I’ve just read a new e-book called Reading the Evidence: Synthetic Phonics and Literacy Learning, edited by Margaret M Clark OBE, a UK Visiting and Emeritus Professor who the About The Editor section says “has undertaken research on a wide range of topics and has developed innovate (sic) courses”.
Its announcement elicited some e-eye-rolling from members of the Developmental Disorders of Language and Literacy Network, and jovial suggestions that others buy, read and comment on it, but apparently I’m the only one with nothing more important to do (sigh).
Also on my reading list at the moment is The Influential Mind, about how to persuade people in the face of Confirmation Bias, our tendency to ignore or dismiss evidence that’s not consistent with what we already believe. Confirmation Bias is why presenting scientific data to the climate sceptic in your life never works. You can watch a video about it here.
Reading the Clark et al e-book thus also became an interesting exercise in thinking about my own thoughts. I’d set aside time to read the book in order to write what I hoped would be a thoughtful, informed response, but my brain kept coming up with other ideas. Did it keep switching off because of my own confirmation bias, or because of the standard of what I was reading?
The rush to e-self-publish
The self-published “Reading the Evidence” e-book was conceived when Clark met Australian education academics Misty Adoniou and Paul Gardner at a Glasgow conference in June 2017, after which they published three UK educational magazine articles, which you can read here. These formed the basis for chapters in the October 2017 e-book.
The e-book’s other contributors are three education academics from the UK (Greg Brooks, Henrietta Dombey and Terry Wrigley) and one from Australia (Robyn Cox). Appendices include statements from UK and Australian literacy teacher associations querying the evidence for (respectively) UK early years synthetic phonics teaching and the Australian Year 1 Literacy and Numeracy Check recommended in September by the Federal Education Minister’s panel of experts.
The opening sentence of the e-book’s acknowledgements is: “I suspect that when my six contributors, three in UK and three in Australia agreed to write for this book they were not fully aware as to the speed at which I work!” Perhaps an editor can improve the book’s grammar and punctuation before its launch at this month’s AARE conference. Page numbers would also be nice, but perhaps most e-books just have a % at the bottom of each page.
Strongly held beliefs
Early in Clark’s first chapter she includes a quote from a 2016 book co-written by “Father of Whole Language” Kenneth Goodman, who argues that phonics is irrelevant because people read for meaning not sound, and that English is too irregular to teach using phonics.
Of course people read for meaning, the question is how, and the answer is by deeply relating sounds and parts of spoken words to letters, and combining this decoding skill with their oral language skills (See this Thinking Reading blog post and/or Mark Seidenberg’s book and/or the work of Stansilas Dehaene for more). So I was surprised by Clark’s e-book Goodman et al quote:
“In this book we take on the formidable misconception: that reading involves the accurate sequential recognition of words and that accurate word recognition is necessary for comprehension (Goodman, Fries and Strauss, 2016:xx)”.
“No, accurate sequential recognition of words is precisely what skilled readers do,” insisted my brain. “Time to do the laundry”. But I soldiered on.
Clark says reading is much more than phonics. Great! Something we can all agree on. However, reflecting on her own researches (sic) about young fluent readers, Clark queries children’s need to read aloud before reading silently, and the need to teach spelling, though she writes, “The fact that spelling was being caught by these children does not of course mean that other children may not need to be taught to spell”. (p7%). “Interesting use of the double negative”, said my brain. “Time to put out the recycling”.
Clark also queries the evidence considered by the US government’s National Reading Panel and the UK government’s independent review of the teaching of early reading. Again quoting Goodman et al, she writes, “A disquieting picture is painted of the power wielded by large commercial organisations to influence government literacy policies, including in many developing countries, often falsely claiming a research basis for the policy” (p8%).
My recalcitrant brain wondered why she didn’t go for the trifecta, and suggest Australia’s government-sponsored National Inquiry into the teaching of reading was also dodgy, but that was possibly because it was Melbourne Cup Day.
