Test to the Teach

making-good-progressWhen Daisy Christodoulou told us not to teach to the test, I assumed she was mainly concerned with teachers spending too much lesson time making sure children understood the intricacies of the mark scheme at the expense of the intricacies of the subject. Personally, I’ve never spent that much time on the intricacies of any mark scheme. I’ve been far too busy making sure children grasp the rudimentary basics of how tests work to have time spare for anything intricate.   For example, how important it is to actually read the question.  I spend whole lessons stressing ‘if the question says underline two words that mean the same as …., that means you underline TWO words. Not one word, not three words, not two phrases. TWO WORDS.    Or if the questions says ‘tick the best answer’ then,  and yes, I know this is tricky, the marker is looking to see if you can select the BEST answer from a selection which will have been deliberately chosen to include a couple that are half right. BUT NOT THE BEST. (I need to lie down in a darkened room just thinking about it).

But this is not Christodoulou’s primary concern.

Christodoulou’s primary concern is that the way we test warps how we teach. While she is well aware that the English education system’s mania for holding us accountable distorts past and present assessment systems into uselessness, her over-riding concern is one of teaching methodology.  She contrasts the direct teaching of generic skills (such as using inference for example) with a methodology that believes such skills are better taught indirectly through teaching a range of more basic constituent things first, and getting those solid.  This approach, she argues, creates the fertile soil in which  inferring (or problem solving or  communicating or critical thinking or whatever) can thrive. It is a sort of ‘look after the pennies and the pounds will look after themselves’ or (to vary the metaphor) a ‘rising tide raises all boats’ methodology. Let me try to explain…

I did not come easily to driving. Even steering – surely the easiest part of the business – came to me slowly, after much deliberate practice in ‘not hitting anything.’ If my instructor had been in the business of sharing learning objectives she would surely have told me that ‘today we are learning to not hit anything.’

Luckily for the other inhabitants of Hackney, she scaffolded my learning by only letting me behind the wheel once we were safely on the deserted network of roads down by the old peanut factory. The car was also dual control, so she pretty much covered the whole gears and clutch business whilst I concentrated hard on not hitting anything.  Occasionally she would lean across and yank the steering wheel too.  However, thanks to her formative feedback (screams, yanks, the occasional extempore prayer), I eventually mastered both gears and not-hitting-anything.  Only at that point did we actually go on any big roads or ‘ to play with the traffic’ as she put it.  My instructor did not believe that the best way to get me to improve my driving was by driving. Daisy Christodoulou would approve.

Actually there was this book we were meant to complete at the end of each lesson. Michelle (my instructor) mostly ignored this, but occasionally she would write something – maybe the British School of Motoring does book looks –  such as ‘improve clutch control,’ knowing full well the futility of this –  if I  actually knew how to control a clutch I bloody well would. She assessed that what I needed was lots and lots of safe practice of clutch control with nothing else to focus on. So most lessons (early on anyway) were spent well away from other traffic, trying to change gears without stalling, jumping or screeching, with in-the-moment verbal feedback guiding me. And slowly I got better. If Michelle had had to account for my progress towards passing my driving test, she would have been in trouble. Whole areas  of the curriculum such as overtaking, turning right at a junction and keeping the correct distance between vehicles were not even attempted until after many months of lessons had taken place. Since we did not do (until right near the very end) mock versions of the driving test, she was not able to show her managers a nice linear graph showing what percentage of the test I had, and had not yet mastered.  I would not have been ‘on track’. Did Michelle adapt my learning to fit in with these assessments?  Of course not!  She stuck with clutch control until I’d really got it and left ‘real driving’ to the future- even though this made it look like I was (literally) going nowhere, fast.  Instead Michelle just kept on making sure I mastered all the basics and gradually added in other elements as she thought I was ready for them.  In the end, with the exception of parallel parking, I could do everything just about well enough. I passed on my third occasion.

I hope this extended metaphor helps explain Christodoulou’s critique of teaching and assessment practices in England today. Christodoulou’s book ‘Making Good Progress?’ explores why it is that the assessment revolution failed to transform English education. After all, the approach was rooted in solid research and was embraced by both government and the profession. What could possibly go wrong?

One thing that went wrong, explains Christodoulou, is that instead of  teachers ‘using evidence of student learning to adapt…teaching…to meet student needs’[1], teachers adapted their teaching to meet the needs of their (summative) assessments. Instead of assessment for learning we got learning for assessment.

Obviously assessments don’t actually have needs themselves. But the consumers of assessment – and I use the word advisedly –  do.  There exist among us voracious and insatiable accountability monsters, who need feeding at regular intervals with copious bucketfuls of freshly churned data.  Imagine the British School of Motoring held pupil progress meetings with their instructors. Michelle might have felt vulnerable that her pupil was stuck at such an early stage and have looked at the driving curriculum and seen if there were some quick wins she could get ticked off before the next data drop.  Preferably anything that doesn’t require you to drive smoothly in a straight line…signalling for example.

But this wasn’t even the main thing that went wrong. Or rather, something was already wrong, that no amount of AfL could put right. We were trying to teach skills like inference directly, when, in fact, these, so Christodoulou argues, are best learnt more indirectly by learning other things first. Instead of learning to read books by reading books, one should start with  technical details like phonics. Instead of starting with maths problem solving, one should learn some basic number facts. Christodoulou describes how what is deliberately practised – the technical detail –  may look very different from the final skill in its full glory. Phonics practice isn’t the same as reading a book.  Learning dates off by heart is not the same as writing a history essay.  Yet the former is necessary, if not sufficient basis for the latter. To use my driving metaphor, practising an emergency stop on a deserted road at 10mph when you know it’s coming is very, very different from actually having to screech to a stop from 40mph on a rainy day in real life, when a child runs out across the road. Yet the former helped you negotiate the latter.

