Synthetic Phonics Teaching and Testing in the UK 

​

Every parent in the UK knows their child is going to undertake the Phonics Screener Check (PSC) at the end of Year One and that Synthetic Phonics is the DfE's phonics approach of choice.

​

Children will be expected to look at around 85 graphemes (consisting of 1 or more letter) and blend them into a word.  The pass rate is 32 out of 40. 

This is actually a very simple task and one that over 85% of children are capable of learning within their first year of school when they are offered specific experiences and are able to learn at their pace. This is the reason why most children do not achieve these skills before the end of their first year of school- it is not that they are not capable, it is that they are not given the optimum conditions.

All children need to do is look at a letter (or letter string) and say the associated 'sound'. This is similar to looking at a dog and knowing to say 'woof'. They also need to be able to 'hold' those sounds together, from beginning to end (all through the word) to 'blend' them into a word unit - a spoken word. This doesn't even need to be a real word. If the letters given are only to represent 1 (or sometimes 2) sounds, then the children only need to be able to recognise and blend that limited number. To do this they need phonemic awareness. Phonemic awareness relates to the ability to isolate, segment and blend sounds and no letters are needed. Children with poor phonemic awareness eg Dyslexic students will have poor phonemic awareness. They can sit in synthetic phonics lessons every day, say the 'letter sounds' when presented with them eg on flashcards, but still be unable to pass the phonics screener check - they can't blend. 

The National Institute of Health (NIH) indicates that nearly all children have the cognitive capacity to learn to read, estimating that only 5% of young readers have severe cognitive impairments that would make acquiring reading skills extremely difficult. While the remaining 95% of students have the capacity to read, not every student will learn to read under the same conditions. The British Dyslexia Association (2018) now reports an incidence of 4% cases of severe dyslexia and 10 % less severe. It also shows that phonics as currently undertaken, whether basic, analytic or synthetic is not the answer to the dyslexic problem. It would appear to be necessary but not sufficient.” (p.1)

In the early stages of reading, phonological awareness is the strongest predictor of reading progress (Stanovich, 1986). Nagy, Berninger and Abbott (2006: 136) note that “an increasing body of research has documented how a variety of symptoms of reading disability can be traced to a basic phonological deficit (e.g., Adams, 1990; Morris et al., 1998; Olson, Wise, Conners, Rack, & Fulker, 1989; Shankweiler et al., 1995; Stanovich, 1986. Adams et al (1998) indicate that a child’s level of phonological awareness at the end of kindergarten is one of the strongest predictors of future reading success. Yet they also note that “more than 20 percent of students struggle with some aspects of phonological awareness, while 8–10 percent exhibit significant delays. Early intervention is crucial and can make a real difference to students with limited levels of phonological awareness.  The NRP found that children as young as four years of age benefited from instruction in phonemic awareness and the alphabetic principle when the instruction was presented in an interesting and entertaining, albeit systematic manner.

 

A critical stage of learning to read is the mastery of decoding (Stuart, Stainthorp and Snowling 2008) which requires an understanding of the mapping of phonemes to graphemes. Ehri’s theory of reading development (Ehri 1992, 1995) initially proposed four phases; pre-alphabetic, partial alphabetic, full alphabetic and consolidated alphabetic. Later in her career, Ehri allowed that there may be a fifth stage, the Automatic-Alphabetic stage – which when most words are known by sight, and when decoding is almost immediate; students can now read fluently.

These phases are cited as general and descriptive; they are neither set nor universal stages of reading acquisition. Rather, they are categories used to clarify the text-based elements of the complex, interactive relationships among the many influences on reading skills. Ehri outlined that they will overlap, and why she terms them ‘phases’ rather than ‘stages’.  It is not until Ehri’s full alphabetic phase, when children have mastered the simple and complex alphabetic code that letters are sequentially mapped to sounds, enabling children to make phoneme to grapheme mapping all through the word (Johnston and Watson 2007) They are able to read phonetically predictable ‘regular’ words with accuracy, and do so with automaticity.

