I saw John Hattie speak recently at a conference on his latest re-imagining of his Visible Learning work. He was an excellent speaker and charming company. I was particularly flattered that he asked me to sign his copy of my What if… book. After he’d finished his presentation he asked me what I thought and I said I’d have to go away and have a think. This is an attempt to tease out a response.
Broadly, I found myself in agreement. Hattie makes the astute point that the 400 learning strategies identified in his most recent meta analysis cannot be directly compared; some are effective at some stages and for some reasons but not for others. This makes a great deal of sense and it fits well with the idea of the expertise reversal effect that finds discovery approaches tend to backfire with novices but start to become much effective with experts.
The model of learning that Hattie proposes is quite complicated:
It suggests that for learning to be effective it must take account of three different areas of input that he’s termed skill (prior achievement), will (habits of mind) and thrill (motivation). The model relies on Biggs’ idea of ‘learning processes’, which outline three approaches to learning: deep, surface and achieving.
An ‘achieving strategy’ aims to get the most bang for students’ buck: what is the maximum they can achieve with the minimum of effort? A ‘surface strategy’ to learning is one that aims “to reproduce information and learn the facts and ideas—with little recourse to seeing relations or connections between ideas”, whereas a ‘deep strategy’ that aims “to develop understanding and make sense of what they are learning, and create meaning and make ideas their own.” In this model it’s fairly clear that whilst achieving and surface strategies might have their uses, a deep strategy is the one we should all be rooting for. After all, who wouldn’t want students to make sense of the concepts they encounter and come up with their own ideas?
The idea of depth resurfaces in the ‘three phases of learning’: surface, deep and transfer. Hattie was very clear that the surface phase isn’t any less important than the deep phase, just that it must precede it. The model emphasises “the importance of both surface and deep learning and does not privilege one over the other, but rather insists that both are critical.” In order to represent these phases, he uses SOLO taxonomy; ‘unistructural’ and ‘multi-structural’ outcomes are superficial, ‘relational’ outcomes are deep, and ‘extended abstract’ outcomes are where transfer occurs. I’m not so sure about this. It turned out that Hattie had read my thoughts on SOLO and we managed to find time to have a short discussion, but essentially we left the matter unresolved.
Essentially, my problem is this: knowing and understanding are the same thing. We usually use the word ‘understanding’ to imply that a thing is known at greater depth but I think ‘understanding’ actually represents greater breadth: the more you know about a subject, the better you understand it. The fact that you can reverse that statement – the more you understand about a subject the better you know it – just highlights the meaninglessness of using the different words. They are essentially the same thing save that we want to imbue understanding with some mystical, ethereal quality. The application of Occam’s razor suggests that there is no need for such a quality when quantity of ‘mere’ knowledge will suffice.
The problem with depth is that it’s a misleading metaphor. It obfuscates the fact that the so-called ‘relational’ stage of learning (making links and connections between ideas) is something that occurs naturally without any effort on our part. When items of knowledge are stored in long-term memory our brains automatically organises them in relation to each other. If we know two ideas are related, we cannot help but connect them. As we know more about a subject, our schema grows and we become increasingly expert. Expertise is simply a function of practising the application of what we know. The more we know, the better we can think. The more we practice thinking, the more automatic basic applications become, freeing up cognitive resources for more complex applications.
This is where ‘transfer’ occurs. It’s well know that transferring ideas between contexts is much more straight forward for experts than it is for novices and, with that in mind, I suggest that rather than teaching for transfer, we’d be better off teaching for expertise. Anders Ericsson, who’s been researching expertise for the last three decades, suggests that we adopt a strategy of purposeful practice which Deans for Impact summarise in this handy graphic:
I should say, that (effect sizes aside) I don’t see much wrong with Hattie’s suggestions for the kinds of strategies teachers and students ought to implement at different stages. My quibble is with the model. If we replace surface→deep→transfer with novice→expert, everything becomes a lot more straightforward.
In my next post I unpick the specifics of research underpinning the model.
I read your account of Hattie’s model twice. It didn’t make much sense to me. However, the way you re-imagine it toward the end of this blog makes perfect sense and is a lot simpler to boot. Do we really need Skill Will and Thrill? Here’s hoping that doesn’t reach my principal who is sweet on Hattie….. or we might all be chanting it at staff meeting next week.
My account didn’t make sense, or the model didn’t? My account isn’t a complete summary so maybe there’s something important I’ve missed, sorry.
sorry, his model doesn’t make a lot of sense to me not your account.