The driving test has two main parts; technical control of the vehicle and behaviour in traffic (a.k.a. playing with the traffic). It is abundantly clear that to play with the traffic safely, the learner must have mastered a certain amount of technical control of the vehicle first. Imagine Michelle had adopted the generic  driving skill approach and assumed  these technical matters could be picked up en route, in the course of generally driving about,  and assumed that I could negotiate left and right turns at the same time as maintaining control of the vehicle. When I repeatedly stall, because the concentration it take to both brake and steer distracts me from concentrating on changing gears to match this slower speed, Michelle tells me that I did not change down quickly enough, which I find incredibly frustrating because I know I’ve got a gears problems, and it is my gears problem I need help with. But what I don’t get with the generic skill approach is time to practice changing gears up and down as a discrete skill. That would be frowned on as being ‘decontextualised’. I might protest that I’d feel a lot safer doing a bit of decontextualized practice right now – but drill and practice  is frowned upon – isn’t real driving after all – and in the actual test I am going to have to change gears and steer and brake all at the same time (and not hit anything) so better get used to it now.

Christodoulou argues that the direct teaching of generic skills  leads to the kind of assessment practice that puts the cart, if not before the horse, then parallel with it. Under this approach, if you want the final fruit of a course of study to be an essay on the causes of the First World War, the route map to this end point will punctuated with ‘mini-me’ variations of this final goal; shorter versions of the essay perhaps. These shorter versions are then used by the teacher formatively, to give the learner feedback about the relative strengths and weaknesses of these preliminary attempts. All the learner then has to do, in theory, is marshal all this feedback together, address any shortcomings whilst retaining, and possibly augmenting, any strengths. However, this often leaves the learner none the wiser about precisely how to address their shortcomings.  Advice to ‘be more systematic’ is only useful if you understand what being systematic means in practice, and if you already know that, you probably would have done so in the first place.[2]

It is the assessment of progress through  interim assessments that strongly resemble the final exam that Christodoulou means by teaching to the test.  Not because students shouldn’t know what  format an exam is going to take and have a bit of practice on it towards the very end of  a course of study.  That’s not teaching to the test. Teaching to the test is working backwards from the final exam and then writing a curriculum punctuated by  slightly reduced versions of that exam – and then teaching each set of lessons with the next test in mind.   The teaching is shaped by the approaching test.  This is learning for assessment.  By contrast Christodoulou argues that we should just concentrate on teaching the  curriculum and that there may be a whole range of other activities to assess how this learning is going that may look nothing like the final learning outcome. These, she contends, are much better suited to helping the learner actually improve their performance. For example, the teacher might teach the students what the key events were in the build up to the first World War, and then, by way of assessment, ask students to put these in correct chronological order on a time line.  Feedback from this sort of assessment is very clear –if events are in the wrong order, the student needs to learn them in the correct order.  The teacher teaches some  small component that will form part of the final whole, and tests that discrete part. Testing to the teach, in other words, as opposed to teaching to the test.

There are obvious similarities with musicians learning scales and sports players doing specific drills – getting the fine details off pat before trying to orchestrate everything together.  David Beckham apparently used to practice free kicks from all sorts of positions outside the penalty area, until he was able to hit the top corner of the goal with his eyes shut.  This meant that in the fury and flurry of a real, live game, he was able to hit the target with satisfying frequency.  In the same way, Christodoulou advocates spending more time teaching and assessing progress in acquiring decontextualized technical skills and less time on the contextualised ‘doing everything at once’, ‘playing with the traffic’ kind of tasks that closely resemble the final exam.  Only when we do this, she argues, will assessment for learning be able to bear fruit. When the learning steps are small enough and comprehensible enough for the pupil to act on them, then and only then will afl be a lever for accelerating pupil progress.

Putting my primary practitioner hat on, applying this approach in some areas (for example reading) chimes with what we already do,  but in others (I’m thinking writing here) the approach seems verging on the heretical.  Maths deserves a whole blog to itself, so I’m going to leave that for now – whilst agreeing whole-heartedly that thorough knowledge of times tables and number bonds  (not just to ten but within ten and within  twenty   – so including  3+5 and 8+5 for example) are  absolutely  crucial. Indeed I’d go so far as to say number bonds are even more important than times table knowledge, but harder to learn and rarely properly tested. hit-the-button I’ve mentioned hit the button in a previous blog. We have now created a simple spreadsheet that logs each child’s score from year 2 to year 6  in the various categories for number bonds. Children start with make 10 and stay on this until they score 25 or more (which means 25 correct in 1 minute which I reckon equates to automatic recall.  Then then proceed through the categories in turn – with missing numbers and make 100 with lower target scores of 15.  Finally they skip the two decimals categories and go to the times table section – which has division facts as well as multiplication facts. Yes!  When they’ve got those off pat, then they can return to do the decimals and the other categories. We’ve shared this, and the spreadsheet –  with parents and some children are practising at home each night. With each game only taking one minute, it’s not hard to insist that your child plays say three rounds of this first, before relaxing.  In class, the teachers test a group each day in class, using their set of 6 ipads.  However since kindle fire’s were on sale for £34.99 recently, we’ve just bought 10 of them (the same as the cost of 1 i pad). We’ll use them for lots of other things too, of course – anything where all you really need is access to an internet browser.

When we talk about mastery, people often talk about it like it’s this elusive higher plan that the clever kids might just attain in a state of mathematical or linguistic nirvana when really what it means is that every single child in your class – unless they have some really serious learning difficulty – has automatic recall of these basic number facts and (then later) their times tables.  And can use full stops and capital letters correctly the first time they write something. And can spell every word on the year 3 and 4 word list (and year 1 & 2 as well of course).  And read fluently – at least 140 words a minute, by the time they leave year 6. And have books they love to read – having read at least a million words for pleasure in the last year (We use accelerated reader to measure this – about half of year 6 are word millionaires already this year and a quarter have read over 2 million words.) How about primary schools holding themselves accountable to their secondary schools for delivering cohorts of children who have mastered all of these (with allowances for children who have not been long at the school or who have special needs)  a bit like John Lewis is ‘Never Knowingly Undersold’, we  should aim (among other things) to ensure at the very least, all our children who possibly could, have got these basics securely under their belt.