Clark goes on to write, “There is not (sic) evidence for synthetic phonics as the required approach rather than analytic phonics” (p9%). “Maybe she’s still getting around to reading the UK and Australian reports,” helpfully suggested my brain. “The bathroom needs cleaning”. Anyway you get the idea of what Clark is on about, so now I’ll turn to the other contributors.
Policing, political imperative and profit
The chapter by Paul Gardner discusses Finnish educationalists Pasi Sahlberg’s theory that there is a GERM (Global Education Reform Movement) which spreads like a virus, with symptoms including, “schools get ill, teachers don’t feel well, and kids learn less”, and “standardisation and accountability” (p21%). He says Systematic Synthetic Phonics (SSP) is being imposed across the globe via a systematic process of “policing, political imperative and profit”.
Wow. I thought I’d simply gotten fed up with seeing children who could barely read or spell, studied the best available evidence about how to help them, experienced success and encouraged others to do likewise. I hadn’t realised I was covertly working for SSP Big Brother.
While I generally disagree with the Centre for Independent Studies on just about everything, so have some sympathy with Gardner’s concerns about its involvement in literacy policy-making, I see its Five From Five project is a valuable and solidly-evidence-based initiative. Everyone (yes! even big business!) benefits when kids are taught to read and spell well.
Much, much more complexity is needed
Misty Adoniou’s chapter says our flirtation with the Phonics Screening Check is a symptom of panic about declining literacy standards, but that “It is not the basics we require in Australian education, it is complexity” (referencing her own article in The Conversation last August).
Under the heading “Why is (sic) synthetic phonics, and the Phonics Screening Check, failing to yield returns?”, Adoniou’s limited knowledge of synthetic phonics programs is on display. She says they can’t account for morphology, so would spell “jumped” as “jumt”. She says they “focus on phonically regular decodable words such as ‘dog’ and ‘cat’, on the premise that this will lead students to be efficient decoders of non-regular words, such as ‘would’ or ‘one'” (p28%).
She’d be most welcome to visit my office to learn how synthetic phonics programs deal with the complexities of English phonology, orthography and morphology, including words with unusual spellings, any time.
Adoniou writes that, “phonological processing is only one component of the literacy puzzle”, suggesting that fast and first phonics and “high stakes phonics assessments (sic)” would make teachers neglect work on comprehension, fluency and vocabulary, because, “When we give all our educational policy attention to only one component skill, it is inevitable the other components will be sidelined” (28%). Just to be clear: nobody is suggesting doing any such thing. Also, as far as I can tell, most teachers can walk and chew gum.
Adoniou predicts a Phonics Screening Check wouldn’t work anyway because, “it fails to test some of the most common phoneme-grapheme correspondences in English, and indeed screens for a very limited number of the hundreds of phoneme-grapheme relationships”. Yup, that’s what screening is. If it were a diagnostic assessment, it would be called a diagnostic assessment.
She writes that the Phonics Screening Check “fails to provide useful information about the one skill it is designed to assess – phonological processing”. Any time she’d like to understand the difference between phonological processing, phonological awareness and orthographic knowledge, and how each one is assessed, I’d be happy to help.
Adoniou argues the Phonics Screening Check’s funding should be spent “giving teachers professional learning to support them to make informed diagnostic analyses and perform tailored interventions for struggling readers” (31%), though $22 million is not nearly enough, since these skills are needed by all early years teachers.
I wonder whether the experienced education academics who wrote this e-book ever lose any sleep over not having taught undergraduate teachers how to do informed diagnostic analyses of literacy skills, or how to intervene effectively with struggling readers. I wonder whether any of them are aware of the 2014 meta-analysis showing phonics is the only definitely effective intervention for struggling readers.
Adoniou worries that desperate parents of struggling readers turn to health professionals for help, giving non-teachers “huge influence” on education policy, and causing “the ‘medicalisation’ of general literacy instruction,” using “synthetic phonics programmes, or speech therapy designed for hearing impaired, or eye exercises and tinted lenses prescribed by behavioural optometrists”. Again, just for the record, eye exercises and tinted lenses are not recommended, and no speech pathologist would target literacy skills using a therapy approach designed for hearing impairment.
