Addendum to: Cambridge University discriminates against children from private schools.

[Marchesman]

"From 2013 to 2023 the proportion of UK state-school admissions rose from 61 per cent to 73 per cent. This increase was made possible by undeniable discrimination against another group of students – those who, whether through a choice made by their parents or a scholarship won by their talents, attended fee-paying schools."

For an insider's perspective on Cambridge University's descent into mediocrity see: "Decline and fall: how university education became infantilised" D. Butterfield, Spectator 26th Oct..

When measuring socio economic status people generally use very broad brackets which don't really take into account the vast difference between those brackets.
For example socio economic grade AB = 20pc of population. This covers those approx £80k household income and above. However a family with 80k income is highly unlikely to have afforded private school. And then when their kid is at university will probably be struggling to fund the living expenses from their income (especially if they have more than 1 child) whereas average private school family income is around 150k. Can you see that a student from those backgrounds one is likely to have to work in holidays (such as my DP son who is state educated and at Oxbridge) compared to his private school friends many of whom don't need to work in the holidays.
The point is simply that those who can afford private school are almost inevitably this highest earners in society and that means their children will have better opportunities, less pressures to take jobs etc. Traditional measures of socio economic (ONS) do not capture this level of detail.
I stand by my assertion is that this study tells us that those who have the greatest privilege in background tend to do the best which is not actually new information

Your "assertion that this study tells us that those who have the greatest privilege in background tend to do the best which is not actually new information," is not consistent with the evidence collected by the university:

"In spring 2019, as part of the development of the 2020-21 to 2024-25 Access and Participation Plan (APP), the University of Cambridge conducted a self-assessment of data on gaps in continuation and attainment between different student groups. The exercise demonstrated the absence of any persistent trends in gaps in attainment due to participation (POLAR4) or deprivation (IMD) markers"

How did they identify and then ignore the ‘lower potential’? I can see that they may have applied slightly too great an allowance for previous educational experience, but (apart from in hindsight) how did they know at the time of offer that these specific students were of lower potential? It’s only nearly 4 years later that the university sees the results of their admissions judgements - so if these were the 2019 results, informing the 2019 application and 2020/1 matriculation rounds, you might expect to see ‘tweaks’ informed by the data in the 2023/4 results at the very earliest - and those results have unique factors of Covid-era education affecting them too.

My dc go to a private school (so no sour grapes here) and the reality is that Oxbridge has been over recruiting independent school students for years. There are an incredibly high number of very bright children from state schools achieving the best grades who merit places. Put simply, trends towards admitting more state school candidates at both Oxford and Cambridge merely reflect that reality.

I see no evidence whatsoever that admitting more state school pupils is adversely impacting quality of graduates when interviewing for roles in the City.

On reflection, I’m also not entirely comfortable with ‘getting a first’ being the ‘gold standard’ measure of ‘having the right potential to have been admitted 3/4 years previously’.

If you were saying ‘98% of all students who fail their degrees at Cambridge were state educated’, that’s an indication that either they were wrongly admitted or that the specific environment is not meeting their needs well.

The measure ‘good degree’ (2:1 or above) would be more valid, as it is the be criterion needed for many high quality graduate schemes etc. Someone gaining a 2:1 or 1st from Cambridge absolutely had ‘the right potential to be admitted and benefit from its education’. My understanding is that the difference is much smaller for that measure? Do you have the full time series for that data? (You have said you don’t ‘like’ that measure because so many Cambridge students meet it - ie very few of the students admitted are genuinely ‘of low potential’)

https://www.cambridgeassessment.org.uk/news/state-school-pupils-do-better-at-university-cambridge-assessment-research-confirms/

State school kids with the same A-level results as private school students do better at university.

I think there are definitely issues raised around teaching and administration etc at Cambridge but I just think that theres a dollop of snobbery there too.

I am definitely not against selective schools, I’m a grammar school kid with a child in private school. There is no doubting that there is more time invested in each child at a private school but it doesn’t surprise me that candidates that look great at A-level from a highly focussed and individualised environment moving into an independent study environment don’t do as well.

@User37482 Nine year old research isn't very relevant I'm afraid. Look at @Marchesman 's more recent data.

Ok, sorry yes hadn’t read the thread, there is an obvious gap. Tbh I think students should be selected purely on ability. It makes no difference to me if they went to a private school, comp or a state school.

I think the issue is how Cambridge admissions procedures fairly and objectively recognise, assess and rank ‘ability’ for every student who applies.

Because while it is obviously correct that there are factors other than ‘raw attainment’ that measure ‘underlying ability’, and thus should be taken into account, the OP feels that ‘adjustment for prior schooling’ may have gone too far.

It’s interesting that, in the OP’s data source, A level score (within the limited range of variation relevant to a University with a very high standard offer) is not strongly correlated to final results.

OP had a 36 page thread last year on the higher education forum about "Blatant social engineering at Oxbridge" where they repeatedly had their arse handed to them. They've now taken their bee-filled bonnet to an audience on the private schools thread where they might have assumed a more sympathetic audience, without much luck.

