In the summer of last year I came across a paper and some articles by Daniel Willingham in the Washington Post which brought to mind a nasty flashback.
A nasty flashback – the heart of darkness:
Back during my second year of teaching, I was asked to work with a Local Authority Literacy Consultant. We would trial some of the National Literacy Strategies materials on guided reading – rarely used at the time and still rarely used now in secondary schools. She was a very kind and well-meaning, experienced teacher: keen to support someone early on in their career with a top set Year 7 group who were eager to be stretched.
The focus of these sessions was to be on developing the students’ “reading strategies.”
The advisor brought in some posters about skimming, scanning, empathasing, questioning, predicting, highlighting, inference, deduction and ‘reading backwards and forwards’ (which was and probably still is a real thing).
It was 2002. It was the future. The posters included some ‘state of the clip art’ images of bright pink and bizarrely orange faced children who were reading books.
If you tolerate this…
We worked our way through team teaching some skimming lessons. They went ok, though I now suspect many of our top set, 11 year olds were probably wondering why we were modeling such a straightforward process in such great detail. We weren’t stretching them, but they tolerated us.
We worked our way through team teaching some scanning lessons. A bit better, but most of these students could scan for evidence in the level of text we were looking at already. Even in the easy peasy days of ‘the noughties’ the Key Stage 2 SATS were at least a little bit challenging and, it turns out, Holes was already part of the EYFS curriculum – high expectations. Again though, we were tolerated.
Then we worked our way through a lesson on close reading. I was teaching close reading. I thought I was teaching close reading. We looked at an extract about the warden who paints her fingernails with poisoned varnish, then attacks another character with them. We gave the students a number of quotations from the passage. We gave them a question: How does Sachar present the warden in this extract? We gave them some stock phrases they could use: this suggests, this implies, this heightens the impression that, this escalates the idea that, this conveys. We modeled how to use these. Then the students tried it in pairs. Then independently. Again, we were tolerated. I thought I was going great guns.
The following week we tried it again with some non-fiction. We were moving into the realms of close reading. It all fell flat. I hadn’t been teaching close reading. I didn’t really appreciate why back then and I still feel I’m working on getting my head round it.
You Oughta Know
So, what did this ‘nightmare’ inducing paper by Daniel Willingham say. Well, here’s the full text and here’s a brief summary:
There is a correlation between listening comprehension and reading comprehension. However, the differences between listening and reading, particularly the demands of decoding letter strings, make reading comprehension more complex than listening comprehension.
When listening, speakers can periodically check our comprehension either by checking our non-verbal cues or through questioning us. Writers of texts can’t do this as they aren’t generally present when we’re reading their texts and can’t amend what they’ve written to suit us if we are confused. Equally, when listening, we can ask questions of the person creating the spoken text. When we read, we can only ask questions of the text, go back and read the text again or seek answers within ourselves.
Limits in decoding skills, vocabulary and subject knowledge therefore act as barriers to comprehension.
Three overarching routines are, as a consequence, important:
- Monitoring your own comprehension to decide when you need to re-read a text
- Making links between the information in different sentences
- Making links between the text and what you already know
There are numerous studies in strategies which support these routines. However, out of these 481 studies of reading comprehension strategies, only sixteen fulfilled both of the following criteria:
1. They had been peer reviewed
2. They showed a causal relationship between the strategy and the improvement
Only eight out of these sixteen strategies “appear to have a firm scientific basis for concluding that they improve comprehension in normal readers.” Just two of these have been studied in enough depth to provide an effect size. There is an issue as the effect size given by the original testers is impacted by the tests being designed by the researchers. When independent tests were used, the effect sizes were smaller. Experimenters tend to use texts and questions which are well suited to their strategies performing well. Despite this, these two strategies have a significant effect size: question generation (0.36) and multiple strategy instruction (0.32). Most research has been into individual strategies rather than comparing the strategies or unpicking which strategies might be best for which students. There is also little evidence of strategies having any impact before 3rd grade (Year 4). When students’ working memories are focused on decoding, there will be little space left for comprehension strategies.