Adoniou writes “these medical interventions” are unsuitable for normally progressing children, and would be like giving all six-year-olds glasses: unnecessary, expensive and harmful. I hope the teachers successfully using synthetic phonics programs in mainstream early years classrooms can respond to this. Time to move on to the remaining e-book authors.
Fast and furious reform
Sydney academic Robyn Cox writes that we are no longer an outpost of the British Empire, then muses on her experience teaching in the UK, where she came to understand “the OFSTED culture and could understand the place of inspections in pursuit of the then Labor government’s mission to provide ‘…a good school for every child is our mission’ (Blair, 2001)” (sic, p34%).
She also states her main point on p34%: “Most particularly I want to argue that this Phonics Screening Checklist (sic) is just another one of many policy initiatives in which the Federal government is seeking to control what goes on in the schools in States by stealth.” “Ah, States Rights arguments, my favourite,” said my brain. “What’s for lunch?”
She quotes media releases about underwhelming NAPLAN results, which she says “prompted a media focus criticising the profession and the re-emergence of a public discussion of the ‘reading wars’. Resulting in the English teaching profession being maligned, judged and questioned.” (punctuation in original).
Recent Federal intervention has resulted in major changes to initial teacher education (ITE), including a literacy and numeracy test for prospective teachers, which Cox says is robustly trialled and is an effective test for its purpose, “Albeit, one which operates on a view of language which is counter to that underpinning the Australian Curriculum: English. Further to this, the clear line of sight between testing ITE students’ literacy ability and their own preparation for the teaching of literacy cannot be ignored”.
Does anyone know what she is talking about? I’m sorry, I have to move on.
There’s evidence and evidence
Margaret Clark’s Chapter 6 says that phonics should be included in early reading instruction, but “There is no evidence to support phonics in isolation as the one best method” (p44%). Well, of course, nobody disagrees with that.
Having described large quantities of research without saying much about its quality, she further concludes, “There is not sufficient research evidence for synthetic phonics as the required approach rather than analytic phonics”. She doesn’t say what might constitute “sufficient research evidence” for her to be persuaded, so I’ll leave that to you to ponder.
Her Chapter 7 is a critique of the 2006 UK inquiry into early literacy teaching (“the Rose Review”), beginning with the suggestion that it heralded the end of teacher autonomy, and introduced in synthetic phonics a “method not dissimilar to those used in Victorian classrooms”. “Ah, the Satanic Mills, my favourite,” said my brain. “Another coffee?”.
Clark argues that the Rose Review’s process and conclusions were both flawed, and ditto for the Clackmannanshire research. She prefers the conclusions of a 2006 report by Torgerson, Brooks and Hall, a narrow review of randomised controlled trials (thus excluding the Clackmannanshire controlled trial), which found no statistically significant difference between synthetic and analytic phonics.
Weighing the evidence
The e-book’s Chapter 8 is by one of the authors of this review, Greg Brooks, who firstly explains that, “I was convinced then, and still am, that theory suggests that synthetic phonics is more coherent than analytic phonics as a strategy for young learners working out unfamiliar printed words.” (p50%).
He writes that there’s plenty of evidence that systematic phonics instruction is better than unsystematic instruction or no phonics, but “still not enough (yet) to prove the superiority of one variety of phonics over another”.
Machin et al from the London School of Economics in 2016 analysed UK student attainment data and found the introduction of synthetic phonics had made an impact at ages 5 and 7, but by age 11 other children had caught up. They wrote, “There are long-term effects only for those children with a higher initial propensity to struggle with reading.” The kind of kids who most concern readers of this blog, in fact.
Brooks points out that no other type of phonics was studied to allow comparison of methods in this research, and that data analysed were from 2005-2010, before the Phonics Screening Check was introduced, so “if anything the findings might suggest that things were improving without it” (p51%).
