you measure how quickly the gas syringe expands. The idea is that they do this experiment, and they figure out for themselves that the smaller these marble chips, the faster the rate of reaction. You can have a discussion then about surface area. But they never did. They never figured this . So they did the whole experiment beautifully. The data is staring them in the face. But they never seemed to figure out the relationship that they were supposed to figure out. And I now understand this is not surprising, 'cause I had too many things to think about. You know, "Where's the acid? How much acid do I need? How do I put this on this stand? How do I clamp the gas syringe in place ?" And I now realize that they're thinking about all those things, and that takes over. We did experiment, and they didn't get it. So, at the end, I'd sort of tell them the answer. And I felt guilty about that, because they hadn't discovered it, I suppose. And, I now know I shouldn't have felt guilty. But I also know that I could have taught that from the outset, in a much more structured way. Where the students would've left understanding the concepts better without wasting time. We could still have done the experiments. Experiments are really good for helping to reinforce concepts, but they're good for helping to REINFORCE concepts. Kids are not little scientists. There's an idea that what little kids need to do is what people who are professionals in that field need to do. 'Cause trained scientists figure things out by doing experiments. The feeling is that little kids need to figure things out by doing experiments. But little kids are novices. They're not trained scientists. Trained scientists have a vast amount of knowledge to draw upon when they're doing that. They know the results of similar experiments, and anything related. And they've got that huge knowledge base. Lots of stuff, we'd say, in long-term memory. So what novices need to learn something new is completely different to what experts do when they have all that knowledge.

Susan Lambert: 7:38

So you just gave a really practical example, from your own experience, a little bit about what Cognitive Load Theory is. So we're gonna unpack that a little bit. And I've got it right here, "A Little Guide for Teachers: Cognitive Load Theory." It's a brilliant book. Congratulations, by the way.

Greg Ashman: 7:55

Oh, thank you.

Susan Lambert: 7:56

I've been through it twice. And I'm going to get through it a third time. There were so many "Aha!" moments as I went through it. Before we actually jump into talking about some of that, I kind of wanna establish a couple of definitions, if you don't mind.

Greg Ashman: 8:10

Sure.

Susan Lambert: 8:11

Because I'm not sure that everybody understands cognition. If we can get the Greg Ashman definition of cognition. And then if we could get a little definition of learning, to sort of ground us in the conversation. So, can we start with cognition?

Greg Ashman: 8:27

Yeah. Cognition is all those little processes around learning. It's about thinking about things consciously. Moving them around in your head. With them eventually sort of transferring into what we might call long-term memory. We haven't talked about that model yet, but it's the process of thinking about ideas, moving them around, processing them. It's not a subconscious process.

Susan Lambert: 8:55

So it's the act of stuff that's happening in your brain.

Greg Ashman: 8:58

Yes! Absolutely. That's what it is.

Susan Lambert: 8:59

And that cognition then produces learning. Is that a good link?

Greg Ashman: 9:04

Yes. So Daniel T. Willingham , a professor at The University of Virginia, has a really pithy motto: "Memory is a residue of thought." So when you think about things, what's left over is a memory.

Susan Lambert: 9:18

Oh. So I would assume that that's learning, right?

Greg Ashman: 9:21

Well, yeah. So, this is quite controversial, 'cause people have lots of philosophical attitudes towards learning, and it kind of depends on definitions a bit. But, in Cognitive Load Theory, we would say, and this is based on Paul Kirschner's idea, that learning is a change in long-term memory. So that's how we define it. Now, people don't like that, 'cause they say , "Well, what about learning to play tennis?" for instance. How is that a memory? It's controversial, because people think of memory in quite a reduced and limited way. They think about memorizing facts. Or, you know, the date of a battle. And they think that's what memory is. But, to us in Cognitive Load Theory, memory is a much more expansive thing. So, the ability to hit a ball in a certain way is actually something that is stored in long-term memory. Because it's your long-term memory that guides the muscles to do certain things. So we would say learning is a changing, long-term memory. But that's in a very expansive way. It's not just the accumulation of facts.

Susan Lambert: 10:23

Yeah. That makes a lot of sense. And, as a matter of fact, I'm gonna read a quote from your book, so that we can sort of dive into this idea of the model. So you say in your book, "Cognitive Load Theory models the mind as consisting of working memory and long-term memory." Can you tell us a little bit about those two elements?

