Archive for category Literacy and Oracy

Writing and Desirable Difficulties: how making writing harder to learn makes students better at it (eventually)

I often wish that I could remember more about what went on in lessons when I was at school. There are snippets that I can recall quite well: I can remember learning to balance quadratic equations, memorizing the periodic table and about the history of WW1. One thing, though, that I cannot remember – no matter how hard I try – is how I learnt to write.

In fact, I’m almost definite that in my school, we were not explicitly taught grammar and sentence structures. One reason that I know this is that as a trainee teacher, I had to learn this myself. The question, then, is how is it that I know how to write? I have a hunch that I learnt how to write through reading, but how do I help people learn to write who – on the whole – read less than I did?

For me, the explicit teaching of the structural content (sentences, features etc.,) of writing has been a vital part of helping my students to write well. In this post, I will discuss the evolution of my practice in this area, critiquing each method I have used, before arriving at one that I consider effective.

1) The sentence variation starter #1: When I first started thinking about improving my students’ sentence writing abilities, the first thing I used, along with many of my colleagues and, I imagine, teachers the world over, was this:


I taught students how to write using simple (S), compound (C) and three kinds of complex sentence (moving the subordinate clause to the beginning (BCX) middle (MCX) and end (ECX) of the sentence). To download my lesson on how to teach students to do this, click here: Lesson S C ECX BCX MCX.

At the start of every lesson, I gave students a simple sentence that related in some way to our content for the day. They had to transform it into the four other types. My hope was that by the constant, repetitive practice, students would begin naturally to use these kinds of sentence in their writing.


When students engaged in extended writing, they would, generally resort to the same patterns (or lack of) as before we began intensive practice of the five sentence types, I think for the following reasons:


  • The fantastic book ‘Make it stick: The Science of Successful Learning’ finds that deep learning comes when practice is done in the same context as performance. For example, if a musician sits at home and practices the scales all alone, even to a high level of expertise, they will be lost in a real life performance situation when mastery requires so much more. Or, to put it another way: ‘If you always practice the same skill in the same way, from the same place…you’re starving your learning on short rations of variety.’ By making the context and conditions of practice exactly the same, every lesson, students were not actually able to use their learning in any other context than as part of a lesson starter.
  • This book also states that ‘Effortful retrieval makes for stronger learning and retention.’ Now, while having the same starter lesson after lesson certainly asks students to repeatedly retrieve the information, is it actually effortful? Is performing the same thing, multiple times per week a challenge? I would argue that the grammar starter, if repeated in exactly the same way is not challenging enough to encourage real, long-lasting learning. David Didau’s ‘Learning Spy’ blog argues that this sort of practice is in fact an example of performance rather than learning: this sort of ‘massed practice’ means that students become adept at performing under certain conditions, rather than making real progress.

So, what did I learn?

  • Repetitive grammar starters, isolated from real writing, are pointless. Students need to practice their grammar in the same arena as I would expect them to perform: in real pieces of writing.
  • Recall of the material needs to be made more challenging: asking students to repeat the same knowledge, in the same way, in the same part of the lesson means that they become trained to perform in a certain way. They actually learn very little.
  • In fact, would I even be satisfied if students were only able to use five kinds of sentence structure? I felt that I needed to demonstrate higher expectations of my students by asking them to master many more structures and features.

2) The Sentence Variation Starter #2: The next thing that I tried was to compile a larger variety of sentence types – a list that I called ‘The Big 15’–and have the students utilise them in a real writing context. The list contained the five sentence types above, as well as others that I took from Alan Peat’s ‘Exciting Sentence Types’ resource. This contained sentences such as:

  • Some; Others – The first part of this sentence begins with the word some. The second part of the sentence is separated with a semi colon and the word others. Some children walk to school; others travel by car.
  • Starting a sentence with a simile – using a simile, followed by a comma, as the first clause of yours sentence. Yawning like a lion, he sat up.