(My teacher husband and I now pause to have an argument about what should make it to the final list.   Shouldn’t something about place value be included? Why just facts?  Shouldn’t there be something about being able to use number bonds to do something?  I’m talking about a minimum guarantee here – not specifying everything that should be in the primary curriculum. He obviously needs to read the book himself.)

Reading

My extended use of the metaphor of learning to drive to explain Christodoulou’s approach has one very obvious flaw. We usually teach classes of 30 children whereas driving lessons are normally conducted 1:1. It is all very well advocating spending as much time on the basics as is necessary before proceeding onto having to orchestrate several different skills all at the same time, but imagine the frustration the more able driver would have felt stuck in a class with me and my poor clutch control.  They would want to be out there on the open roads, driving, not stuck behind me and my kangaroo petrol.  Children arrive at our schools at various starting points. Some children pick up the sound-grapheme correspondences almost overnight; for others it takes years. I lent our phonics cards to a colleague to show here three-year-old over the weekend; by Monday he knew them all. Whereas another pupil, now in year 5, scored under 10 in both his ks1 phonic checks.  I tried him again on it recently and he has finally passed.  He is now just finishing turquoise books. In other words, he has just graduated from year 1 level reading, 4 years later.  This despite daily 1:1 practice with a very skilled adult, reading from a decodable series he adores (Project X Code), as well as recently starting on the most decontextualized reading programme ever (Toe by Toe – which again he loves) and playing SWAP. He is making steady progress – which fills him with pride – but even if his secondary school carries on with the programme[3], at this rate he won’t really be a fluent reader until year 10. I keep on hoping a snowball effect will occur and the rate of progress will dramatically increase.

Outliers aside, there is a range of ability (or prior attainment if you prefer) in every class and for something as technical as phonics, this is most easily catered for by having children in small groups, depending on their present level. We use ReadWriteInc in  the early years and ks1.  Children are assessed individually by the reading leader  for their technical ability to decode, segment and blend every half term and groups adjusted accordingly.  So that part of our reading instruction is pretty Christodoulou-compliant, as I would have thought it is in most infant classes.  But what about the juniors, or late year 2 – once the technical side is pretty sorted and teachers turn to teaching reading comprehension.  Surely, if ever  a test was created solely  for the purposes of being able to measure something, it was the reading comprehension test, with the whole of ks2 reading curriculum one massive, time wasting exercise in teaching to the test?

I am well aware of the research critiquing the idea that there are some generic comprehension skills that can be taught in a way that can be learnt from specific texts and then applied across many texts, as Daniel Willingham explores here.  Christodoulou quotes Willingham several times in her book and her critique of generic skills is obvioulsy influenced by his work. As Willingham explains, when we teach reading comprehension strategies we are  actually teaching vocabulary, noticing understanding, and connecting ideas. (My emphasis).  In order to connect ideas (which is what inference is), the reader needs to know enough about those ideas, to work out what hasn’t been said as well as what has been. Without specific knowledge, all the generic strategies in the world won’t help. As Willingham explains

Inferences matter because writers omit a good deal of what they mean. For example, take a simple sentence pair like this: “I can’t convince my boys that their beds aren’t trampolines. The building manager is pressuring us to move to the ground floor.” To understand this brief text the reader must infer that the jumping would be noisy for the downstairs neighbors, that the neighbors have complained about it, that the building manager is motivated to satisfy the neighbors, and that no one would hear the noise were the family living on the ground floor. So linking the first and second sentence is essential to meaning, but the writer has omitted the connective tissue on the assumption that the reader has the relevant knowledge about bed‐jumping and building managers. Absent that knowledge the reader might puzzle out the connection, but if that happens it will take time and mental effort.’

So what the non-comprehending reader needs is  very specific  knowledge (about what it’s like to live in a flat), not some generic skill.  It could be argued then that schools therefore should spend more time teaching specific knowledge and less time elusive and non existant generic reading skills. However, Willingham concedes that the research shows that even so, teaching reading comprehension strategies does work. How can this be, he wonders? He likens the teaching of these skills as similar to giving someone vague instructions for assembling Ikea flat pack furniture.

 ‘Put stuff together. Every so often, stop, look at it, and evaluate how it is going. It may also help to think back on other pieces of furniture you’ve built before.

This is exactly the process we go through during shared reading. On top of our daily phonics lessons we have two short lessons a week of shared reading where the class teacher models being a reader using the eric approach. In other words, we have daily technical lessons,  and twice a week  we also have a bit of ‘playing with the traffic’ or more accurately, listening to the teacher playing with the traffic and talking about what they are doing as they do it.  In our shared reading lessons, by thinking out loud about texts, the teacher makes it very explicit that texts are meant to be understood and enjoyed and not just for barking at and that  therefore we should check as we go along that we are understanding what we are reading (or looking at). If  we don’t understand something, we should stop and ask ourselves questions.   It is where the teacher articulates that missing ‘connective tissue’,  or ‘previous experience of building furniture’  to use Willingham’s Ikea metaphor, sharing new vocabulary and knowledge of the how the world works, knowledge that many of our inner city children do not have.  (Although actually for this specific instance many of them would know about noisy neighbours, bouncing on beds and the perils of so doing whilst living in flats.)

eric

For example, this picture (used in ‘eric’ link above) gives the the teacher the opportunity to share their knowledge that that sometimes the sea can get rough and that this means the waves get bigger and the wind blows strongly. Sometimes it might blow so hard that it could even blow your hat right off your head. As the waves rise and fall, the ship moves up and down and tilts first one way, and then the other. (Pictures are sometimes used for this  rather than texts so working memory is relieved from the burden of decoding).

