What *does* improve children's writing?

//What *does* improve children's writing?

In my last post I discussed evidence that suggests grammar teaching does not lead to an improvement in children’s writing. Although it seems implausible that grammar teaching would not be positively correlated with writing outcomes, there’s a lot of evidence that is strongly suggestive that what I prefer to believe may not in fact actually be true.
I’ve written enough about cognitive bias to know that I am predisposed to look for evidence that supports my preferences and dismiss evidence that contradicts them. The point of evidence is that it forces us to confront the extent to which our intuitions map against objective reality, and to that extent at least I feel the need to explore the idea that I might be wrong.
So, although I struggle with the thought that grammatical knowledge would not lead to greater proficiency in writing, I’m willing to accept that grammar teaching may be counter productive for all sorts of reasons. The grammatical concepts may not be carefully sequenced, the teaching itself might be focussed simply on learning meta-linguistic terminology, the teaching could be focussed on performance rather than learning, and so on. Another issue could be opportunity cost. It might be that time spent teaching grammar is time that could be more profitably spent on some other activity more likely to improve children’s writing.
Students tend to get a lot of writing practice in school and so, for the opportunity cost argument to work, it would have to be the case that what children spend their time doing in English lessons and elsewhere is more likely to lead to improvement. If, for the moment, we accept that grammar teaching doesn’t improve children’s writing, there are at least three questions we should consider:
1. Is children’s writing as good as we would want it to be?
Broadly, no. Every teacher will know that although some children write wonderfully well, in many cases written responses are wooden, technically inaccurate and do a poor job of expressing complex thoughts. My suspicion is that there is probably a Matthew Effect in operation; those children who we might describe as ‘word-rich’ start their school career with a linguistic advantage – they have heard more words, reach more and thus have greater vocabularies and more experience of the written word than their ‘word poor’ peers. A 1988 study suggests that 9 year olds at the 90th percentile read almost 2.5 million words a year outside of school, whereas a child at the 10th percentile will have been exposed to around 50,000 words. It takes the bottom 10% a whole year what the top 10% read in two days!Those who begin school with this sort of advantage require little in the way of teaching – their ability to write seems (at least to some extent) imbibed osmotically from the sheer quantity of written material they encounter. But what of everyone else?
I’ve suggested before that our ability to write is linked to our ability to speak. We can only write down that which we are able to say. Some children are very good at making connections and explaining their thoughts but struggle to write them down. They are limited to only being able to express their thoughts in spoken language – they’re unable to mentally translate into written language because they do not possess the vocabulary and structures of written language. They know that if they wrote down what they have said aloud it would no longer make sense.
Another problem is the embedding of bad habits through poorly constructed writing practice. I call this the capital letter problem. Children often know that their writing is inaccurate or clunky. They know they should use capital letters, spell correctly and avoid punctuation errors, but they don’t know these things well enough. Anything that occupies our attention reduces our capacity to think. This capacity is strictly limited and, if we’re juggling complex ideas, we won’t have much attention left to think about accuracy. Skilled writers will have automatised these basics so that they quite literally don’t have to think about them.
I think we can agree that, on average, children’s writing is not as good as we would want it to be.
2. What sorts of things do English teachers do to improve children’s writing?
English – and by extension writing – tends to get a pretty generous allocation of curriculum time in both primary and secondary schools, and children also practice writing in many other curriculum areas. All this represents a huge opportunity for improving children’s writing. So, what do teachers – especially English teachers – spend their time doing?
Greg Ashman has recently written that the teaching of writing is backwards. Typically, students are asked to write something and then, when they make mistakes, teachers spend their time producing a written commentary on how to correct these errors. He suggests that it would be more sensible to give children precise instruction on how to write before asking them to do the writing. This may seem obvious – and many teachers will of course be working hard to model and scaffold writing tasks effectively – but in my experience of working with teachers in a great many different settings, this is all too familiar. We do this because although we recognise good writing when we see it, too few teachers understand how the ‘skill’ of writing is actually composed of many thousands of items of propositional knowledge. These items become invisible as they’re automatised and so it can be almost impossible for a skilled writer to recognise the components of good writing.
All too often, English teachers expect students to get better at writing by practising writing extended pieces. Obviously this has a place, but it will disproportionately advantage word rich students who already know how to write well. This is bit like expecting people to get better at tennis just by playing tennis or to get better at the piano just by bashing the keys. In many areas of complex skill we know that we have drill the basics. Athletic prowess results from acquiring often minute technical skills; fluent musicianship stems from repeating scales and arpeggios over and over. It makes sense to think that fluent writing is most likely to arise from a similar process, but exactly what this process might look like is something that has not been codified and shared amongst English teachers.
3. What are the most effective methods for improving children’s writing?
I’m happy to accept that being able to parse or diagram a sentence bears very little resemblance to the skill of writing. Although it feels intuitively true that mastering the basics of English grammar could only be useful, let’s assume I’m wrong. What else then does the research advocate?
The picture is a little bleak. The paper by Wyse and Torgerson which concluded that grammar teaching is ineffective says that, “Although the numbers of robust experimental trials relevant to effective teaching in schools have increased, our analysis of trials in relation to the teaching of writing suggests that there are still too many studies that are not of sufficient methodological quality. In particular, too many studies are weak in relation to allocation of pupils to groups, and the measures for writing remain a challenge.”
With this caveat they make the following recommendations:

