Saturday, 14 April 2018

The case for a 'theory of learning'.

Why we need a shared theory of learning.

“If we know how they learn,
we can see how to improve our teaching?”

Why Learning Theory?

 “A theory…is a model that accurately explains large groups of observations….and allow us to make definite predictions about future events”  Stephen Hawking

(We are using the term 'theory' in the scientific sense of something derived from the evidence, not in the everyday 'that's only a theory' sense.  Alternative words are:  Shared model, shared understanding, shared conceptual framework.)

Professionalising teaching

Teachers’ universal complaint is the endless flow of initiatives from government.  Within a few years, most of these either fade away, or are found to fail.   The level of “policy failure” is very high.

Why is it that, of all the professions, education is the one where government, journalists, parents and gurus all feel they are better-informed that teachers?  Why is it that engineers, midwives, archaeologists, meteorologists etc are all consulted for their expertise, while teachers are lectured?

What the other professions have, which teaching lacks, is that shared model which forms the foundation of their professional understanding.  Wherever in the world you study midwifery, using the evidence-based medical model, you will be taught roughly that same material.  

The precondition for a shared theory is wide agreement on the evidence.  In the past there has been little agreement, but more recently the level of agreement almost all those who look at the evidence about learning has increased dramatically.  All sorts of books, blogs and websites are offering very similar ideas.

What makes a good theory?  

Any theory is a simplification of reality.  Good theories:

  • explain a high percentage of observations
  • make predictions which can be tested
  • are not easily falsified

The role of neuroscience

Most observers agree that we are very far from the point where we can ask neuroscientists for advice on how to teach.  However, neuroscience - the study of the brain - can be used as the basis for an explanation of the evidence.

The evidence-sources we can use are:
  • classroom experiments
  • cognitive sciences

 Let's 'give it a go'.  

Would you agree with the following draft statements?  They are brain-based explanations for the evidence we see in learning.  (This is not meant to be comprehensive, just for starters!)

Principles and Implications

Learning happens when new connections are made between neurons in the brain of the learner.  For the connections to be long-term memories, the same pathway needs to be activated several times.  Unless this happens, the synapses gradually reset to their original state and no memory exists (forgetting).
- Spaced repetition is vital for this process.

For the new learning to be understood, it must link to existing, prior knowledge.
- If the prior knowledge is missing, the new learning may only be retained by rote.

Sense data enters the student’s brain all the time from many sources.
- The student needs to maintain attention on the learning material.

The information which is attended to enters Working Memory.
- The student’s WM is easily overloaded.

Accessing secure long-term memories uses less WM space.
- Mastery of the basics is essential for higher learning.

Repeated pathways create long-term memories whether they are correct or incorrect.
 - Feedback is needed during the learning process to avoid misconceptions.

The brain has a huge capacity to process visual material with little effort.
- Teaching materials should combine visual/images and words.

If you would be interested to help develop this model/shared understanding/theory, please contact me at

Other sources

The theory we build will not be our own invention!  The theory is simply the patterns we see in the evidence and several people have 'had a go'.  


“I’ve found an exception:  this disproves your theory.”  If this theory can explain the majority of the evidence - then it is good/useful. Simply finding exceptions does not invalidate the theory.  If a competing theory can explain more observations and make better predictions, then we can say our theory is defunct.

“I don’t believe in theory – I’m only interested in getting the job done”. Many people are sceptical of the need for theory: However, whenever we apply any teaching method or policy, we are basing these decisions, consciously or not, on a set of values and assumptions.  If we do not examine these underlying elements we will not find out if we agree with them, or whether there is any evidence that they are valid.

“Don’t you mean “Teaching Theory?”  We share the view that “There is no such thing as teaching, only learning”.  “Teaching” is the word we give to a number of different activities which may end in learning.  It is Learning Theory which informs teaching practice, not the other way around.

Draft Mike Bell EBTN April 2018

Saturday, 27 February 2016

Six steps to outstanding learning.

‘6-steps’ summarise the evidence

With evidence available from so many sources, it’s difficult to see the wood for the trees.  By grouping the effective methods into the six steps many teachers apply anyway, we can see they make more sense.
It’s not just PAR: ‘Present>Apply>Review’, it’s OPPARR

Orient>Prior knowledge>Present>

You can find the comparison Hattie, Marzano and EEF on the website and we are working on including a range of contributions from cognitive science.
Learning involves making new links between neurons in the brain.  If the brain hasn’t changed, learning hasn’t happened.
Do you use steps similar to these six?

Step 0: Setting the scene – ready to make links

Improving behaviour has a high effect-size.  Do you have clear rules, applied without too much discussion and backed up by senior staff?
Students who have a Growth Mindset achieve, on average, one grade higher than those with a Fixed Mindset. How does your school/college promote this?  Are you good at 'Not labelling students' ?