When teaching children knowledge is extolled as the next panacea, it’s not that I don’t agree, it’s just that I reckon people really underestimate quite how basic some of the knowledge we need to impart for our younger children. I know of primary schools proudly adopting a ‘knowledge curriculum’ and teaching  two hours of history a week, with two years given over to learning about the Ancient Greeks.  I just don’t see how this will help children understand texts about noisy neighbours, or about what the sea is like (although you could do that in the course of learning about Ancient Greece if you realised children didn’t know), or, for that matter, what it is like to mill around in bewilderment.  The only kind of assessment that will help here is the teacher’s ‘ear to the ground’ minute by minute assessment – realising that -oh, some of them haven’t ever seen the sea, or been on a boat. They don’t know about waves or how windy it can be or how you rock up and down.   This is the kind of knowledge that primary teachers in disadvantaged areas need to talk about all the time.  And why we need to go on lots of trips too. But it is not something a test will pick up nor something you can measure progress gains in.  The only way to increase vocabulary is one specific word at a time. It is also why we should never worry about whether something is ‘relevant’ to the children or not. If it is too relevant, then they already know about it – the more irrelevant the better.

I don’t  entirely agree with the argument that since we can’t teach generic reading skills we should instead teach lots more geography and history since this will give  children the necessary knowledge they need to understand what they read.   We need to read and talk, talk talk about stories and their settings -not just what a mountain is but how it feels to climb a mountain or live on a mountain, how that affects your daily life, how you interact with your neighbours.  We need to read more non fiction aloud, starting in the early years.  We need to talk about emotions and body language and what the author is telling us by showing us.  A quick google will show up writers body language ‘cheat sheets’. We need to reverse engineer these and explain that if the author has their character fidgetting with darting eyes, that probably means they are feeling nervous. Some drama probably wouldn’t go amiss either.  Willingham’s trio of  teaching vocabulary, noticing understanding, and connecting ideas is a really helpful way of primary teachers thinking about what they are doing when they teach reading comprehension. What we need to assess and feedback to children is how willing they to admit they don’t understand something, to ask what a word means, to realise they must be missing some connection.  None of this is straightforwardly testable. That doesn’t mean it isn’t important.

Writing

Whereas most primary schools, to a greater or lesser degree, teach reading by at first teaching phonics, the teaching of writing is much more likely to be taught though students writing than it is through teaching a series of sub skills.  It is the idea that we ensure technical prowess before  we spend too much time on creative writing that most challenges the way we currently do things.

Of course we teach children to punctuate their sentences with capital letters and full stops right at the start of their writing development. However, patently, this instruction has limited effectiveness for many children.  They might remember when they are at the initial stages and when they only write one sentence anyway – so not so hard to remember the final full stop in that case. Where it all goes wrong is once they start writing more than one sentence, further complicated when they start writing sentences with more than one clause. I’ve often thought we underestimate how conceptually difficult it is to understand what a sentence actually is.  Learning to use speech punctuation is far easier than learning what is, and what is not, a sentence. Many times we send children back to put in their full stops, actually, they don’t really get where fulls tops really go.  On my third session doing 1:1 tuition with a year 5 boy, he finally plucked up the courage to tell me that he know he should but he just didn’t get how you knew where sentences ended.  So I abandoned what I’d planned and instead we  learnt about sentences. I told him that sentences had a person or a thing doing something, and then after those two crucial bits we might get some extra information about where or why or with whom or whatever that  belongs with the person/thing, so needs to be in the same sentence.   We analysed various sentences, underlining the person/thing in one colour, the doing something word in another colour and finally the extra information (which could be adjectives, adverbs, prepositions, the object of the sentence – the predicate minus the verb basically) in another. This was some time ago before the renaissance of grammar teaching, so it never occurred to me to use the terms ‘subject’ ‘noun’ ‘verb’ etc but I would do now. It was all done of the hoof, but after three lessons he had got it, and even better, could apply it in his own writing.

What Christodoulou is advocating is that instead of waiting until things have got so bad they need 1:1 tuition to put it right, we systematically teach sentence punctuation (and other  common problems such as verb subject agreement), giving greater priority to this than to creative writing. In other words, stop playing with the traffic before you’ve mastered sufficient technical skills to do so properly.  This goes against normal primary practice, but I can see the sense in this. If ‘practice makes permanent’ as cognitive psychology tells us (see chapter 7 of What Every teacher Needs to Know About Psychology by Didau and Rose for more on this), then the last thing we want is for children to practice again and again doing something incorrectly. But this is precisely what our current practice does. Because most of the writing we ask children to do is creative writing, children who can’t punctuate their sentences get daily practice in doing it wrong. The same goes for letter formation and spelling of high frequency common exception words. Maybe instead we need to spend far more time in the infants and into year 3 if necessary on doing drills where we punctuate text without the added burden of composing as we go. Maybe this way, working memories would not become so overburdened with thinking about what to say that the necessary technicalities went out the window. After that, we could rewrite this correctly punctuated text in correctly formed handwriting.  Some children have genuine handwriting or spelling problems and I wouldn’t want to condemn dyslexic and dyspraxic children to permanent technical practice. However if we did more technical practice in the infants  – which would mean less time for writing composition – we might spot who had a genuine problem earlier and then put in place specific programmes to help them and/or aids to get round the problem another way. After all,  not all drivers use manual transmission, some drive automatics.