Orthographic mapping (OM) involves the formation of letter-sound connections to bond the spellings, pronunciations, and meanings of specific words in memory. Being able to read words from memory by sight is valuable because it allows readers to focus their attention on constructing the meaning of the text while their eyes recognize individual words automatically. If readers have to stop and decode words, their reading is slowed down and their train of thought is disrupted. Reading words from memory by sight is especially important in English because the alphabetic system is variable and open to decoding errors. This ‘sight word reading’ happens automatically without the influence of intention or choice.

As the PSC check assesses skills against the full alphabetic phase it fails to assess reading against a developmental framework of reading development. If a child fails to pass the test, according to the designated pass rate, the PSC does not offer teachers a way to ascertain at what stage the student needed more support or to plan for an intervention. This is especially relevant to dyslexic learners when early identification of phonemic awareness deficits is key.

 
The Phonics Screening Check is not a check of phonics decoding, it is a check of phonics decoding of a limited set of graphemes (around 25% of all graphemes used within the English written language)
So when the DfE claims that 'the purpose of the phonics screening check will be to confirm that all children have learned phonic decoding to an age-appropriate standard' this is a little misleading. It is also limiting - what is an 'age- appropriate' standard? 
If we want children to be able to read by 6 (before the end of Year 1) then they will need a lot more than the ability to recognise and blend those graphemes. 

​​

And although teachers are told that any variation can be accepted for the nonsense word testing, this relies on teacher knowledge. One might also question whether children who know more phoneme to grapheme combinations might be at risk of failure, if not giving the specific pronunciations of pseudo words. In 2020 I worked with 5- 7 year olds to explore the pseudo word ‘sut’ used in the 2019 PSC materials.

 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/809986/2019_phonics_pupils_materials_standard.pdf
The explanatory notes outline the expectation that it be pronounced to rhyme with cut (the s representing the same sound as in ‘sip’) and states that ‘all regional pronounced for the u are acceptable’. But what if the child is a competent reader and pronounces the word with a ‘sh’ at the beginning – wanting to demonstrate their more advanced knowledge of the code. (as in sugar) Those children found over 100 ways to pronounce the word, analysing how many graphemes each letter could represent. Most phonics programs assume that children acquire knowledge of phoneme to grapheme patterns through direct instruction, ignoring the reality that there are simply too many letter-sound relationships in English orthography for children to acquire by direct instruction, probably between 300 and 400 (Gough & Hillinger, 1980). Much, if not most, of what children learning to read in English come to know about its written orthography is acquired through implicit learning, especially knowledge of context-sensitive letter-sound correspondences that depend on position-specific constraints or the presence of other letters (Bryant, 2002; Tunmer & Nicholson, 2011; Venezky, 1999). In contrast, those combinations taught and tested within UK newly validated SSP programmes are fewer in number and are largely context-free, involving one-to-one correspondences between single letters or digraphs and single phonemes.

​

The Ofsted inspection handbook states that the sequence of reading books should show a cumulative progression in phonics knowledge that is matched closely to the school’s phonics programme and that children should read and re-read books that match the grapheme-phoneme correspondences they know. This expectation is consistent with the English programmes of study: key stages 1 and 2 National curriculum in England which states that pupils should read aloud accurately books that are consistent with their developing phonic knowledge and that do not require them to use other strategies to work out words. This does not align with Ehri’s theory of reading development, or others such as The Simple View of Reading, in which there is an explicit acknowledgement that word recognition alone is insufficient to produce good readers.

