Less Load, More Learning: First Principles of Cognitive Load Theory (with John Sweller)

What’s the best way to choose how you’ll teach something so it actually sticks?

Design your next lesson so learners don’t just follow along—they understand, remember, and apply their new skills.

By grounding your instruction in Cognitive Load Theory, you’ll gain a practical compass for sequencing content, trimming unnecessary load, and accelerating real mastery.

Our guest, Dr. John Sweller, pioneered Cognitive Load Theory during more than four decades as Professor of Educational Psychology at the University of New South Wales. His research has reshaped classrooms, training programs, and learning technologies worldwide.

WHAT WE COVER IN THIS EPISODE

• Why learners often absorb less when they start by solving problems—and what to do instead
• The expertise‑reversal effect: why novices and experts need opposite instructional treatments
• How to recognize when learners look active but aren’t actually learning
• The modality, split‑attention, and redundancy effects—and how they guide interface and content design

Practical ways to balance intrinsic, extraneous, and germane load so learners stay challenged without being overwhelmed

Find The Design Psychologist on your favorite podcasting platforms (or share this link with a friend): https://designpsychologist.buzzsprout.com/2395044/follow

If this show’s been useful or thought-provoking for you, I’d love it if you would do me a quick favor and let the Apple audience know! I know it takes an extra step—but it really helps new listeners discover the show, and it makes a big difference for us as we grow.

Just open Apple Podcasts, search for The Design Psychologist, tap the show, scroll down to the bottom of the page, and hit “Write a Review.”

Transcript

Welcome to The Design Psychologist, the show that helps you use psychology to design
better experiences. I'm Thomas Watkins, your guide to becoming a more powerful
psychology -informed designer.
If you have to teach something, what's best for picking the way you're going to
teach a lesson. Most of us have taught things to people at some time or another.
We often figure it out as we go along, but can we do it more intentionally?
In other words, can we design the ideal lesson using the science of psychology?
Imagine you're trying to teach some skill to someone else, maybe working,
or creative writing, or how to build a spreadsheet,
or ballet dancing, or maybe playing chess, or fishing. Think about how you adjust
the level of what you're teaching, depending on what the person seems to need. You
might start off showing things you think they can handle. But if you're going too
slow, you say, "Oh, they've got that part already. Let me jump ahead. And if you're
going too fast, you suddenly realize, oh, I've got to take a step back and lay
some more foundation before I get to the advanced stuff. But what if it's a lesson
plan you had to put together? How do you pick what goes in there?
For example, how much text should you use? When should you use pictures or diagrams.
And how about practice questions or video material? And as you pick out these pieces
you're going to use, how do you assemble them and sequence them? What if we don't
have to rely on our gut instinct or trial and error every time we teach something?
What if there are scientific principles about teaching rooted in how we remember
things and how we attend to things. Principles that could help us structure our
lessons more effectively from the start. This is where the science of cognitive load
theory comes in. For decades, it's empowered many teachers across the globe to be
more effective at optimizing instruction. They do this by matching the learner's
capacity for learning something, and it helps them retain and build on what they
learned later. Today, I'm happy to share a conversation I had with John Sweller,
who I'll introduce momentarily. Together, we explore questions like, why do people
often learn less when solving problems, and what should we be doing instead? What is
the expertise reversal effect and why do experienced learners need the exact opposite
of what beginners need and what's going on when people look like they're learning
but aren't and what's the modality effect and how can principles like that and split
attention and redundancy help us design better lessons and better learning tools and
interfaces. By the end of this conversation, you'll see how understanding memory and
cognitive load can dramatically shift the way you teach or even how you communicate.
And you'll also walk away having taken the first steps toward designing experiences
that truly support real learning. So let's have a listen as I introduce and speak
with today's guest.
Welcome to The Design Psychologist. We've got an extraordinarily special guest today.
It's John Sweller, who is a cognitive scientist, and he's the one who's behind the
extremely influential cognitive load theory, often abbreviated as CLT.
This groundbreaking work by John in the late '80s in beyond laid a strong foundation
for teachers, educators, people putting together learning materials and really
understanding how working memory limitations affect learning and therefore how
instructional methods should be adapted to account for this. CLT has shaped the whole
world of education from lesson plans to educational media approaches and across all
different kinds of subject matter, math, science, and other areas, literature, music
to some degree, athletics. And these ideas in CLT,
cognitive load theory, are very special in that they've been heavily backed up in
the lab and very tested and also very applied. So it's one of those few areas that
it really reaches across that spectrum. He is emeritus professor of education
psychology at the University of New South Wales in Australia. And the book we'll be
focusing on today is Cognitive Load Theory that John Sweller co -authored along with
Paul Ayers and Slava Kalyuga.