Step 1: Prior knowledge – existing links

1.1 Assessing prior knowledge
Since new learning and memories are built on what is already known, do you assess the prior knowledge of your students before the start of a new unit?  What methods do you use to do this?
1.2  Filling gaps in prior knowledge
Can you assess students before the start of the course to give time to fill in the gaps?  What do you do when a gap is identified? 
Effective ways to deal with missing prior knowledge include: early intervention, phonics, small group, one-to-one or peer tuition.  
There are several interventions designed to help slow readers catch up: reciprocal teaching, repeated reading, vocabulary and phonics all apply here.  The meanings of words need to be secure in long-term memory.

Step 2: Presenting new material – showing links

2.1: Linking to prior knowledge
Do you use Similes and analogies to create links between the new material and what the student already knows?
2.2: Not just words
Students can receive their first contact with new material either from your teaching, or by reading a book, watching a video or demonstration etc.    Of the four, reading (visual words) is the most problematic for some students.   Do you use Graphical or Tactile Methods?
2.3: Big-picture and fine-detail
Do you give students both levels?  Do you use Advance organisers give the big-picture at the start of a topic or  Summarising  to pull a big-picture from the detailed learning they have done.
2.4 Limits of working memory
Do you present new material in short chunks so working memory is not overloaded?  Do you keep instructions short for less able students?

Step 3: Setting a challenging task – creating links

3.1  Challenge
If the task is too easy, it will simply exercise prior knowledge.  If it is too hard, the student will fail.  In both cases, no learning can take place.  Do you set a challenging task: which the student can achieve with a bit of struggle and feedback.  
Do you think your tasks are hard enough?
3.2 Worked examples and modelling
Do you give your students model answers or worked examples so they know what a good answer looks like?
3.3 Goals
Are Goals and learning objectives clear so that students can focus on what matters?
3.4 Promoting thinking
While a few students can 'think things through' for themselves, most need help or training. Do you give students more complex tasks which link the material such as Hypothesis testing, problem solving, note-making or summarising or do you focus on simple factual recall)?
Do you use collaborative or cooperative methods so students have to articulate their thoughts and decide whether another opinion is better than their own? 

Step 4: Providing feedback – checking the links

4.1  Giving feedback
Feedback is essential to check that the learning is not mistaken (that the brain is making the right links).  It needs to happen during the process, not after it.  It occurs near the top of all three lists and should be considered essential.  Sometimes the term 'Assessment for learning' is used.
What kinds of feedback do you give? Verbal or written? By the teacher, by peers or the student themselves using assessment criteria or mark-schemes.  Does it include what is correct (the medal) and what needs improving (the mission)?
4.2  Making an improvement
If the student does not act on the feedback, little new learning takes place.  Do you require students to act on the feedback?

Step 5: Repetition – securing the links

Repetition is vital to secure long-term memories.  Do you use spaced practice? 
Mastery learning, repetition and homework (not at primary level) all give opportunities for repetition.  
Once new knowledge is understood, do you require sufficient practice to secure long term memories?

Back to Step 1

Assessment of the new learning then shows whether a secure foundation (Prior Knowledge) exists for the next part of the learning. 
Because long-term memories form over several weeks, learning needs to be in the form of overlapping cycles where the practice with recent material overlaps with new learning. 

Mastery learning:  a technique where students keep repeating a piece of vital learning until they achieve 80% in an assessment.  This is repetition combined with a recognition of the need for secure prior knowledge..

Saturday, 5 December 2015

Should we get students to learn by rote?

The question as to whether we should teach students to learn things by heart (by rote) continues to create debate in education.  What does the evidence show?

The importance of secure memories

We know for sure (from neuroscience, testing etc) that our brains have a very limited Working Memory capacity.  Although there is dispute about the number of items we can hold in our heads at the same time, the number 7 seems to be an average  while some students can hold 9 items.
Working memory is the space where we think about things, work things out, mull over holiday plans, do mental arithmetic etc.
If the sentence is too long, it does not all fit in working memory.

This diagram shows a 5-slot working memory (WM). The first thing to see is that complex instructions can overload working memory.  Those working with less-able students know they need to break long sentences into shorter ones. (Young children's books always have short sentences.)  So, how do we ever think about something more complex?

The answer is that long-term memories can be drawn into just one slot, thus freeing space for the rest of the thinking.

Here is a trivial example:  If you are asked to remember the left-hand list


 and you have these acronyms stored in long-term memory, you will be able to remember it much more easily than the right-hand list which has the same letters.

This is sometimes called 'chunking', but this simply means that the unit is stored so securely (and the meaning understood) that it can be drawn into one slot in WM.

In mathematics, this helps resolve the question as to whether we should teach students their 'times-tables'.  If, when faced with the problem 8 x7 (or verbally, 'eight sevens') the number 56 pops into your head without any thought, then more complex problems can be tackled than if WM slots are taken up with the process of working out that part of the problem.