Christodoulou mentions her experience of using the ‘Expressive Writing’ direct instruction programme, which I duly ordered. I have to say it evoked a visceral dislike in me; nasty cheap paper, crude line drawings,  totally decontextualised, it’s everything my primary soul eschews (and  it’s expensive to boot). However, the basic methodology is sound enough – and Christodoulou only mentions it because it is the ones she is familiar with. It is not like she’s giving it her imprimatur or anything.  I’m loathed to give my teachers more work, but  I don’t think it would be too hard to invent some exercises that are grounded in the context of something else children are learning; some sentences about Florence Nightingale or the Fire of London for example, or a punctuation-free excerpt from a well-loved story.  Even if we only did a bit more of this and a bit less of writing compositions where we expect children to orchestrate many skills all at once, we should soon see gains also in children’s creative writing. Certainly, we should insist of mastery in these core writing skills by year 3, and  where children still can’t punctuate a sentence, be totally ruthless in focusing on that until the problem is solved. And I don’t just mean that they can edit in their full stops after the fact, I mean they put them (or almost all of them in ) as they write. it needs to become an automatic process. Once it is automatic is it easy.   Otherwise we are not doing them any favours in the long term as we are just making their error more and more permanent and harder and harder to undo.

Certainly pupil progress meetings would be different. Instead of discussing percentages and averages,  the conversation would be very firmly about the teacher sharing the gaps in knowledge they had detected, the plans they had put in place to bridge those gaps, and progress to date in so doing, maybe courtesy of the ‘hit the button’ spreadsheet, some spelling tests, end of unit maths tests, records of increasing reading fluency. Already last July our end of year reports for parents shared with them which number facts, times tables and spellings (from the year word lists) their child did not yet know…with the strong suggestion that the child work on these over the summer!   We are introducing ‘check it’ mini assessments so that we can check that we we taught three weeks ago is still retained. It’s easy, we just test to the teach.

[1] Christodoulou quoting D. Wiliams, p 19 Making Good Progress?

[2] Christodoulou quoting D. Wiliams, p 20 Making Good Progress?

[3] I say this because our local secondary school told me they didn’t believe in withdrawing children from class for interventions. Not even reading interventions. Surely he could miss MFL and learn to read in English first? As a minimum.  Why not English lessons? I know he is ‘entitled’ to learn about Macbeth but at the expense of learning to read? Is Macbeth really that important? Maybe he will go to a different secondary school or they’ll change their policy.

Advertisements
Test to the Teach

How to deepen understanding of reading by teaching children how to enjoy it better.

 

Or possibly…

How to enjoy reading more by understanding what you read more deeply

Like many primary schools this year, reading is the focus of our development plan. Despite my frustrations with this year’s ks2 reading sats, which I’ve written about here, our children did ok. Slightly above average.  And we are an inner city, deprived, multi-cultural, multi-linguistic, multi-story carpark kind of school. But that’s 20% below how they did in maths.  Now that gap may have more to do with test design than inadequacies in our teaching, but it still infuriates me.  If we can do it in maths we should be able to do it in reading…

So I spent the summer reading about reading. Like many others, I reconsidered what I knew about reading by reading Doug Lemov’s  ‘Reading Reconsidered’. See for example, Nick Hart’s recent interesting blog on reading, and thatboycanteach’s recent post and Rhoda Wilson’s work – who did away with guided reading long before it became a thing.  It is clear that skilled shared close reading of texts is back in the ascendency whereas the carousel-style guided reading lesson is on the wane, particularly in ks2, or where readers can decode fluently. See for example, the Destination Reader approach developed in Hackney primary schools.

However, thatboycanteach’s post – which was incredibly useful – set twitter mulling over ‘reading comprehension’ as a thing, particularly with regards to fiction.  Why can’t children just read and enjoy a story – why analyse to death, why answer turgid sats style ‘inference’ questions?  While I fly no flag for the ks2 reading sats paper, I would respond that inference in general is a vital life skill. Children who cannot infer, who cannot see beyond literal surface features, are children with social and communication problems. Socially skilled children, in contrast, are excellent at spotting the subtle signs that indicate to the initiated that all is not what it seems. Whether in the playground or in the pages of a book, the ability to ‘read’ people is an incredibly useful skill.

Nor are enjoyment of reading and analytical close reading necessarily in opposition.  The whole point of shared reading is that we share with children texts that would be too hard for them to read on their own. We do the heavy lifting, thereby giving them access to richer and more substantial texts than pupils would be able to cope with on their own. It seems to me that primary teachers have three main jobs with regards to reading. First ensure that they can get the words off the page easily and with sufficient fluency to enable at least a superficial understanding of what they have read.  Secondly, to try and inculcate a love of reading. Without this we will produce a generation who can read, but don’t. Indeed, during my initial teaching training way back in 1988, the main rationale for the ‘real books’ approach that I was trained in was to avoid creating literate children who only ever read under duress.  (I saw the light re phonics some 8 years later, but that’s for another blog another time).  Hooking children on books will sometimes mean seeking out the ‘gateway drug’ book that will set our dear pupil on the path to a happily text-filled future. The Recruit, perhaps, a book so racy (by the very low bar of primary education), at our school you are not allowed to borrow it until we have cleared it with your mum first. (Seriously though, if you are considering using it, just stick to the first book in the series.  As with the Harry Potter books, the main characters’ age by one year in each book- though with a good deal more precocity than described in Hogwarts – by book 3 we are deep in year 9 territory – don’t be fooled by the relative tameness of the first book). Thirdly, and just as importantly, we are here to teach children how to ramp up the readable. In our hands lie the tools that will enable children to unlock hitherto seemingly impenetrable texts. By our efforts texts once dismissed as ‘boring’ and ‘irrelevant’ can come alive for pupils. Doug Lemov is explicit in his reason for insisting that students work with hard texts; this is what they will be faced with at college. If they are not exposed to hard texts until they go to college, we shouldn’t be surprised if many young adults drop out when faced with a long and academically demanding reading list. I’m not quite as sold on the ‘university or bust’ credo, but certainly that is one reason why we should present children with rigorous, aspirational texts.  And certainly the reading demands of the new GCSE are much higher, so in primary we need to make sure the groundwork for that is rock solid.