Greg Ashman: 10:44

Sure. So, first of all, it's really important to note that it's a model. And all science really progresses by producing simplified models of the world. And those models are good to the extent that they can make predictions about what's gonna happen. There are things that are not included in the Cognitive Load Theory model that definitely do exist. That's not a problem with the model . So, for instance, there's a thing called sensory memory, which we've got lots of evidence for. It's a buffer where you temporarily store information that you get from your eyes or from your ears. It doesn't really have any implications yet for Cognitive Load Theory, and the predictions the theory makes. So we don't include it in the model. But that's not a flaw with the model. You want to keep things as simple as you can to have this predictive path . There are two main components to the model that we use. One is working memory. Working memory is the thoughts you are conscious of having. It's what you know you are thinking about at any given time. And it's extremely limited. You can process about four items in working memory at a time. Now, what is an item? That's a really important question, and probably one of the central questions of Cognitive Load Theory. We'll set that aside for now. Long -term memory is almost the reverse. It's effectively limitless. It must be possible to fill up your long-term memory. I assume so, but no one has ever done it. So, it's an effectively limitless store of information. It doesn't necessarily mean you can easily retrieve things from long-term memory. It doesn't really fit within Cognitive Load Theory, but there's a whole science to retrieval, and making sure that things are easy to pull out of your long-term memory. But the two things interact with each other when we're learning anything new, academic knowledge. And we might talk about the distinction between biologically primary knowledge and biologically secondary knowledge. I'm talking about biologically secondary knowledge, academic knowledge, when we're learning something new. Academic knowledge that has to pass through working memory and into long-term memory. And so you come up against this kind of bottleneck where the working memory can only process four items at a time. So it can become easily overwhelmed. And if you overwhelm it, then you disrupt that process of passing from working memory into long-term memory.

Susan Lambert: 13:07

So once it's in long-term memory, outside of the model is the idea of retrieval?

Greg Ashman: 13:12

Yeah.

Susan Lambert: 13:12

OK. That makes sense. I'm getting older, so my retrieval is not as good , but that doesn't mean things aren't in my long-term memory. Is that right?

Greg Ashman: 13:21

Yeah. Most people now in the field think that you don't actually lose anything. It's all there. It's just about retrieving it. Even wrong ideas. So people used to think that if you had a misconception, say about physics, which is a classic one, that what you had to do is disrupt that and change that in long-term memory into the correct view. But that's not what a lot of people think now. They think that the misconception stays there forever. And what you have to do is give an alternative, which also goes into long-term memory. And then, over time, the alternative correct conception outcompetes the misconception. Becomes more useful to us. And so we retrieve that more easily, and more readily. But the misconception is still there. You don't lose anything.

Susan Lambert: 14:12

That's fascinating! It's also terrifying actually, as educators. It puts a lot of responsibility to ensure that what we're getting into long-term memory for helping our students is actually not misconception then. Is that fair?

Greg Ashman: 14:27

Yes. It turns out there's no rule. There's no sort of structure that insists that things in long-term memory are consistent with each other. So you can have opposite ideas coexisting at the same time. At that point, it's about retrieval.

Susan Lambert: 14:42

Wow. That, that is fascinating. And answers all kind of questions about all kind of topics for me. But we won't go down that rabbit hole. Alright , so we've talked a little bit about this model. Let's talk a little bit about load. So what exactly is load? I think you talk about two types of load in your book.

Greg Ashman: 15:02

Yeah . Load is the number of items you have to think about at any one time. And again, we're just pausing that key question, probably a big question about what an item is. And there are two sources of loads in a learning situation. One is intrinsic, and that's the stuff that you absolutely have to think about. You can't avoid it if you're gonna do this task. These are the things that you have to manipulate. Now, sometimes you can break a long task down into a smaller task. For instance, writing a paragraph. We might actually want to focus on the sentence level, because for a learner at a particular time, a whole paragraph might just be too much. Might overwhelm working memory. We might wanna start by focusing on sentences. We do have some agency over the intrinsic load, in the sense of what we ask people to do. It is inherent to the task. Extraneous load is load that's imposed by the learning situation. It doesn't need to be there. For instance , a good example would be you project a PowerPoint slide. And you've got some information that you want students to read. And you've got, I don't know , a jumping kangaroo at the side of the slide. That's something that students need to look at, process, maybe decide it's not relevant. But that's all taking a load in my experiment. All the stuff about measuring out the acid and all that, if what I wanted them to learn was the principle of surface area and rates of reaction, all of that stuff is extrinsic load . If I wanted them to learn about measuring volumes of acid and how to do it, then it would be intrinsic. So it sort of depends a little bit on what your task is. It's almost like the Zeroth Law, like it's an assumption that Cognitive Load Theory doesn't ever really state, but it's necessary. The theory is that we have a clear objective about what we want students to learn. You might think, "Well, of course we do. Surely we do." But I think sometimes we don't. I think sometimes we operate on a theory that if students are doing a literacy-type activity, they're engaging with books. Or they're doing some reading or something. Or that somehow, just by doing something literacy-based, they will learn something. A good example from my context is we used to have a little maths game that we played called Greedy Pig. And the teachers would go, "I've done Greedy Pig with my class today." And another teacher would say, "Oh, I better do that with my class. 'Cause I don't want 'em to miss out on doing Greedy Pig." It's a probability game. But no one could ever really say what they wanted students to learn as a result of playing that game. No one could articulate that. The theory was if we do something mathy, where kids are doing something with maths, that they will learn maths. So you can see that in order to be able to identify what is intrinsic load, you have to be very clear about what it is you want the students to learn.