My ‘Big 15 sentences’ sheet can be downloaded here: The Big 15

Alongside this list, which students stuck into the front inside cover of their books, I created a lesson starter called ‘At the Bus Stop’ (actually, I stole this from a primary school that I visited). As the class was working on descriptive writing at the time, at the start of every lesson I asked students to imagine meeting a different person at the school bus stop. They had to describe this encounter using a specific set of the big 15 sentences, selected through random numbers (1, 14, 5, 6, 3, 11 etc.,). This ‘Slow Writing’ approach was heavily influenced by David Didau’s blogging, found here and here.

At the Bus Stop

I then asked students to read and assess each other’s descriptions, before giving them time to revise and edit their work. This is an example of a student’s piece of writing for the above task, with revisions made on the board:

Bus Stop Student Example


This was a far more successful experiment. There were clear improvements in student writing across the board, and students actually enjoyed their starter far more than the previous one; it also linked in much better with our lessons. There were, however, still some issues with this approach:

When I assessed students on their descriptive writing, while there was a marked improvement in sentence variation, some students used a disappointingly small number of the sentence types, and all had issues ‘blending’ these with descriptive features. While I had made improvements in asking students to practice in a more realistic context, I had not concentrated on actually asking them to learn and retrieve the key knowledge.

Another quotation from ‘Make it Stick’: ‘Effortful retrieval both strengthens the memory but also makes the learning pliable again, leading to its reconsolidation.’ My students had become reliant on the Big 15 sheet as a crutch in their writing, and I, as their teacher, had not known enough to concentrate just as much on the students’ memorisation of the sentence types as I had on their ability to use them. Retrieval of the knowledge was effortless; hence, learning was limited.

What did I learn?

  • Again, I needed to make recall of the material more challenging. I could have included periodic tests on the sentence types and I could have removed the Big 15 help sheet after a certain period of time.
  • Another insight from ‘Make it Stick’: Rather than mixing up the writing styles that students were practicing (interleaving descriptive writing and its constituent sentence types and features with other writing styles, for example), I concentrated only on one writing style, leading – or so says the  evidence presented in the book – to suboptimal learning: ‘…the research shows unequivocally that mastery and long-term retention are much better if you interleave practice than if you mass it.’

3) My current project:

Inspired by the ‘Learning from My Mistakes’ post, ‘Death to the Sentence Stems: Long Live the Sentence Structures’, from which I took nearly all the sentence structures to follow, I decided to make my expectations of student writing much more challenging.   I created sets of cards – one set for descriptive writing, and one set for argumentative writing – containing all the success criteria ( not only sentence structures but language features, structural elements like connectives and key expectations, like using clear topic sentences) for each style. I also attached points to these, so for example in descriptive writing:

  • A Personification, 5 commas and 3 tos sentence:Harsh white walls frown at the monotone uniformed prisoners, men with bleached faces and no eyes threaten, guns hoover, thunderously muted, waiting for someone to move, to think, to breath.’ is worth 4 points.
  • A Comma sandwich (MCX): ‘The sun, which had been absent for days, shone steadily in the sky’ is worth only 1. These are awarded based on the difficulty (or flashiness) of the sentence.
  • Using onomatopoeia, or any other descriptive feature, is worth 1 point.
  • Failing to meet certain fundamental expectations for the style, like using a mixture of short and long sentences, or clustering descriptive features to create vivid imagery, deducts 5 points.

While in argumentative writing:

  • A Not, nor, nor sentence: Nobody, not the postman, nor the housekeeper, nor Jim himself knew how the letter had got onto the doormat. is worth 3 points
  • Use of imperatives or any other argumentative feature is worth 1 point.
  • Failing to use connectives to link ideas and paragraphs appropriately results in a loss of 5 points, amongst a variety of other key stylistic expectations.