And we'll be talking about those ideas, why they're important and how we should
think about applying them. John, welcome to the show.
I've spent my career educating teachers in teacher education and that's pretty much
been my career and of course I've carried out the research associated with cognitive
load theory during that time. The vast bulk almost all of that time has been spent
at the University of New South Wales. Now, when I'm learning about this stuff, I'm
seeing the areas that it's been applied to and I think we've probably advanced a
long time since before people were thinking about this. What was education or what
is education like when people aren't considering this cognitive load theory and
thinking along these lines? Where do folks tend to go wrong and do teachers tend to
go wrong when they're not thinking about this. Right. That's an important issue
because for the last, I guess, almost two generations,
to a substantial extent, I feel we have gone wrong.
And we've gone wrong because we've ignored what's known about human cognitive
architecture.
The important thing about human cognitive architecture is that it's only been,
I guess, relatively recently that it's all been put together sufficiently to have
real instructional consequences to really start telling us, look,
this is how we learn, how we think, how we solve problems, and given that this is
how we learn, think and solve problems, this is how we ought to design instruction.
And that design of instruction is substantially different to what's been going on
until relatively recently. Things started changing,
I guess, five, ten years ago, more or less around the world.
And those changes are substantial, they require educators and others who are concerned
with presenting information to act in a different manner to the way in which we've
been acting for about two generations. Now, when we talk about instructional design,
we're talking about, want to make sure we understand what all that word includes the
printed up materials that a teacher uses or the lesson plan or the methods that a
teacher employs while teaching a class. Is it all of that? A whole lot of it, yes.
Yes. What a teacher does. It's more than just a teacher for that matter.
I mean, obviously, The teacher is essential to this, but it's anybody who's
presenting information to other people, anybody at all. And of course,
the job of teachers, the major job of teachers is to provide information,
but they're not the only professionals who have to provide information. Right. And is
it schools, as you're saying, is anyone who provides information In terms of who's
caught on to this the most, is it largely school teachers in elementary,
up to high school, college? Is it spread evenly across the board? Yeah. Yeah.
It's educators in general, all the way from preschool to university.
Gotcha. Okay. So thinking about cognitive architecture, starting at the And your book,
Cognitive Load Theory, starts off by talking about categorizing knowledge. And you
know, I find this interesting because there's so many different ways, depending on
what a book is about, how knowledge will be categorized. You know, you've got
different systems, semantic knowledge versus episodic knowledge, or you've got, you
know, kind of thinking fast and slow, the different ways that we behave. And, you
know, as we've realized through studying cognition, there's different ways and
different types of things going on underneath the hood. You went with, you went with
a direction to biological, primary, biological, secondary. Could you talk about that a
little bit? Sure. That's a really important distinction that was put forward by David
Geary, who's a professor in the United States.
Probably the best way to think about that is in terms of learning to listen and
speak versus learning how to read and write. We all learn how to listen and speak
and
need to go to school to learn to listen and speak. Humans have been listening and
speaking since we became humans and it's an immensely complicated procedure that we
find very, very easy because we've evolved over countless generations to learn how to
listen and speak and that means it doesn't have to be taught and we don't teach
people how to listen and speak. To give you a very obvious and very simple example,
if I was to ask most people, how do you teach a person to organize their mouth,
their tongue, their lips, their breath, their voice, in order to speak.
How do you teach somebody to do that? Most of us would look somewhat startled, say,
"Well, I don't, you know, we don't teach that." And indeed, we don't have to teach
it. It's arguably the most complicated thing we do, but we don't there's no there's
no curriculum in any school which says okay this is how you organise how to speak.
We've evolved to do it and all you have to do is drop in a in a normal in an
abnormal society and you're going to learn to listen and speak. In the normal course
of events that's simply going to happen. Some people may have some minor deficiencies
and we have speech pathologists who are trained to fix up that. But for the vast
majority of us, we're not taught how to listen and speak. And listening is even
more complicated. I mean, all from a physical perspective, all anyone's hearing right
now is modulating sounds, nothing else. Of course,
we know there's much more than that. Nobody taught us specifically how to listen to
speech. And I'm talking about our native language now.
Second language is something else again. So that's an example of biologically primary
knowledge. Right. And the idea, right, is that we've evolved this out of evolutionary
psychology, that if we've had to do something for hundreds of thousands of years or
even millions, that that gets written and hard -coded deep within us,
but then there's things that we've only been doing for maybe a few hundred years,
and that hasn't made it down there. Exactly. And the contrast is learning to read
and write. Humans invented how to read and write about, I don't know,
about 5 ,000 years ago, relatively recently. But during almost all of that time,
the vast majority of humans never learnt to read and write. It became common about,
I don't know, about 150 years ago when mass education began. With that education,
almost nobody learns to read and write. I said earlier on that we tend to get
things wrong in our instructional procedures.
