Look at school stats? OK, then.
Let's look at First Language. It's defined as
A pupil's first language is not English when the pupil:
- Has been exposed to a language other than English during early development, and
- Continues to be exposed to this language in the home or in the community
Great.
Trouble is that this could be the language of a grandparent who did child care for 3 months and now contact is a monthly Facetime. The person caring could be deceased since that period, so they barely understand more than Latte/Milch/Wara/Nofosuo/whatever and ciao/hallo/pele/akyem but they still hear it spoken by people.
It could also be 'OK, your Mum spoke Twi but we've always spoken English to her'.
They may have been exposed to it, but that doesn't mean they speak or understand it at all.
Is the First language the father's German, the mother's English or the Grandparent's German/English/French or Creole (and which variant)?
Add into that the Hostile Environment made people scared, so many parents said First Language was English and, yet again, that giving this information is purely at the discretion of parents.
You could have quite accurately, an EAL % of 68% of 1200 kids, but only actually have 7, 3 of whom are siblings, who need EAL support (the MIS gives the ability to record proficiency, but this is not collected by the DfE).
If you look at FSM provision - which is a binary yes or no - how do you know whether the % is due to long term poverty, the effect of transitional protections meaning that most have only been entitled because a) they had a couple of months on Income Support before starting a new £58k job 5.5 years ago, b) Covid hit so they stayed eligible until the end of that school phase or, with the coming academic year, because they have a disability/serious illness and therefore qualify for a couple of quid in Universal Credit?
If you're looking at Faith, it isn't a compulsory collection item for a non-religious school and is only compulsory in those due to the religious organisation's requirements in a completely different census. Not all faiths have a structure and the levels of bureaucracy that the RC church does - Islam requires nothing more than the Shahada, Hinduism requires nothing, Paganism nothing - and there can be atheists stating Catholicism because they did the baptism or a dispute amongst parents as to which faith or none to add. To the people of Faith, it can make a massive difference - some branches of Thomasian Christians are counted as Catholic, others as Anglican, there's also a section for CofE. What do you put down for Presbyterian? Bear in mind that there is Scottish Presbyterian and there are Black Majority Churches that use the same. How about Quakers? They don't believe in many of the tenets of Christianity that others do, so which category? Or do we use membership of Churches Together in England (or Churches Together in Britain and Ireland? Do we include specific churches that are members of the Evangelical Alliance? And if we do, under what category?
If you're talking about progress and attainment stats, do you understand the actual differences and limitations in the data? That it doesn't include privately educated or EHE children or those withdrawn from SATs except in the measures where it does? Or where SATs have no relevance? Or who is excluded or included? Are we using the 2019 comparisons or have you noticed that the dataset was updated 8 weeks ago to account for entirely new qualifications and a pandemic?
Look at SEND. Are they EHCPs or K codes? Do they have more because it's an undersuscirbed school, is it because they're more inclusive, is it because the LA directs them to take children whose needs cannot be met, is it because they have a great SEND department who don't ignore compliant, academically able children?
If you don't understand school data which is very strictly defined and standardised for all school required to submit it in one year but not from year to year, you are not going to able to find the truth in junk data.