Susan Lambert: 17:58

Oh, that makes so much sense. It makes so much sense about why this idea of clear learning objectives is so important. Not just to the teacher to craft the instructional opportunities, but for students to actually have learning opportunities related to that.

Greg Ashman: 18:16

Absolutely. You see administrators walk around all the classrooms at the start of the period, and check whether teachers have written the learning objectives on the board, and check whether the students are writing them down. All of that is performative. The only thing that matters really is that the teacher has an idea of what they're doing. And maybe has communicated that to the students as well, because that helps keep them on the bus. All the writing down, all the formulaic ways of constructing them, all of that is not necessary to the project.

Susan Lambert: 18:52

Extraneous load, maybe?

Greg Ashman: 18:55

Yeah, well, could well be.

Susan Lambert: 18:58

< laugh> All right . So you keep talking about this. What is an item? Let's go back and talk about what is an item. I can only hold four, is that right? Working memory is about four items.

Greg Ashman: 19:08

Yeah. OK. So a good way to explain this, which I used in the book is 3x = 18. So, 3x = 18 is a little bit of algebra. And if you look at 3x = 18, you can immediately identify four , maybe five, symbols. So the three, the x, the equals, and the 18, or is it a one and an eight? Well, again, if students are familiar, which they would be by the time you're doing algebra with them, with the Base-10 system and how to represent numbers, then they will know what 18 means. But much earlier on, when they aren't familiar with that, the 18 would be two items for them to process. 'Cause they haven't gotten that knowledge yet. So you start to get the idea that an item kind of depends on what you know already. If we go back to the 3x = 18, you say, "Well, OK, so there's four items, let's call it four." But there's not just four, because there are relationships between them as well. They're not just a list of four names to remember. They're in a relationship with each other. For instance, if you decided to divide the 3x by three, then we know that the equals means equivalent to. Well, that's another thing that students have to learn. So that's another item, the what that means. So we'd have to divide the right-hand side by three. You can see that for someone who is unfamiliar with these terms, the fact that the x represents an unknown number, a lot of people struggle with that concept initially. We've already overwhelmed the four items of working memory. But anyone that's done algebra to any level, when I said 3x = 18, probably less than a second later they would know that x is six. And so, for them, what we would call relative experts. That is important, the idea that it's about being a relative expert. I think sometimes when we talk about novices and experts, people think of kindergartners and PhD grads. It's not that. You can be a relative expert in something in grade 3. So, for relative experts, 3x = 18 is imposing no load on working memory. Why? Because you can draw on long-term memory. And, in fact, what people are doing when you tell 'em 3x = 18 and they go, "x is six," they're pulling that schema from long-term memory and solving it almost entirely in long-term memory. Almost unconsciously. They're not having to think through all the steps. So it doesn't impose any load. Something that would overwhelm the working memory with too many items for a novice, all those limitations disappear when you are pulling these schemas. These webs of connected ideas. Ideas from long-term memory. That tells you something else that's very important, which I think people miss. People who like to opine on education and they say, "Oh, we don't need students to know facts anymore, because they can look things up on the internet. Or they can use ChatGPT." These arguments are not new. They've been around. People used to say it about libraries, or television. But of course information that is out there cannot be used in the same way as a schema in long-term memory. When I say 3x = 18, and you immediately know x is six, you've processed all that unconsciously. In order to use information out there, you would have to go through a conscious, deliberative process of researching, which is much more taxing on working memory. So, those two kinds of information, information that's out there and information that's in long-term memory, are completely different. They're qualitatively different. And it really suggests to educators that this idea that kids don't need to know anything anymore, they just need to practice skills, is really quite a pernicious and damaging idea. They DO need lots of knowledge, in long-term memory, if they're gonna be able to think in a sophisticated way, and develop critical thinking skills. Because knowledge is what you think with.