One of the reasons we got things wrong in our instructional procedures is We didn't
know about that distinction between biologically primary and biologically secondary
knowledge. Learning to read and write is biologically secondary. We invented,
we devised, we developed schools precisely in order to teach biologically secondary
knowledge and without educational establishments, The vast bulk of biologically
secondary knowledge just wouldn't be acquired. And we know that just by looking at
the history of reading and writing.
We need to be explicitly taught. And one of our problems was, a whole lot of
people, and some of them are still saying this, said, "Look how easy it is to
learn complex things outside of school and how hard it is to learn things inside
school. We're doing it all wrong in school, all we have to do is mimic what
happens outside of school and people will learn easily, effortlessly, automatically.
Just put them in the right environment and they'll pick things up. And there was a
concerted movement to try to make schools sort of mimic what goes on outside of
schools. And in a way that made sense, because if you don't know the distinction
between biologically primary and biologically secondary knowledge, it makes sense.
We do learn things outside of schools much more easily than within schools. But of
course that ignores the fact that schools were invented, devised,
developed precisely in order to teach things which, if they're not taught in schools,
tend not to be taught at all. And the result is, you know, you might get one or
two percent of the population knowing how to read and write, and the rest are
illiterate. Schools were designed to change that. And of course, this doesn't just
apply to reading and writing. I just used that as an example. It applies to every
single thing that's taught in schools. They're taught in schools precisely because
without schools, they're not going to be learned by the vast majority of the
population and we would end up with a society vastly different and indeed vastly
inferior to the sorts of societies we have around the world now. So that's the
distinction between biologically primary and biologically secondary knowledge. It's a
really important distinction and it's one that needs to be kept in mind. Don't try
to mimic in schools what happens outside of schools. People prefer to learn stuff
outside of schools, because the stuff they learn outside of schools is stuff that
we've
evolved to learn, they don't need to be taught. And this theory goes step by step
kind of building as it goes along to fully understand. So by the time you're in
the middle of the theory, you're really understanding what's going on. And we've got
this foundation of biologically primary, biologically secondary. And there's just
automatic things. No one has to teach us how to, you know, pick fruit from the
tree once you know what the fruit is. You know, you just, there's automatic survival
things. But then there's this higher up stuff that you just have to learn. And then
one of the areas where this becomes really relevant is when you're learning and
you're getting instructed, a teacher, educator is talking to you, your brain is
working, and your brain might be working on irrelevant stuff, irrelevant stuff,
and we've got to maximize that. So could you walk us through,
you know, just sort of the extrinsic and the different kinds of load, wherever you
want to start with cognitive architecture, and I'll jump in. - Okay, okay. Okay,
Let's start with how we acquire information.
There are two ways we can acquire information. The first way is we can work things
out for ourselves. And humans are good at that. One reason why humans have become
the dominant mammalian species on Earth is because we're pretty good at working
things out, problem solving in other words. So we can acquire biologically secondary
information by problem solving. It's a long,
slow, difficult process, but we're good at it, and we can acquire information that
way.
But there's a second way of acquiring information. And that pretty much involves
doing what you and I are doing right now, transmitting information between us.
We are better than any other species on earth at transmitting information between us.
We're extremely good at it. We're the only species on earth that can transmit
enormous amounts of information from one individual to another individual.
When we do it by speaking and listening or reading and writing, we can transmit
information between us. It's a vastly more efficient way of acquiring information than
working it out yourself. you can spend literally not exaggerating years attempting to
work something out that somebody can tell you in literally a few minutes or even a
few seconds. There's an immense difference. So the first point to note about the
acquisition of information is the best way to do it is to get it from other
people. Now, can you always get it from other people? Well, obviously not. Sometimes
you don't have access to other people or to books or internet or written information
that other people have produced. Sometimes you've got to work things out yourself and
then you've got to engage in problem solving and try to work things out yourself.
So we've got to engage in problem solving and as I said, humans are good at that.
But as a procedure, it's not a good way of obtaining information. You only engage
in problem solving when you can't get information from other people because that's
such an immensely efficient way of getting information.
It can be easy, it can be quick, we've evolved to be able to this.
