Tag Archives: teaching children to read

THIS is a BORING book!

I’ve just watched a great 2016 BBC4 documentary called “B is for book”. It follows a group of London children from their first day at school for a year, and explores how they learn to read.

The kids live on a public housing estate in Hackney, and most speak languages other than English at home.

The film is not currently on the BBC website, but a few people have put it on YouTube. The version I watched is here, and you might like to keep it open in a new tab while you read, so you can quickly find and watch the interesting bits I describe below.

You’ll love all the children, but I was most entranced by a little boy called Stephan. An honest child with a low tolerance for Educrap, he looks and behaves a lot like a little boy I worked with last year, also a twin from public housing inclined to slide under the table.

At 19:42 on the video clock, the two children having the most difficulty learning to read in the film, Maria and Stephan, are asked, “What’s the hardest word you know how to spell?” First, they do this:

Continue reading

Decodable texts and lesson-to-text match

A state election looms here in Victoria, and parent-run group Dyslexia Victoria Support (DVS) is petitioning politicians to provide decodable books to all kids starting school in 2019.

Decodable books provide the reading practice for phonics lessons. They include sound-letter relationships and word types learners have been taught, plus usually a few high-frequency words with harder spellings needed to make the book make sense, which are also pre-taught.

Decodable books would replace the widely-used predictable/repetitive texts, which encourage children to guess and memorise words, not sound them out.

At the moment, children might be learning about “i” as in “sit” in phonics lessons, but take home a predictable text that might contain words like “find”, “ski”, “shield”, “bird”, “friend” or “view”. Instead of helping kids practise the sound-letter relationships they’ve been taught, their home readers can undermine this teaching.

DVS’s campaign hit the statewide media this weekend, yay, with an article called “Dull, predictable: the problem with books for prep students” in Fairfax newspapers.

NSW already funds decodable books

The NSW state government has recently given public schools $50 for each child in their first year of school to buy decodable books.

I completely agree with these reactions from AUSPELD President Mandy Nayton and NSW SPELD‘s George Perry:

Several Victorian teachers have recently told me scarce funds are the main barrier to their schools buying decodable books for all their beginners.

Following NSW’s lead in Victoria would not be expensive. 56,766 children are now in their first year at Victorian government schools, so $50 for each child equals $2,838,300. An extra $1,219,050 would extend the offer to the 24,381 preps in non-government schools. Chickenfeed in the context of the state budget.

Does decodable text work?

Professor Tim Shanahan of the University of Illinois Center for Literacy, an internationally recognised authority on literacy-teaching, recently wrote a blog post called “Should we teach with decodable text?”. His point is summarised most succinctly in one of his comments: “No matter what you’ve been told or which group of kids you are concerned about, there is no research evidence supporting the idea that decidable (sic) text works.”

Myself and many others found this odd, as decodable texts aren’t meant to be used in isolation, they’re part of explicit, systematic phonics teaching, and Shanahan was on the US National Reading Panel which said phonics teaching is effective.

More recently, a 2012 systematic review found that “decodability is a critical characteristic of early reading text as it increases the likelihood that students will use a decoding strategy and results in immediate benefits, particularly with regard to accuracy”.

Shanahan admits the lack of a clear, agreed definition of decodable text means the available research is kind of a mess: “with little evident agreement about what decodable text is, what it should be compared with, and what outcomes we should expect to derive from it.” He says, “Perhaps the best study of the problem was conducted by Jenkins and colleagues (2004).”

I was keen to read this article, though since I’m not affiliated with a university, accessing it cost me $AUD59.30 (no, I’m not joking, I agree with George Monbiot that scientific publishing is a rip-off). No wonder there’s a knowledge gap between researchers and practitioners. Anyway, I coughed up, downloaded and read it.

Whatever the text type, learners were prompted to decode

At-risk first-graders in this study were divided into three groups, a control group and two experimental groups who had one-to-one tutoring using the phonics-based program Sound Partners for 25 weeks, 4 days a week for 30 min per day. One group’s reading practice was with decodable storybooks, and the other group read books with uncontrolled spelling patterns. Both groups outperformed the control group, but not each other.

The article says that during reading practice, “If students hesitated for more than 5 sec, or misread a word, the tutor prompted them to use previously taught phonic skills (e.g., isolated specific letters in the word and coached the student to figure out the word), or supplied a letter sound or word, as needed.”

Nobody told these kids to look at the picture and guess, think what word might go there, or otherwise use multi-cueing type strategies. Even when they were reading non-decodable books, these children were encouraged and assisted to decode by tutors who understood phonics. Perhaps this matters as much or more than book type.

I wonder whether book type affects how consistently parents encourage and assist young children to decode, rather than guessing words.

Establishing a useful definition of decodable text

To some researchers, decodable text is text containing only the most regular spelling patterns in the language, i.e. it’s a property of the text alone. From a practitioner’s perspective, this is not an interesting or useful definition, it’s far too simplistic.