That is a matter of opinion. Since you said on this thread that "simply reducing it to all state v all private delivers a nonsense data set" (which reflects exactly what Cambridge has been doing) I am inclined to question your judgement.

What I originally posted was this:

"Despite resistance from some tutors, Cambridge University’s Access and Participation Plan 2020-21 to 2024-25 includes a target to increase the proportion of UK state sector students that is entirely separate and independent of aims for POLAR4 quintiles 1 and 2. Formulating admissions targets for the University of Cambridge’s Access and Participation Plan (2020-21 to 2024-25) | Cambridge Admissions Office
The university's own research in 2011 had "found no statistically significant differences in performance by school type, and there was no evidence of the phenomenon observed at other UK universities of state sector students outperforming their privately educated peers" https://www.cao.cam.ac.uk/sites/www.cao.cam.ac.uk/files/ar_gp_school_performance.pdf Subsequent data shows that students from independent schools performed better in examinations than students from state schools by 2015/16, at a level that is highly statistically significant: https://www.informationhub.admin.cam.ac.uk/university-profile/ug-examination-results/archive
Therefore, APP 2020-21 to 2024-25 makes no attempt to justify the state school target on the basis of student performance. In fact the only justification given is: "We recognise that school type is not a characteristic used by the OfS or contained within its Access and Participation dataset; we recognise too that the state versus independent binary masks a range of educational experiences…[however] each of the under-represented groups identified within this Plan appear in far greater numbers in state maintained schools, as do students from low income households who are not identified by any of the measures currently available to us."
The result of this can be seen in https://www.cao.cam.ac.uk/files/attainment_outcomes.pdf
In final degree examinations: "The per cent mark remained lower for the three secondary school types: • Comprehensive (estimate = -0.70, SE = 0.19, t = -3.63, p< 0.001); • State grammar (estimate = -0.98, SE = 0.19, t = -5.22, p< 0.001); • State other (estimate = -0.87, SE = 0.20, t = -4.32, p< 0.001)" To put this into context, these are the figures for students with "cognitive or learning difficulties (estimate = -0.88, SE = 0.33, t = -2.67, p< 0.01)"
Regarding the acquisition of a First: "The probability of the outcome remained lower for the three secondary school types: • Comprehensive (coefficient = -0.20, SE = 0.06, z = -3.13, p< 0.01); • State grammar (coefficient = -0.30, SE = 0.06, z = -4.81, p< 0.001); • State other (coefficient = -0.24, SE = 0.07, z = -3.57, p< 0.001)"
Selection according to potential? Really?"

If you can find anything on that thread that constitutes having my "arse handed to me" I would be interested to see it.

"That is a matter of opinion. Since you said on this thread that "simply reducing it to all state v all private delivers a nonsense data set" (which reflects exactly what Cambridge has been doing) I am inclined to question your judgement."

As you would know if you hadn't selectively quoted it, the thing being "simply reduced" was not admission statistics or quotas but your judgement of finals firsts as the marker of outcomes between "all state" and "all private". Aside from the general reductiveness of the metric, all privates are not equally likely to get Firsts. There are a few specific schools that will completely skew the numbers. The data for that is, as far as I'm aware, unavailable, but I doubt it would help your argument of general private superiority at Cambridge if it was.

"If you can find anything on that thread that constitutes having my "arse handed to me" I would be interested to see it."

As there's quite a lot to go through, we can probably let others judge that for themselves if they're that invested in the subject. (I'm not; you clearly are.)
www.mumsnet.com/talk/higher_education/4819030-oxbridge-blatant-social-engineering-not-admission-according-to-potential?page=1

You are evidently not "invested" enough in the subject to pay any attention to the metrics that Cambridge used, and your opinion that "there are a few specific schools that will completely skew the numbers" is baseless.

One wonders if you go through life offering uninformed opinions or if you reserve them for discussions about private schools.

I've been very restrained about not offering my absolute opinion on you, but you make a conclusion pretty easy despite me only having the info of some weak but obsessive Rees-Mogg cosplay on a couple of grievance-filled Mumsnet threads.

So not just private schools. Thanks for clearing that up.

Going to imagine I have as much direct and parental experience of private schools as you, and rather more of Oxbridge Of course, that doesn't mean much considering the opinions of the thankfully ex-Cambridge prof you so love.

I'd rather not derail the thread further, so will bow out now. Take the last punch, if you like.

I have now read the stats and I still don’t understand the argument.

Setting aside school type for a moment, female undergraduates got fewer firsts and so did certain groups of BAME student, if I have read these stats correctly. I would like to hope very few people would even try to suggest this is because those candidates are less bright.

When I was a student (more than 3 decades ago) there were relatively few women getting firsts in almost all (possibly even all) tripos subjects at Cambridge. It was suggested back then that this is because a certain level of self confidence is required to exhibit the kind of standout bravura performance needed to get a first, rather than the kind of steady, conventional and safe approach that tends to get a 2:1. Most tripos subjects have also traditionally been dominated by exams, with much riding on one week of old-style finals, which it has been suggested can tend to suit typically male working styles more.

