Ofsted’s new “similar schools” model attempts to bring greater context to inspection. Comparing schools serving very different communities has long frustrated school leaders, so a model that groups schools according to factors such as disadvantage and SEND has the potential to make comparisons fairer and more meaningful – if it is well implemented and managed. The inclusion of English as an additional language (EAL) in the model is a positive step. Alongside the revised inspection toolkit, which now asks inspectors to consider the extent to which teachers assess pupils’ English language proficiency, this signals that EAL learners are beginning to feature more prominently in accountability measures. But if the similar schools model is to fulfil its potential, using a binary EAL measure that only tells us whether a learner uses EAL or not is not enough on its own. Multilingual classrooms are now the norm. The proportion of schools with EAL pupils has grown significantly, with these children making up at least 5 per cent of the roll at two in three schools. In 2013, this was the case in less than half. Bigger difference Through more than a decade of research with the University of Oxford, The Bell Foundation has consistently found that for EAL learners, proficiency in English makes a much bigger difference to their academic performance than anything else. It matters about four to six times more than their gender, socioeconomic disadvantage or ethnicity combined. Understanding a child’s proficiency in English can therefore provide valuable context and help schools identify how best to support EAL learners to access the curriculum and achieve their potential. Knowing that two schools in the same region have 40 per cent of EAL pupils without additional information as to their proficiency in English tells us very little. One school may serve large numbers of children from second or third-generation multilingual families who were born and educated in England and are fluent in English. Another may have larger numbers of children who have recently arrived, including refugees and asylum seekers who are new to English and are adapting to an unfamiliar education system. On paper, these schools may look similar. In reality, they are serving very different pupil populations: different languages, educational backgrounds, and levels of English proficiency. Treating EAL as a binary characteristic obscures the diversity in multilingual cohorts and the information needed to understand their educational context. In addition, evidence shows the proportion of EAL learners in a school has little impact on the achievement of other learners. Recognition With the revised inspection toolkit now recognising the importance of recording language proficiency, inspectors must consider whether teachers assess pupils’ English language proficiency accurately and regularly. They must also consider how they develop academic vocabulary systematically and provide appropriate language support for learners in the early stages of acquiring English. This is an important recognition of the role language plays in accessing learning. There is, however, a tension between these new expectations and the information available to support them. With no nationally consistent approach to assessing or recording English language proficiency, schools in England are not currently required to record it in the school census as they are in Scotland, Wales and Northern Ireland. The information needed to build a fuller picture of EAL learners is therefore not routinely available. There are also early questions about implementation. Looking at recent inspection reports suggests that the experiences and outcomes of pupils who use EAL are not always explicitly visible within inspection reporting. This underlines the need for Ofsted to strengthen EAL-specific training and guidance for inspectors, so new expectations are applied consistently and reflect the diversity of learners using EAL. Data challenge It also means continuing to develop how inspection considers the context in which schools are working. There is also a wider data challenge. A nationally consistent approach to assessing and recording English language proficiency would give schools, inspectors and policymakers a much richer picture of EAL learners and provide a stronger foundation for meaningful comparison. The diversity of multilingual learners also reflects differences in language background, educational experience and prior schooling. For the similar schools model to work as intended, the data used to identify comparable schools needs to capture more of this context. There is now an opportunity to build on the inclusion of EAL in the similar schools model, recognising English language proficiency as an important part of the picture, while developing a more nuanced understanding of the pupils and communities schools serve. This would help identify genuinely comparable schools, give inspectors better context for understanding outcomes and ultimately make the model more useful for school improvement.