Lies, damn lies and entrance statistics

Shock! Horror! Oxford University appears to be discriminating in favour of white students and against ethnic minorities:

In medicine, 43 per cent of white students who went on to receive three or more A* grades at A-level got offers, compared with 22.1 per cent of minority students with the same grades, according to data obtained by The Guardian newspaper.

Now, let's bear in mind that as A-level grade inflation has taken hold, three A* grades is no longer the talent indicator that even three A grades might have been 25 years ago:

In a 2007 report Robert Coe compared students scores in the ALIS ability test with equivalent grades achieved in A level exams between 1988 and 2006. He found that students of similar ability were achieving on average about 2 grades higher in 2006 than they were in 1988. In the case of maths it was nearer to 3.5 grades higher.[24]

i.e. an 'A' in 2006, which is presumably near an A* in 2013, was a 'C' in 1988. But that 43% vs 22% relative percentage seems to be damning, surely?

Assuming, just for the moment, that Oxford University medical professors aren't rampant racists, what could be the explanation? An Oxford University spokesperson hints:

"We do know that a tendency by students from certain ethnic groups to apply disproportionately for the most competitive subjects reduces the success rate of those ethnic groups overall."

Chinese and South Asian (Indian, Pakistani, Bangladeshi) students are disproportionately likely to apply to become doctors; you can tell this because they are disproportionately (compared to % of UK population) likely to be medical students. This is inversely true for Afro-Caribbean students who are rarely found in the professional medicine track. What the Daily Mail fails to supply is the raw stats on applicant and successful applicant numbers broken down by ethnic group, which could then be analysed based on the ethnic distribution of 18 year old students. I would be very interested in those numbers, which I suspect would tell a very different story to the DM's angle.

I invite the reader to inspect the list of Oxford University physiology lecturers. It's not obvious that they are prejudiced against South Asian or Oriental medical experts. Out of 53 staff, 6 match that ethnic background (11%) which is much greater than the 5% you'd expect even given the most favourable reading of the UK ethnic breakdown.