Some children are fluent before the PSC (UKLA 2012) and some are reading even before starting school. Some may still fail the PSC. As there is no baseline check of decoding skills prior to the PSC data cannot be produced to ascertain ‘how far travelled’ for each student, to explore factors impacting on progress. Teachers now include pseudo-words in their phonics teaching (Walker et al. 2015) and phonics is a distinct area of the curriculum rather than being one element of teaching reading and spelling as reported outlined within The National Reading Panel (2005) Phonics as a distinct curriculum has become a new ‘regime of truth’ (Foucault, 1980) and subsequently facts about reading for pleasure or even reading for meaning, lose their power (Clark, 2016)

We can clearly see that when looking at the new Validation process for Synthetic Phonics programs. In the ‘Essential Core Criteria’  (https://www.gov.uk/government/publications/phonics-teaching-materials-core-criteria-and-self-assessment/validation-of-systematic-synthetic-phonics-programmes-supporting-documentation#essential-core-criteria) the term ‘read’ or ‘reading’ is only used 5 times, and in every case as it relates to decoding of a single words, and that those words consist of the graphemes taught systematically, along with a few ‘exception’ words – and only as they encounter them within their ‘decodable' readers.  If children are to become fluent readers, then is more needed to achieve this according to the science of reading. The DfE states that children must be taught to ‘read printed words by identifying and blending (synthesising) individual phonemes, from left to right all through the word’ (DfE 2020) and ‘should not encourage children to guess unknown words from clues such as pictures or context’.  Ehri (2014) instead outlines the various strategies used when faced with unfamiliar words. Readers might use their knowledge of the writing system to apply a decoding strategy, and in languages with regular grapheme-phoneme relations decoding is straightforward. However, the English writing system includes multiple ways to pronounce graphemes and to represent phonemes, so readers must be flexible and expect variability. This is helped by a meaningful context. The PSC only tests a group of ‘regular’ graphemes, and therefore does not assess alternative strategies used when faced with unfamiliar words, or provide meaningful context so as to ascertain reading behaviours – the behaviours exhibited during the learning to read learning journey.  Another strategy for reading unfamiliar words is by analogy, which involves finding in memory the parallel spelling of a known word and adjusting its pronunciation to match letters in the unknown word (e.g., reading thump by analogy to jump). As beginners accumulate a larger store of written words in memory, this strategy becomes more useful.  Is this expected by the end of year 1, ie after 2 years of explicit instruction? The PSC again does not assess how students use existing code knowledge.  Prediction is also a strategy known to be used to read unfamiliar words. Initial letters, for example, plus context cues in the sentence, the passage, or pictures to anticipate what the word might be. Once a word is predicted, then its pronunciation is matched to the spelling on the page to verify that the sounds fit the letters.

And what of irregular words? Rules are usually used to produce patterns and “exception” or “sight” words whose pronunciations violate the rules must be memorized. The ‘rules-plus-sight words’ approach remains the basis of phonics instruction. Teaching phonics by teaching rules and memorising exceptions, however, leaves out the statistical patterns that permeate the system and drive the fast, implicit learning process.(Seidenberg et al 2020) In her 1995 paper, Ehri puts considerable emphasis on the concept of the way that ‘sight words’ are memorized and claims that this is a not a rote memory process; instead it involves making systematic connections between the spelling of the words and their pronunciation. However this is not what happens within synthetic phonics programmes, where these words have been called ‘tricky’ and teachers told they must be learned by ‘sight’ through rote learning. If using Jolly Phonics around 50% of the written English language would be considered ‘tricky’ They may pass the PSC and decode books in which these words are restricted, but how does this align with Ehri’s developmental stages, and the skills used when figuring out words not covered within this narrow curriculum. 