Susan Lambert: 23:15

Yeah. I love this quote, and then I wanna go back and ask you another question, but this quote from your book, "So, if we want to enhance critical thinking, building knowledge in our long-term memory may be our best bet." And we're gonna unpack that a little bit. That's what you were just saying, right?

Greg Ashman: 23:30

Yes, absolutely.

Susan Lambert: 23:31

Yeah. And now, guess what? My working memory, just, shoot, that question that I was gonna ask you is GONE.

Greg Ashman: 23:38

<Laugh>.

Susan Lambert: 23:38

Oh! I know what it is! <Laugh>, I got it back. I caught it before it left. What you were describing, in terms of this 3x = 18, sounds a little bit to me, because I'm in early literacy, like what we would call automaticity with word recognition .

Greg Ashman: 23:53

Absolutely. That's exactly what it is.

Susan Lambert: 23:55

OK. So you're talking automaticity. And automaticity is important in all kind of ways then for people.

Greg Ashman: 24:01

Yes. Absolutely! So early readers initially, they're having to figure out, is this a "b," is it a "d?" How do these things blend together? We don't do that when we read, but it's not that we do it differently or we use a different set of skills. It's because all those things have become automated. We're just drawing on schemas in long-term memory that we're unaware of. The other thing to note about that is, because it's so effortless to pull things out of long-term memory and use them , we underestimate the value of that. You get this thing called the Curse of Knowledge. and this is something that teachers struggle with, because something is so obvious to the teacher that there's this empathy gap between the teacher and the student. 'Cause they can't imagine what it would be like not to know that. And so the on-ramp to a new idea is too steep, because they go through it too quickly. Because to them, the steps and the connections and the links are all obvious. It also leads to people , you've probably seen in articles by, I don't know , professional mathematicians, saying, "Oh, all this practice of times tables and drills. That's not what professional mathematicians do. We need to get kids doing investigations, and now really love maths." And, of course, the person doing that knows all their multiplication facts. They can solve simple linear equations very easily. And because they can do that, they kind of undervalued it. They don't realize the impact that knowing all that stuff has had on them, as an in individual, and the value of it. They can then make these general, sweeping claims about how that's not actually important. And I think we do suffer from that a lot in education. That people tend to want to talk about abstract nouns, and very highfalutin ideas. And, really, a lot of it is about the details. And about the little steps. About being able to tell a "b" from a "d." And we need to put our minds to the best ways of teaching those things to students. Yeah, the abstract stuff is fine, but it's those little details that really matter.

Susan Lambert: 26:12

You talked a little bit about biological primary and secondary. Can you unpack that for us? Because I think it's sort of related to how we help design instruction, but you're the expert here, so I wanna hear from you <laugh>.

Greg Ashman: 26:27

Absolutely. This is a slightly controversial idea, because it draws on evolutionary psychology, which is a field which is itself a little bit controversial. But I think the basic idea most people can accept fairly easily. A guy called David Geary came up with this distinction between biologically primary and biologically secondary knowledge. And the easiest example is literacy. Humans have been around for hundreds of thousands of years. Well, we dunno exactly, but anatomically modern humans, so humans that look like you and me have been around a hundred thousand years or so. And probably before that, they've been talking to each other. So, it's unlikely that language just suddenly appeared one day. It's likely that it grew from more rudimentary forms of language in ancestor species of humans. What that means is that our ability to speak and to listen to others has evolved with us. It's something that has taken place over an evolutionary time period. Presumably, humans who are better at speaking and listening were better able to pass on their genes to the next generation. It provided an evolutionary advantage. So we've evolved this capacity to speak and listen. No one takes little two-year-olds, sits 'em in classrooms, and stands at the front and says, "This is where you put your lips to make a "p" sound." No one does that, because we're primed to do that through evolution. What we say with biologically primary knowledge is that all the stuff I was talking about with items, and working memory, and all that, it doesn't apply. Because we don't consciously learn it. Long -term memory is already pre-primed to pick this knowledge up. Go straight in there without us noticing. We don't have to think about it. So, this model of the mind that Cognitive Load Theory puts forward, it's not relevant to biologically primary knowledge. What it is relevant to is biologically secondary knowledge. So, the related example is writing. The earliest writing we have ... well, it appeared separately in different human cultures ... but the ancient Sumerians, 5,000 years or so ago , they started trading with each other. And if I gave you two cows, I'd put two little stones, I'd cover it in clay so that I know you couldn't mess with it, and that's my record that I've given you two cows. And then people started putting impressions on the outside of the clay. And then they realized they didn't need to put the stones in the clay. They could just use the impressions. And then from the impressions developed a system of writing. So that's the earliest writing in Europe, the ancient Sumerians. But that was only a really short period of time ago, really. For most of that time, it's only been a very elite class who could read and write. If you look at mass literacy, that's only a couple of hundred years old in, say, western Europe. So even if being able to read and write produces an evolutionary advantage, there hasn't been enough time for us to have evolved to pick up that knowledge. So instead we have to process it in working memory. We call that biologically secondary knowledge. We're not primed to pick up that knowledge. It's not natural. So we have to process this in working memory. And Cognitive Load Theory applies to all this biologically secondary knowledge . If you think about it, school is various variations on reading and writing. Maths is a very specialist form of writing as well. The business of school is this biologically secondary knowledge. And you could say we invented school in order to pass on this biologically secondary knowledge. And because we are not pre-programmed to pick it up, unfortunately, to learn it is quite effortful. It takes a lot . And it looks different. And we do have to put kids in classrooms, and get 'em to practice it over and over and over again. Because it's not something that they'll easily pick up. When I thought about Cognitive Load Theory initially, I thought, "Well, why would the brain have this limit on it, of these four items that you can process in working memory? Why would we have evolved that way? I mean, that seems like a bit of a limitation. Surely if you were to design the mind, you wouldn't do that." But actually, it's really clever. Because there's a lot of information in the environment. And if it all just went straight into our long-term memory, like our ability to speak and listen did, our long-term memory would be rapidly filling up with all sorts of rubbish. So it works as a filter system, pointing out what's important to learn, what's the cultural knowledge that you really want to keep hold of. So that's briefly the difference between biologically primary and biologically secondary knowledge.