Others see decodability as a property of the relationship between text and learner knowledge, so that as children learn more of our spelling code, more books become decodable to them.

However, learner knowledge is usually part of the dependent variable – the thing being tested and measured – in scientific reading research. It thus can’t also be part of the independent variable, which the experiment changes or controls while measuring the effect on the dependent variable.

To help practitioners and avoid research quagmires, the definition of decodability really needs to focus on lesson-to-text match. Practitioners are in control of what they do, not how children respond.

I’d like researchers to define decodable texts as reading practice activities including the phoneme-grapheme correspondences, word types and high-frequency words which have been systematically and explicitly taught, and specify what percentage of words (say, 5%) fall outside this.

Kids should be reading texts they can’t read yet?!

Tim Shanahan’s blog post includes the statement that “even beginning readers should be reading more than decodable texts”. Which seems a bit like saying “even beginning readers should be reading words they haven’t been taught how to read yet”, but perhaps this is yet another problem with definitions.

His definition of reading might be like the Australian Curriculum’s: “To process words, symbols or actions to derive and/or construct meaning. Reading includes interpreting, critically analysing and reflecting upon the meaning of a wide range of written and visual, print and non-print texts”. In which case he might mean that children should be read stories and taken to art galleries and interpretive dance performances. All good, though this definition of reading wouldn’t pass a pub test.

However, I imagine his definition of reading is more like the International Literacy Association’s definition, since he is a past President (along with “father of Whole Language” Ken Goodman and Marie Clay of Reading Recovery fame): “The process of simultaneously extracting and constructing meaning through interaction and involvement with written language”. This could mean listening to others read, I suppose.

He might also mean that beginners should be encouraged to attempt words containing spellings they have never been taught. If he means mainstream kids and about one word in 20 during independent reading, or one word in ten during supported reading, that’s probably fine.

The proportion of untaught spellings which makes a text too difficult is probably highly learner-dependent. Kids with extensive experience of reading failure can, in my experience, quickly melt down and refuse to continue once they start making mistakes.

Self-teaching and signal-to-noise

Shanahan’s over-arching concern seems to be that beginners not be overly shielded from the full complexity of our writing system. Much of what skilled readers know is self-taught. Once we can read at a basic level, we gather data about how words are written through the process of reading (see the Self-Teaching Hypothesis, and this dense but interesting journal article on Reading as Statistical Learning).

Shanahan writes, “Presenting students with lots of decodable text, text that’s much more regular that normal text, might mess up some of these cognitive calculations”.

Well, yes, once kids have the basic knowledge and skills that allow this self-teaching to commence. Until then, it’s all noise and no signal. I’m reminded of the day one of my clients surprised his dad by asking, as they sat on the train at Alphington Station, “Why is there a ‘ph’ in Alphington?”. They’d been through that station many times before, but he only asked about the “ph” after he could sound out the rest of the word.

I’m always delighted when my clients start independently reading environmental print (shop signs, chip packets, car stickers, menus, whatever) aloud. This seems to be an important reading milestone, and one I wish researchers would investigate more closely (or maybe they have and the results are behind an expensive paywall, where I can’t see them).

Shielding beginners and strugglers from knowledge of the full horror of English spelling simply isn’t possible. Today’s kids are surrounded by spelling complexity for most of their waking hours. We can hardly ask them to avert their gaze from every word containing (a) spelling(s) they haven’t yet been taught.

Teaching needs to help beginners and strugglers discern the signals among all the noise. Decodable books strip the noise back, making important signals at first a lot clearer and easier to learn, and they encourage kids to sound words out rather than guessing them.


The definition of decoding (or “glamping with David Hornsby”)

Last May, educational consultant David Hornsby spoke at Sydney University about “The Role of Phonics in Learning to be Literate”, and his talk was recorded and put on YouTube.

I recently heard from the Developmental Disorders of Language and Literacy Network that the video was being shared widely by teachers on social media. So I took a look.

Mr Hornsby begins by explaining that his definition of decoding is “making meaning or comprehending”. Here’s his slide saying this, in case you think I made it up:

In the solidly-evidence-based Simple View of Reading, the term “decoding” usually means word identification, whether this is achieved by sounding out words, or instantly and without conscious effort, once a familiar word’s spelling, pronunciation and meaning are bonded in memory (via orthographic mapping, or in dual route models, via the lexical route).

Sometimes researchers use “decoding” to mean only sounding out words, but in both its broader and narrower definitions in the reading research, decoding’s key focus is linking print to speech.

Decoding is not about semantics, syntax or context, which fall on the Language Comprehension side of the Simple View of Reading equation (Decoding multiplied by Language Comprehension equals Reading Comprehension). Continue reading

Related Posts Plugin for WordPress, Blogger...