Evidence-Informed Ideas Every Teacher Should Know About.

I love the idea of ‘evidence-informed wisdom’. I honestly can’t remember where I first encountered this but, essentially, it’s the idea that, as teachers we are faced with making hundreds of decisions a day – largely about how to question, how to motivate and how to adjust explanations, feedback,  and the pace and depth of learning sequences.  To do this well, our best hope is that our decisions are well-informed by wisdom – the wisdom gained over years of experience combined with ideas from our training and from our engagement with the body of knowledge that is out there about learning theories, our subjects and the general principles of good teaching practice.   The better-informed we are, the more wisdom we accrue and the better our decisions are likely to be.

There is a LOT to absorb but I’m finding that, currently, there are a few key ideas that resonate particularly strongly with me on my travels.  This is not an exhaustive list by any means – before you all tell me the things I’ve missed off – but it is a list that might help cut to the chase, especially if you are just getting started with research engagement and the world of cognitive science.

Rather than attribute each idea, I’ll admit that the true original sources are sometimes quite difficult to pin down to specific documents as people cite other citations and so on.  I’ve only done this when it is obvious to me and I’ll happily add further links if people point me in the right direction.   However, you will find nearly all of this referenced in the documents in my collection of research papers here:  Teaching and Learning Research Summaries: A collection for easy access.

From the papers in that collection, I might just highlight three key resources for anyone wanting to get stuck in:

  • The classic Barak Rosenshine paper ‘Principles of Instruction’
  • The superb book Why Don’t Students Like School by Daniel T Willingham
  • The excellent summary book Psychology by David Didau and Nick Rose.

So, here are some key evidence-informed ideas, I think every teacher should know about:

Learning is about knowledge in long-term memory, not immediate performance

  • Implication: Teach for the long-term; make it your explicit intention that whatever knowledge, skills and understanding (aka ‘knowledge’) you are teaching, students should be able to demonstrate that they’ve learned it at some point in the future – not just immediately.  This means you’ll need to check back later – with all of them. It means you also need to teach them how to remember things and practice things without leaving it to chance

Memory is strengthened by retrieval practice.

  • Implication:  Learning does not miraculously ‘sink in’.  It’s essential to use and to teach retrieval methods routinely: low stakes quizzing, key practice routines, mental rehearsal methods,  well designed assessment regimes that support the accumulation of knowledge over time – and strategies for revision that replace weak ideas like ‘going over your notes’ with self-quizzing and elaboration.

Learning builds in schemas;  knowledge allows you to accrue more knowledge; you can’t build a house without foundations.

  • Implications:  Firstly – you can’t have too much knowledge; it all helps; it’s all good. Secondly, you need to invest heavily in building secure foundations, checking students’ prior knowledge and activating it before giving them more. Weak schemas usually explain common misconceptions – so anticipate them and tackle them head on rather than waiting to see if they form.  Finally, you need to go back to the basics if that is where your students are at.

Direct instruction is most important with novice learners, especially those with weak prior knowledge and low confidence. 

  • Implication:  Take care to plan instructional sequences so that key ideas are explored very directly, deliberately and carefully, avoiding overload or under-guided learning that allows misconceptions or fixed mindsets to take root through repeated confusion.  Use lots of modelling and checking for understanding before moving to a guided practice phase; scaffold the support taking account of students’ emerging confidence with the material.  Deep-end problems need care – students are unlikely to work out complex ideas for themselves unless they have secure prior knowledge at the right level.

We remember what we think about – memory is the residue of thought. (Willingham)

  • Implication:  Plan lessons so that students will spend maximum time thinking about the specific learning at hand – not extraneous material, distracting product-making activities or excessive additional references or additional layers of complexity.  Eg if you want students to gain fluency with a word, they should use it, practice saying it, writing it, organising it amongst others – not, say, paint a picture that might be related to the word.  They will remember what they do and think about. You don’t want them to remember how to paint if confidence in using a word is what you’re after. Science theory and practical work each have an important role – but don’t do one if you want students to learn the other; they will learn the one they think about the most at any given time.

Screen Shot 2018-03-19 at 20.22.13.png

Responsive teaching – a two-way interactive process – is essential: more effective teachers ask more questions to more students, in more depth, checking for understanding, involving all learners. (Wiliam, Nuthall, Rosenshine) 

  • Implication:  As you can’t see learning, you need to be constantly seeking feedback to you from your students about the extent of their understanding.  This means asking lots of questions to multiple students, with probing exchanges to establish what they’ve learned and how well you’ve been teaching them.  One student’s response is never going to be enough – at least, that’s a huge and risky assumption. Avoid the classic pitfall of rhetorical questions like ‘is everyone ok with that?’

The optimal success rate for learning is high but not 100%: successful learning stems from early success; growth mindsets are reinforced by success through effort, not constant failure – and not a constant false confidence through under-challenge. 

  • Implication: Aim for 80% success rate. You need to adjust your teaching to build confidence if a student’s success rate is too low or increase the challenge if they are getting everything right.  If a student is struggling, getting bogged down, go back to getting them to practise things they can already do and then try to move on building on confidence, not simply battering away with things they can’t do.

Learning is most effective when cognitive load is optimized. 