Susan Lambert: 31:16

Within your long-term memory, you say many people think of it as like a computer system, right? To index, and all this stuff. But it's really more like a web of information.

Greg Ashman: 31:30

Yes. One criticism of Cognitive Load Theory that I've seen is that it models a mind like a computer, but the mind is not a computer. Well, Cognitive Load Theory does not model the mind as a computer. So that criticism is a bit of a straw man. It's interesting. The early theories of working memory, like the Working Memory Model from Baddeley and Hitch in the 1970s, do take inspiration from computers. And they have this thing called a central executive, which is supposed to be controlling what the working memory pays attention to. And that is a problem. That's a real philosophical problem, because if the central executive is controlling what the working memory pays attention to, what's controlling the central executive? Well, in a computer, it's a computer program. But what is it in a human? It's this old philosophical problem of if you imagine the mind with a little person in there pulling all the levers, then what's controlling the little person? Do they have a mind and is there a little person there pulling all the levers? It's the ... what's that Disney movie where the little people are in the mind? Anyway, that's a bit irrelevant. But Cognitive Load Theory doesn't do that . Cognitive Load Theory says that there's no central executive determining what to pay attention to. Attention is directed, but through an interaction between long-term memory and the environment. So, a good example of this is if I said to you, "Don't think of a white bear." And you think of a white bear, even though I said don't think of a white bear. Because the stimulus from the environment, me saying that, has triggered a schema in long-term memory, probably for a polar bear or something. And involuntarily off you go. You're thinking of it. That's what's controlling what we pay attention to. Not some kind of central processor. And that removes that problem, that paradox, of, "Well, what's controlling the central processor?" Fundamentally, the mind is not like a computer. Cognitive Load Theory says the mind is not like a computer. But, of course, you don't have to accept Cognitive Load Theory if you don't want to.

Susan Lambert: 33:27

Well, but I'm pretty sure our listeners will, because this is a fascinating discussion. What you're talking about, again, makes me think of my role as a teacher. I was a teacher. I planned lessons. I planned instruction. I didn't know anything about Cognitive Load Theory. I'm too old to have used PowerPoint, so I didn't have any dancing elephants or dancing kangaroos. But I did probably bring in some extraneous load to my students that I didn't recognize. So, how then can educators use Cognitive Load Theory to both plan instruction and support student learning?

