Sunday 13 May 2018

Learning Theory Pt 4: What is learning?

(This page is draft.  Please help us improve it by adding your comments (giving the paragraph number) at the bottom of this page.  The aim is to have a bare-bones introduction to the question 'what is learning?'.)

What is learning?

1  Looking from the outside

There are many definitions of learning, see these ten.
This one, by the authors of 'Make it Stick', is typical: “Acquiring knowledge and skills and having them readily available from memory so you can make sense of future problems and opportunities.”
That's what we see when we observe someone from the outside, but it doesn't explain how they do it.  To understand that we need to look inside the brain and see the actual physical process of making a long-term memory.

2  Looking from the inside

Learning in the brain happens when long-term memories are formed.  The links between brain-cells become permanently changed and the memory-pathway can be accessed at any time (so long as it is linked to prior-knowledge pathways)
"Cells that fire together, wire together."
The basic unit of the brain is the neuron, or brain-cell.  The process defined above implies a permanent  change in the brain with the new learning stored in an accessible way.  Looking at how the brain does this we see that the main change is not to the number of cells, but the connections they make.
Cells communicate with each other at chemical connections called synapses. Learning happens when some synapses are triggered into a new state - long-term potentiation.

3  Forming memories

Neurons:  brain cells

The way that the brain stores information turns out to be quite different from the way a computer stores it.  If you had a video camera and a bunch of hard drives, even the newest ones, you’d probably fill them up if you videoed your life over the weekend.  But the brain doesn’t ever get full, and the reason for that is that it stores the information in quite different ways. 


It is made up of brain cells; information is collected from other cells by the dendrites (left) and transmitted to others on the right.  When you're very young these cells have potentially 500 connections, so learning involves creating pathways which are preferred. Some of those 500 just fade away, but the preferred ones get stronger: and that’s what we call learning.


3.1  Chemical connectors - synapses

Cells are not directly connected, they have chemical-junctions called synapses. This diagram shows three brain cells joined together, (greatly simplified).  At the junction there’s a gap: a synapse.  The link is chemical with neuro-transmitters moving across the gap.   The reason why the brain has evolved (or was created) in this complicated way (rather than actually joining) is that it’s the way we learn. If you were born wired up, you wouldn’t be able to learn anything. At birth, a lot of the human brain is just sitting there, ready to be connected.

We can tell we have chemical links by observing what happens in our own brain.  If you wake up in the morning feeling a bit dull and you have some coffee, the caffeine boosts your neuro-transmitters and you wake up a bit.  Later on you're feeling wide awake and you have a few too many drinks, the alcohol suppresses your neuro-transmitters. Your speech may get slurry and you may lose your inhibitions (as the synapses from that area cease functioning!).

3.2  Long-term potentiation: the memory mechanism

This is the mechanism by which long-term memories are made. The synapses (gaps between brain cells) for the memory pathway become permanently changed, which means that the signal can flow easily along the path.

We can think about this in terms of gates. The signal arrives at the sending side of the synapse and causes neurotransmitters to be released. These neurotransmitters attach to the receiving side, triggering the signal in that cell.

There are two types of receptor on the receiving side: 
  • a 'small gate' which opens easily from the outside
  • a 'big gate' which can only be opened from the inside.
When the pathway is used repeatedly, the big gate is opened and remains permanently open. In other words, a long-term memory is formed.

This video below gives a technical explanation of this process.

This video gives a simplified version.



3.3  Short term memories

When we first start to learn something, we may appear to have formed long-term memories because we can answer questions later in the lesson.  However, these traces can fade completely if they are not backed up by repetitions to induce long-term potentiation.
These short-term memories are therefore just the early stages of long-term ones (not something different)

4  What are memories? 

Memories are not stored like they are on a computer. If you took 100 pictures all with blue in them and stored the images on your computer, the blue would be contained in every single picture.  However, we’ve got a part of our brain which detects different colours: blue, green, red. Every blue object in your memory will be connected to the blue area;  the computer is much less efficient than the brain (but much more accurate!).

This diagram shows the visual memory of a child who has a yellow cup (after seeing it several times). The small group of cells, represented by the yellow spot, is connected to the shapes that make up the cup, and to the colour of that cup. The ‘memory’ is simply a spot with connections, and so it takes a very small amount of space.   In a photograph you’d have all the information about the shape of that cup, but, in a brain-memory, you just have the links to the information.
Of course, no memory is only visual.  The child has physical sensations of holding and drinking and also an emotional response.  The full network of the memory has many components.



There isn’t one place in the brain for memories – they are simply links, so, for example, the memory for actions is in the same place as the action, while the names of things tends to be in the lower, left back part of the brain.  However, you have no sense yourself that this is the case!

4.1  Memory as a hierarchy

 You could say we have very basic shapes at the lowest level, and more complex objects are remembered just as the links to these basic shapes.  Clever eh!?