And can I emphasise that these two ways of acquiring information, either through
problem solving or from other people, they're biologically primary skills. We were
acquiring biologically secondary information, but the skills we use to acquire the
information, they're biologically primary. So they're the two ways of acquiring
information. Now next step, once we have started acquiring information,
it needs to be processed. It needs to be put into a form which we can use and
the structure that's used to organize the information into a usable form It's called
working memory.
And when working memory is dealing with novel information,
information we previously didn't know, we previously didn't deal with, it's extremely
limited, very, very limited, to limitations. First limitation is in terms of capacity.
we can only acquire about seven items of information at any given time.
We can process, in other words, think about use, deal with in some way,
two, three, maybe four units of information, extremely limited in capacity.
Dealing with novel information. They'll talk about dealing with familiar information in
a few minutes. So that's the first limitation, limited capacity, but there's a second
limitation which is equally as severe. Working memory is sometimes called short term
memory and the reason it's called short term memory is that we can hold information
in working memory, novel information in working memory, for know more than about 18
seconds. After that, it disappears.
Now, we can keep it in working memory longer by rehearsing. In other words, if you
come across a telephone number, for example, which you're not familiar with and that
you need to dial through, you can keep it in working memory somewhat longer by
going over And you may notice you automatically do that,
you keep repeating it to yourself, but if you don't repeat it to yourself, you'll
still stay there, but only for about a few seconds, 18 seconds or so, then it's
gone.
So working memory is extremely limited in capacity and duration.
That's an important part of human cognitive architecture that anybody who's presenting
people with information needs to keep in mind. Now, what happens next?
Well, once information has been processed by working memory with all the limitations
of working memory that I've just discussed, it can then be transferred to long -term
memory. Long -term memory is a different structure to working memory and its most
obvious characteristics when compared with working memory is that the limitations of
working memory completely disappear. We assume, as a matter of I guess logic and
rationality, there must be limits to long -term memory but if there are limits we
don't know where they are. Long -term memory is absolutely enormous and of course it
it holds information for very long periods of time.
Completely different to working memory. Let's go to the last and really critical part
of human of architecture, which is the part I really need to emphasise.
Once information has been transferred to and held in long -term memory,
it can be transferred back into working memory in order to govern action that's
appropriate to the environment in which we find ourselves. And it's at that point,
it's almost miraculous, you have an enormous change in working memory.
I've spent a lot of time just a few minutes ago emphasizing the limitations of
working memory. Those limitations only deal with novel information that we haven't
previously dealt with. When Working memory deals with familial information,
information that's been stored in long -term memory. Those two limitations of duration
and capacity completely disappear. We are transformed. Once information has been stored
in long -term memory, we become different people. I'm not saying anything novel to
anybody when I say education is transformational. We all know that, but this explains
why we can do things, we can think about things in ways which are utterly
impossible until we've got that information in long -term memory.
Right, it feeds back. It feeds back into the things that you're learning now. I'd
like to throw an example out there if you were learning music, for example. The
first time as a novice, you're trying to figure out that if someone tried to teach
you how to play something, it's like, okay, the white keys and the black keys and
put my fingers here and there and it's too much. But as you train your long -term
memory, that's the one that sticks with you more. Now when you learn something new,
you're seeing much less, or your brain is juggling much less information,
new information. Yeah, yeah, yeah, that's right. The old information,
which may have consisted of enormous numbers of elements of information which working
memory had difficulty
And now that it's coming from long -term memory, all those elements are treated as a
single element. Let me give you an example.
If I showed you in written form the statement,
"The cat sat on the mat," okay, so I show you the written statement, "The cat sat
on the mat." Now, that written statement consists of a whole lot of straight and
curved lines with spaces and if I showed you that information for let's say two
seconds, then removed it and told you, look I've just shown you a whole lot of
shapes, can you reproduce them? Of course you can. A trivial task.
You may not do it with 100 % accuracy, there may be little bits that you get
wrong, but the vast bulk of it is correct. Now, if I showed somebody that same
statement, who had not learned the English alphabet, then the experiment again,
here's a whole lot of lines and curves and straight lines and shapes. Here they are
for two seconds, now reproduce them. How much of it would you get right? Almost
nothing. Almost nothing. And you know, as an obvious example,
if you can't read Chinese and you read the statement and you saw the statement the
cat sat on the mat in Chinese, you thought for two seconds and taken away and say
reproduce that how much would you reproduce almost nothing? That's what I mean when
I say we are transformed. I mean, that's a simple perceptual example, but This
applies to absolutely everything that you learn once once stuff has gone in the long
-term memory you're a different person and That's why that human cognitive architecture
is so important. The purpose of education is to provide people with information
that's to be held in their long -term memory and then they can do things.