Greg Ashman: 34:07

Well, there's lots of things. So, if you are doing a PowerPoint, there's lots of little effects that Cognitive Load Theory has discovered that can help you design them to be better. One thing that a lot of people do, you go to any presentation that someone's giving, and they'll put up a slide with writing on it and they'll either read that out or they'll start paraphrasing what it says. Both of those things are a bad idea, because you're then putting your audience in a position of should they read the slide themselves? Should they listen to you? Is it the same? Is it different? And you can quickly overwhelm people's working memory. So, if you're gonna put up a slide with lots of information on it, give people time to read it. Then when they've read it, you can start paraphrasing it, if you want. That's a fairly trivial example. I think the broader principle, though, of Cognitive Load Theory is to keep in mind this idea of not overwhelming working memory. Break it down, and then break it down some more. One of the things I say to my math teachers, and I dunno whether there's a literacy equivalent of this, is when they're writing out the method for solving a maths problem, leave a space between each line. And then, when they're done, fill in the space with the step in between. Because there's always steps that they kind of miss. And I think that's a good discipline, because that helps you breach this empathy gap. Because you are a relative expert, it's hard to imagine what it is that the students don't know . So, break it down, and then break it down some more. Make sure the on-ramp is really shallow. Another example, again from math ... I apologize to keep going to math <laugh> ... but what I used to do is demonstrate to students how to do a particular type of problem, and then I'd give them a problem to do that's slightly different. Because I thought, "Yeah, well, you don't want to give them one that's exactly the same." Give them one that's exactly the same first. At least then you can see whether they've comprehended what you've done, you know? Make the on-ramp to the learning shallow, not steep, and keep checking, all the time, where the students are at. Then, you're not gonna overwhelm working memory. However, having said that, everything I've just said applies to things that are a bit complex, like learning to read, or write paragraphs, or do maths. If you do literally want kids to remember a list of words, and there might be occasions where you might want to do this, say if you're teaching French and you just wanna teach some French vocabulary words. That won't overwhelm working memory, because each single word is just one item. You can take them one at a time. That's just one thing to think about. And so, sometimes in those situations you might wanna intentionally increase the cognitive load. Maybe introduce a little bit of difficulty. Because the sweet spot is hitting this three or four items. And if you've got too few, you might not generate as much learning. The same applies to relative experts. If you've got kids who are pretty good at doing something, you don't wanna break that down and reteach it. You just want to get 'em to practice that thing. An example might be, you've got students who can construct sentences and you teach 'em paragraphs. What you don't want to do is when you teach 'em the paragraph, go back to the sentence level and repeat how to construct sentences. 'Cause they can do that. So, you would get them to do that bit themselves. So that they're practicing the thing that they can already do. But then, you are explicitly teaching the thing that they can't. You break that down into the smallest steps possible. People will be thinking, "Well, how would I know what it is that they can do and what it is that they can't?" I think Cognitive Load Theory implies a lot of formative assessment. So you're constantly checking what the kids can do. Because in order to break it down to a level that's appropriate for them, so that they're practicing the things they can already do, but you're breaking down into small steps of things they can't, you need to know what those are. If you came in to one of the classrooms at our school, you'd see kids constantly writing things on mini whiteboards, so that the teacher can check what it is they can and cannot do. And make those decisions.

Susan Lambert: 38:12

Fascinating. Now, at the top of our discussion, I think what you said is that "there's a good and bad way of explicit teaching."

Greg Ashman: 38:24

Yeah.

Susan Lambert: 38:24

You get the essence of what I'm asking?

Greg Ashman: 38:28

Yes. Cognitive Load Theory suggests that for novices, you want to teach things explicitly. For relative experts, they can do things that maybe look a little bit more like inquiry learning, solving problems, figuring things out for themselves. 'Cause they've got the knowledge and long-term memory to do that. But for novices learning something new, you want to explicitly teach it. And there's a big body of work on explicit teaching, stretching back to at least the 1960s, where researchers went into classrooms, looked at what the teacher was doing, recorded what the teacher was doing, and then tried to correlate that to the gains in learning the students made during being in that class. And they came up with a set of principles, which is neatly summarized in a document called Rosenshine's "Principles of Instruction." But various other researchers looked at the same evidence and came up with similar ideas. Thomas [Good] and Jere [Brophy] called it active teaching called it active teaching. It's the same thing. And , basically, explicit teaching is about breaking these things down into small bits, but it's a process. The definition of explicit teaching I use is that more concepts are fully explained and procedures are fully demonstrated before students are asked to apply those concepts or procedures. You could do that by giving a one-way lecture for an hour in a lecture hall, and that's probably not the most effective form of explicit teaching. And it's certainly not what comes out of this research that is summarized in Rosenshine's "Principles of Instruction." What you want to do is be very interactive. This, again, ties in with this idea of formative assessment. When I was teaching explicitly in a kind of default way, when I couldn't get constructivist teaching methods to work, what I would do is model three or four different problems for a period of 10–15 minutes. Then I would set the students some seat work to do on their own . That's not optimal, because by the time I've set them the seat work to do, they've forgotten my first couple of explanations. They'll be raising their hands and saying, "How'd you do this one?" And I'm getting frustrated thinking, well , I'll just show you one just like that. What you want to do instead is model something and immediately get the students to do it. And I use many whiteboards for this. And maybe then do another one. Yeah . And then maybe do another one. And gradually you're releasing control. So you might give a hint the first time you have to do it on their own and then gradually withdraw that. Or you might give a scaffold. So here's your first step, then your second step, then third step, do do the first step. OK. Do the second step. OK. Do the third step. Right . Good. Now the next one, do all three steps yourself and, and so on. And you gradually releasing it to the student and do all of that before you move on to the next problem type. What you're doing is building it in a much more robust way. But of course, that is highly interactive. And Rosenshine's "Principles of Instruction" also include things like daily and monthly review. This goes back to this idea of retrieval. You train your brain that something is important and needs to be retrieved by retrieving it. If you do the classic thing, the first term we're gonna do this topic, and then the second term we're gonna do a different topic, and we're not gonna return to this until the next year, then a lot of it won't be gone from long-term memory. But you won't be able to retrieve it very easily. What you want to do is space that out. When you're doing the next topic, have a starter question that's based on the previous one. And that sort of stuff. I would urge anyone that's not familiar with it to look at Rosenshine's "Principles of Instruction," because it is kinda the distilled wisdom of the craft knowledge of lots of effective teachers.

