Listening to the voices of children: An illuminative evaluation of the teaching of early reading in the light of the Phonics Screening Check

The Phonics Screening Check (PSC) was introduced in England in 2012 for children in Year 1. There have been criticisms in relation to its reliability and appropriateness as an assessment tool for early reading although supporters of the PSC see it as a valuable tool in securing progress in reading. The DfE funded evaluation (Walker et al, 2015 p.8) concluded however, that it “did not find any evidence of improvements in pupils’ literacy performance, or in progress, that could be clearly attributed to the introduction of the PSC”. This article reports some of the findings from a doctoral study that sought to illuminate the voices of those most affected by the PSC: children in Year 1 and their teachers. The study used an illuminative evaluation methodology (Kushner, 2017) and focused on a range of schools in a large city, selected for their diversity in relation to attainment data (PSC and reading) and socio-economic status. The findings demonstrate the negative backwash from the assessment process which has influenced the way that phonics is taught and so raises some questions for teachers and policy makers about the approach to the teaching of early reading in the light of the PSC.


Introduction and background
There is little disagreement about the importance of developing children as skilled readers who are able to read for purpose and pleasure. Voices from a variety of research and policy perspectives agree that being a reader is an indicator of future socio-economic success and that the motivated reader is more likely to be a higher attaining reader (Organisation for Economic Cooperation and Development, 2009) and one that reaps a wide range of other benefits (Hempel-Jorgensen et al, 2018).
Whilst there is a consensus about the value of being a reader, this is not reflected in agreement on the best approaches to its teaching and assessment. One foundational element of current English reading policy is that of 'first and fast' systematic synthetic phonics, as set out in the Department for Education 'core criteria ' (2010) and later highlighted by Walker et al, (2015) and Ofsted, 2015. The focus on phonics was enshrined in the National Curriculum (2014) and then enshrined in the accountability measures, including the PSC, introduced in 2012. It was introduced as a mandatory assessment tool for children in Year 1 (usually aged 5 or 6) designed to test children's abilities to decode both real and pseudo decontextualised words at an 'appropriate level' (DfE, 2010) using only phonic knowledge to read each word. Whilst individual school results are not made publically available, the data from the check form part of a schools' suite of data and this is used as a scrutiny and accountability tool by Ofsted. Whilst there is much debate about the focus on systematic synthetic phonics, this was not the focus of this study. This study aimed to understand the impact of the PSC on the processes and practices of the learning and teaching of reading. The government commissioned report by the National Foundation for Education Research (NFER), suggested that the PSC had no clear impact on literacy attainment (Walker et al, 2015). Caution was advised, in the more recent 'Progress in International Reading Literacy Study (PIRLS): National Report for England', (McGrane et al, 2017 p.146) in drawing causal links between the slight rise in England's international comparative scores in reading and the introduction of the PSC. More recently Clark (2018) reported on an independent survey of head teachers, teachers and parents (Clark and Glazzard 2018) which sought views on the usefulness and impact of the PSC. Responses from all groups suggested the PSC was neither useful nor effective. With these reports in mind it is particularly pertinent that a study should consider the qualitative outcomes (intended or unintended) of the PSC in addition to a focus on quantitative data. Added to this, there has been a wide range of criticisms of the Check. The PSC was criticised on its introduction (United Kingdom Literacy Association, 2012;Davis, 2013, Clark, 2013 with concerns raised about the test's fitness for purpose i.e. as a test of early reading skills (Darnell et al, 2017); its effectiveness in its identification of children in need of additional reading support and its appropriateness for children with English as an additional language and the more able reader (Davis, 2013) and its 'validity' and 'sensitivity' (Duff et al, 2014;Grundin, 2018). Clark (2017) has continued to raise 'unresolved issues' in terms of 'validity and value' whilst Dombey (2011 p.23) claimed the check would "distort the process of learning to read".
My study was prompted by one such possible distortion: a head teacher's anecdote following the first PSC in 2012. Her school is an inner-city primary with high numbers of black and minority ethnic (BME) pupils. She recounted how some children became quite distressed when trying to blend the real word 'nigh'. The children had failed to spot the trigraph 'igh' and so attempted to sound and blend each letter, resulting in the word 'n-i-g-hur'. To the children this sounded like an insulting and inappropriate word. The dilemma for these children was an unintended consequence of the PSC and I hypothesised that the PSC may have raised further unintended issues for teachers and children.
Whilst the debate in relation to teaching, assessment and accountability continues there seems to be one voice that has not been heardthat of children. The group that is at the heart of the debate, those learning to read, have not been listened to and so their insights missed. The United Nations Convention on the Rights of the Child (1989) and the statement by the High Commissioner For Human Rights (2005) enshrined the ethical, moral and legal rights of children to have their views heard about decisions that affect them (Mortari and Harcourt, 2012) and so there is also an ethical and moral imperative to seek the views of the child in relation to an aspect of education that holds the key to their social, educational and economic empowerment. This article will therefore focus on the children's data and findings from the study.