The spots in the ‘basic objects’ level in the diagram may represent your memories of table and chair, for example, and the spot in ‘complex objects’ could be your memory of ‘kitchen furniture’. So, everything is linked together in some sort of a hierarchy.
The knowledge in a student’s head (and in yours) is also hierarchical.  Sometimes students struggle for years.  Often the reason why, for example, they still can't do maths, is because the prior knowledge links were missing. The classroom research suggests that linking to prior knowledge is a good idea and a successful thing to do.   This means that the difference in students’ results between the teachers who link to prior knowledge and those who don’t, could be a whole grade.

4.2  Grandmother cells 

There is growing evidence that memory is structured as links. Researchers have found individual cells which respond only to one thing (for example: a picture of Jennifer Aniston on her own).  The term 'grandmother cell' is being used to label these cells - they are connecting point for all the components of a memory: the person, what they said; where you were, your emotional response etc



Further evidence comes from the observation that damage to certain small areas of the brain can result in the person being unable to make long-term memories.  This is not because the long-term potentiation has stopped working, but because 'grandmother' cells can no longer be formed.

4.3  Multi-sensory teaching

Because memories are networks of connections, the more the merrier!  If you only teach in words, your student will have mostly word-memories.  Adding some visual images (sometimes called 'dual-coding') adds an extra dimension.
Doing something practical or linking the learning to something exciting will also help.
However, as we will see in the section on Working Memory, it is important not to provide too much information at once.  Also, although doing something while looking at it and listening to the teacher's explanation is effective, expecting your students to read and listen at the same time is not.

4.4  Long-term memories are persistent.

Once a pathway is formed it is almost impossible to get rid of it.  memories you formed as a child, which you might not even think about for years and years, are sometimes immediately available years later.   People who have traumatic experiences would be really pleased to get rid of those memories!  
This can also tell us why people hang on to misconceptions and why it’s important to nip them in the bud as quickly as possible. (see the 'Feedback' section later on)

However, there is still sufficient space for us to carry on learning!

5  Forming long-term memories in the classroom

Sometimes we may feel that repetition might be a waste of time, because the students appear to already know it, because they’ve succeeded in the lesson. However, our own experience shows that students often seem to have forgotten what they appeared to have learned last lesson;  without repetition, the brain will almost ‘reset to zero’.  To secure the memory-pathway. you don’t re-teach the same material, the learners have to do the processing (which exercises the pathways).  It could be a plenary or questioning at the end of the lesson, it could be a recap or quiz at the beginning of the next lesson, or it could be a homework task which is repetition.
"Learning takes place through the active behaviour of the student: it is what the student does that they learn, not what the teachers does."

5.1 Performance v Learning

These terms are sometimes used to label the difference between the short-term memories in the lesson - where the student may appear to have learnt (because they can 'perform', but cannot perform a week later.  'Learning' requires repetitions.
Performance in the lesson is no guide to whether long-term learning has taken place.

5.2  Spaced v massed practice.

Normal teaching practice would be to teach your unit lesson by lesson, and the students apply the knowledge, and then you revise all the material at the end of the topic/unit: that’s called ‘massed practice’. By contrast, ‘spaced practice’ is: teach a bit, test, teach a bit more, test, review the bit you were taught first, retest etc. That works really well.
We know that building memories in the brain is completely different to computers, computers remember everything and they remember it first time. The brain has evolved, or was created, to forget everything that it only meets once.
You probably all have the classroom experience:  you teach something and the students even seem to understand it, and then next week they don’t even remember having done it. The memory seems to have been completely erased. Well, it’s not that it is erased, it was never formed.  
Now we know quite a lot about the memory forming process, we can make some 'rule-of-thumb' guidelines:

  • Teach the new learning (activating prior knowledge, presenting the material, setting a challenging task, and providing feedback).
  • Ensure at least three spaced repeats
  • Repetition 1 within 24 hours.
  • Repetition 2 within 3 days.
  • Repetition 3 within a week.

There needs to be a gap between the 1st activity and the 1st repeat so that a certain amount of 'forgetting' has taken place.  This can either be just a gap of time (the suggestion is at least 20 mins), or it can be the student doing something different for a while (see 'Interleaved Practice')
If first repeat is too soon, the brain effectively treats it as one visit. So if you did something at the beginning of the lesson and the students had to do something with it at the end; that would be a first repeat.  The first repeat is the most important because, if it was a few days before the next repeat, then the student may have completely forgotten.The brain is reset to zero, and those pathways are available for a completely different learning.
If we want to secure memories we have to structure the learning by giving our students the opportunity to process the same information on three separate occasions.  One at least 20 minutes apart, maybe for homework, certainly by the beginning of the next lesson, and then build in two more repeats before those pathways have faded.

5.2.1  ..but some of my students seem to learn without me doing any repeats!

The research suggests that all people learn in the same way. The reason why some people learn faster is because they actually do the repeats themselves, in a ‘mulling over’ process.  They tend to be students who are interested in the topic or have friends also interested.