One of the peculiarities, you know, we all have biases associated with particular
terms. If you go to most educators and say, "Is learning a good idea?" Everybody
will say enthusiastically, "Oh yeah, of course it is. That's the purpose of
education. Of course we want learning." If you go to the same person and say, "Is
memorising a good idea?" But best they'll pause and in many cases they say,
"Memorising?" No, no. We don't want people to memorize stuff we want them to
understand and you know things like that. No difference between memorizing and
learning. If you haven't memorized something, you haven't learned anything.
But we've got the term memorizing when applied in education has negative connotations.
The term learning and when applied to education, well that's all But they're the
same, can't just deal with it to me. But I guess that comes from the history of
where people look at some lesson plans that seem to only emphasize the rote
information versus things that go beyond that. But then people kind of over interpret
that and say, well, memorizing things is wrong. Like, well, no, you need, that's the
basics of getting something your head. Yeah, yeah. Let's be clear what the difference
is between 'wrote memorising' and 'memorising with understanding'. 'Wrote memorising' is
the same as memorising with understanding except you have even more 'wrote
memorising'. In other words, you connect A with B with C.
Once you make that connection, we say, 'Oh, you've now understood it. Well,
all that's happened is you've memorized the relation between A and B and C.
So you just keep on memorizing and you've understood it.
No, I mean, in your book, you spell out the studies and it builds and builds and
builds
It becomes more and more apparent how you apply this stuff. And one of the things
is intrinsic versus extrinsic cognitive load. And we're thinking in this situation of,
you know, an educator is teaching their administering information through their words
and through maybe pictures and other things. And there's only so much, as you've
pointed out with working memory, it's limited. So what is a person paying attention
to in any given moment and what are they actually getting in through this filter of
their perceptual and cognitive system because we're trying to get stuff to stay.
So what's the extrinsic versus intrinsic and how should we be thinking about that?
Okay. Okay.
that's a complicated issue. I'll try to explain it. Give a simple version.
Yeah. Yeah. Yeah. Um, we, we acquire elements of information.
Now, what, what's an element? An element of information is pretty much just something
that you need to learn. Some elements interact.
You can't learn A without learning B. You can't learn B without learning C.
The example I normally comes from, you know, just simple arithmetic.
If you have to learn that two plus three equals five, you can't understand
understand, that's an equation, 2 + 3 = 5. You can't understand that equation
without taking note of each of those individual elements.
You have to know what the 2 means, you have to know what the plus means, you have
to know what the 3 means, you have to know what the equals means, you have to
know what the 5 means. In other words, and you have to consider all of those
simultaneously. If you just understand the two or the three or the equal sign or
whatever, you can't understand it two plus three equals five. You can only understand
that once you hold all of those elements simultaneously in working memory.
So element interactivity, each of those is an element and they interact with each
other. You can't just learn one by itself and then the next one and then the next
one. Other things may have lots and lots of elements but those elements don't
interact. You can learn each one individually.
If you're learning a second language, as an adult for example, you can learn the
translation of the word cat independently of learning the translation of the word
dog. They don't interact. In other words, you've got a whole lot of vocabulary
elements and ultimately one way or the other, you have to learn what each one of
them means, what each one of them is, what the translation of each one is. Element
interactivity is low. So you can divide information into high element interactivity
and low element interactivity information. Now, if elemented activity is high,
you've got a working memory problem because you have to hold all of the 2 plus 3
equals 5, you have to hold all of that simultaneously in working memory.
And, you know, we said a few minutes ago, working memory when dealing with novel
information very limited. It's not limited for...
No, just to jump in there with like an analogy it'd be maybe like juggling and
It's how many objects are you handling at the same time? And if you haven't learned
fully and committed to memory what two is what three is what five is and all of
it these things and What plus is and then you have to remember oh, yeah adding
means you're putting them together Versus subtraction that's the one where you're
taking it apart and then equals is it so then there's So many things, but as we
have long -term memory, those condense into singular objects.
And so we're not juggling as many things as we learn more and more about an area.
- That's a very good analogy. Yeah, that's exactly it. A professional juggler who's
doing what he or she really knows how to do, it's all one action to them. Okay,
for
You and me, or well certainly for me, each one of those dozens of actions,
they're a separate action and I've got to think about each one of them. And
unfortunately gravity is getting in the way and the ball is falling down to the
ground while I'm trying to figure out what do I do now. And it's the same with
purely cognitive activities such as learning simple arithmetic but it's exactly the
same or the way up to higher studies in universities so the same procedure applies.
So that's elemented activity. Now there are two ways in element in reactivity can
vary. One way is the way I just described now. It can vary depending on the degree
of
natural interaction. That's called intrinsic cognitive load. It's intrinsic to the
material. Nothing you can do about that other than not learn that material.