The most recent meta-analyses of high-quality research studies on writing suggest that, rather than emphasise grammar, the following practices could be selected as a priority for teaching writing in primary/elementary education: (a) an increase in the amount of time that pupils have for writing; (b) adoption of a process approach to writing; (c) creation of a classroom environment that is appropriately supportive of pupils’ attempts at learning to write better; (d) development of pupils’ writing skills, strategies and knowledge, including ways of planning writing; (e) a use of assessment for learning techniques; (f) a use of computers as part of the process of writing; (g) a use of writing meaningfully across different subject areas (Graham et al., 2016; Wyse, 2017). The robustness of the evidence underpinning these practices is built not on single studies but on multiple RCTs and experimental trials.

Let’s briefly consider each suggestion in turn.
(a) an increase in the amount of time that pupils have for writing
As noted, children already spend an awful lot of time writing. If current practices are ineffective, simply doing more of the same is unlikely to result in an improvement. My advice to teachers is to ask students to write less. Specifically, if teachers don’t explicitly value the process of writing, then students are likely to learn that how they write is relatively unimportant. I’d rather see children doing shorter more focussed writing exercises such as those advocated in my Slow Writing posts.
(b) adoption of a process approach to writing
The idea here is to focus on the process rather than the product of writing in order to encourage a more metacognitive approach to the act of writing. This is essentially the idea behind Slow Writing and one I’m happy to endorse. However, what children write is as important as how they write and if we entirely neglect the product, that is also likely to have negative consequences. I see the ‘process approach’ as involving modelling and scaffolding and, like all scaffolding, teachers need to have a plan for removing it to prevent children becoming dependent on it.
(c) creation of a classroom environment that is appropriately supportive of pupils’ attempts at learning to write better
I’m not really sure what this means. If it means decorating the physical environment with helpful reminders and motivational quotes, then – as I explain here – it’s probably not a good idea. If it means creating an environment where children are encouraged to behave well, work hard and value effort, then it’s almost certainly a good idea.
(d) development of pupils’ writing skills, strategies and knowledge, including ways of planning writing
It’s somewhat tautological to suggest that the best way to improve children’s writing is by developing their writing skills. It’s obviously right that children need to develop the knowledge to write well, the question is, how? I’d agree that it makes sense to devote some time to effective planning but again, do we know what effective planning actually is? Does it look the same for all types of writing? Should students to taught to use different kinds of graphic organisers? Committing curriculum to planning may be useful, but will also be limiting. Eventually children need to develop effective mental representations of what good writing looks and feels like so this approach must address how students will internalise the idea of planning.
(e) a use of assessment for learning techniques
Again, it depends on what you mean by assessment for learning. I’ve written extensively on how we get AfL wrong and so it should go without saying that we would need to avoid these pitfalls. In the worst case, teachers can end up justifying the kind of backwards teaching Ashman has written about as an AfL technique. Let’s not do that.
(f) a use of computers as part of the process of writing
As a professional writer, I always use a computer to write. As a classroom teacher, I rarely did. If the aim of English teaching is – in part – to support children’s performance in national exams, then they won’t be helped by spending too much time using computers. Unless they learn to hand write fluently, they will be at an enormous disadvantage. I suspect that underlying this recommendation is the notion that using computers is inherently motivating. This is a mistake. Computers are a useful tool, just as a pencil is. It may be true that some children – especially those with terrible hand writing – will prefer to use a computer, nobody gets better at something by avoiding doing it. Also, there’s the additional problem that computers can be an absorbing distraction. If my experience is anything to go by, I spend more time stopping children using Word Art than on improving their writing.
(g) a use of writing meaningfully across different subject areas
This is something that happens as a matter of course in most primary settings and is definitely something that secondary schools ought to be better at. Non-English teachers sometimes feel resentful of the idea that they ought to be developing their students’ writing skills, but this is terribly misguided. Writing is different in different curriculum areas. It would be foolish indeed for geography teachers to rely on their English teacher colleagues to teach the kind of writing that works best in geography. It’s also worth considering that time spent thinking how best to express ideas about content will not only improve students’ ability to write, it could also improve their ability to think about the content they’re learning. Who wouldn’t want that?
In conclusion, if teachers are doing all of the above exceptionally well then maybe they have a solid argument for not teaching grammar at any point. I’m not and never have argued that grammar teaching should replace any other part of the writing process. My position is a backlash against the fact that for decades, schools in England have systematically avoided teaching grammar and as a result are now ill-equipped to do it well.