6  What this theory can explain

This model (theory) of how memories are formed, and how they are links to existing memories, helps explain a great number of the things we see in classrooms:
  • students failing to learn due to lack of spaced repetition
  • students who 'can't do maths' (or another subject) due to lack of prior knowledge
  • students appearing to understand something one day, but to have completely forgotten by the following week
Using evidence-based methods, even without a theory, will improve the learning.  However, understanding this theory will help you choose the most appropriate method to use to solve the problem you see in your classroom and so significantly improve the learning of your students.  They can also understand why some of their students are failing to learn and be better equipped to help them.

7  In summary:

  • Learning involves making long-term memories.
  • Memories are links between neurons made by strengthening synapses.
  • The links require spaced repetition.
  • Memories are networks of links.  If prior knowledge is missing, memories cannot be formed.
  • Once formed, memories and habits are very hard to change.  Feedback is essential to ensure the right links are made.


Thursday 3 May 2018

Learning Theory Pt 3: The cognitive science basis of the theory

The cognitive science basis of a shared theory of learning.

1  Existing theories

There are already many theories of learning:  behaviourist, constructivist, Piagetian, cognitive load etc (see this list)  When a discipline has many theories it often indicates that the subject is not understood.  As the evidence mounts, patterns emerge and the number of possible ways to interpret it falls.
This process happened in medicine during the 19th century.  At one time there were a whole range of theories which competed to explain the origins of disease: imbalance of humors, possession by spirits, bad air, sins of a past life etc.  Each had their own experts.  As the evidence mounted in support of physical-reality, evidence-based explanations such as blood circulation, germ theory etc, gradually the old theories lost effect.  A similar process is now taking place in education.

2  The basis of a unified theory of learning

Learning happens in the brain, so, any explanation of the learning process has to be able to describe the learning process as changes to the brain.
Before the arrival of brain scanning techniques such as fMRI etc, educationalists had to make educated guesses about what was happening in the brain.  Now we know so much more, it is now possible to link learning with other knowledge about the brain discovered by the cognitive sciences.
If we use the common language of the brain and cognitive psychology, we will now be sharing our model with neuroscience-trained educators such as Speech and Language Therapists and other brain-based therapies such as mental health practitioners.  By joining the club of other evidence-based theories we gain respect and credibility.

2.1  Sources of evidence

"The potential for the neuroscience of learning to form a foundation for teacher training is one area that offers further possibilities."
Foreword by Baroness Susan Greenfield CBE in  Neuroscience for Teachers: Applying research evidence from brain science. Crown House Publishing. 
How the Brain Learns. David Sousa

3  The language of the theory

Examples




  • Neuron
  • Neurotransmitter
  • Synapse 
  • Pathway
  • Cortex, visual cortex, frontal lobe etc
  • Plasticity
  • Functional magnetic resonance imaging (fMRI)
  • Working memory, phonological loop, visual/spatial sketchpad
  • Cognitive load
  • Long-term memory
  • Long-term potentiation
  • Executive function

  • 4  Limitations of neuroscience

    It is not possible to take an insight from neuroscience and, from it, make recommendations for the classroom without extensive trials and evaluations.  This is the same as in medicine and other evidence-based professions: When a new drug is being developed, although a great deal of chemistry, biochemistry etc is known, it is still impossible to predict the effects of a new drug without extensive testing and trials.  However, this does not stop the use of biochemistry to explain the effects of the drug, once the evidence is available.
    The use of cognitive neuroscience to explain effective teaching does not mean that it can be used directly to give teachers advice on how to teach.
    Teachers should apply 'what works' using the evidence from experiments.  However, we can use the language and ideas of the cognitive sciences to explain how they work and to guide in us in sensible directions to develop new ideas and techniques that may work.

    5  Neuro-myths

    These are ideas which are common in education and are often presented with a brain-based explanation, but which have no evidence to support them.  Examples include 'left-brain: right-brain', learning styles, brain gym etc.
    Some people are concerned that using cognitive science explanations will make neuro-myths even more likely.
    However, there are plenty of myths in education which don't rely on pseudo-neuroscience explanations - and we can use very basic science to show why all the myths are myths.

    6  Level of description

    Clearly there is a fear that, by using the language and ideas of cognitive sciences, that the teacher will be overwhelmed and spend time learning complex facts and concepts which will never be useful in their classroom practice.
    However, all theories can be represented at different levels of detail and technical language.

    6.1  Example from surgery

    Any heart surgeon knows the arteries and veins which supply the heart muscle with blood/oxygen in a very high level of detail, yet they all probably started their understanding with a simple diagram (source) like this one while at secondary school.
    It is at the appropriate level of detail for the student to start their learning.  Later on they may start to use a diagram (source) like this one:
    As their skills develop, the surgeon uses more and more complex versions of the explanatory model.

    6.2  Appropriate theories for teachers

    A similar approach is available for the science of learning.  The level of detail required is one which allows teachers to make sense of what they see in their classrooms and can be taught during initial teacher education.  One example is this very simple model of memory:


    We start a draft in the following pages.

    7  Theory into practice

    Please don't judge the 'theory' on this page till you have seen some concrete examples.