How clever to go to Oxbridge...? (Navel-gazing discussion).

[justlumpingalong]

Discussion with DH and some friends last night, triggered the boat races being on telly today... We reckoned you needed to be top 5% of your birth cohort, academically, to stand a chance (most of us went there, so clearly it must be a near-genius thing...)....but then we started picking it apart.

Our analysis went a bit like this:

Oxbridge take, very roughly, about 1% of kids turning 18 each year. But their recruitment and selection processes would have to be pretty awesome to find the top 1% and only the top 1%....and clearly these systems are not that great at all.

Apparently half of all schools don't put anyone forward for Oxbridge, so if you're at one of those schools, you're out. So, Oxbridge now need to find the top 2% from the remaining schools.

However, half of all kids don't apply to University at all. This may be a sensible, self-aware strategy from many people - aware that their skills lie elsewhere, but I'm willing to be a fair few of our 2percenters get lost that way. Let's say that Oxbridge now need to get the top 3% of those remaining in the pool to fill their places.

Now remove all the kids who think that Oxbridge is 'not for the likes of me'. I reckon there's loads of them - maybe half of all potential applicants? So now Oxbridge is fishing for 6% of the remaining pool.

Now get rid of anyone who wants to study Drama, or Dutch, or Dentistry, or one of the myriad other subjects that Oxbridge don't offer. And forget anyone who is burning to go the Big Smoke, or must follow their girlfriend to Hull, or wants somewhere you can do fell running. I reckon we're now down to fishing for 10% of the remaining candidates.

Now to interviews. What is the ratio of interviews to offers? 1:4? So, to fish the top 10%, even assuming a perfect shortlisting strategy on Oxbridge's part, you've got to invite the top 40% to interview.

And then we have interviewing error. All the evidence that has ever existed on utility of interviews for selecting the best candidates for anything, shows that they're crap. Let's be generous and suppose that, with aptitude tests thrown in, the efficiency of this final stage of selection is 50%. That would mean that offers are made to kids who are, largely, in the top 20% of their cohort. That, roughly, gives an IQ of 113.

So, ignoring all debate about what 'clever' means, what do you reckon?

Ps go gently - bit hungover...

[TychosNose]

expat96 ime Professors at all top research unis in the uk are more likely than not to have a degree from Oxbridge (undergrad or postgrad).

[expat96]

quit2dis:
You're exaggerating about well, well within the top 1% in maths ability: Caltech undergraduates aren't better than Oxbridge undergraduates in STEM.

I'm not sure I'm exaggerating. Let's make sure we understand each other.

• Are you claiming that the average Caltech undergrad is worse than the average Oxbridge undergrad in maths knowledge and ability? I would very enthusiastically dispute this as only a minority of Oxbridge undergrads specialize in STEM whereas all of Caltech's do.
  

• If, on the other hand, you are claiming that the average Caltech math or physics major is worse than the average Oxbridge math or physics student, I would agree that is probably the case in the first year of undergraduate study. If I understand correctly, pretty much all Oxbridge math and physics first years will have at least 2 strong STEM A-levels. Due to the fragmented high school system in the US, not all Caltech admits will have had the opportunity to take courses at this level. I have no opinion about the comparison at the completion of undergraduate studies, bearing in mind that the normal US course of study is 4 years vs the UK's 3 years.
  

• If, as I infer, you are claiming that the average math or physics graduate student at Cambridge who did their undergraduate degree at Oxbridge is better than the one who did their undergraduate degree at Caltech, I will defer to your judgment. I will, however, point out that the proportion of Caltech math and physics majors going on to Ph.D programs may be very different to the proportion of Oxbridge math and physics students going on to do Ph.Ds. I would also point out that there may be a different propensity for the best students (the kind that are likely to gain graduate school places at Oxbridge) to cross the ocean when they all have very good departments much closer to home. Therefore I would question whether your observations provide much information about the relative quality of the full undergraduate populations at these schools - or even the STEM subset.
  

• Is it your opinion that Oxbridge maths and physics students are not well within the top 1% of the UK in maths ability?

[expat96]

To clarify, in the last line of the previous post I am referring to Oxbridge undergraduate maths and physics students.

[quit2dis]

I am further assuming that the children of Oxford professors will be more familiar with Oxford and are more likely to have attended open days, etc., as their senior schools are likely to be geographically close, so will have a higher propensity to apply.

But in the UK students don't want to stay in their home town. It is not the norm that children of Oxford academics even want to go to Oxford (as opposed to Cambridge, LSE etc).

