Schema-building: A blend of experiences and retrieval modes make for deep learning.

Learning is complicated so it can be useful to use conceptual models to help understand and discuss it. For me, a powerful concept is the idea that we organise ideas, knowledge, the things we learn, in patterns of connected information called schema.

Screenshot 2020-01-04 at 15.25.46
Image source BURAK ARIKAN


Schema-building is at the heart of this diagram by Oliver Caviglioli for my Rosenshine booklet:

brain image

It’s important to bear in mind that any model is a deliberate simplification – but there’s an assumption that people can and will broaden back out to real-world complexity.  For example this diagram was produced in the context of a book on instructional teaching; it’s not a universal model for all kinds of learning.  There is undoubtedly and unashamedly, an emphasis in the book on the kinds of learning that are difficult and require significant levels of teacher instruction; conceptual ideas that require deliberate, careful chunking, sequencing, modelling and practice.

However the ‘Learning’ label still includes all manner of learning modes:

  • listening to an explanation
  • reading books; studying diagrams and images
  • handling equipment or artefacts
  • witnessing an event or a demonstration
  • experience such as seeing a waterfall and perhaps feeling the spray on your face
  • making a circuit or a pot or solving a puzzle

Similarly ‘remembering’ is a label for a wide set of processes that involve drawing on our memory of things we know or have experienced;  it doesn’t only mean the simple act of conscious factual retrieval.   As Dan Willingham explains beautifully in his book, Why Don’t Students Like School, understanding is remembering in disguise.  (If your first-glance reaction to that is to be doubtful, I’d strongly suggest reading the book directly because the statement is not self-explanatory but, in DW’s hands, it makes perfect sense.). So ‘remembering’ includes ideas  about thinking and understanding.

With this in mind, I think it’s really important that when we talk about teaching by breaking things down into chunks or we discuss retrieval practice techniques, we always consider that information and ideas arrive and form in complex ways and should be assessed and evaluated in a variety of ways.  If we think too reductively in terms of simple knowledge inputs and simple knowledge outputs, we’re likely to be missing huge swathes of the warp and weft that enable us to form sound schema for the concepts we’re dealing with. Warp_and_weft.jpg

In addition to this, there are questions about developing students’ agency in relation to learning; helping them to determine their own learning goals and gaps as part of building metacognition and self-regulation. It’s not enough for teachers to own the knowledge checking process – always setting the quiz questions and framing the terms of how knowledge is encountered and tested. Ultimately students need to learn to do this themselves, generatively constructing and evaluating what they know and don’t know and learning to explore a knowledge domain with purpose.

Where students have knowledge gaps, there could be all kinds of underlying reasons:

  • gaps in experience – insufficient exposure to situations, phenomena, events, places so that abstract ideas can’t be rooted firmly in a foundational schema and are either not understood or are swiftly forgotten.
  • gaps in memory  (including gaps in understanding  – ie knowing how ideas connect): not having encountered knowledge before or  failing to absorb knowledge into memory,
  • gaps in confidence and fluency – where knowledge has been encountered but hasn’t been subjected to sufficiently intense or varied practice such that students are continually tentative or forgetful or prone to errors.
  • limited or faulty schema: where knowledge has been arranged in a schema that contains misconceptions, preventing students making sense of new information.

In nearly all the situations I encounter where students are struggling, it is possible to drill down to find that one or more of these reasons explains their problem.  In one way or another, it all boils down to them not knowing something they need to know.

However, experiential, tacit knowledge can’t be imparted: it’s about providing opportunities for it to be gained:

Screen Shot 2018-06-28 at 18.09.11

Formal, declarative, procedural knowledge usually requires sound instructional teaching, especially at a novice stage – but can be gained from reading and other forms of learning, followed up with lots of practice.  You can’t do too much practice.

Screen Shot 2018-06-28 at 18.08.55

But the whole is greater than the sum of the parts. There’s a schema-fabric that is woven from lots of parts, none of which are fully meaningful without the others.

Screen Shot 2018-06-28 at 18.08.40

If we’re going to succeed in addressing students learning gaps, our approaches have to have the richness that the material warrants.  Quizzing, for sure, can play a role in supporting fluency and recall of some factual knowledge.  But, if students:

  • are asked to practise explaining something verbally in pairs
  • have to manipulate a piece of apparatus using their knowledge in order to demonstrate a phenomenon or take some measurements
  • have to close their books and engage in a dialogue from memory using the language they’ve learned
  • have to produce a timeline or essay plan exploring how a character in a novel develops as the story progresses
  • engage in a paired elaborative interrogative question exchange, devising and exchanging questions and answers

…they could all be far richer retrieval practice experiences.  Remembering takes many forms. When they’re more complex, less easy to define and control, messier, harder to give precise feedback on.. they’re probably going to deepen learning more than a daily quiz produced by the teacher can for the very reason that they are more complex, linking more ideas together. For some students, an overemphasis on quizzing processes may continually misfire because their experiential platform for the knowledge just isn’t adequate or they can’t apply the knowledge in new contexts due to lack of practice.

The solutions lie in planning a curriculum rich in experience but with a strong instructional focus and sound formative assessment processes where the conditions of practice are varied, embracing simple and complex forms, and where students are learning to evaluate their knowledge for themselves.  It’s so important that we don’t use simplified models of learning to fuel reductive policies such as the absolute requirement for daily teacher-led quizzing and to remember that the underlying complexity is ever-present.

Related posts: 

The #1 problem/weakness in teaching and how to address it.

Tight-Loose and the Folly of Prescriptive Lesson Structures.

10 Techniques for Retrieval Practice


  1. This really “weaves in” with the #1 issue you discussed previously – helping every child that you teach to build effective schemas and constantly checking that they “get it”. Excellent.

    Liked by 1 person

  2. I think retrieval practice is falling victim to hype. The overemphasis on quizzing means that only substantive knowledge is being quizzed and that disciplinary knowledge is largely being ignored.


    • Steve Higgins (2018) suggests that we must take account of what he calls the ‘Bananarama Effect’-“It ain’t what you do it’s the way that you do it”; anything new needs practice. Research conducted by the EFF identifies some common pitfalls with retrieval practice but argues that, with some thought and practice, these can be avoided.
      I think there is sufficient research based evidence to indicate the efficacy of retrieval practice.

      Liked by 1 person

      • Certainly. I wasn’t implying retrieval practice isn’t effective, just that the hype of low stakes quizzes has meant it’s normally being focused on in a singular specific way. It’s good that people are focusing more on it though!


  3. […] How well can your students explain the water cycle? Do they understand the concepts, the sequence and the terminology? Can they mentally track water in its various states at a particle level?  Can they use the word ‘precipitation’ or ‘condensation’ in context, un-prompted?  If not, then it’s likely that more intensive practice with the words and concepts is needed – linking diagrams and models to words generatively (from memory) practising saying the words; using various quiz formats; rehearsing telling the whole story; explaining more scenarios where the water cycle comes into play.  (This is where the variation in schema-building comes in as per my last post: Schema-building: A blend of experiences and retrieval modes make for deep learning.) […]


Leave a Reply

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

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

Facebook photo

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

Connecting to %s