Methodology
The methodological approach used was that of illuminative and democratic evaluation as outlined by Parlett and Hamilton (1977) and House and Howe (2005) and developed by Kushner (2000; aiming to incorporate "the views of insiders and outsiders [and] gives voice to the marginal and excluded" (House and Howe, 2005 p.81): in this study, that group is the children who take the PSC. Kushner (2017 p.20) suggests that evaluation can challenge the "single narratives" and "one-dimensional explanations" given by those who make policy and describes evaluation as shifting attention to "the way the programme is seen and experienced". This qualitative approach to seeking the voices of those most affected by a policy practice, has a growing place in evaluation research.
The research used a mixed methods approach: a questionnaire was completed by 59 teachers in 14 schools across the city and from this a smaller group of 7 schools were identified. These schools were purposively sampled using a range of socio-economic indicators, PSC and reading attainment to ensure a range of schools in the study: this is set out in Table 1. All schools were graded 'good' or 'outstanding' by Ofsted: the quality of teaching and learning therefore was judged to be good and so we can assume the quality of the teaching of reading was part of this judgment. Focus groups of Year 1 children were selected by teachers to represent a range of attainment and gender in each school and teachers were asked to select children who would be happy to engage and talk with someone with whom they were not familiar. Following these focus groups, the data had a primary analysis and this, along with the questions raised by the questionnaire data was used to formulate the discussion points for the teachers' focus groups. Researching with children presents its own challenges and so developing methods that enabled children to have a voice were carefully considered. Marwick and Smith (2014) had introduced playful approaches to data gathering in their conference presentation to the United Kingdom Literacy Association and from this I identified their use of the story of 'Beegu' by Alexis Deacon as a vehicle for positioning children as the experts of learning to read and as teachers of reading. Engaging with a text as part of the research process, enabled an approach that was similar to daily classroom practice but was also emblematic of my personal positioning on the teaching of reading: viewing children as active agents and critical participants in the learning process, where active engagement is required and views reading as a meaning making process. Using a Beegu knitted toy and the text, children were told that Beegu had seen them reading in school and was so interested she wanted to learn to read herself. Children were invited, as experts, to suggest ways that Beegu could learn to read and become a reader. I did not ask children directly about the PSC because many children were not aware they had taken the PSC and raising the test as something to discuss could have made children unnecessarily anxious. The study design therefore focused on eliciting children's responses to their teaching of reading which, according to the evaluation by Walker et al (2015) and the teachers' quantitative data from this study, had been adapted for the sole purpose of raising PSC scores.
The transcripts from the focus groups were analysed using a multi-layered approach to thematic analysis (Braun and Clarke, 2006) which involved the identification of patterns 'through a rigorous process of data familiarisation, data coding, and theme development and revision' (Braun and Clark, 2018), the data being the words of the children. The tensional themes of connection and disconnection were identified.
Street's (2016) ideas on the importance of considering the 'telling' case or voice was also used to select data to foreground in the study. All of the qualitative data was made available in the appendices of the study in line with Kushner's (2017 p.161) view that data are more important than findings as it is the data that "represents the peopletheir hopes, fears, aspirations and failures." Mitchell (1984) introduced the idea of the telling case or voice and identifies it as one that offers new insights. Street (2013, p.40) defines the telling case as one that shows "telling instances of behaviour that elucidate, contradict, or expand relationships presented in earlier fields of study". The data that follows represents some of the 'telling' voices of children in this study. All of the schools' and children's names are pseudonyms.

Phonics as separate from reading: disconnection and connections
When Beegu suggested that she had seen children learning phonics and she wondered if this would help her to learn to read, in all seven schools there were children who did not see phonics as a tool or a skill that supported their reading. They identified phonics almost as a separate 'subject' and as a lesson where particular activities were expected: these were not talked about as being 'reading' and the purpose was dislocated from their application in reading. When asked if phonics would help 'Beegu' to read many children suggested that it would help her to write but not to read. The following telling exchange exemplifies the interactions of the focus groups.

Fig Tree School
Researcher  (2017) research, with a focus on ability groupings, also found that phonics was seen as a separate subject and was disconnected from the process of reading. Some might argue that this disconnect is not significant: for example Ehri and Snowling, (2004) and Chall, (1995) might identify this as part of the linear approach to learning to read. Rose (2006) too suggested that the skills of reading precedes their application. Hall (2013) however, identifies one difference between highly effective and less effective teachers of literacy: the contextualisation of the learning activity. Denton and West (2002) found in their larger scale study, that a single focus on phonics resulted in improvement in alphabetic knowledge but did not have an impact on fluency, engagement or comprehension and so this disconnect of phonics from continuous text reading is problematic.
The phonics lesson then, for some children, was merely an academic exercise rather than one connected to learning to read. Phonics teaching had its own technical language that children used to describe their learning:

Acorn school
Researcher: What is phonics? The disconnection of phonics with reading and its connection with a subject about sounds led to one child in Fig Tree School suggesting that the purpose of books was to enable children to get better at their sounds.
Frank: All the books are good because they help with good sounds.
It is also important to note that most children talked about 'sounding out' and many children made the connection between phonics or 'sounding' and their reading when telling 'Beegu' what she needed to do to learn to read:

Chestnut School
Felix states that to read you need:

Phonics, because then we can see the word and what it makes and the sounds
Children in Chestnut School also provided an explanation of phonics that identified a connection with word reading and also connections between reading, writing and phonics. Pete makes a distinction: that phonics is connected to reading but that the phonics lesson is limited to word reading or 'finding the words' as Pete puts it here, but that real reading involves something more than this -his emphasis on the word 'read' indicates he is aware that reading involves more than just decoding the word.
It was also noted that the children in Chestnut School expressed their ideas and connections in longer and more sophisticated articulations. It is possible that in some of the other schools children made the connection between phonics and reading but were not as able to articulate this level of meta-cognition. Chestnut School was in an area of high socio-economic status (SES) with very low deprivation indicators and whilst SES is not explored in this article, it was discussed in the thesis findings.

Connections: Phonics is about writing
Clearly, reading and writing are reciprocal processes and the guidance for phonics teaching found in 'Letters and Sounds' (2007)  The teachers' data, which are not discussed here, reveals more about phonics teaching practice in relation the teaching of the alternative graphemes, which tended to be done through the teaching of spelling, and this begins to make sense of children's ideas. Again in the same way as seeing word reading as disconnected from continuous text reading, the children here show their understanding of phonics as a way to learn spelling, viewing spelling as writing, rather than writing as a purposeful and contextualised activity.

Disconnection and connection: the case of alien words.
The question of pseudo words or alien words as children and teachers generally referred to them, was not raised by children in all of the children's focus groups.
Teachers suggested that many children were not aware that they were taking or practising for a test. I felt therefore, that ethically it was not appropriate to ask children directly about the PSC or about pseudo words unless the children raised the issues themselves as I did not want to raise new concerns about being tested. In the four out of the seven focus groups where they were discussed children provided some wide ranging explanations; from enabling them to communicate with aliens to being just something that teachers had to do. Claire, in Dogwood School explained them as: Claire: I know, Miss X said that um they are just to help you with your sounds they are not for any other use they just help with soundsa bit of a waste of time Here are a sample of children's ideas across different schools about why they were taught alien words: What is evident is that children are trying to explain something that does not immediately make sense and that these explanations reveal the possible negative impacts on children's understanding of reading, reading instruction and the reading (and writing) process.