[quit2dis]

My statement is that the average level of a final year Caltech maths or physics class (and the level of the students in it) is not higher than a final year maths/physics class at Oxbridge. In my experiences of both, the level at Caltech is usually lower. Oxbridge final year covers material which is not done until grad school in the US, but the overall pace is higher and deeper at Oxbridge.

The levels are more comparable in grad school. Caltech students get longer to do PhDs and hence tend to have a big advantage over the UK PhDs for this reason.

I dispute the "well, well, within the top 1%" as I don't think you have evidence for this. Lots of very bright kids don't go to Caltech - they go to other top schools instead such as Princeton for maths and many bright kids through socio-economic circumstances go to their local state universities instead of top private schools.

I think Oxbridge STEM students are within the top 1% for maths ability. I would not be prepared to split them more than this, as there are many very strong STEM students who for whatever reason do not apply to Oxbridge.

[expat96]

quit2dis, may I ask what is the basis for your comparison? You may well be right. My experience with Caltech is from over 25 years ago; my experience with Oxbridge is confined to my work colleagues, who all have been through a further filter so are unlikely to be representative of the populations at the schools.

You are absolutely correct that many STEM inclined kids will go to MIT, Princeton, Harvard or many, other schools. You are also correct that the top students at Berkeley, North Carolina, Michigan or many other state schools are as good as the students at the top private schools. But keep in mind that the United States is larger than the UK. There are approximately 5MM 18 year olds in the US right now. 1% of them would 50,000 kids. Caltech enrolls about 200 US students each year. I'm not suggesting that it enrolls the top 200 each year. I am suggesting the vast majority of its students will be from "well, well inside" the top 50,000 in math ability.

You are also correct that I don't have direct evidence for this statement; I no longer have access to Caltech's maths achievement test scores, which would be perhaps the most relevant data. The only figures which I can still access come from their most recent Common Data Set, which indicates that their 25th percentile scores on the SAT Math is 770, the ACT Math is 35 and the ACT Composite is 34. Officially, all three scores are 99th percentile, as the testing boards cap the percentile ranks at 99. Unofficially, the 25th percentile ACT Composite score (the only one for which I can easily find the summaries) is 2.6 standard deviations above the mean. That corresponds to roughly the 99.5 percentile of the students in the US who are thinking about applying to college. If you believe there is a selection bias in the people taking the exam, then the percentile rank of the whole population should be higher. And this is for the composite score; when I was there Caltech had an unusually high spread between the math and English scores (I let you guess in which direction). So, feel free to continue to dispute my statement; as I wrote, I can't prove it directly. But I'll keep believing the circumstantial evidence.

[Sadik]

" I went to a highly selective London private school on a scholarship. I was at the top of the year in that school, but when I went to Cambridge I definitely felt that I was in the bottom half (although I did get a 2:1). I've since had a successful career and worked with lots of intelligent people, but never felt out of my depth the way I did at Cambridge."
Maybe it depends on subject /college/ cohort? I went to a not-particularly-academic-at-all comp (over 90% left at 16, of those who did stay on only a handful took 3 A levels). No-one I'm related to as far as I know stayed at school past age 16. I expected to find Cambridge terrifying - I'd say there were a handful of super-bright people, but plenty didn't seem any smarter than people I've met elsewhere in life. (And some were distinctly lacking in common sense )
I'd say that on the whole people I knew at Cambridge from more 'normal' backgrounds (ie not public school) were mostly hard workers with an aptitude for exams rather than exceptionally clever per se.

[teacherwith2kids]

"I'd say that on the whole people I knew at Cambridge from more 'normal' backgrounds (ie not public school) were mostly hard workers with an aptitude for exams rather than exceptionally clever per se."

I do think that depends on subject and cohort. My college friends, and those doing my science subject - yes I would agree, and tbh the main differentiator by the end of the course was work ethic, because some of the brightest on entry simply had no experience of having to work really hard.

The mathematicians i met through a club, several of whom were Maths Oympiad types - now they WERE exceptionally clever

[Dellarobia]

Sadik Maybe you're right that it depends on subject / college etc. I'm very interested to hear that some of you had a different experience to me (in terms of how bright you felt your fellow students were). Until now I had assumed that nearly everyone at Cambridge is super bright!

[Sadik]

It might help that I went to a college which is consistently at the bottom end of the ranking tables, and that at least back then was majority state school pupils.

Also, at school, I spent a lot of the time feeling distinctly behind the curve (great at exams, not so great at social skills), whereas Cambridge is/was well supplied with the socially inept.

[teacherwith2kids]

Della, I do also think it depends on how bright, or rather how suited to the Oxbridge experience, one turns out to be oneself?