Willingham’ view, which he’s expressed in a range of publications since is that, based on the evidence he’s seen, that “Reading strategy programs that were relatively short (around six sessions) were no more or less effective than longer programs that included as many as 50 sessions.”
Willingham describes reading strategies as tricks rather than skills. They are shortcuts to a surface understanding of a text. They have impact – though it’s unnecessary to spend weeks practicing them. However, as every text is different, comprehending each text actually requires the reader to hold the vocabulary and the background knowledge to unlock that text.
Where do we go now? Where do we go?
Every other year or so the makers of shaving razors produce an updated version of their product.
The risk here is that I suggest a way forward, then keep adding extra reading razors to enable students to read close, closer than close, closer than you could ever imagine. If I do this, apologies. This is my current best attempt and it’s heavily influenced by Reading Reconsidered by Doug Lemov.
Firstly, let’s go back to the holes in my Holes lessons. At the time, I thought that the reason these worked, whilst the subsequent non-fiction lessons failed was because the novel was pitched at the right level and the non-fiction was pitched too high. Perhaps counterintuitively, I now think the opposite. There was less to say about the Holes extracts than the non-fiction, the students weren’t learning any new vocabulary from the section of Holes which would help them in the short or the long term and they were unused to having to struggle with texts so, when it came to the non-fiction, with its odd ways, the tricks they’d learnt weren’t enough.
Instead, I now believe the focus needs to be on:
1. Exposing students to challenging texts on a wide range of topics in order to increase their background knowledge
2. Selecting texts which exemplify excellence and basing your planning around these
3. Implicitly and explicitly teaching tier two vocabulary and subject specific terminology
4. Increasing the amount students are thinking and writing about both the content and crafting of texts
This is how we do it:
Each Key Stage 3 lesson will begin with a ten minute Fluency Fix session. This will incorporate:
- Explicit vocabulary teaching using a mixture of tier two words which relate to the core text and other texts that students will be studying that half term.
- The learning of quotations from the core text to be used in the end of term exam.
Over the coming term in Year 7, alongside studying Beowulf, students will have a weekly close reading lesson and a writing lesson. This term’s close reading sessions will focus on narratives. In subsequent terms, we’ll move on to non-fiction. For the next six weeks, we’ve chosen extracts from the following texts and put them in this Wild Adventures Anthology:
- Heart of Darkness – a narrative featuring a journey through challenging terrain
- Lord of the Flies – a narrative featuring two contrasting characters in a challenging landscape
- Treasure Island – a narrative featuring a threatening character
- Jaws – a narrative featuring a threatening creature
- Witch Child – a narrative featuring a wonderful discovery
- Metamorphosis – a narrative featuring an increasingly desperate situation
We’ve selected these as they incorporate a range of complex vocabulary. They’re also great for exploring both language and structure.
In order to try to maximize the impact of these text choices, we’ve prepared teaching scripts in the style of those in Reading Reconsidered. We’ve used the comments function on Word to highlight where questions should be used to draw out meaning from the text. Here’s the document for the Heart of Darkness session. These scripts incorporate a number of readings of the text – conveying to students that they shouldn’t expect to understand the text on their first reading. We’ve mixed the following types of reading:
1. Contiguous reading – working through the text from start to finish
2. Leapfrog reading – jumping through the text to explore a specific image, theme, character
3. Line by line reading – analysing a part of the text in great detail
During the first reading, opportunities are taken for implicit vocabulary teaching.
Following on from this, the other readings feature text dependent questions, moving from establishing the literal meaning of parts of the text to analysing the deeper meaning of the language or structure used by the writer.
We’ve used this grid from the Reading Reconsidered training to devise these questions. This was a real eye-opener for me as I’ve had a tendency in the past to jump to analysis too quickly, before students have understood the literal meaning. We’ve also built in some of the question types Andy Tharby has provided in this post on an approach to improving analysis. Finally, in some of the more analytical questions, we’ve used more tentative language to open up a culture of error and exploration.
I’ll keep you posted with how much closer this gets our students.