Susan Lambert: 41:59

That's fascinating. Thank you so much. And we'll see if we can find that and link our listeners in the show notes. Alright, next topic. What about productive failure? Or productive struggle? Isn't it good for kids to struggle through some of this kind of stuff? Doesn't it help them actually learn more?

Greg Ashman: 42:16

This is an interesting idea. Productive failure. Productive struggle. That's technically slightly different things. When I gave you that little saying of Willingham's earlier, "Memory is a residue of thought," this implies that you've gotta think hard about something in order to build these memories. And, to an extent, that's true. So I've said to you, there's about four items that we can process in working memory at a time. If we overwhelm working memory with too many items, then that leads to little learning. But also, if we underwhelm it, if we've only got one item to think about, that might not transfer into long-term memory. We want to operate at a sweet spot. If you've got things that students already know how to do, then you might want to deliberately introduce difficulty. There's this idea of "desirable difficulties" that's come out of the work of a psychologist called Robert Bjork. And they did a series of experiments where they'd give you word pairs . You might have a state and its capital. So, you might have Texas and Austin. And then you might have a different state, New York and Albany. And they give one group of people just the whole list of states and capitals. And then the others, they'd go, Texas "A" and then a line, and then the students had to fill in the answer themselves. They found that students who in the second condition, where they had to construct some of the response themselves, remember these state-capital pairs better than the students who were just given the list. And they call that a "desirable difficulty." And yes, that makes sense, because in terms of Cognitive Load Theory, just looking at the list is not much to process. You're deliberately increasing it a little bit. You get into this sweet spot where you've got a few items to process. I don't think there is a conflict, really, between that concept and Cognitive Load Theory. But you can see how it could confuse people, that they could get the message that you've always gotta introduce, deliberately, difficulties. Well, if you're teaching kids 3x = 18, and you're teaching them about that for the first time, you definitely do not want to introduce any "desirable difficulties," 'cause they're not desirable. You want to break it down as much as possible. Productive failure. Productive struggle. This is the idea that you want kids to really grapple with the concepts. Productive failure is what I did my PhD research in. And this is a very specific kind of productive struggle. What you do is you give students a problem to solve. And the classic example is, and I'm a bit off my territory here 'cause I dunno baseball but, you give a load of hitters. Are they hitters that hit the ball in baseball?

Susan Lambert: 44:53

Batters? The batters.

