How much can we learn from measuring EAL students’ English proficiency? A lot, actually.

In October 2018, Oxford University, The Bell Foundation and Unbound Philanthropy published a research study that used data collected as a part of the now rescinded requirement for school to assess and report English language proficiency in the school census. NALDIC published its response to the findings of the report here. In this post, one of the report’s authors, Annina Hessel, provides a summary for NALDIC of the research and its findings. 

Annina Hessel

In the January 2017 census, teachers were for the first time asked to rate all their EAL pupils on their proficiency in English, using a five-point scale from New to English through to Fluent. The new scale promised to be a valuable tool for teachers and policy-makers to learn more about EAL students, beyond the usual simple yes/no distinction from non-EAL students, offering many possibilities: within schools, teachers would be able to use the information to guide their decision-making on who needs language support, and how much. On the national level, the scale would make it possible to answer central questions such as: how much do EAL students differ from each other in their English proficiency, and what predicts if an EAL student is more or less proficient in English? And very importantly: can differences English proficiency explain something of the larger heterogeneity amongst EAL students across subjects, and if so, how much can it explain?

Despite its potential, the data gathered through the new scale has not been used to answer any questions on the national level, beyond one graph published by the Department for Education. We asked the Department for Education for permission to study the national data, but the request was refused as the Department for Education had chosen to exclude proficiency in English from the National Pupil Database. As an alternative, we established a data-sharing project with a number of Local Authorities in England who agreed to share fully anonymised pupil level results. Our dataset consists of the results of 140,000 EAL and non-EAL students, from 1,569 schools across six Local Authorities from different parts of the UK. Through careful analysis of the information from all these schools and students, we found answers to the following questions:

How much do EAL students differ from each other in their English proficiency, and what predicts if an EAL student is more or less proficient in English?

Overall, EAL students differed widely from each other in their English proficiency. While across all EAL students, almost one third was rated to be fully fluent in English, there was still about one fifth of who teachers reported to be completely new to English, or to be in the early stages of their learning of English. In the context of an English-language school system as we have it in the UK, this is not a trivial observation. Almost all teaching is delivered through the medium of the English language, be it through texts, video or audio materials, or in classroom discussions. If a sizeable proportion of students in the classroom can only access the learning content to a limited degree, these students are also less likely to perform to their full potential.

Of course, just knowing that students differ in their abilities and knowledge will not help anyone to plan their lesson or plan language support funding – we would also need to know who is likely to need support. We thus checked which EAL students were more likely to be more or less proficient in English. We found that what mattered most for EAL students’ English proficiency was not their gender, or if they were eligible for free school meals (an indicator of the family’s socio-economic background), but their age: older EAL students were more likely to be more proficient in English. On the flipside, there were many more EAL students with little English language skills at Reception or Key Stage 1, indicating that language support would be most valuable in those early years of schooling. That such targeted language support would be crucial was supported by the answers we found to our next question:

Can differences English proficiency explain something of the larger heterogeneity amongst EAL students across subjects, and if so, how much can it explain?

To explore this issue, we looked at EAL students’ grades across key stages and subjects, and linked it to their English proficiency. We found that the grades of EAL students with lower English proficiency differed greatly from their peers with better English language skills. As the figure below exemplifies for reading grades at age 7, there was a strong and steady link between EAL students’ English language skills and their grades. That is, EAL students with stronger English proficiency also did better on national assessments, and that was true across subjects and age groups. This link was so strong that English proficiency explained four to six times as much of EAL students’ grades as other central factors combined (we used gender, free school meal eligibility, and ethnicity).

These numbers further underline that EAL students’ English language skills have real implications for their school career – and that language support is worth the time and effort that it takes. At the same time, the story told be these numbers also has an additional twist that transcends identifying support needs: when you look at EAL students who were rated competent and fluent in English, you can see that these students did better than the national average. This shows that being an EAL student needn’t be a problem, quite the contrary. EAL students are part of the diverse country that the UK is, and their diversity enriches their own as well as the school’s lives. The high performance of EAL students with strong language skills underlines that it is possible to succeed in school while reaping the benefits of growing up with more than one language. What is important is that EAL students be adequately supported in their language learning to allow them to do exactly that. Measuring EAL students’ English proficiency is an important first step towards such adequate support structures, and resources such as The EAL Assessment Framework (2017) provide a great starting point for teachers to do so.


For a long time, practitioners and researchers have seen the limitations of only distinguishing between EAL and non-EAL students when wanting to understand a student’s abilities and needs. Using a simple five-point scale, teachers were able to provide a measure that makes it possible to gauge a student’s need for support and likelihood to succeed. I think this shows how much we can learn from measuring EAL students’ English proficiency – which is a lot, actually.

Annina Hessel is postdoctoral Research Scientist in the Department of Educational Psychology at the University of Göttingen. She did her doctorate in the Departments of Education and Experimental Psychology at the University of Oxford.

Follow Annina on Twitter here @AnninaHessel

na_logo-naldic is a publication of NALDIC, the subject association for EAL. Visit to become a member.