​
Synthetic phonics became a legal requirement in state-funded primary schools in England in 2007, advocated for by the Rose Review (Rose 2006) and has remained, for the most part, unchallenged by political parties. The Labour government published a non statutory programme Letters and Sounds (DfES 2007) to further promote ‘synthetic phonics’ and in 2010 this approach was again emphasized by the coalition government, publishing The Importance of Teaching (DfE, 2010). Self-assessed commercial synthetic phonics programmes were promoted on the DfE web site between 2010 and 2014 and schools were encouraged to purchase them over non listed programmes, eg through match-funding.  
In Sept 2021 The Teachers Standard (DfE 2011) required teachers to demonstrate their understanding and adherence to the underlying ‘alphabetic principle’, and to ensure compliance, a phonics screening check (PSC) has been undertaken at the end of Year 1, since 2012.  The test is to measure if students can ‘decode’ a set of real and pseudowords, based on around 85 high frequency graphemes, to an appropriate standard.  In 2016 the Conservative government published ‘Educational Excellence Everywhere’ (DfE 2016) and most recently the government has invited programme developers to submit their ‘systematic synthetic phonics’ programmes to be validated; this time to be reviewed by a panel of ‘independent evaluators’ ‘Validation involves a process by which ‘a programme has been self assessed by its publisher and assessed by a small panel with relevant expertise, and that both consider it to meet all of the most recent Department for Education (DfE) criteria for an effective systematic synthetic phonics (SSP) programme.’  (https://www.gov.uk/government/publications/phonics-teaching-materials-core-criteria-and-self-assessment/validation-of-systematic-synthetic-phonics-programmes-supporting-documentation)

The government remains committed to the “phonology first” hypothesis (Bowers and Bowers 2018a) in which grapheme-phoneme correspondences are taught in an organised sequence as opposed to incidentally.  There is ample research relevant to developing effective phonics instruction, demonstrating for example the advantage of direct instruction over indirect methods (e.g., Stuebing et al., 2008). However, the research literature does not provide detailed guidance about specific instructional solutions. The synthetic phonics approach introduces learners to the mapping of graphemes to phonemes (the smallest sound units) usually from the very beginning, and students practice blending those combinations in order to ‘decode’ the word, regardless of whether they understand the words.  The PSC ‘is not a test of reading in its fullest sense, where meaning counts as well as decoding’ (Ellis and Moss 2014, 242); the words are presented without context. Indeed, ‘pseudo’ words are included in the test of single words to ensure that children are not identifying the words from memory or using other clues eg from context. The PSC is designed to assess student competence with regards to recognizing and blending this initial grapheme teaching sequence, using the graphemes included in the old Letters and Sounds programme, with no assessment of comprehension, and with an assumption that the real words are words a child already has in their spoken vocabulary. For example, if the word was ‘flow’ and the child pronounced it to rhyme with ‘brow’; not adjusting it as they did not know the word.  English has a very complicated alphabetic code, which is often referred to as ‘opaque’ whereas alphabetic languages, such as Finnish, German, Hindi, Italian and Spanish have very simple alphabetic codes. The phoneme to grapheme (sound-symbol) relationship in these languages is reliable, which is why they are known as ‘transparent’ alphabetic codes. Furthermore, words can be segmented and graphemes blended differently depending on accents and dialect, an issue I encountered when teaching around the world. For example, presented with the word ‘ant’ within a phonics test, Australian children would generally pronounce the word ‘ant’ with different phonemes than children in England. For a test to be deemed valid, there is an assumption that the word is pronounced with the intended ‘sounds’, but this can mean the child ‘translating’ the word into British English and the assessor understands that this has occurred. Because phonemes are produced with combinations of oral gestures (e.g., lip closing, tongue raising against the roof of the mouth, lip rounding, jaw raising or lowering) and gestures lower in the throat (e.g., vocal cord vibration), an individual phoneme can be described as the set of gestures required to say it (Fowler, Shankweiler, & Studdert-Kennedy, 2016) They may differ depending on dialect or accent. The words being segmented by the teacher while teaching synthetic phonics may be segmented differently by the student.