We know from research on school assessment that changes to assessment styles swing percentages in relation to male and female achievement.

There are clearly far more factors in the mix in relation to attainment at the university than raw talent. I’m unconvinced that drawing a straight line between achievement at university and ability either in relation to students from certain school types, gender or other diversity characteristics is possible.

Any attainment gap opening based on school type could well be because of a mixture of prior educational input, lack of confidence/struggling with the unfamiliar environment. What these stats are perhaps really suggesting is just that more efforts are needed to support students from lower socioeconomic backgrounds and enable them all to attain at the level of their potential.

I sent my own dc to private school and I and other private school parents have to accept that in having done so we have purchased privileges for them. They have smaller classes and more individual attention. It is to my mind entirely fair for those privileges to be taken into account when it comes to Oxbridge admissions.

@MissHavershamReturns
In attainment terms you haven't purchased privileges for your children because "exam differences between school types are primarily due to the heritable characteristics involved in pupil admission".
https://www.nature.com/articles/s41539-018-0019-8
So the premise that an adjustment for school type is fair at the point of entry to an academically selective university is problematic. Pupils in private schools are a selected high attaining group as are those in grammar schools and in socioeconomically (or academically) selective comprehensive schools. It is necessary to appreciate that Cambridge set one target for POLAR4 q1/2 admissions and another one for state school admissions. Type of school refers to state vs private, and not for example the type of school that has never sent anyone to Cambridge - contextual admissions are entirely separate.

Applicants from private schools go through the same process as state applicants and those with the most potential in the private pool are admitted. Because private school numbers were targeted for reduction, admission for them has become progressively more selective than before, while admission for applicants is less selective than before. Inevitably a gap has opened between the performance of students from the two types of school. Had a target been followed to reduce admissions from any other demographic, students in it would demonstrate a similar improvement in performance.

Socioeconomic status has no demonstrable effect on attainment at Cambridge, nor is this about giving a leg up to the disadvantaged, levelling playing fields etc etc. Comparing 2017 and 2021 the number of POLAR4 q1 and q5 acceptances from comprehensive schools increased by 19 and 135 respectively.

If by fair you mean meritocratic you need to go back to Parks 2012, Academic Performance of Undergraduate Students at Cambridge by School/College Background - which is still being cited by the university to support the claim that "school background is not a factor in Cambridge degree success":
https://www.cao.cam.ac.uk/admissions-research/school-background

There are fuller explanations from various people in the "Cambridge University discriminates against children from private schools thread" that was started after academics proposed that the proportion of privately educated students at Cambridge be reduced to 10%. This thread was not intended to rehash it, but simply to flag up Butterfield's article which describes the mental health issues etc that followed the abandonment of a meritocratic admission system.

And finally, controlling for other factors, gender is not a significant factor in terms of firsts, or percentage marks at Cambridge - except for "good honours" where males do worse.

But @Newbutoldfather 's point was that accepting lower performance at degree level may be necessary to ensure the really top ability people get through that gate, regardless of background.

If they are never going to achieve potential - even if that's due to very real personal circumstances - then there's no point in the discrimination at University entrance.

Or put another way: it's a reasonable goal for a university to aim to advance knowledge, which would be the further end of that funnel: acclaimed academics /Nobel prize winners.

If lowering the bar for state school students results in more of these top achievers, then that's an acceptable thing to do... even if you get a lot of 'false positive' contextual offer students doing less well at degree level.

If, however, you end up with fewer acclaimed academics /Nobel prize winner... because so many lower ability state students (or those for whom the barriers to achievement remained, as you say) were offered places that they displaced private school students who would have reached those heights... then you can no longer use the reasoning that you don't want to miss out on the really exceptional people.

Each student who doesn't get a place may have been that truly exceptional Nobel prize winner. Whether that's a state school student who didn't have the facilities to get good A levels, or whether it's a private school student rejected due to discrimination.

That's why we need the information about results to go further down the funnel, if we want to say that 'exceptional achievement' is the end goal, rather than degree attainment.

Not convinced that argument stands up.

If you look at the 20 schools which sent most pupils to Oxbridge last year: 10 of them are private (441 offers total) and 10 of them are state (455 offers total).

www.keystonetutors.com/news/oxbridge-which-schools-get-the-most-offers#

Why do you think the private super-selectives influence Cambridge results any more than the state super-selectives?

Those very top state schools are just as likely to have drawn top students from the private sector as the other way round: being free is a big draw!

Much more likely to fail to recruit top achievers from private schools. Firstly because an applicant with all A stars is substantially more likely not to have been given an opportunity to demonstrate the ceiling of their ability. Secondly, we know that school type does not statistically affect attainment.

As discussed, and as the university finally worked out, contextual offers are the way to go, rather than blanket targets for school type.