I don't have the kind of job that marks me out as 'academic' - I teach 9 year olds, not university students. DD fell around laughing when she came with me to my old school's open day and met several of my classmates, who regaled her with stories of how clever I had been. On the other hand, I have a silver cup somewhere that was awarded for best finals results in my (mixed, ancient) college - so perhaps my perception of my fellow students as being 'normal and a bit like me' (I still remember the relief, the homecoming, of 'now THIS is where my people have been all my life' on arrival at my college, that I hadn't had in my academic girls' school) is a result of where I stood in that particular pecking order at that point?

No claims for my brains now - pregnancy x 2 did for that. I always say that I made 2 more brains, and it took 2 brains-worth of activity and sharpness out of my own.

[Dellarobia]

Whereas I do have the kind of job that makes people say "you must be clever" when they hear what I do - but I certainly didn't come anywhere close to winning a uni prize (I'm seriously impressed by yours btw!).

[quit2dis]

There are approximately 5MM 18 year olds in the US right now. 1% of them would 50,000 kids. Caltech enrolls about 200 US students each year. I'm not suggesting that it enrolls the top 200 each year.

This is a bit misleading, though, since Princeton enrols 1000 students per year, and so do MIT and other top private schools for STEM. Meanwhile the top state schools have way higher enrolment, although admittedly only a relatively small fraction will major in STEM. Many top schools are competing for the top 1%, and they all have different selection requirements.

BTW you have also now committed to a specific definition of top 1% in maths ability - you are equating it with SAT/ACT maths performance. This is indeed the definition used for college entrance, but it is a questionable definition i.e. SAT/ACT doesn't measure just potential, but is arguably affected considerably by prior education and socio-economic circumstance. The OP had in mind an abstract definition of intelligence which was not affected by education and circumstances.

[expat96]

I have no intention of misleading.

If you are familiar with US universities you will know that there is a "magic circle" of schools which are at the top of the pyramid: Harvard, Yale, Princeton, Stanford, MIT and Caltech. Several revealed preference surveys consistently indicate that these schools often lose cross-admits to each other, but much less often to other schools. These six schools in aggregate enroll ~7,000 students each year. ~15% of these 7,000 are from overseas, leaving ~6,000 places for US students. Again, there are a little over 5MM 18 year-olds in the US, meaning that these schools have places for approximately 0.12% of the population. These schools are also all private with large endowments which allow them to give generous financial aid which, in the US, covers living expenses. To be clear, these schools do not enroll the 0.12% cleverest kids in the US; as I made clear earlier, the first four, at least, explicitly seek something other than pure academic excellence.

For comparison, Oxbridge together enroll over 6,000 students each year, of which ~5,000 are UK residents. There are a little over 900,000 18 year-olds in the UK, meaning that Oxbridge have places for approximately 0.55% of the population.

You misrepresent my preferred metric for maths ability. I explicitly stated that I would prefer something like the math achievement test or, even better, the GRE Math test. However, there will be a serious selection bias in the people who take the GRE Math test, and I can't get school aggregate data for the math achievement test. I made clear that the numbers I provided are the best proxy I can get at the moment. You can accept them or not; I have already written that I can't do better.

You have not responded to my query about the basis for your comparison between Caltech and Oxbridge STEM students. Let me be more explicit: Does your experience include teaching or TAing undergraduate math or physics classes at Caltech? If not, again, what is the basis for your comparison?

[deouynt2]

I went to a comprehensive school, achieved AAB at A-level and went on to study at Trinity Hall, Cambridge. My experience of the entrance process was to check that I could actually handle the Cambridge education. I think I got in mainly because of my performance at interview. I found studing at Cambridge tough - very tough - and I went through hell to get a 2.1 out of it. Having been to Cambridge doesn't seem to have done me any favours in later life. Oh and my IQ score is around 110 (according to some online tests).

[dairymilkmonster]

I studied medicine at Oxford and my husband did too.
DH: extremely bright, not that hardworking, very logical expansive mind with exhaustive knowledge and fab memory. went to private school with high expectations but he didnt really need what was on offer. Insane exam results all through uni with minimal work.
Me: Bright but not extremely so , obsessively hard working. well, i was pre kids. I got through with hard slog. I was at a comp where the teachers had no idea how to apply to medical school ( was in late 90s...hope better now) so had to fend for myself, but my parents helped with research. At oxford i was always studying whilst others had fun, but i got my medical degree in the end and had an amazing education.

Plenty of each type in oxbridge! Most of the people like me came from private schools in my experience but not exclusively.