Greg Ashman: 44:54

Batters, OK. So, you give them lots of statistics for batters. And you then ask them who's the most consistent? And they try and solve it in various, different ways. And then when they've done that, you teach them the mathematical way of working out consistency, something called "standard deviation." And the idea is that that's better, because although they failed, productive failure, they failed to solve the problem in the traditional way. They're somehow more primed to now be taught the proper way to do it. And there's some experiments that have been done that seem to show this effect. And so it leads people to say, "OK, well before you teach students how to solve a problem, let them grapple with it and have a go at solving it themselves." I think that that's dangerous, because I think struggling with a problem is not intuitive, that would not be particularly motivating. It could lead to a lot of frustration, particularly for students who have, say in the area of maths, low math self-concept. So they don't think of themselves as particularly good mathematicians. So you then get them to struggle with maths. It's not gonna help. So I thought, "I'm gonna investigate this question. Because, to me, it doesn't make much sense that this idea of struggling with the concept would be particularly useful, because it's gonna overwhelm working memory." And I was a bit critical of the way the experiments had been designed that did seem to show this effect, because you could have different forms of explicit teaching that the practice kids did after they'd been been shown how to do it varied. So what I did is I did a series of experiments, and the problem that I was using was how to find the efficiency of light bulbs. We then moved on to different kinds of devices, but calculating efficiency was basically the idea. It's a science concept. What I did is I got all the students in a lecture theater, and I randomly assigned 'em to one of two conditions. The first condition, the students read a text that was assigned text, but unrelated to everything that we were doing. And they did that for 20 minutes. While they were doing that, the other group of students were given lots of data on light bulbs and they were asked which is the most energy-saving light bulb. And they went about trying to solve this in various ways. Often they'd go, "The one that gives out the most light , the most energy saving," and so on. Then I'd stop both groups after 20 minutes. And I'd teach all of them together. And, this is critical. By teaching them all together, they couldn't have gotten different instruction. Slightly. So I taught them all together about the concept of calculating efficiency. Demonstrated how to do it. It was interactive. They had little calculators and they had to hold up the calculators to show me the calculations they'd done. And then I flipped the condition. So, the ones that had solved problems now did the reading, and the ones that did the reading now solved problems. And I found that the students who had the explicit teaching before the problem-solving basically outperformed the students who had the problem-solving before the explicit teaching. That's consistent with Cognitive Load Theory, because calculating the efficiency of light bulbs is ... we'd say there's a lot going on there. There's lots of elements. High in element interactivity, we'd say. And so, in that context, the explicit instruction first was superior. So, this idea that it's always good to get students to struggle, I think it fails the common-sense test to some extent. Struggling is not necessarily a particularly motivating experience. But also, in terms of Cognitive Load Theory, if it's something that's complex, if it's not just learning a list of names, then you probably do want to actually explicitly teach it from the outset.

Susan Lambert: 48:26

<laugh> Very wise <laugh>. That's a fascinating description. I really appreciate that. And we could go on, and on, and on with many more concepts. But we are going to point our listeners to this great book, "A Little Guide for Teachers: Cognitive Load Theory." Before we leave though, I'm wondering, is there any more advice that you want to leave our listeners with? Or any more thoughts?

Greg Ashman: 48:53

Well, most of what people say about teaching methods is not grounded in evidence. And I would suggest that people should seek out the evidence. When people come into schools and talk about various teaching methods, I would ask for the evidence. It's a surprising thing to learn about a profession that a lot of these practices are just really based on what someone thinks might be a good idea, and not necessarily evidence. There's also a lot of how people wish the world WAS, rather than dealing with how the world IS. And there's a lot of that behind teaching methods. People think they want students to have agency. They want students to have an important role. They , they want the classroom to be centered around the students. They don't like the idea of a teacher as an authority, bestowing wisdom on the students. They want students to be democratic. And students to be active participants. And so, for all the right reasons philosophically, because they want kids to have agency and be able to do things themselves, they put them in situations that are not optimal for learning. They make them do investigations, and solve problems. And it's well intentioned , but we know from cognitive science basically that the students are not going to learn as much. So, think about the teaching methods that you are being presented with. Ask about the evidence. And question whether this is really the optimal way of teaching literacy, or whatever it is. Or whether it's more based on wishful thinking.

Susan Lambert: 50:30

That's great advice! And one of the reasons that this podcast is out there in the world, and I think that people are resonating with it, is because it's a way for them to actually get access to the evidence. Well, Dr. Ashman, it's been a fascinating conversation. I wish we could go longer , but again, we will point our listeners in the show notes to the resources we talked about, and I'm going to encourage them to go out and get this little book on Cognitive Load Theory, because it's just brilliant. So thank you again for joining us. We really appreciate it.

Greg Ashman: 51:01

Thank you, Susan. You're too kind.

Susan Lambert: 51:05

Thanks so much for listening to my conversation with Dr. Greg Ashman. Dr. Ashman is deputy principal at Ballarat Clarendon College in Victoria, Australia. He's also an honorary fellow at the Australian Center for the Advancement of Literacy at Australian Catholic University and a part-time professor at Academica University of Applied Sciences in Amsterdam. He's the author of "A Little Guide for Teachers: Cognitive Load Theory," as well as the Substack "Filling the Pail." We'll have links to those as well as other resources in the show notes. Share your thoughts on this episode in our Facebook discussion group, Science of Reading: The Community. Science of Reading: The Podcast is brought to you by Amplify. For more information on how Amplify leverages the Science of Reading, go to amplify.com/ckla. Next time on the podcast, we're closing out this knowledge-focused Season 8. Subscribe to Science of Reading: The Podcast to make sure you don't miss it. And if you've liked Season 8 so far, please consider rating us, and leaving us a review. Thank you again for listening.