If you're going to learn that material, you're going to learn how to add two plus
3 equals 5 and you're going to learn how to add. You have no choice but to look
at all of those elements. So that's intrinsic cognitive load and that's important.
We have to take that into consideration when teaching. Teaching somebody how to
memorise the symbols of the chemical periodic table or learning words of a second
language, it's not the same as learning how to put things together which have to go
together. But there's another way in which element interactivity can change.
And that depends on the way in which we teach. You can teach in a way which
increases element interactivity or you can teach in a way which decreases element
interactivity. And most, not all, but most of the cognitive load theory effects,
that's to say the instructional effects, they depend on that type of element
interactivity. So for example, you can teach people how to do simple arithmetic or
mathematics, or indeed any subject you care to mention by having them work things
out themselves. In other words you may recall when I was talking about human
cognitive architecture you can either work things out yourself or you can get
somebody to show you how to do it.
Element interactivity is decreased by showing people how to do something and That
leads to the major, one of the major cognitive low theory effects which is called
the worked example effect. You can show people how to solve problems by giving them
a worked example. Now that applies in every single area.
Obviously it applies in areas like mathematics. Instead of getting people to try to
work out a problem, you show them what the solution is and ask them, study that
solution. But in humanities areas, it applies equally as well. If you give somebody
an essay question, you're giving them a problem to solve. And sure,
when you're testing them, that's what you need to do. But if you want to teach
people how to write essays, the best way to do it is by giving them worked
examples. We've got lots of data demonstrating that. In other words,
the worked example in, say, humanities areas, in an area like history area,
like what were the causes of World War One or something like that, you don't get
somebody immediately to write an essay. You not only explain to them explicitly and
directly what the causes are of WWI, you also tell them,
"Okay, here's a question you'd like to have to consider at some stage if you're
doing an exam or something.
Here's the question, what are the causes of WWI? Here's an essay which lays out a
good way of answering that question. Here's another essay which also lays out the
answer to that question. What I'm saying here is that it doesn't matter what area
you're teaching or what area of information you're asking people to process,
you can always give people worked examples and it's better to,
in the first instance, study those worked examples. Now,
is that all you do? No, at some point, once you've studied worked examples, then
you've got to practice it yourself. - So isn't that an area that a lot of people
misunderstood, I believe, right? Because your research was showing that worked examples
worked better when people were trying to learn something, then I of a lot of people
said, "Okay, great. Let me just show how to do this in a diagram or some kind of
example." And then people will know how to do it. But I think a lot of folks
missed the next step, which is that they have to go through and actually do it as
a guided exercise. Yeah, yeah. And that's really,
really important. And it leads to another cognitive low theory effect called the
expertise reversal effect. What happens is, okay, right at the beginning when people
are just beginning to learn, you give, let's look at it the way we run an
experiment. You give one group of people
problems that they have to solve. You give another group of people exactly the same
problems and with the worked example when you tell them study those problem
solutions. And what you'll find is when you're dealing with novices who just started
to learn, the people who are given worked examples do
Now, people are getting better and better after a while as you start running the
same experiment with people who are a bit more knowledgeable. You find that
difference between worked examples and problem solving decreases. Worked examples are
still better, but they're not massively better as they were right at the beginning.
Then get people who are even better than again, and the difference has disappeared.
People giving worked examples and problems to solve, they don't differ. Get even more
knowledgeable people. Worked examples, they're just a nuisance at this point.
You still need to practice, but you need to practice solving the problems. So at
that point, people who solve problems are doing better than people who are studying
worked examples. Hence the expertise reversal effect as they become more expert, as
you become more expert, you switch from studying examples to actually doing it
yourself. And it's a strong effect and the reason that effect works is that There's
another effect called the redundancy effect. It's redundant to study worked examples
beyond a certain point. At the beginning, yeah, you need to study worked examples.
Once you get a reasonable amount of knowledge, you need to start practicing it
yourself because studying a worked example at that point is redundant. It's an
irrelevant activity. It's an activity you shouldn't be engaged in anymore. It's an
activity. The activity you should be engaged in is actually solving the problems,
making sure you really know how to solve the problems. There's another effect called
the testing effect, which is closely related here. One of the best ways of learning,
not only just studying the worked examples and studying the information, but then
testing yourself. Now, depending on the context, some students won't test themselves
and we tend not to because it can be a fairly difficult process to test yourself.