2017-12-01T13:22:27+00:00December 1st, 2017|writing|


  1. Hannah December 1, 2017 at 1:34 pm - Reply

    I wonder if encouraging children to read widely and more often may in fact have a positive impact on their writing?

    • David Didau December 1, 2017 at 2:11 pm - Reply

      It seems highly that reading widely does have a positive impact on writing, but if children can’t decode fluently they’re unlikely to read widely even if encouraged.

  2. Tom Burkard December 1, 2017 at 4:04 pm - Reply

    I take anything written by Torgerson or Wyse with a healthy pinch of salt. Both were dismissive of the 2006 Rose Review, and both wrote reports strongly critical of synthetic phonics. A 2006 report by the former bordered on dishonesty. I certainly agree that their current paper has very little of interest to teachers who want to improve pupils’ writing, and I strongly agree with your contention that children are made to write too much. In most instances, all they are doing is developing a distaste for writing and overlearning a lot of bad habits.
    If there is any way of teaching children to write well, we certainly haven’t discovered it yet–and it’s not for a lack of trying. With expository writing, a thorough understanding of the subject of the composition is essential, and time practising writing detracts from the time available for extending and enriching that understanding. Assuming that extensive reading helps children write well–I wouldn’t bet on there being much more than a strong correlation here–the best we can do is to teach them enough about a subject so that new information strongly enriches existing schemata.
    Of course we value good writing, if for no other reason than our exam system demands it. Given that we don’t know how to teach this skill, we should be re-thinking our exams rather than spending huge amounts of time that has little effect other than demotivating pupils and putting them off education. Essay questions are possibly the most inefficient and unfair means of measuring pupils’ learning yet to be devised.

  3. Jen December 1, 2017 at 4:18 pm - Reply

    While I certainly would encourage reading, and it does seem that students that also have a book out in a spare moment tend to be my best second language students, I have to point out that even those that do are not guaranteed to write better unless they are actually paying attention to and processing grammar/sentence structure. Perhaps those who really like to read a lot are the one who are inclined to process at a deeper level? However, reading is a receptive skill and writing is productive. The former would surely enhance the latter–I’ve yet to see a good writer who was not also a good reader. However, reading for getting the gist of a piece doesn’t necessarily mean that that student will write well. In cognitive science, is not attention the very first step to in-taking knowledge? Most people will not pay attention to grammar/sentence structure if they do have their attention called to it.

    • David Didau December 2, 2017 at 12:39 pm - Reply

      I completely agree. That said, those who read a lot are much more likely to acquire tier 2 vocabulary then those who don’t. They also learn all sorts of general knowledge which operates in the background to improve both reading comprehension and writing. I also think it’s entirely plausible that frequent readers will learn a lot of implicit information and syntax but, of course, they are not at all likely to learn anything explicit. That which we don’t know explicitly cannot be thought about.