The trouble with this is the 3 different jobs is that each one requires a different kind of book, with the pupil exposed to each in different ways. The ‘getting the text off the page’ stage requires books with restricted vocabulary – with words that the pupils already have the knowledge how to decode phonetically. It goes without saying that at the beginning when the child has only a little knowledge of phonics, these aren’t going to be the most stimulating works of literature ever encountered – though hats off to them behind Project X Alien Adventures – the fully decodable scheme our children go nuts for. Even without these heirs to Biff and Chip, in my experience children enjoy the success that quickly comes using phonetically decodable readers, even when the early books have to try and weave words like ‘pin, ‘mat’ and ‘sad’ into a compelling story line (although sitting on a pin would make one sad, it is true). The ‘getting hooked on books’ stage requires thrilling storylines, some daring-do and/or humour, possibly suspense, characters you care about. Or possibly hate, but in an invested way. But not necessarily nuance, higher level vocabulary, ambitious sentence structure. We are in ‘BeastQuest  and Wimpy Kid territory here. Books children will eagerly read but you would never, ever in a million years read out loud to your class.  They are the literary equivalent of a Pot Noodle. The third stage requires the sort of book CLPE advocates in its literature collections within its Core Book Lists.  For example, the utterly brilliant Varjak Paw or the Lady of Shallot – poems generally in fact. Usually children do not read these independently, at least not at first. They may be the end-of-the-day class story, or used in shared or guided reading. They comprise the literary equivalent of smuggling extra vegetables into pasta sauce.

I fear there is a triple constraint going on here.  In the same way that holidays can be cheap and exotic but not relaxing, or relaxing and cheap at the expense of being exotic, or relaxing and exotic but ever so expensive, it appears we can only have 2 out of any 3 here. So a text can be decodable and exciting (Alien Adventures) but not literature. Or it can be exciting and literary (Frank Cottrell Boyce’s Unforgotten Coat) but you need to be a fluent reader to access. Or I suppose it could possibly be both literary and decodable but not immediately gripping – again many poems are perfectly readable in terms of the individual words, and no one doubts their literary pedigree, but while we may perhaps entice our children, or some of our children, to love them, they are unlikely to become avid readers on a diet of these alone. They are, perhaps, the literary equivalent of stilton.

What this all means is, organising the teaching reading to a class of children with different dietary needs is difficult. The guided reading carousel has the advantage that the guided group’s focus group book can be catered for the group’s particular needs while the other children do something else, some of which includes reading self-chosen books in the hope that this will help them develop a love of reading. But it is a model that devotes a relatively short time – if any – to the teacher making the expert reader’s thoughts visible and explicit; to explaining vocabulary, inference and meta-cognitive comprehension strategies.  If your model does do this, fantastic. But does it also give children time to read the kind of books that they want to read too? And what about the ones who still aren’t fluent? Are you going to tell one that everybody is fluent in your school? Like you don’t ever have new arrivals or children with statements?

In an ideal world, we’d spend the first 20 minutes either learning to spell or learning to decode, followed by 45 minutes of high quality shared reading for all where the teacher modelled their thinking out loud as they read an aspirational text – gradually transferring more and more work to the class as the lesson progressed and the children thought hard about the texts they were studying’ including written answers. Then we would do some longer writing linked in some way to the shared text for an hour, with more teacher modelling, explicit grammar teaching, peer feedback, 10 minutes handwriting practice and finally 30 minutes to read our own self chosen – though guided by an expert librarian or knowledgeable teacher – text, as well as going to assembly, having a break and doing an hour’s maths. And having 30 minutes of story time at the end of the day. That leaves 5 hours a week left for the rest of the curriculum – and that’s without swimming – that time sink that swallows 2 hours time per every 30 minutes actually, er, swimming. Or nativity plays. Or trips. Or having half the class out at a hockey tournament.

So that’s not going to happen.

Having sought out and studied various models this term, it seems the triple constraint comes into play and forces us to make choices about what we are not going to do.  Which one of teacher modelling of reading, giving children time to read for pleasure and teaching children to actually read (the words) are we not going to do?  We are not very good as a profession at owning our choices though and tend to downplay the thing we are not doing – like it wasn’t that important anyway. So old style guided reading carousel is strong on children reading for pleasure and the mechanics of reading, but had less capacity for quality teacher exposition.  Models that have a lot of teacher exposition either dispense with devoting any curriculum time to reading for pleasure, or can’t cope easily with the children in the class who are not yet properly fluent. With this latter group, schools either seem to say – it’s so important these struggling readers are exposed to teacher talk around challenging texts so they must be included; they can miss other lessons for decoding practice. Or they say, decoding practice has to come first, and if they miss foundation subjects then they will miss the very knowledge of the world they need to be able to make inferences, so the decoding children decode elsewhere while the rest of the class do shared reading.