It's much easier to keep studying worked examples. Test yourself and you find out,
"Oh right, I thought I knew how to do this I don't really, I need to go back and
study this. If students don't test themselves then it's a good idea for their
instructors to organise their instruction in such a way that they give students lots
and lots of tests not just to tell the instructor what the students know or don't
know but to tell the students what they know and don't know. In other words, you
use the test in order to improve your learning, in order to tell you,
"Okay, I need to practice this. I need to look at this again. I need to think
about this again. This other thing I don't need to think about again because I
really know it well." That's the testing effect. We're going through some great
effects here and one of the great things for the audience about this book is that
the theory is set up in the earlier chapters of what's the architecture? How does
this stuff work? But then there's this set of effects and that's how cognitive
psychology kind of is. And in a practical example, you don't ever have one right
silver bullet answer to all things should be done this way. It always is about
balancing the effects. Like so For example, your John just mentioned the redundancy
effect and you might think, well, isn't redundancy good? Well, sometimes it is, but
there's a lot of times where they're both pieces of information are competing for
the same working memory resources, so they're kind of boxing each other out and
overloading us. And so it's really about understanding these effects And then picking
as an expert, when do you use and utilize what effect? Exactly.
Yes. Yes. And the redundancy effect is really a central effect. You find a lot of
people when they're giving talks, they'll talk and they'll simultaneously
on a slideshow, they'll put their words up and the audience,
we really can't listen and read at the same time. It's redundant.
And what all of the audience invariably do, they decide, "Oh look, I'm not going to
read that stuff, I'm just going to listen." Or alternatively, they're going to say,
"I'm going to read that stuff and I'm not going to listen." And of course, they're
trying Of course, you're trying to read this stuff and you got this guy in the
background speaking and interrupting you or you're trying to listen and there's this
written stuff keeps coming up and attracting your attention and you can't do both.
First is showing a picture or something. So if you were to show a picture, that
would be different, right? because words on the screen and words going into your
ear, it's both using your limited
audio working memory, right? But then if I show a picture, they can complement each
other potentially, and even that you have to think through. I can under some
circumstances, but only if the picture is really necessary. If you have a picture
which is just telling people exactly the same as what you're saying the same problem
arises but some pictures really important if you if you give somebody a diagram you
know think of a geometric diagram obviously you have to put it up on the screen
and you can you can refer to you can say Look, angle a, b, c equals angle x,
y, z for whatever reason, that's perfectly acceptable.
Although even that depends on the knowledge of the learner. If the learner is
somebody who knows enough to say, "I know a, b, c because x, y, z," so if you
show them that and put it up on the screen, That's just redundant. - Okay,
then let's, in the time we have left, let's punch through a few more of these
effects like modality effect relates to that. - Yeah, yeah. Look,
before I talk about the modality effect, let's talk about the split attention effect
because that really leads to the modality effect. Split attention effect occurs when,
if you've got a diagram and you've got some text and if you can't understand a
diagram without the text and if you can't understand the text without the diagram,
you have to organise them so that each line of the text makes it clear what part
of the diagram you're referring to. Let's go back to a geometry diagram. It's one
thing to say angle a, b, c equals angle x, y, z. But if you say that,
and the learner has to go to the diagram, they've got to hold that statement,
angle a, b, c equals x, y, z in working memory, then they've got to go to the
diagram and search where the hell is angle a, b, c? Oh, there it is, and where is
x, y, z? Oh, that's over there. Now, why am I looking at these two, oh okay,
they're supposed to be equal. You've got to switch backwards and forwards between the
two. And you can have a lot of, this is just geometry, but there's a lot of areas
where that's a, you know, if you give somebody a graph or something, the same sort
of thing applies. That's called split attention. You've got to split your attention
between both of them. You've got to remember what's been, what was written, and
you've got to remember what was in the diagram and you've got to put them together
mentally and what are you using to put them together mentally? That limited
processing working memory. You can't do it. Much easier if you either put the
statement in the right place on the diagram so that you can see exactly where angle
ABC and XYZ is or if you can't put it in the right place for physical reasons,
put an arrow which points to them so that people know which bits are being talked
about. Now in that situation, one way of getting around this is to put the diagram
up and instead of presenting the information, angle A,
B, C, X, Y, Z in written form, present it in oral form, say it instead of writing
it, because we have partially separate working memories for listening and for seeing.
And you can use, if you use both of them, You can, to some extent,
not a massive extent, but to some extent, you can increase your working memory
capacity because you can listen while you're looking at something.
They're partially separate, not completely separate, they're partially separate. So
that's the modality effect. You may be better off saying angle ABC equals XYZ
instead of writing it. That's the modality effect. Now it's easy to assume,
oh okay, that's what we ought to do with all diagrams and text. It isn't.
