Kirsten Snook is a Teaching and Learning Adviser for English at Herts for Learning.
Confession time. I’m quite an analytical person. Always have been. I like to know the parameters of things, what’s expected of me and how to get there…maybe we all do. It is this that led me to looking into readability of the test papers and trying to find out what is being expected of the children and therefore us as teachers.
At YR-Y2 we have the book-banding system; it’s not perfect (and it’s certainly not the whole diet) but it is a finely-graded system that enables us to gauge whether the book is likely to be readable, too hard or too easy. It also means we can base our assessment of whether a child is reading and comprehending as they are supposed to on text of sufficient complexity. In their 2014 report ‘Ready to Read?’, Ofsted found across a sample of Stoke-on-Trent schools that books were not being matched closely enough to children’s developing reading ability, and this meant struggling readers had to resort to too much guesswork and other unreliable strategies. Closely matching texts to readers, in YR-Y2, is imperative if they are to develop healthy strategies, apply their phonics, and become more fluent and comprehending active readers. But we know all this, right?
What about life beyond book bands? What do we use then to ensure books are matched appropriately to readers? And do we need to?
Frequently schools view Lime band as a launch pad, a sign that the reader is about to fly the book-coop and be able to choose their own books which interest them and that they understand. Yes, we need to equip readers with the knowledge and skills to be able to select their own material, but we also need to know if what they are selecting is going to pave the way towards age-relatedness. This is where my search for the test developers’ Tool of Choice came in.
Having pasted/typed the passages and then turned them into simple ‘text files’, so that they were devoid of formatting, I ran each text used in the sample and actual test papers, at both KS1 and KS2, through a total of 10 tools. The one that appeared to be most consistent, logical and showed a stepped approach across the text extracts was Lexile. This is an online tool (https://lexile.com) where you can enter a text title and it will tell you what Lexile measure it comes out as (this has been analysed by a tool but also tested out on children and mediated), or you can paste a section of text into it and it will analyse it for you there and then. What is crucial about feeding text in manually is all the guidance you will see on these websites about the sample you enter (for example passage length, where to take passages from in the text, removing headings/subheadings, etc). This is why some have tweeted that when they’ve analysed the papers using varied tools or processes they have come out at a reading age of 15-16, or even 18-20, and that considerable variation can be seen between differing paragraphs.
This then gives you a rough idea of readability of that analysed passage. The software crunches your passage in terms of: word frequency (how rarely-occurring/Tier 2-3 is the vocab that has been used), sentence length (average number of words per sentence) and average number of syllables in words. It can even tell you whether any of those aspects are above what is expected for the age-range, so whether there is a spike in an area of challenge (e.g. the texts themselves aren’t harder than last year, but the sentence lengths readers are required to cope with demand a high degree of fluency, in all its shapes and forms).
My methodology was to use as large a sample as possible (e.g. ‘Wild Ride’ from ‘The Last Leopard’ I used all 398 words), ensure any rules about punctuation, formatting etc were followed (so as not to skew the analysis) and to record all the results as they came back in. From this I could see the more reliable patterns emerging around Lexile, which were also borne out by ATOS, another widely-used tool. Now, I’m not claiming that the Lexile measure is the ‘right’ one to use, but evidence would indicate that it is strangely in-line with a spiralling increase in text complexity seen on the test papers.
An interesting finding was that, at KS2, sentence lengths are coming out high for their age. Teaching implications? Teach the children to parse the meaning in phrases; model and expect fluency; ensure readers have ample opportunity to practise on easier or familiar texts. BUT proportion of polysyllabic (3-syllable+) words used appears low for their age. Don’t be lulled though… the vocabulary stretch is coming via Tier 2-3 language rather than longer words…breadth rather than depth of vocabulary. Depth may be hiked later on, or even next year.
This is a perfect example of how a readability measure can give you a good starting point, but you as the human need to then do some qualitative analysis.
For me, it’s like the Simple View of Reading, but for text selection – we must bear both in mind. E.g. would you keep giving a poor comprehender a book without words? Of course not. You could use the picture book to teach the explicit skill of inference on its own via the booktalk but then they need to apply this on text, at lower slopes first and build up. They won’t get better at doing this on denser texts without increasing the density of text, gradually. (Test developer materials from STA refer to this as the ‘lexico-grammatical density of the text’.)
Considerations for choosing a text:
Knowledge of child: Will they have this background knowledge? Will they pick up on subtleties of authorial intent/shades of meaning? What is their decoding like (both fluent and nearly fluent), both phonically regular and irregular words? What about attention span or working memory?