  4. Michael Pye December 2, 2017 at 12:38 am - Reply

    This is why I read your blog David. A much stronger and clearer argument that I really enjoyed reading and thinking about.

  5. Tricia December 2, 2017 at 9:18 pm - Reply

    I enjoy your posts David and the responses they generate…especially around literacy, teaching and learners. Let’s go another step back… Before focusing on grammar and structures children need to learn HOW to write. Speaking with, developing thinking processes, reading to, developing wide vocabulary are desirable prerequisites in developing the potential capabilities of ALL learners. Learning to write and teaching writing are explicit and developmental processes. That which learners bring to the learning and the Oral discourse is their foundation to build upon. For learners entering school, learning that oral language and thinking can be turned into print, and learning to know how to do this using a print code, is complex and multidimensional. Knowledge of the learner, knowledge of literacy learning, knowledge of developmental writing progression, knowledge of individual learner’s developmental progression, are key factors of consideration to initiating and developing learner capability. Beginning learners first years at school are critical- the importance of literacy teaching in their first years of formal schooling incomparable. First we develop individuals’ ability to make the connection between oral and print, guiding them to express their voice, writing for authentic purposes. Looking to the grammar and structures according to the purpose and audience of the writing, will occur as explicit guided feed forward – feedback guided instruction… within the appropriate developmental progression of literacy learning.

    • David Didau December 3, 2017 at 1:11 pm - Reply

      Isn’t grammar and structures part of HOW to write? I’m not making any claims about what to teach first, but beginning with links between oral and print before they know anything about print seems a bit previous. Might it not be better to begin with some letter formation as part of phonics instruction so they build up a thorough knowledge of phoneme/grapheme relationships?

  6. adyer2014 December 3, 2017 at 8:22 pm - Reply

    Essay writing and argument building are essential skills of the humanities-based curriculum at university level and hence have been brought down through most level three courses to level two and even level one. Students not only have to master and manipulate content but also have to develop meta-cognitive sequencing skills. English creative writing also demands a “voice” – again, difficult to master. I would agree that our exam system militates against those students who intend to leave full-time education at 16 or 18 and until this changes we have no choice but to prepare all learners to jump through the linguistic and extended writing hoops to prove academic worthiness. There must be a better way in the 21st century to demonstrate mastery across subjects

    • David Didau December 3, 2017 at 10:47 pm - Reply

      Do you think there’s a better way to demonstrate mastery of writing other than by completing written assessments?

  7. jen December 4, 2017 at 2:55 am - Reply

    Although I strongly recommend reading the entire (short) book–this article gives a good introduction to The War Against Grammar by David Mulroy. He is a Stanford PhD and Professor Emeritus at the University of Wisconsin–Milwaukee. Reading his own words is better than my attempt to paraphrase them!

  8. chestnut December 4, 2017 at 9:29 pm - Reply

    They also need adults who speak formally, such as teachers who can speak in coherent sentences. I am certain that children who are involved in proper discussions at home at the table for example are also at an advantage. It is not just about vocabulary but how children put a sentence together. It isn’t just about capital letters but where to put the full stop.

    • David Didau December 5, 2017 at 8:33 am - Reply

      Yes, of course. No one ever claimed otherwise

  9. Anna Hasper December 10, 2017 at 8:52 am - Reply

    Hi David, after all those years I still enjoy your posts and a lot of what you write seems common sense to me as a EAL teacher of English! Thank you for putting this into words so clearly. Whereas I very much support the suggestions you’ve highlighted in your blog, I do think one key area is missing: using model text or mentor text. I know that many argue using these can take impact learners’ creativity, however, struggling writers can benefit hugely from model text. They more or less make visible what we like them to do and are a great tool to make Ss notice particular language structures used or genre specific features. It’s maybe more part of the product approach but combining process & product has been very helpful for many of my learners. And yes… reading. Definitely key, but how we struggling readers to read more that’s indeed another challenge!

    • David Didau December 11, 2017 at 10:41 am - Reply

      I completely agree: modelling is essential.

  10. […] 3. What *does* improve children’s writing?, by David Didau […]

Constructive feedback is always appreciated

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: