Research School Network: Reading Comprehension If putting in miles and miles on the question-and-answer treadmill is not going to improve our readers, what will?

I’ve learned a great deal working as an Evidence Lead in Education (ELE) over the last couple of years: about ​‘best bets’ for improving outcomes, about how to put evidence to work strategically, and about the challenges facing an evidence-informed practitioner trying to effect and manage change.

One of the most significant barriers to making effective use of evidence is the universal truth that, as teachers and school leaders, our enthusiasm is plentiful but our time is scarce. Now I know this is hardly revelatory but it does present a particularly tricky balancing act to Research Schools (and anyone else who wants to mobilise evidence and translate it into classroom practice).

Research can help us to identify ​‘best bets’ for how to use our time most effectively. Engagement with research has great potential for making professional development more powerful and ultimately improving outcomes for children. However, this potential for progress is not without potential pitfalls.

For evidence to be effectively ​‘put to work’, it needs to be communicated to enthusiasm-rich and time-poor teachers succinctly enough that it can be assimilated amongst countless competing priorities and ideas. At the same time, when it comes to education research, ​‘the devil is often in the detail’ so teachers need to go beyond the headline information to think deeply about the recommendations and how they might work in the specific context of their schools.

To demonstrate this, I’m going to ​‘zoom in’ on one specific report — ​‘Improving Literacy in Key Stage Two’, which received a timely update in November 2021. While the recommendations remain the same, the updated guidance provides teachers and school leaders with updated research, further exemplification, and new tools to enhance literacy provision in their schools.

Much like the guidance report for ​‘Teacher Feedback to Improve Pupil Learning’, the updated literacy guidance is structured around a series of vignettes. These vignettes are helpful because they represent the common dilemmas facing schools. As such, they will hopefully prompt the kind of reflective discussions that represent the ​‘best bets’ for school leaders to focus their time on. I’m going to focus here particularly on the vignette for the third recommendation: Teach reading comprehension strategies through modelling and supported practice.

The dilemma presented in this vignette is one that brought back memories for me: I’ve been that teacher and had that discussion. The year was 2016 and my Year 6 class had recently completed the infamous ​‘Way of the Dodo’ SATs paper. While the text did not make any of my class cry, as was reported in national newspapers as happening elsewhere, I had seen them struggle more than a teacher would want to see their class struggle. The number of children who did not meet the expected standard came as a surprise and so an issue was identified: a number of our children were having difficulty with their reading comprehension. Being determination-rich but time-poor, I did what I often did in those days: I jumped to conclusions. First, I completed a question-level analysis of the children who didn’t meet the expected standard and deduced that they struggled to answer vocabulary and inference questions. As inference made up the largest proportion of the assessment domains (36%), I decided that I needed to give children more practice answering inference questions. My reading lessons would predominantly involve sharing a text — either through independent reading or following while listening to me read — then answering a worksheet of questions about the text as practice for the way they’d be assessed.

The test results that followed this change in practice did show a slight improvement; however, the progress of low-prior-attaining children, a group that contained many of the disadvantaged children in my class, remained stubbornly low. Furthermore, my experience of these reading lessons and ongoing assessment of the children’s reading capabilities was telling me that the extra inference work was not having the desired impact.

Thankfully, it was around this time that I did some reading of my own that strongly influenced my thinking, most significantly Daniel Willingham’s ​‘The Reading Mind: a Cognitive Approach to Understanding How the Mind Reads’. The first edition of the ​‘Improving Literacy at Key Stage Two’ guidance report had also been released and several blogs, books, and articles were appearing that served to highlight the importance of explicit vocabulary instruction and reading fluency practice. Each of these gave me a fresh lens through which to view my comprehension issue.

With the benefit of hindsight and a deeper dive into the evidence, I can see that the reason why I’d failed to solve the problem was a mistake in my initial interpretation of what the problem actually was. This time-poor teacher had jumped to the conclusion that the answer to my comprehension struggles was ​‘more comprehension’, treating assessment focuses like inference as if they were generic skills that could be developed with extensive practice. I was labouring under the assumption that answering inference questions in one context would support the ability to better make inferences about other texts in the future.

To explain why this is not the case, I’m going to turn to Christopher Such, whose book ​‘The Art and Science of Teaching Primary Reading’ is one I can’t recommend highly enough:

‘The factors that underpin a reader’s language comprehension abilities are all related to the specifics of whatever text is being read: the text structure knowledge that supports comprehension is the text structure of that text; the vocabulary required is the vocabulary of that text; the worldly knowledge required is that which relates directly to that text. So, is language comprehension a skill that can be transferred to a different context? No, it is not.’

If I wasn’t actually teaching children transferable skills, what was I doing? The reality is that I was spending a great amount of time assessing comprehension and not enough time modelling to children how to be strategic readers. By allowing the assessment domains to warp my practice, I’d made decisions that were mostly counterproductive and inefficient at best. It is the opportunity cost of these choices that stopped me from effectively developing strategic readers in my class. Although my intentions were good, my choice of a skills-based, test-driven approach meant that I invested a great deal of time in activities that provided little value beyond that which can be gained simply from the act of reading. This had a knock-on effect on the breadth of texts we had time to read together, the depth of our book talk, and the development of children’s background knowledge. This last point is further exacerbated when you consider that low-prior-attaining children often completed extra comprehension practice during wider curriculum lessons, thus missing out on opportunities to build the background knowledge that was inextricably tied to their ability to comprehend texts they might encounter in the future.

If putting in miles and miles on the question-and-answer treadmill is not going to improve our readers, what will? To answer this, we need to think carefully about what the evidence is saying. The ​‘Improving Literacy’ guidance report recommends that we, ​‘Teach specific reading comprehension strategies, modelling their use, and providing regular opportunities for pupils to practice the strategies.’ The use of the word strategies here is significant as it refers to what Willingham would refer to as a ​‘bag of tricks’ which, once taught, do not develop further with the kind of extensive practice I had been undertaking in my reading lessons. The teaching of comprehension strategies is best thought of as metacognition, itself a concept that is prone to misconceptions and a concept that can be hard to describe what it looks like in the classroom.

At the risk of failing to heed my own advice and oversimplifying a complex idea, I’ll explain what I mean based on a basic definition of metacognition from the foreword to the guidance report ​‘Metacognition and Self-Regulated Learning’:

On a very basic level, it’s about pupils’ ability to monitor, direct, and review their learning. Effective metacognitive strategies get learners to think about their own learning more explicitly, usually by teaching them to set goals, and monitor and evaluate their own academic progress.

One recommendation from the metacognition report that has great potential to maximise reading lessons is recommendation three, ​‘Model your own thinking to help pupils develop their metacognitive and cognitive skills’. In the reading classroom, this means teachers choosing rich and challenging extracts of larger texts for a ​‘close read’, providing a running commentary of the processes an expert reader goes through to understand what they read. This will involve sharing how their understanding is updated sentence-by-sentence, noticing whether they understand what they’re reading, drawing connections between sentences, and relating the sentence content to background knowledge and knowledge of other texts. I’ll give a little sample of what I mean using an example from B.B Alston’s epic adventure story Amari and the Night Brothers:

To begin with, I’d make sure that children understand the vocabulary of the extract — after all, not knowing what the words mean is a complete blocker to making sense of the text. Some of the words here that might need explanation and interaction are principal (the head of my school is called the headteacher), squad (drawing attention to the difference between squad and group), state (a word with other meanings used more frequently in this country) and stalk (making comparisons to a predator stalking its prey). I’d also discuss phrases like ​‘giving (someone) an earful’ and ​‘the feeling is mutual’ that might prove to be a barrier to understanding. Even though this story takes place in a familiar context, I’d also activate children’s background knowledge by reminding children that being sent to the principal’s office usually means someone has done something exceptionally good or exceptionally bad.

After that, I’d ask children to follow along as I read the text aloud without stopping, allowing the children to hear a model of fluent reading. Then I would tell them that I was going to read the text closely to monitor whether I had fully understood the situation:

After a couple of lines, I’d update the children on my thinking, noticing the sentence ​‘Again.’ and speculating about comparisons to the main character in another book we’d read, Ghost by Jason Reynolds. As we read further, I’d update the children further that I’d noticed the phrase ​‘stuck-up Little Miss Princess’ and the way it suggests our character and Emily Grant have clashed before. I would identify my uncertainty about the phrases ​‘tiny little shove’ and ​‘Wasn’t my fault’, speculating whether this might be a case of a biased or unreliable narrator.

There’s plenty in these next paragraphs that I could use to model comprehension strategies to my class. I could ask myself the question ​‘Why might the author have chosen the word squad to describe the group?’ and share ideas of what it leads me to believe about the characteristics of these children. I could feign confusion at the insertion of ​‘You’ve really done it this time, Amari’, pausing and rereading the sentence a couple of times before attending to the italics to understand that the ​‘you’ in question is the narrator herself and not a different character called Amari. I could ask a quick question to check that we’ve understood that Quinton is Amari’s brother and speculate whether she might be like other characters with ​‘successful sibling syndrome’. I’d certainly pause at the sentence ​‘There’s not much to cheer about any more’ and use it to ask questions and make predictions — Did something happen to Quinton? I think that this might be the reason why Amari is at the principal’s office so often. At the end of the chapter, I would summarise my reading and add it to a cumulative summary for display.