I’ve come across two approaches – one via Destination Reader and one in the approach advocated by Great Bowden Academy (of which more shortly) where the reading lesson is divided into a teacher led introduction modelling a particular learning objective with the whole class looking at the same text, followed by a time where the children seek to apply the same objective (finding places where the reader needs to infer the meaning from textual clues, for example) in the text of their own choice.  This has the obvious advantage of children engaging more deeply with self-chosen texts but the drawback that the teacher – who cannot possibly have detailed knowledge of every book currently being read – has to be very skilled at questioning the children on texts s/he does not know well. If I am reading them correctly, both Nick Hart and thatboycanteach keep with the same text throughout the lesson, with children analysing the text themselves and after an initial input phase. Personally, this approach seems more fruitful because they teacher will know the text under study really well and will have chosen the text deliberately as being rich in whatever technique(s) they are teaching, and so will have more to offer the children.  It also allows for children to do more writing about texts.  The downside being – when do the children get time during the school day to read their own choice of text? Maybe they all read at home…I’d love to work in a school where they all read at home without a huge song and dance and bribes and threats and even then…

We haven’t started doing either approach yet so I’m not going to the stake for what is, at this stage, a hunch, and we are going to do a mixed approach (I think, things are still tentative at this stage) where in years 4-6, 3 days a week are teacher led using one text, and the remaining two have no teacher input and are old fashioned independent reading of self-chosen books, but with the teacher working with a group on whatever aspect of reading they deem necessary. Those who need decoding practice come to the first teacher led lesson, but then have their TA led reading intervention the rest of the week. The children will sit in mixed ability pairs.  The text the teacher will use might be a revisiting of the current class end of day story – so the text can be analysed in the kind of way that would ruin a good storytelling, or a book the class are using in literacy for their writing –  a version of Beowulf say, (with this easier version used in some lessons), or The Wedding Ghost perhaps.  Or they might use an extract from a non-fiction text based on their humanities or science topic, or a poem.  In case you haven’t come across it, Opening Doors to  Famous Poetry and Prose by Bob Cox is an amazing resource, and he has recently published two versions for younger children (the one I’ve linked to is aimed at ks3, but usable with year 6). It is more about using famous – and demanding texts as a springboard for writing, but obviously includes unpacking the texts themselves.

Which brings me, (finally, the reader gasps) to my title.

How to deepen understanding of reading by teaching children how to enjoy it better.

Or possibly

How to enjoy reading more by understanding what you read more deeply.

Is it possible to teach children how to enjoy reading? Surely they either enjoy it or they don’t. maybe we can facilitate that enjoyment but setting up the necessary pre conditions, decoding, fluency, vocabulary development etc., but whether or not a child enjoys a text is surely an intimate matter shared between the author and the reader?

Actually yes you can teach children how to enjoy reading. We ourselves were taught how to enjoy it. I know this is not true for every teacher, but I think I’m pretty safe in assuming most of us grew up in a family where young children were read to. If so, we probably don’t remember our mum or dad or whoever saying things along the way such as…’oh no, Little Red Riding Hood’s gone of the path. Her mum told her not to. I think that means something bad is about to happen!’ (hamming it all up). Or maybe that happened for you at primary school. Either way, without expert reader input, you would not have realised to feel scared for LRRH because we realised she was now in danger – a danger, what is more, that our protagonist is oblivious to.

An aside. When my husband was training to be a teacher his tutor relayed the true tale of the delights of teaching in an inner city classroom. We are in year one, and the teacher is, with great theatre, reading the story of the three little pigs to the class. There is at least one little boy who has never heard the story before, but whom is greatly taken up with the drama of the plot. The teacher reads ‘and he huffed and he puffed and he blew the house down.’  To which our apprentice reader exclaimed ‘fucking bastard!’  (Tick whatever box you now have to tick to record that you now have evidence that the child can understand character motivation). I’ve dined out on that story, and I wasn’t even there.

Anyway, we teach children to enjoy stories by fantastic hammy story telling and by the asides we make that point out the secret clues that alert us to something hidden within the text, a kind of literary Easter egg.   Well, dear reader, there exists a book that describes exactly what these clues are and tells you how to teach them to children. And that book is, Notice and Notes: Strategies for Close Reading by Kylene Beers .

I came across this book courtesy of Great Bowden Academy at the Institute of Education’s annual Research and Development conference.  At this conference members of the research and development network come together from their various schools and share with each other the small scale research they have been doing that year. We are not talking randomised controlled trials here. But thought provoking and useful nonetheless. Our school team were there to share what we had found out  about growth mindset and how it  could be used with teaching assistants – which you can read about here if you wish. I was lucky enough to be on a table with the Great Bowden team, and their project had been about improving pupils’ reading fluency, engagement, comprehension and enjoyment.  A major part of their approach had been to teach the children about the strategies for close reading outlined in Notice and Note. This involved whole class explicit teaching of close reading techniques, a lot of the teacher thinking out loud and pointing out the clues, or signposts as Beers calls them.  Later in the lesson the children read their own texts and looked out for with whatever was specifically the specific aim of the learning objective for the day or ‘spotted and jotted’ signposts.  This sounded intriguing, so I bought the book, and read it over the summer holidays.

It really is very clever.  What Beers has done is identify 6 tropes frequently found within stories that give the reader a heads-up ‘stop, take notice here, this is important’ information. It is full-fat, text level inference we are talking here.  The first, and most useful one is, ‘contrasts and contradictions’.  We teach the children that when you’re reading and a character says or does something that’s opposite (contradicts) what s/he has been saying or doing all along, you should stop and ask yourself, ‘Why is the character doing that?’  The answer could help you make a prediction or make an inference about the plot or the conflict in the text.

