These Oxbridge courses are a joke...

I have extracted the Oxford and Cambridge data from that source (thanks again!). But can a statistician explain what this is actually saying...



Specifically which data point shows the correlation?

Oxford says that their research shows that GCSE performance is a better indicator of potential for their courses than AS results - I doubt they'd be making that up since it's in their interest to get the best applicants!

Actually, I have just found some data. If you compare the correlation coefficients between Oxford and Cambridge in this data, you'll see that there seems to be a stronger correlation between GCSEs and degree performance at Oxford than at Cambridge, if I'm interpreting this correctly.

I have extracted the Oxford and Cambridge data from that source (thanks again!). But can a statistician explain what this is actually saying...



Specifically which data point shows the correlation?

I don't think I've ever seen Oxford actually say this.

Interesting find.

@Noble. ?

"Logistic regression model 1 [GCSEs] predicted whether or not students would get a 2:1 or above with 69.5% accuracy while model 2 [AS] predicted with 68.4% accuracy. Neither of these figures is especially high, given that you would expect 50% accuracy by randomly guessing the outcomes with no consideration of student prior attainment. Hence, although GCSEs appear to more accurately predict whether a student will get a 2:1 or better in their degree, neither is especially accurate. Model 3 returned a prediction accuracy of 70.1%; only a gain of 0.6% accuracy on using GCSE results alone. "

^ most illustrative section for me.

Interviews or admissions tests, viral variable for both unis, are not included in that study. Looking at the two in isolation had the potential to be misleading.