In both sets, there were roughly 35 cards overall, with around 20-25 sentence structures, the rest being made up of features or key stylistic expectations. A full set can be seen below; the black cards are the key expectations, which remain laid out at all times. The others are either structures or features, which are shuffled and placed in separate piles.

The set of cards for descriptive writing can be downloaded here:Cards – Describe , and for argumentative, here: Cards – Argue.

Cards laid out

How did I use the cards (the steps):

  1. Familiarsation: I began to use the cards in paired writing activities. A fantastic discussion of the benefits as well as a successful structure for paired writing can be found on the Reflecting English blog. Students worked in pairs, each directing the writing of the other by selecting two cards: one sentence structure, and one feature. The students would work to combine these in their writing, asking their partner to check their attempt for correctness before requesting another pair of cards. The image below shows a game being played; the two bottom cards are in active use; when finished with they are placed to the back of the piles.

Cards during a game

  1. Memorisation: After familiarising themselves with each card through the paired writing process (this took a couple of lessons), the students then went about memorising the cards. This was done through using them as flashcards, with students looking at the title of the card and having to recall and write an example. Cards that they could complete went on one pile, and those they couldn’t on another, continuing until all cards had been successfully recalled.  From ‘Make it Stick’: ‘Something as simple as a deck of flashcards can provide an example of spacing. Between repetitions of any individual card, you work through many others.’ I think, therefore, that this method, giving the sheer number of structures and features to commit to memory, fulfills the criteria for a ‘desirable difficulty’, especially when you interleave two different writing styles, which I did.  My students found this hard!
  1. Testing: At this point (after leaving a gap of a couple of days), I removed the cards completely and introduced low-stakes testing to aid recall of the structures and features. Previously, I would have continued with students continually using the cards in their writing until I felt they were comfortable with them. Evidence says this is incorrect; again from ‘Make it Stick: ‘Today, we know from empirical research that practicing retrieval makes learning stick far better than reexposure to the original material.’  I used a Quizlet page for the testing (writing to describe here; writing to argue here. This is the easiest tool I know of for quickly making randomised tests). Students then used the cards to relearn material they had forgotten. I repeated this step a couple of times, interleaving the describe and argue tests to build the level of difficulty.
  1. Writing without prompts: I then gave students a writing task to complete without the aid of the cards. This served as another test with an even higher level of difficulty as they didn’t even have the names of the sentences and features to act as prompts, something that students found fairly taxing. Knowing that they were trying to achieve a certain score in their writing helped to provide impetus to recall the material, even when effortful.
  1. Modelling, peer assessment and improvement: Following this, I showed the class a model that I had written using the cards, asking the class to score this:

Teacher model assessed

The idea of scoring the model was to show the students the quantity of sentence structures and features that a high quality paragraph might contain. There is obviously no hard and fast rule here: the idea of scoring writing can seem a little formulaic but I wanted the students to see that good writing is, in the main, packed with interesting sentence structures and features (students scored my paragraph at around 19 points).

Next, I asked students to read their partner’s work, picking their best and worst paragraphs and scoring them in the same way. Here is an example of a student’s ‘worst’ paragraph (it achieved 4 points):

Student Model - BEFORE

Finally, student re-wrote their paragraphs, attempting to bring them closer to the score achieved on my model. Again, this was done with no prompting from the cards, and achieved a 16 point increase, demonstrating, I think, the power of adding some sort of a ‘stake’ to a test: the simple desire to want to beat their own score increased effort and improved recall. This is also just the sort of effortful recall that leads to long term memorisation.

Student Model - AFTER

Now, it goes without saying that I repeated these steps quite a few times, especially the testing, modelling and peer assessment stages. I was also very careful to interleave each writing style –both with each other and with other completely different topics. I hope this meant that I was leaving enough of a gap for optimal ‘forgetting’ to occur before returning to the material. This method has, without a doubt, done more to help my students produce excellent writing than any I’ve ever used before.