It took me a while to write but in practice, all of the above can take place in the space of a few minutes. The implication of this is that it frees up time that would have otherwise been spent writing answers to comprehension questions, which means I’m able to read further into the story and allow more time for the kind of rich, discursive questioning that is too often sidelined by a skills-based approach. I wouldn’t advocate dissecting all of a book in this way and would spend most of my sessions on activities like reading to the children (or asking them to read independently), practising repeated oral reading to develop fluency (I won’t go into detail about this here but start with recommendation two of the Improving Literacy report and the new Reader’s Theatre tool) or asking children to share their thoughts about what they had read. Occasionally I might ask children to respond to their reading through their writing, but this wouldn’t be the default approach. To use activities like these to their full potential, first we need to create time by stopping activities that offer little value.

The more the children become familiar with the processes involved in comprehension monitoring, the more the emphasis can shift from modelling this metacognitive talk to guiding and supporting the children’s own monitoring through questioning and scaffolding. This gradual release of responsibility that is so central to the metacognition guidance will require careful assessment and management but if it can be executed well, it should go a long way towards achieving the aim of producing readers who use comprehension strategies independently and habitually to access increasingly challenging and rewarding texts.

I’m not claiming to be an authority on comprehension strategies, nor am I even claiming that the above is an exemplary model of comprehension monitoring in action (for this, I’d recommend the aforementioned Christopher Such book or Teaching English by the Book by James Clements). I know I need to continue to develop my practice, refining and improving this approach as I become more confident and experienced with it. What I can say with some confidence is that it is an example of an approach that has brought my practice far closer to the kind of work that the evidence has identified as a ​‘best bet’ and holds far greater promise than my previous skills-based approach to reading lessons.

Over the course of Key Stage Two, if complemented by quality vocabulary instruction, reading fluency practice, and a curriculum that builds rich background knowledge, the cumulative impact of this modelling should have a profound impact on children’s ability to comprehend what they read and not just on their performance in assessments. If every teacher in a child’s journey through Key Stage Two can develop this metacognitive approach to reading with a diverse offer of rich and challenging texts, there’s no telling what we might achieve.

While it has transpired that I have more to say about this subject than I initially expected, I would like to briefly come full circle and explain how my story of changing reading practice relates to education research.

Firstly, it highlights the importance of a robust exploration of the problem when seeking to effect change in schools. Because of the busy nature of our jobs, we are prone to jump to conclusions as I did after ​‘The Day of the Dodo’. It’s why I’m convinced that, of all the reports the EEF has produced, ​‘Putting Evidence to Work – A School’s Guide to Implementation’ is the most important for school leaders to understand.

It also shows the dangers of superficial compliance with evidence: it would have been easy for me to have skimmed the Improving Literacy guidance poster, seen ​‘Teach reading comprehension strategies through modelling and supported practice’ as well as words like prediction, questioning and summarising, and assume that my approach to reading was backed by the evidence. The ambiguity of terminology here doesn’t help — reading strategies and skills are terms that I’ve seen used interchangeably, so it’s difficult to know for sure that when we’re talking about reading strategies, we all mean the same thing. It’s an issue that we see repeated often in education discourse. For school leaders, ensuring their team of teachers share a common language and understanding of the distinguishing features of terms like these is as challenging as it is important.

No EEF report is an island. I was a subject leader who wanted to improve the quality of reading comprehension in my class and then my school. To do so required engagement with, and reflection on, recommendations from the ​‘Improving Literacy in Key Stage Two’ guidance, as well as those included in ​‘Metacognition and Self-Regulated Learning’. Managing change should be underpinned by the process outlined in the ​‘School’s Guide to Implementation’ and the successful translation of the evidence into practice at scale with a team of teachers will likely be determined by the degree to which the mechanisms of ​‘Effective Professional Development’ are followed.

In the busyness of school life, finding the time to thoroughly engage with evidence is a significant challenge, but if my experience of teaching reading has taught me anything, it’s that when our time is limited, making sure we use it effectively is even more important.

Rob Laight
_{Assistant Head Teacher at The Coppice Primary School
and an Evidence Lead in Education at Billesley Research School.}