For example, in the book Holes, (where the setting is a kind of character), the book opens with the sentence, ‘There is no lake at Camp Greenlake’.  An obvious (to the expert reader), and intriguing contradiction that should raise an eyebrow. However, in our experience, many children have had to have this pointed out to them. Indeed the year 6 teacher tells me she often spends most of a lesson on this very sentence, making explicit how the author is contrasting the name of the place Camp Greenlake, with the reality (no lake, and no green either, we will soon discover). The name is a contradiction.  And that contradictory name – that’s presents as one thing whilst really being something else entirely, is central to the plot.  Camp Green lake purports to be a suitable environment for the rehabilitation of delinquent teens, but in reality, (spoiler alert!) serves only as a vehicle serve the avarice of the Governor. But by labelling this specific technique ‘contrast and contradiction’ we are giving children a generalisable strategy that they can use with other texts. And that is what is key here about this approach. Consider, for example, within the Harry Potter oeuvre, the contradictory character of Snape. He seems to be a baddie but every now and then he does things that make you wonder; and Dumbledore trusts him, so maybe he is a goodie after all?  But then he does something really bad – like killing Dumbledore – which is in complete contradiction with being one of the good guys, but then again… it’s in the tension of the not knowing that the enjoyment, the thrill of reading lies.  Or the character of Edmund in The Lion, the Witch and the Wardrobe. We know he really did meet the faun Mr Tumnus yet he lies and said it never happened. Why? And what might happen as a result!  More subtly, in Wonder, in the excerpt mentioned by thatboycanteach, where Summer comes up with a list of (inclusive) rules  about who is, and is not allowed to sit on the same lunch table as August and herself, this contrasts (somewhat painfully) with the exclusive and rejecting actions of everybody else; this alerts us early on that Summer is going to be a key figure in the protection and  maybe eventual acceptance of Auggie.  I’m trying to get our year 6 teacher to teach children the word ‘juxtaposition’ before they leave; – if only to impress their secondary school teachers!

The next signpost is the Aha moment. When a character suddenly realises, understands or finally figures something out, you should stop and ask yourself, ’How might this change things?’  For if a character just figured out a problem, then you have discovered the central conflict in the story. Or if the character understood a life lesson, you probably now understand the central theme of the book. For example, when Felix in Once finally loses his naivety and makes the horrified realisation that the Nazi’s actually hate Jews, that’s an Aha moment. It serves to amplify his resilience; the life lesson of this amazing book. Or this blog describes several great picture books that have great Aha moments in them. It’s American, so I didn’t know quite a few of the books, although some were familiar. But it made me think of all those great books we use to teach acceptance, whether of different others or of oneself; Giraffes can’t dance, for example. Gerald the giraffe has an Aha moment when he realised he can dance. Or Elmer learning to accept his multi coloured true self in a world where everybody else is grey. Which again reminded me of the CHIPS project (Combatting Homophobia in Primary School)  series of book based lesson plans.

The ‘Again and Again’ signpost is when you notice a word, phrase or situation mentioned over and over again. It is a signpost telling you to stop and ask yourself, ‘Why does this keep happening again and again?’ Your answer might tell you about the theme and conflict in the text or might foreshadow later events.  Pernille Ripp, our American blogger again has some picturebook examples. In Holes, the whole recurring family curse motif is central to the denouement.  Two phrases from ‘the Lion, the Witch and the Wardrobe’ spring to mind; ‘always winter, never Christmas,’ and ‘he’s not a tame lion’.

In some ways similar, we have Memory Moment, when the author interrupts the action to tell you about a memory. When you come across a Memory Moment then stop and ask yourself, ‘Why is the author telling me this now? Why might it be important?’ The answer might foreshadow later events, or might tell us about the theme and conflict in our text. The family curse in Holes in a Memory Moment used Again and Again, and indeed does foreshadow the climax.

Then we have ‘the Words of the Wiser; when a wiser, and usually older character takes our protagonist aside and offers serious advice. When we read this we should stop and ask ourselves ’What’s the life lesson and how might it affect the character?’    They tell us about the central theme of the text; about how character and plot interact. Sometimes I get Words of the Wiser muddled up with Aha moments. The key difference being Aha moments are discovered by the character themselves and often have the text saying something like ‘I suddenly realised that…’, whereas in Words of the Wiser, someone else gives advice. Someone like Yoda or Dumbledore.

Finally, we have Tough Questions, when the protagonist asks themselves something very difficult. This should cause us to stop and ask ourselves ‘What does this question make me wonder about?’ and will help us understand the central conflict of the story and help us think about what might happen later in the story. In some ways these are the mirror images of Aha moments.  The protagonist asks themselves ‘how will I ever fit in?’ whereas in the Aha moment the character might discover they don’t need to change to fit in; it’s the tight limits on what constitutes ‘fitting in’ that needs to change, not the protagonist. Or the tough question ‘how will I ever survive?’ might have a later aha moment ‘I will survive because I now realise that…’

The book is gives a model, almost scripted lesson for each signpost.  As the book is American and pitched towards middle school, while these lessons are great for teacher inset, they are not immediately usable with British primary aged children. But the general principles definitely are. The internet is also awash with useful bookmarks and posters to help you implement this approach; this one for example has further links to a whole raft of resources. It’s a much bigger thing in the US as the demands for close reading in the Common Core change how reading is taught in the US. Here in England the new curriculum is also causing us to rethink old ways.

I think the ‘Notice and Note’ approach is really useful in explicitly showing students the breadcrumb trails authors want readers to find in their work.  It is difficult for us expert readers to remember what it was like before we just knew that certain types of event or phrases were vital clues to the whole book, let alone give them names or number them in a list. But as we all know, many students find inference really hard -possibly because they didn’t have parents making a metacognitive commentary – ‘oh no, how will the pigs survive now they can’t live with their mummy?’ – as they share a book at bedtime. But better than that, it really does help some children finally ‘get’ why reading can be pleasurable.  Once you know the author is playing with you, you can play along too.

It’s not the whole answer. It doesn’t solve the triple constraint or tell us how to plan our lessons. There won’t be any questions about it on the ks2 sats paper. But it might help them to learn to love books. And children who love reading are much, much more likely to do well academically aged 16. One study from the Institute of Education found that

‘The combined effect on children’s progress of reading books often, going to the library regularly and reading newspapers at 16 was four times greater than the advantage children gained from having a parent with a degree.’

So worth a punt, I reckon.

How to deepen understanding of reading by teaching children how to enjoy it better.