Benefits of this method:

  • At every stage, students found it really, really hard, but this is a good thing.  As Didau writes: ‘Deliberately choose the harder, more difficult option. Learning isn’t easy. But as Hattie reminds us, “A teacher’s job is not to make work easy. It is to make it difficult.”
  • Rather than just sentence structures, language features or any other criteria, it encompasses all of these together. This means that students have to be able to recall and master multiple things at once, as well as worrying about content, something that expert writers can do but that is a hugely challenging thing to grasp.
  • Unlike with grammar starters, students practice their writing in a ‘performance context’: they create real pieces of writing, without aids or prompts, just as they would in an exam situation.
  • It is easy to test students on their abilities, and easy to judge real progress being made over time.
  • It creates pieces of writing that score highly in exams. Examiners like interesting structures and features, combined in unique ways.
  • It privileges knowledge over skills: students need to learn the content before they can use this in writing. Put simply, there is no such thing as the skill of writing; it is just a case of mastering a vast network of different areas of knowledge. This method recognises that fact.

Once again, for download, my cards for argumentative and descriptive writing are here: Cards – Argue, Cards – Describe.

Further reading: in no particular order, here are some of the things that I read which helped me in my thinking about writing.


, , , , , , , ,


Rigorous Analytical Writing: Is P.E.E fit for purpose?

Rigorous Analytical Writing: Is P.E.E fit for purpose?

In my relatively short teaching career (two schools; two countries; five years), I have been exposed to a vertiginously long list of different strategies, acronyms and procedures for helping students to write analytically: do you PEE? How about PEARL? Or PETER? PETEETER? What about EPLWE?

My personal journey through this quagmire is of interest, at least to me, because it highlights some important issue related to my progress as a teacher of English and to the teaching of the subject in general.   In this post, I want to survey the different methods for structuring analytical writing that I have used, listing the reasons why, thus far, I have been dissatisfied with all of them. Through this, I hope to uncover some fundamental reasons why all of the above – and more – are unfit for purpose, and why a different approach to the teaching of analytical writing is required.

Much of this post, in particular the ‘different approach’ to analytical writing just mentioned, has been influenced – or borrowed wholesale – from David Didau’s ‘The Learning Spy’ blog, as well as his book ‘The Secret of Literacy’. Specific posts will be highlighted below, but just to give credit where due.

A (brief) survey of analytical writing method to date

Point, Evidence, Explain (P.E.E)

I started, as I imagine many new English teachers did – before they knew any better – with good old reliable P.E.E. For those who aren’t haven’t heard of this, students make a point, answering a question about a text; give evidence, by quoting a line from the text; and then ‘explain’ the link between the evidence and the point.



  • P.E.E is easy to remember. With the faint scent of urea adding a minor frisson, students are able to remember the acronym easily which, I suppose, means that they are able to learn how to use the structure more quickly.
  • If followed to the letter, PEE is really easy to do, which might make it an appropriate tool for those beginning to learn about how to write analytically.
  • It cross cuts schools, and subjects, more of which will be discussed below.


  • It is far too easy, lacking in the necessary rigour and detail to make students into expert analytical writers. When you think about it, asking a student merely to explain the link between the evidence and the point is hardly all there is to writing analytically. Where is the fine-grained detail? Where can the student explore the nuances of language? The grade ladder below demonstrates that if students merely ‘explain’ the links between their evidence and their point this is – I think at most –a C grade response.

Grade Ladder

This leads to the next problem.

  • For any teacher who uses PEE for anyone other than students who have never written analytically before, you quickly realize that asking students to ‘explain’ is not enough. (For a different argument for the same view, see Reflecting English’s excellent post on the subject.) For that reason, the term ‘explain’ is then expanded to include language analysis, evaluation of writer’s intention and effect on reader, and a whole raft of other tools.
    • This means that P.E.E quickly loses any ‘sticky’ benefits mentioned above: students don’t only need to remember P.E.E, but in actual fact a whole list of other procedures when analysing language
    • For those of us who struggle to get students to really understand what command terms mean (explain, explore, analyse etc.,), this is perpetrating a fairly large scale confusion of the term ‘explain’. Using P.E.E is a catch-22: if we take the basic meaning of the term ‘explain’, then P.E.E is not fit for purpose as it only really allows a very basic level of analysis. If we expand it, then we are misusing it with potentially damaging consequences if students are asked to ‘explain’ something in another context or possibly an exam.

P.E.A Lasagna

In a brief and possibly slightly confused attempt to supplant P.E.E, in my NQT year I devised the P.E.A Lasagna method of analytical writing. To be fair to it, the year 10 English Language group that I used it with did well in their exams, however I did not pursue this approach into my third year as a teacher. Please see the below images for an explanation of the approach. I developed this to help students improve in a few key areas:

  1. The point / white sauce idea came from the fact that students were not linking back to their point in the rest of their paragraph. They would write the point, and then go off on the tangent. The idea here was that this should remind them that they should always be making links back to the central point
  2. The ‘double-blocking’ of the pasta sheets/evidence was because I wanted students to be inserting multiple short sections of the text in their paragraphs, rather than simply having one large chunk at the start
  3. Finally, I wanted students to attempt to attempt to pick quotations apart in their analysis, hence the language features / words split.

P.E.A 1P.E.A 2P.E.A 3  P.E.A 4


  • Compared to P.E.E, this is a lot more rigorous in its requirements of the students. Unadulterated, it allows for higher achievement than if students simply followed the basic P.E.E structure to the letter.


  • Although at the time I thought this was easy to remember, I have since changed my mind. While my students did get the hang of it, I’m not sure the silly metaphor helped in this regard, and it potentially distracted students from the key learning, especially before lunch.
  • A central problem for this method is that it is not universal; a really good analytical writing method should cross-cut across all subjects, with only minor changes. This could never be used in any subject bar English. If students are to get really expert at analysis, as with any area of know-how, they should be putting this to use across the curriculum.

Finally: PETER, EPLWE, PEARL and the rest

Since my early flirtations with the two above methods, I have used a whole raft of other tools, all of which suffer from many of the pitfalls described above.

EPLWE (Evidence, Point, Language, Word, Effect) was developed for a particular exam (Edexcel Literature) in which – according to an examiner’s report – it seemed as if students were required to write lots of short, shallow paragraphs rather than anything with any depth


  • For this, it worked. Students were able to churn these out at around one every three minutes.


  • It resulted in a situation where students were using one type of analytical writing method in literature, another in language and yet another in history. Not ideal.
  • It also had two Es, explain and effect, which made it difficult to learn.


The same can be said for PETER (Point, Evidence, Technique, Explain, Reader response) and PEARL (Point, Evidence, Analysis, Reader Response, Link). Both of these have their good points and, to be fair to them, students have and will continue to perform well in examinations using these methods. But – and this is a big but – none of the above are really fit for purpose. I will recap the reasons why:

  • Most rely on a confused and erroneous use of command terms, leading potentially to student error
  • Most are only suitable for one small area of our own subject, sometimes only for one exam. As such, students are never in a position to develop a level of comfort and expertise
  • Apart from PEE, which I have seen used on other subjects, none of the above are truly suitable for other academic areas that require analysis of evidence, such as history.
  • And finally, the most damaging reason of all: anyone who  “does” analysis professionally, i.e., academics, journalists and so on, do not use any of the above methods. As such, all of them are gross over-simplifications and are symptomatic of a production line, low expectations approach to teaching. If we really want students who are good at our subject, we should be teaching them to analyse language as a “junior version” of how it is done in the real world, by experts.

The solution

After reading the following blog posts by David Didau (@learningspy), here and here, I am now convinced that there is another way, a way that avoids many of the issues listed above. For the last 6 months, I have been putting this into practice with my students.