For better or worse, ‘understanding’ now implies a ‘deeper’ acquaintance with a subject than merely ‘knowing’ about it. I’ve never been too happy about using the word ‘deep’, either. As someone from the Education Forum once remarked, it merely signifies the presence of hippies. The schema model provides a much clearer picture of how learning works; the more knowledge we have, the more new information is likely to make connections. Considering that this implies exponential growth, one can only weep when you consider that primary education has become a knowledge-free zone, with kiddies’ reading restricted to hungry caterpillars, friendly giants and the like.
[…] model of learning. I’ve already written about my concerns with the metaphor of depth here. In this post I want to explore what his meta analyses reveal about the best approaches to take […]
I think that there is something perfectly sensible about the idea that all is knowledge, and that it is incrementally accumulated. However, some questions which arise for me:
1) Much of Christendom learned to sing and chant in a language they didn’t understand, and then came the came the Reformation. What was the difference between the knowledge they had before, and what they had after?
2) Somebody knows full well that they shouldn’t “Drink and Drive”, but then they are involved in a drink drive accident which kills their best friend. Has this simply ‘added a little’ to their existing knowledge?
3) Where does this leave the notion of threshold concepts? Don’t they represent some.kind of qualitative transformation in knowledge?
1. That’s a classic example of not knowing enough 🙂
2. Knowing something in the abstract isn’t enough to change our behaviour. Often we need direct knowledge.
3. Not sure. Increasingly I’m feeling doubtful about their epistemic status.
Thanks David – interesting answers
David “Occam” Didau, you have rare gift of keeping it simple, but not too simple.
You are also a myth buster.
“Deep Learning” is one of my new despised “wank speak” terms.
I intensely dislike statements like “With so much content in the curriculum when have we got time to do “deep learning”.” When my response is, “By mastering all the content you have done the deep learning”, I am viewed as a heretic.
Deep learning is a big deal at the moment.
I will just still with “learning” and be done with it.
Trouble is, if you say the ‘ deep learning ‘ is just more content, then you need explicit specification of that content, otherwise the kids miss out on it! For example, many people still think that if you’ve told them the formula for the area of a circle and got them to work out 20 areas, then you’ve ” done the content “. I think this kind of ‘bad trad’ teaching may be what the ” deep ” thing is a reaction against? If you say that learning how to solve a variety of interesting problems using that knowledge (and combining it with other stuff and contexts that they are already familiar with) is just ‘ more content ‘ then ok, it’s more content…but most people would understand ‘ more content’ as ‘ now go to the next topic, do not pass Go, do not engage brain ‘.
The ‘deep’ thing is at least a way of justifying one’s lesson to the non-specialist SLT observer who says it is “lacking pace”….Sorry for ranting…
No, a useful rant. This is, I think, exactly where the novice-expert journey provides such a useful framework. Novices have ‘inflexible knowledge’ whereas experts’ knowledge is flexible.
Hi David.
Thanks for this great blog. Can I ask about timescales? To what extent can pupils move from surface to deep to transfer in the space of a lesson? I have seen many lessons where teachers have rushed through these stages just to “prove” progress. What do you think? Can it be done in a lesson?
I don’t think it’s really possible to move from surface to deep to transfer in the space of a lesson. Conceivably a pupil might experience an epiphany *during* a lesson bu that will be due to the steady accretion of knowledge over time.
[…] How helpful is Hattie and Donoghue’s model of learning? Part 1, by David […]
Hi David,
So we have Hattie’s Skill, thrill, will and Fullans Deep Learning. One could be forgiving fir thinking there is a real cash cow in terms of publishing and PL to re frame and package for dummies what we know quality t&l is?
I still don’t understand how expert’s knowledge is formed. From what i understood, expert has a vast and highly organized knowledge that allows them to spot pattern and easily recall. If we defined expert’s knowledge as such, the opposite of it would be inert knowledge, a knowledge that isn’t organized and that cannot be recalled. How do someone proceed to acquire knowledge in a field without transfer problem ? How could someone know his learning will led to an organized and functional knowledge ?
I think you are making an unhelpful assumption:
“If we defined expert’s knowledge as such, the opposite of it would be inert knowledge” – no, the opposite would be no knowledge.
Knowledge self-oraganises. Schema coalesce automatically as we learn more about a domain. We don’t have to know our learning will become more organized, it just happens.