  • Implication:  Plan learning sequences carefully so that concepts flow well, building steadily and cumulatively, allowing students to secure confidence through modelling and practice of each step rather than a whole sequence -eg with multi-stage problems in maths.   Also notice when students are beginning to develop some fluency so that you do not then overload them with extraneous support – let them use the expertise they’ve developed.  Guide practice, but not too much.  Also, strip out unnecessary layers of distraction – visual noise, actual noise, filler material.

Learning Styles are bunkum

  • Implication:  Don’t ever teach in a way that might help ‘visual learners’ or any such equivalent for kinaesthetic or auditory learners.  This is a false premise.  Teach everyone according to the needs of the material at hand, not the expressed preferences of learners, if they have them.  It’s a dead duck, false, wrong, misguided.  Call it out wherever you find it.

Bloom’s Taxonomy was never a triangle with ‘regurgitating facts’ at the bottom and ‘creativity’ at the top.

  • Implication:  Never teach in a way that relegates knowing things to the bottom of the pile, placing creativity and ‘synthesis’ at the top, or get overly bogged down in ideas about ‘higher order thinking skills’ as if they are separate from knowing things.  They aren’t.   Re-think your sense of Bloom’s taxonomy to view knowledge that is the foundation of all else – and knowing things for the sake of it is good. Because there is always a sake and knowing things never stifles creativity; one fuels the other.

Cooperative learning or group work can be extremely effective – but most often it is not because of how it is done. (Slavin)

  • Implication:  Don’t do group work for the sake of it or lazily without taking account of the conditions in which it is effective; group work is not inherently a good idea or a bad idea; it depends on what you want students to learn and how well you organise the groups.    In general, students only achieve well in groups if their collective success requires each individual to be successful (so they have to help and challenge each other) and they all have a clear role.  Without those stipulations, you are likely to get outcomes worse than if they worked individually.  Also stick to smaller groups (pairs and 3s) because it is increasingly hard to deliver on the group accountability and roles the more students are in the group.


As a follow-up I have now written a blog that  pulls some of this together  into what it might look like in a more integrated form:


I have also set up a youtube channel with a series of short videos exploring each of these ideas:  https://www.youtube.com/playlist?list=PL_WHYo5KULldZZ5UTWZEB3q265wO_T6Ko


  1. “Direct instruction is most important with novice learners, especially those with weak prior knowledge and low confidence. ”
    Babies manage just fine learning to walk by themselves mainly; some take longer than others, of course, I might debate your last point too, on the same basis of serendipity. Group work is vital (economically for UK plc) – not everyone can be on song so a bit of experimentation/repetition/… Finally is 11-a-side better than 2 or 3 ?


    • Whilst you are quite correct that humans can learn some things with no teaching required, like walking, talking, recognising faces etc. this is primary biological knowledge. We’ve evolved over millions of years to learn primary knowledge. Everything else on the other hand is secondary knowledge; reading, writing, maths – we can’t learn that without someone actually teaching us. If you are interested in this difference it is explored as part of the basis for Cognitive Load Theory.

      Liked by 2 people

      • Srinivasa Ramanujan FRS was an Indian mathematician who lived during the British Rule in India. Though he had almost no formal training in pure mathematics, he made substantial contributions to mathematical analysis, number theory, infinite series, and continued fractions, including solutions to mathematical problems …

        Is your theory of Cognitive Load culturally agnostic ?


      • Surely Ramanujan’s success as an autodidact has little relevance to his culture – there have been great autodidacts throughout history and throughout the world, but we cannot doubt that they are the exception rather than the rule. A class of teenagers without a teacher will not all become Fellows of the Royal Society, wherever they happen to come from.

        Liked by 1 person

      • Srinivasa Ramanujan actually had an excellent understanding of basic arithmetic before lodging with a couple of college student at around the age of 11. He increased his mathematical knowledge by reading a trigonometry book. He was able to do this because he had a very good grasp of English. I would say that before he started learning and solving problems he had built up a sound understanding of mathematics through direct teaching and practice. He was able to solve these problems because cognitive load was reduced and thus his working memory was able to cope with the rigours of new problems.

        Liked by 1 person

  2. Tom, I see from this post that you have reviewed a similar list of evidence-sources as the ones we have used for EBTN and come to similar conclusions.

    Would you be interested to collaborate on what you might call “a theory of learning”?

    Members of all other respected professions share an understanding of the process they are involved in (engineering, archeology ety) and knowledge of the evidence from which it is drawn.

    The contents of this article are the foundations of a ‘theory of learning’.

    In the absence of a shred theory, educationalists can say what they like (as happens now). This allows government to push the profession around because we are not speaking with one voice. Other professions are not pushed around because they are acknowledged as experts in their field and would anyway simply say ‘No’ if told by government to do something they knew to be stupid.

    Your recent summary of the new Chartered College magazine suggests that they too are in the same evidence-ball-park.



  3. Reblogged this on THE CHA WEEKLY READER and commented:
    Renowned edu-blogger Tom Sherrington summarises a few of the most well-evidenced and consequential findings of educational research and highlights their implications on the classroom practice of an individual teacher. Essential reading. ~5min

    Liked by 1 person

  4. […] Evidence-Informed Ideas Every Teacher Should Know About  (Tom Sherrington) I guess as a response to the problem that I (and many other teachers) set out about having time, Tom has produced a masterful summary of educational ideas, presented clearly and succinctly but with more references and links than you could shake a stick at. All the documents he references, and more, are gathered here for your convenience, just to add to your guilt for watching another episode of The Crown on Netflix whilst you could be reading Rosenshine. […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s