The method requires students to work through a list of five key areas as they write, speak and think about language. These areas, again in a visual created by Didau, are pictured below:

Reading Skills Ladder

Students begin by ‘identifying’ their answer to the question and providing some evidence for their opinion. They then ‘explain’ the link between their evidence and opinion. So far so P.E.E. Following this, students enter the world of real life academic English by of ‘exploring’ tentative alternative explanations for their evidence. Next, they pick and ‘analyse’ specific elements of language, before finally ‘evaluating’ the links to context by looking at writer’s intention and reader effect.

Here are some examples of my students’ early attempts to use this method:

Analytical Method 1 Analytical Method 2


  • This method is challenging and rigorous, requiring a high level of knowledge about the text that students are analysing. They must know language features and word meanings, have a wide-ranging knowledge of the context of the text and most importantly understand the text well. This is a good thing.
  • This approach – or at least so I think – would be suitable for any subject where students write analytically. The point is that there are different areas of knowledge that come to bear in analysis, and this method asks them to use them all. As such, whether students are analysing a poem or a historical source, they can still apply this procedure.
  • Command terms are used accurately.
  • While going through each level, each time might not be quite what an academic does when looking at a text, I think that, broadly, this is much more in line with an “expert” approach to analysis. It is training students, albeit at a lower level, to become comfortable with an approach to analysis that could take them to university and beyond.
  • Finally, it allows a stepped approach to the teaching of analytical writing. When first becoming familiar with the idea of giving an opinion about a text and supplying evidence, students can use only the first few levels. They can progress up the scale as they become more familiar with the different areas of knowledge required at each.


  • This is not the easiest approach to remember, although using Didau’s idea of zooming-in and zooming-out really does help. It does need constant practice to get right, and, as such, introducing students to it just before a big exam might not be a good idea. I think they need a while to get it right.
  • This is not the kind of method where you can get students writing analytically early on in your study of a text – or at least not with any great success. Students need to build up a strong knowledge base of language techniques, contextual information and detailed knowledge of the text itself before they can hope master it.
  • On its own, this does not do enough to help the students build an advanced analytical vocabulary. To combat this, I have created a set of sentence starters that fit within each area, as well as different words that students can use instead of ‘shows’ and ‘suggests’. I provide a small set of examples of above, although the full set can be downloaded here: Analytical Writing Stems

Prompt 2Prompt 1Prompt 3Prompt 4Prompt 5Prompt 6Prompt 7 Prompt 8 Prompt 9

These are just a small selection of the whole group, which I have arranged around my classroom so that students can look around and choose different ways to phrase their analysis. Those with a sharp eye might also spot that I have related each level to the SOLO taxonomy. Controversial, I know, but this links with the way that I like to discuss students’ acquisition and use of knowledge in my lessons.  I should also note that, while I have no link to give, these were influenced by an idea of The Learning Spy’s shared at a CPD event.

The idea with this is that not only are students engaging in an analysis that is varied in the areas of knowledge it uses, but also that they are phrasing this in a sophisticated way. An absolutely fantastic blog post from Reflecting English that deals with many similar issues to this can be found here.

What about the skills-knowledge debate?

As far as I am concerned, teaching students this – and any other – method of analysis as a “skill” is a waste of time. As I mentioned above, this method is ineffective if students are not equipped with the requisite knowledge about the particular text(s) that they will be analysing. A recent, excellent post on this subject, found here on Chris Hildrew’s ‘Teaching: Leading Learning’ blog put it in a way that I like: this, and many other things like it, is not a skill but ‘know-how’. To again purloin some ideas from David Didau, teaching students an analytical writing method is an example of ‘procedural’ rather than ‘propositional’ (fact-based) knowledge – a difference only of degree, however. Students can learn how to follow this particular system, and this knowledge will serve them well, but only if they pair this with some – much more important – factual knowledge of language and the text itself. For a vastly clearer discussion of this topic, please see the two posts above.

And again, for download, here are my Analytical Writing Stems.

Thanks for reading.