He says the classic 1967 research by Jeanne Chall didn’t recommend one phonics method over another, but that in her 1989 revised edition, “Chall had concluded that synthetic phonics is superior to analytic. But we need to acknowledge that this still lacks experimental confirmation” (p51%).
Then, out of the blue, he calls the Year 1 Phonics test an “abomination” and asks “why has all moral sense of the fitness of things educational deserted those who advocate it?” (p52%). But explains this no further. My brain demanded another coffee.
The development and consequences of the Phonics Screening Check
Margaret Clark again writes the next chapter, first describing the UK Phonics Screening Check: an individual, mandatory test that asks children in the middle of their second year of schooling to read 20 real words and 20 made-up words to their teacher, with an originally public pass mark of 32, leading to a spike in children getting exactly 32 items correct.
Initial concerns Clark raises are:
- “The pass/fail decision resulting in many children aged between five and six years of age and their parents being told they have failed”;
- the inclusion of 20 pseudo words (this is SOP in psychological assessment, as it allows assessment of word attack skills), and possible effect of “coloured alien figures” beside each (presumably to make sure children knew they were pseudowords).
- the fact that older children (of course) tend to pass the test more easily than younger children,
- the “demand” that children who “failed” resit the test the following year,
- “possible effects on some successful readers who might yet have failed this test”,
- “the lack of any diagnostic aspects or suggestion that other methods might be appropriate for some children who have failed” (“whatever that means”, said my tired and cranky brain).
Clark is also concerned that phonics experts were involved in test development, but says she doesn’t know if any of them were concerned about the same things she was. “Perhaps not,” said my unruly brain.
In 2012 58% of UK children passed the Phonics Screening Check. Clark gives us to understand that lot more children have been successful since then, but her focus is figures showing younger children do worse than their older classmates.
Clark writes, “Why spend money on developing such a pass/fail test, and why test all Year 1 children (about 600,000) rather than extend the use of diagnostic tests such as Reading Recovery, providing as it does diagnostic information and proven intervention strategies with long term effects?”.
“OK”, my brain said, “Game over. Not only is Reading Recovery not a diagnostic test, but a recent large-scale study in NSW found no evidence it has more than a short-term positive impact.”
Terry Wrigley’s chapter heading indicates it’s about political interference in education. He writes that schools in the UK were already teaching phonics as part of a mix of methods in 1990, and that the words “systematic” and “synthetic” phonics sometimes get confused, with solid evidence backing the former but not the latter. He writes that, “Phonics teaching does not need to involve rigid reading schemes which deprofessionalise teachers and marginalise other aspects of reading including the children’s pleasure” (p63%)
I agree, of course, but Wrigley’s example suggests his preference is anything but systematic. He describes incidental phonics while reading a story about Bella blowing bubbles in the bath, which he contrasts with some rather dry sentences from a beginner’s decodable reader: “Kit is in Don’s cot. Pod tops Sid”.
Not all decodable books are like that. Today I gave a beginner one about hens escaping from their pen, with sentences like, “The hens got in the van! ‘Get the net!’ said Mum”, till finally the last kid catches the last hen: “Meg got a hen. The hen got Meg.” (picture of confused-looking child with chook poo on her dress, always gets a laugh).
This child would be simply unable to read Wrigley’s preferred book, because it has digraphs, consonant blends, vowel variations and two-syllable words. Asking him to read it would be setting him up to fail. He read about 95% of the words in my book independently, and enjoyed this success. As we teach him more sounds and spellings and word structures, he will gradually be able to read harder books, till he can tackle the Bella in her bath book by himself. Till then, we’ll just read children’s literature to him.
Wrigley says that Year 2 reading scores on the UK’s equivalent of NAPLAN have only gone up a couple of percentage points (87% to 89% at L4) between the years 2008 (when synthetic phonics was first introduced) and 2015 (after which the test was made a lot harder). However, a couple of pages later he says that there has been a reduction in children failing the Year 2 tests, from 15% in 2011 to 10% in 2015. It all happened after synthetic phonics teaching was mandated, but he says only 3% of it occurred after the Phonics Screening Check was introduced. I’d still be pretty happy with that sort of reduction in national failure rates.