Unfortunately, sometimes for example, you may find that the spoken,
the language -based material is simply saying exactly the same as the diagrammatic
material. We found that experimentally when we looked at the flow of blood in the
lungs and body. You could say blood flows from the left ventricle of the heart into
the aorta and you can have a diagram of that with an arrow demonstrating that and
then the statement it doesn't help you you don't need the statement it's redundant
but if it's written there people have to read it because they don't know they don't
know it's redundant until after they've read it and while they're reading it they're
not learning they're not looking at the diagram which is the best way of teaching
where the blood flows. So you get a redundancy effect and if you have to
simultaneously consider the statement, written statement and the diagram,
unnecessarily consider both of them, they're saying exactly the same thing, it's
redundant. I might add that redundancy takes a variety of forms.
Some people like to listen to music while they're learning,
no evidence that that's actually effective. We feel as though it's effective, but it
actually isn't.
When you're presenting information to people, don't include a cartoon to attract their
attention. It will attract their attention, but the attention is attracted to the
cartoon. It's not attracted to what you want them to learn. It's redundant, in other
words. So the redundancy effect, it's a major effect. Be careful what information you
present to people who are trying to learn something. You know, don't forget the
limited working memory. Every bit of information you present has to be processed by
working memory. And if it's redundant, if it's
It's going to take away information from stuff that you really need people to be
attending to. A few of the other effects that I've got on my list. Just self
-explanation effect and imagination effect. Okay.
When you have learnt something, try to explain it to yourself.
That will help you learn. Try to imagine. In effect,
that's a imagination that's sort of like self -testing. You know, I've learned this.
Have I really learned it? Let me see if I can go through that process I've just
learned. Well, yeah, I can go through that process, so I should be okay.
Or, good heavens, I thought I knew that process. I don't know it. Try to imagine
what you've just learned. If you can imagine what you've just learned, if you can
go through the process in your head, you're probably okay, you've probably learned
it. - And it also means you don't necessarily need to look at it on a page anymore
once you've learned it. You now have the magical power of within your head,
you can
Generate images of it exactly and even more magical the more you do that the easier
it is to do and the better you've learned it. So you know it's magical in those
two ways yeah imagination is a good way of not only testing whether you have
learned something. But it improves what you've learned you've learned it even better
and better I But the ideal is that you have completely automated what you've
learned. Once you've automated what you've learned, you don't have to think about it
anymore. And what I mean when I say don't have to think about it, I mean you
don't have to use working memory that limited duration,
limited capacity structure, you don't have to use that anymore. It happens
automatically. It's like reading the cat sat on the mat. You don't have to look at
that and say, "Oh, gee, look at all those funny shapes. What does that mean?" You
know, in a split second, you've read the cat sat on the mat. It's automated.
It doesn't use up working memory. Yeah. One of my favorite effects to the guidance
fading effect, because that relates to the expertise reversal effect a little bit.
Exactly, yeah. Provide people with a lot of guidance initially,
then reduce it, reduce it, reduce it, and eventually, they still need to practice,
but they don't need your guidance anymore. That practice can come by themselves.
Absolutely. So, yes, We've been speaking with John Sweller, who is the author of
Cognitive Load Theory, the theory itself and the book, and highly recommended for
folks who, because there's a bunch of books that have been written about how to
apply the theory to different areas of teaching and education. But this book,
really good for giving yourself a strong understanding of what does the research
literature say, how do these effects actually work? And once you have that strong
foundation, then you're now armed with the ability to apply it to a lot of
different areas. Any final thoughts, John, about this theory and how folks should
apply it or think about it? Challenges? Thank you for organising this discussion.
Talking about experts who know what they're doing. You're an expert who knows what
he's doing. That was fantastic. Thank you very much. Excellent.
Thank you for joining me and thank you for this conversation.
So, to wrap things up, this conversation was a rare opportunity to hear directly
from one of the most impactful voices in the history of learning science.
Dr. John Sweller helps us understand why learning is hard and what we can do to
make it easier and more effective. We talked about how our working memory is limited
and how those limits must guide the way we present information and design learning
tasks, and structure learning environments. We explored the different types of
cognitive load and why separating the essential from the extraneous is one of the
most important things we can do as teachers and learning designers and instructional
designers. We also looked at the power of worked examples and schema formation and
the importance of reducing split attention. These are all tactics grounded in the way
the brain actually works. So the takeaway is this. Ultimately, cognitive load theory
gives us a kind of instructional compass. It helps us guide learners without
overwhelming them and to challenge them without losing them, and design in ways that
actually support growth and not just activity. Because when we teach in ways that
align with how people actually learn, we don't just improve outcomes,
we expand access. We give more people a real chance to understand,
to grow, and to thrive.