Poorer comprehenders are sometimes just too overloaded to be able to connect or infer fully. Sometimes they are so used to reading being a struggle, a wade through mud, that they have lost sight of the bigger picture. Sometimes they are stuck in a rut of decoding for decoding’s sake. They need to have opportunities to gobble up sheer mileage of text reading, so that they can once again feel what it’s like to repair inconsistencies and retrieve, connect, infer etc. You may have seen me quote this before: “Easy reading makes reading easy” (Rasinski). Conversely “The rich get richer and the poor get poorer” (Stanovich). Stronger readers get loads of practice at their lower slopes, they are usually given text to read (cross-curricularly, English, etc) which is easy for them to read fluently and so they get much better at it, acquire more vocabulary, get more specified grapheme-phoneme representations in their visual memory which helps their spelling (Rasinski, Stuart & Stainthorp)… and so it goes on.
Knowledge of task: Are they reading this for the first time/re-reading? Are they finding answers to something/reading for enjoyment/ reading against the clock/reading in a pair? Is it cross-curricular?
Imagine you are focusing on a targeted group at the independent reading stage of a whole-class Guided Reading session. The group need to access the text sufficiently to be able to get at the LO. Two major scenarios spring to mind:
– If LO/group target is mainly unpicking vocabulary, then the text may need to be at the challenging end of their range. They will need to slow down fairly regularly. This is like driving in country lanes – you end up going from 1st to 2nd to 1st gear quite regularly. Your attention is very much on the immediate curves in the road, any rabbits jumping out at you, changing gears in response, the fine detail etc. – definitely not on the rural views!
– If LO/group target is to develop an aspect of fluency or comprehension, e.g. prosody (pitch, juncture, stress) to reflect and understand authorial intent, then maybe a text at the easier end of their range is best. This is like motorway driving – they need to be so automatic at gear changing, steering, use of mirrors, where pedals are etc that they can anticipate what other drivers might do next (e.g. pull out suddenly) and have a good big picture of the road far ahead (can see peripheral views!). This expectancy is something that reading fluency is linked to too; it is like a vessel for fluency. Expectancy allows the thinker to unconsciously call up what in those situations can happen (i.e. be aware of others’ intentions/authorial intent)…e.g. big gap between me and car in front = car may fill it. If the reader never experiences that, then they miss out on development opportunities around making full inferences. They also are more likely to ‘gobble up’ the odd new bit of vocabulary as, without the working memory being too overloaded, they can focus on the maybe ten newish words to them, the pronunciations, phonological representations and meanings in context.
So, lastly to:
Knowledge of text:
What quantitative elements? What qualitative elements? What does the child need to develop next? What support is available for what you want them to do with the chosen text? What will fire and inspire? What will give them the will – and the thrill – to find that intrinsic motivation to develop the skill….? For a struggling reader at Year 6, ‘Skellig’ is possibly going to do that much more than say ‘The Iron Man’, yet both are of similar Lexiles. Very different books, very different outcomes wanted yet very similar decodability. And this is what I mean. The numbers give you a starting point. They give you knowledge and with that power – power to make healthy, informed decisions or equally power to leave the decision-making up to a system of numbers alone. I for one would rather I had the knowledge to be aware that whilst Skellig contains fab qualitative elements for Y6 comprehension, it doesn’t pave the way quantitatively for the lexico-semantic density, stamina, fluency etc of Y6 SATs. To borrow Penny’s metaphor, it doesn’t quite give me the Boxing Gloves of Power to be ready for anything.
You will base text choice on the following interconnecting cogs:
So what does this mean for teachers at the chalk face?
In terms of who or what determines whether a text is right for a child, the answer is exactly the same as at YR-Y2…you do! A finely-graded system is helpful for starting the decision ball rolling and certainly for ensuring that assessments are based on sufficiently high expectations, but a system is just that. You ultimately have to mediate the decision that that system informs. No, it’s not all about sentence length etc but one thing’s for sure: if you can’t parse and digest a longer sentence (or polysyllabic word) then that is definitely going to be a barrier to unpeeling the layers of hidden meaning and authorial intent. Whilst ‘Wild Ride’ had an average sentence length of 20-21 words, it also included a 54 word sentence and even one of 61 words. Now that’s some parsing! Anyone who is not sufficiently fluent, and who has not spiralled up in degrees of complexity (lexico-semantic density again!), is likely to come royally unstuck at sentences like that. This information is intended purely to assist your decision-making around text choices for your children in the hope that with your Boxing Gloves of Power you are, in the words of Traction Man, ready for anything.