 

 It is worth noting that the term ‘systematic’ was a key finding with regards to the teaching of phonemic awareness and phonics within the National Reading Panel (2000) Systematic phonics instruction was found to be effective, however no statistical differences between synthetic phonics and other types were found.  Neither the Torgerson et al. (2006) meta-analysis nor that of the National Reading Panel (Ehri et al., 2001) found evidence for a difference in effect size across different methods, with both concluding that the key ingredient of a successful phonics program is that it is systematic In a report dated April 2007, professors Dominic Wyse and Morag Styles concluded that the evidence "supports" systematic phonics; however, the Rose Report's assertion that synthetic phonics should be the "preferred method" is "not supported by research evidence". This criticism is based on the way the research was conducted and how the results were interpreted.  The addition of the word ‘systematic’ may have given ‘synthetic phonics’ more credibility, however this abbreviation is unfortunate within my own practice. I had launched the Speech Sound Pics (SSP) Approach within Australia in around 2011, as an alternative to synthetic phonics. Although SSP is now widely used in the UK to refer to the government’s preferred approach to teaching phonics, the majority of teachers in Australia who refer to SSP are referring to The Speech Sound Pics (SSP) Approach. It was offered as an early intervention for the issues experienced by neurodiverse learners (eg those at risk of dyslexia’ 

Seidenberg (2020) suggests that ‘existing systems – from formal curricula to informal practices by individual teachers – should be examined and augmented in a way that moves them closer to what we know about how learners learn.’ Currently, however, these findings from action researchers (teachers) are deemed anecdotal by researchers and generally ignored.

As the hallmark feature of dyslexia is a problem with word decoding one might view the PSC as a valid ‘check’ after two years in school, and potentially flag students needing extra support. In 2012 The United Kingdom Literacy Association (UKLA 2012) recommended that the ‘check’ not be used with all children, as many would already be reading. However few seemed to question how the PSC would benefit, improve or even be used to impact on reading acquisition for those with dyslexia.  The predominant cognitive explanation of dyslexia is that it arises from a phonological deficit affecting the processing of phonemes in words (see Vellutino et al., 2004 and Hulme and Snowling, 2009, for review ) and specifically relates to the ‘phonemic awareness’ subset.  Without phonemic awareness, a child cannot begin to connect the sounds of our spoken language to graphemes ie letters or letter combinations.  The National Reading Panel (2000) found that many difficulties learning to read were caused by inadequate phonemic awareness and that it was the systematic and explicit instruction in phonemic awareness that directly caused improvements in children's reading and spelling skills. Phonemic awareness skills are exclusively auditory, whereas phonics maps the auditory patterns (speech sounds) with the corresponding visual representations (graphemes). Phonics uses letters, while phonological and phonemic awareness do not.  For many dyslexics, this is a very challenging task as evidenced in their great difficulties in decoding and encoding, and so is the PSC useful for dyslexic learners? Does it simply identify a symptom of deficits, and how would teachers differentiate between an undiagnosed dyslexic learner (with poor phonemic awareness) from a child with good blending skills, who has simply not covered enough of the graphemes? As a trainer of hundreds of teachers working with children with learning challenges across Australia the phrase ‘too little. too late’ is often used when discussing the PSC, newly introduced in many Australian states, and the ‘at risk’ children so many of us are supporting. Our concern is that in the absence of early assessments to isolate skills the needs of neurodiverse learners are not being met. 

​

I cannot see that even the very best teachers can meet the needs of every child in a class, as the range of experiences and skills are too diverse. Therefore, with regards to this somewhat mathematical element of the learning to read and spell journey I am committed to creating technology that enables learners, especially under quality direction and guidance, to learn to recognise and blend those 85 graphemes within a matter of weeks or months (not years) ie constrained skills. They will also be able to read and spell hundreds of high-frequency words. That is currently my priority, and especially as so many around the world need to do this remotely due to the pandemic. Ultimately, the goal is that every child does not just read, but chooses to read - for pleasure. Around 1 in 4 children in the UK are not achieving this. 

Has this proven effective for teachers in Australia?

 Testimonials

​

The team at this school used Jolly Phonics, a synthetic phonics program, until they conducted a pilot in 2018.  As of 2019 they have used the Speech Sound Pics (SSP) Approach