When EAL isn’t EAL: incentives and process in school data

Hearing Jim Al-Khalili being interviewed about his own life this morning reminded me to finish this. He began by talking about his upbringing in Iraq, with an English mother and Iraqi father, and speaking English at home, and then toning down his RP accent when he came to the UK in ’79, as a refugee from Saddam Hussein. Still, if it were now, he may well have been noted as EAL by the school…
 

Take my friends, J and G. Neither speaks English as a first language, one is from the Americas and the other is from the Middle East. Being bright and ambitious, they had the opportunity to pursue academic study in an English-speaking institution, married, became academics, moved to the UK and then had children. These children are trilingual, but you can probably guess which they speak most of, given that the parents also have English in common. Or another friend, D who is English, and her partner, who is from one of those ex-Empire places where a local language is present, but there’s some officialdom done in English, and elites that use English. Their children have also been in the UK since birth. Another family, arriving in the UK as students, learnt to operate in English in high level jobs, but still speaking Urdu at home, then had a child who went to the local nursery and then to primary school.

Each of these children is noted as an ‘English as Additional Language’ pupil, in school returns and assessments. None of them, however, have any trouble with their English. Like the category of Pupil Premium (or Free School Meals), the administrative category is a crude measure of the extra barriers a child may have, and consequently a crude measure of additional assistance they might need. None of the children above need extra help, whereas a child who arrives alone aged 14, seeking asylum and with little English, is going to need a whole lot of help (for more than one reason).
​
The problem here is that schools have a double incentive to make the category inclusive. First, there’s any funding associated with the EAL tag: it’s not a lot, but all money counts. More important, though, is the assessment of schools through various attainment or progress statistics. The aim, from government and via Ofsted, is to ensure children with EAL don’t fall behind and stay behind, and so schools get feedback on the difference between the EAL cohort and the others (presented in various aggregate formats). A school does well if the EAL cohort gets similar SATS grades to the other kids, and this is highly likely if a good proportion of those with an EAL tag are the children of professionals with better English (especially re. vocabulary) than many native speakers.

Indeed, we know that the vocabulary of the most deprived children is much lower than that of better off children, and that vocab aged 5 is the best predictor of whether deprived children can ‘buck the trend’ and eventually escape poverty (Blanden 2006). This suggests that some of those kids who don’t get a great deal of communication may well have similar problems, even if parents are native speakers, and this is why we have Pupil Premium. I think there is an argument for rolling the two together.

Like Pupil Premium, then, EAL statistics should be taken with a pinch of salt. Some schools think it gives others an unfair advantage in the data, as some pupils are noted as unlikely to do well and then their (actually likely) success bumps them up. But there will be other schools, for one reason or another, who will get a cohort of EAL children with far bigger problems, and they will be below average on this particular measure. Few of these statistics come with the kind of comparison that really helps (see EEF for an exception): not only does the kind of EAL matter, but also any other barriers that are also present. I haven’t seen any research that matches each EAL child to an equivalent with the same other problems and attributes (e.g. social background, learning difficulties, school type) AND distinguishes between the level of the EAL barrier too. And, of course, what the statistics have never shown is what would happen if the EAL assistance wasn’t there.

That said, in this instance schools are likely to direct help where it is most needed. If a child tagged EAL doesn’t need help anymore (or never did) it is holding them back to insist on language lessons. Presumably, this is also the case for Pupil Premium, but because the funding is ring-fenced (see Allen 2018), it creates an incentive for doing extra work with children who don’t need it, while restricting how much can be done for those who do.