Chestnut School
It seems that some children think they may encounter alien words in a real book. Their strategy would then to be to identify the word as 'not real', as Isla suggests, and so they would accept this and continue reading. This runs counter to needing to comprehend 'in the moment of reading' (Tennent, 2015) in that the reader needs to continually 'check-in' with the meaning making process as they read. If a decoded word does not make sense, then the reader needs to 'hear' this, stop, re-read and redecode to check for an error. If there was no decoding error, the reader needs to find out what the word means, in the classroom, by asking another child or adult. Alien words seem to disconnect further decoding and reading as a meaning making process.
Mark seems to conflate mis-spellings with alien words. These different misinterpretations seem to be a consequence of an assessment tool i.e. the assessing of phonic knowledge and skill, becoming a curriculum objective. Moss (2017 p.62), suggests that "the assessment tools themselves simply become the curriculum". The approaches taken to the teaching of alien words as a curriculum objective further compounds children's confusion and gives rise to possible negative implications for the development of readers.

The reading context and environment
What I had not anticipated in my research design was that children in a few schools, recognised that Beegu in the story was an alien and that this was a little confusing for them when talking about 'alien words'. What was also noted here was the engagement that children had with the text, leaning in when listening to the story, commenting about Beegu's plight and poor treatment by the adults in the story. Children were engrossed by the story. Some children in particular interpreted the questions within the context of the storythey had empathised so meaningfully with Beegu that they approached some of the discussion from Beegu's point of view. The Dogwood School example below exemplifies this. The consequence of high stakes testing seems to be demonstrated here in the teachers' words demonstrating the negative washback that can be associated with testing.

Conclusions and recommendations
The data presented here are a small part of the data gathered in this study and so the conclusions and recommendations in this article only relate to the children's data presented. Further recommendations and findings I hope will be published in future articles.

1.
There is evidence in this study of negative washback in relation to the narrowing of the reading curriculum; the teaching of phonics in such a way as it becomes disconnected from reading; the teaching of pseudo words to prepare for the PSC, rather than pseudo words being an assessment tool (see point 3 below for a further discussion of this point) and the negative implications for the classroom climate created by the pressure of the test. It is therefore recommended that the PSC, if it continues to be used, should be an optional tool available to teachers to supplement school assessment practices. Ofsted currently inspect assessment practices: if outcomes in a school were not in line with national expectations Ofsted could direct schools to use the PSC if the school did not have an equivalent assessment tool in place.

2.
If the PSC stays in place, the pseudo words in the test are problematic. It is recommended that these words are replaced with real words and so encourage teachers to extend children's application of phonic skills to a wider vocabulary. This too however, has possible negative implications and so further research is needed in this area. Whilst this may be resisted, as it is stated that the purpose of the pseudo words is to ensure that no child can 'sight read' the words and so by-pass the application of phonics skills, it is important for policy makers to balance the negative unintended outcomes of using pseudo words with any positive outcomes in relation to reading attainment (of which there is no current evidence). Gibson and England (2016) and later, Darnell, Solity and Wall's (2017) research concluded that there was little or no difference between using real or non-words in phonics assessment and so the replacement of non-words with real words would not compromise the assessment of phonics skills and knowledge. Because of the high stakes nature of the PSC, pseudo words are being taught as the curriculum rather than being used as an assessment tool.
This practice contributes to the negative unintended outcomes of the PSC.

3.
Professional development for teachers that encourages critical reflection on teachers' practices and policy needs to be promoted. Professional development needs to provide not just technical or performative training in relation to the teaching of reading and in relation to increasing PSC scores, but to develop teachers' understanding and knowledge of the research that underpins practice and so provide the tools for critical analysis of policy, curriculum and practice.

Final comments
This research began from the starting point that phonics is an essential but not sufficient tool in ensuring children developed as life-long readers. The assessment of phonics therefore, is also an essential component of practice to enable a teacher to map the skills and knowledge of the young reader. However, this study has demonstrated that the PSC, because of its high stakes nature, is having some negative impacts on the teaching of early reading and so on children's understanding of the reading process.