UK Phonics Screening Check scores have gone up from 58% in 2012 to 81% in 2017, which Wrigley attributes to teaching to the test, and says hasn’t helped reading for meaning. He also wants the test discontinued because it doesn’t diagnose individual children’s areas of difficulty, but that’s not the purpose of a screener, it merely identifies kids needing further attention.
Like others in this e-book, Wrigley thinks the Clackmannanshire study of very successful synthetic phonics was overrated, and the West Dunbartonshire study, which used both analytic and synthetic phonics plus a lot of other strategies, was possibly more successful. However, he thinks that the author of the UK’s national inquiry favoured synthetic phonics, ignored evidence that didn’t support it, and set about “finding substitute evidence” that supported his view (p68%). Schools started being forced to use synthetic phonics by school inspectors, and Phonics Check data was used in making decisions about school closure and privatisation. He likewise queries the process and outcome of the US National Reading Panel.
His conclusion is almost as long as the rest of his chapter, and says that UK kids’ reading is not improving, its Education Minister is abusing his power, the privatised school inspections agency is coercive, synthetic phonics fits the UK’s nostalgic conservative mindset and the US’s Christian Right’s “need to uphold a divinely rule-bound social order”, and that all this is alienating and stultifying for children.
Teachers are given commercialised training and required to deliver a packaged curriculum, instead of adopting developmental approach where (here he quotes Jonathan Glazard 2017 p178) “blending and segmenting at the level of the whole word is a logic (sic) place to start developing this skill. Children can be asked to blend and segment compound words (tooth brush/tooth/brush)” and then syllables, onset and rime and finally phonemes. This runs so completely counter to everything I’ve read about what’s effective that I will have to save commenting on it for another blog post. He’d rather let teachers choose their own assessment tools “as appropriate to the individual child”.
Neglected lessons from successful classrooms
Henrietta Dombey says that the introduction of synthetic phonics in the UK has led to headteachers urging early years staff to “stop spending time on books and concentrate instead on reading” (??), money being spent on synthetic phonics materials and a focus on phoneme-grapheme correspondences.
She says an exclusive focus on synthetic phonics is not based on good evidence. Correct. Young children learning to read need to work on phonemic awareness, phonics, vocabulary, comprehension and fluency, and of course keep building their oral language skills. Nobody disputes this. But that doesn’t mean the instructional chaos of “balanced literacy“.
However “balanced literacy” is what Dombey advocates. She writes that “Teachers in the most effective schools constantly adapt to the children in their classrooms, rather than faithfully follow a plan” (p75%).
She also writes that “It has been known for well over a hundred years that fluent readers do not process texts one word at a time, much less one letter at a time (Cattell, 1886)”(p76%). My brain wonders what she thinks reading scientists have been doing since then. Then she also quotes Ken Goodman and I’m sorry, that’s me, over and out.
I can’t tell you for sure whether I didn’t learn much from this book because it doesn’t say much that is interesting or useful, or because of my own confirmation bias.
What’s the purpose of this e-book?
Clark’s stated rationale for the e-book is “to enable readers to make their own judgement as to whether the claims being made for synthetic phonics as the method of teaching reading are justified and indeed whether it is true that academics are, as (UK Education Minister) Nick Gibb claims, against phonics” (italics in original, 13%).
Australia’s education ministers are meeting in Hobart on December 8, and the proposed Year 1 Phonics Check will be on the agenda. I discovered this from the Australian Literacy Educators’ Association facebook page which adds, “Now is the time to lend your teacherly voice to this important moment in time,” and encourages teachers to buy Clark’s e-book.
What exactly teachers are supposed to do after reading it is not made explicit by ALEA, but Clark says that the (ALP) WA Education Minister has been persuaded to tell the (Liberal) Federal Minister she won’t be implementing the Phonics Check in WA, and suggests that other state ministers might also be so persuaded.
If you’d like to apply a little countervailing pressure, see this petition.