There are, of course, potential problems with  rote learning.  If the student has no idea how to work out 8x7 (or other simple multiple) then the memory will be meaningless.  Also, if failure to remember compulsory times-tables leads the students to believe that they are 'no good at maths', then the effect can be negative.

Another problem with times-tables is that the students may need to chant the whole table until they arrive at the one they need. One solution is to teach number-bonds by rote rather than tables.   The table shows that there are only 30-40 which need to be recalled as there are repeats (8x7 is the same as 7x8) and some trivial ones (eg 10x8=80).










In language teaching we never debate whether we should teach the meaning of words by rote or not!  We simply show the child a cup and give them the word 'cup' and get them to repeat it until they are fluent.  We do not try to explain why it is called 'cup'.  Equally, when teaching French, we simply require the student to know that the French for 'fish' is 'poisson'.

The need for secure knowledge is better understood once we see the way memories are stored in the brain.  They are not like memories on a computer.  All the information about this page is stored in the file of this web-page.  However, in the brain, the memory of something is just a set of links to what is already known.

If we teach about the Pope to someone who knows nothing about it, we may link to their knowledge of king and church.  However, earlier in their life they would have learned about 'king' by linking it to perhaps father' and 'country'. However, once the students has learned about 'Pope', they do not have a separate memory of Pope (as your camera has a separate memory of the photos you have taken), it has a huge network of links.  When we later want to use 'Pope' in a more complex sentence like "The Pope left the Vatican and celebrated mass in the Cathedral.", we are relying on these networks for vatican, cathedral etc to be secure so we can understand the sentence in our Working Memory.

Summarising the evidence

  • Working memory is used for thinking and is limited
  • Secure memories are vital for thinking.
  • Understanding can only occur if the new knowledge is linked to existing knowledge.
  • Learning by rote things which are understood is valuable.
  • Learning simple facts by rote is not damaging. 

Thursday, 3 December 2015

Reduce the power of the Education Minister?

(This is a draft designed to be circulated to organisations and individuals in UK education.  Please suggest ways to improve it factually or persuasively.)

It's not the Minister's policy, but the Minister's power

UK Ministers of Education
Instead of campaigning against the latest directive from the Secretary of State for Education, we should campaign for the powers they have to be reduced.

For 30 years there has been a stream of directives from The Dept. for Education.  
  • SATs are on, and then off, then, maybe, on again.
  • Schools should be specialist, then academies
  • GCSEs should be modular, then linear.
  • A-levels are split into two parts, then re-united.
  • Schools can let students choose their GCSEs, then we get EBacc
Each time we get a new initiative, teachers (and unions) complain, campaign, sweat, get stressed, resign...

Every time the minister claims they are trying to help students' learning.  They are not evil people; they genuinely believe their own thoughts. All the evidence, campaigning and resigning that teachers do, simply reinforces their belief that they have to make the change because teachers are resistant to the 'obvious' benefits of the policy.

So, perhaps we should stop campaigning on individual issues and get to the nub of the problem.  The Secretary of State for Education has not always had these powers.  Many of them date from the 1988 Education Reform Act.

Could we agree on a political objective which would remove the powers and give them to an independent body (perhaps modelled on the Office for Budget responsibility? 

Here is an extract from a July 2014 NAHT document 'Owning what is ours: a manifesto for education'
"How we change the system is almost as important as what is being changed. Too often change is chaotic and hasty, which limits our ability to make it stick. The profession and government become distracted by conflict over principles rather than engaged in discussion around implementation. 
We propose an ‘office of educational responsibility’. This office will go beyond existing proposals for evidence collection into planning and managing a five-year reform programme. 
This programme would be agreed in advance and subject to rolling review. New proposals for change will need to be submitted to the office for analysis against three tests: evidence of impact, value for money and capacity to implement. It needs to be difficult for ministers to depart from the programme. A high profile chief education officer, coming from the profession, could lead the office. 
Politicians would set principles, policies and outcomes. The profession would determine methods and, subject to representation via the office, be able to implement defined and tested policy in good faith."

Brian Lightman, ASCL leader, suggests an independent body to make decisions about the school curriculum and called for the end of schools having to follow the personal "whims" of policymakers. (Mar 2015)

If a number of teaching organisations could agree what we wanted, perhaps we could persuade the present Sec. of State to close the door on future rapid change to existing policies by creating such a body and handing over her powers.

Although it should be modelled on the Office for Budget Responsibility, this body should not have a similar name.  The problem in education is not irresponsibility; it is the constant changes based on little or no evidence or experiment. 

Dreaming?  Why not?

Unless we can get these powers changed we will be stuck with a continuous round of 'initiatives' as new ministers make their mark on the job.

Mike Bell

Draft Dec 2015