Contextual GCSE Scores

Does anyone know how these are calculated?

e.g.

"Candidates were called for interview if their TSA score was above 57.75 or their contextual GCSE score was above 0.00 (i.e. a positive score). Candidates not meeting either of these criteria may have been invited for interview after careful consideration of other factors".

Keble geography admissions feedback.

Reply 1

ombtom

17

Original post by Pars12

Does anyone know how these are calculated

http://www.education.gov.uk/schools/performance/archive/schools_08/s3.shtml

Reply 2

Doones

21

Original post by ombtom

http://www.education.gov.uk/schools/performance/archive/schools_08/s3.shtml

That's the government's calculation. I'm not sure if Oxford uses the same method.

Reply 3

Doones

21

Maybe @BrasenoseAdm can help...?

Reply 4

fluteflute

19

This might help slightly (not that I understand it):
https://www.whatdotheyknow.com/request/a100_medicine_admissions_statist_3#incoming-726239

Reply 5

Doones

21

Original post by fluteflute

This might help slightly (not that I understand it):
https://www.whatdotheyknow.com/request/a100_medicine_admissions_statist_3#incoming-726239

Easy peasy. Not...

But I *think* it means the candidate gets a positive score if they are outperforming their peers in the context of their school.

Original post by jneill

Maybe @BrasenoseAdm can help...?

Contextualised GCSE A*s estimates, using leastsquares, the expected number of A*s a candidate will have using the
proportion of pupils obtaining 5 or more GCSEs at grades Cand higher at the candidates’ schools as a single predictor.

The dataset consists of all University applicants with 6 ormore GCSEs. Actual A* performance (Y) is compared with the predicted
score (Y-hat). The data is then normalised as follows: whereY-hat = Y, a candidate is scored at 0. Over- or under- performance is expressed
in terms of ‘standard deviations’ from the slope of the linear regression line.

Reply 7

Doones

21

Original post by BrasenoseAdm

Contextualised GCSE A*s estimates, using leastsquares, the expected number of A*s a candidate will have using the
proportion of pupils obtaining 5 or more GCSEs at grades Cand higher at the candidates’ schools as a single predictor.

The dataset consists of all University applicants with 6 ormore GCSEs. Actual A* performance (Y) is compared with the predicted
score (Y-hat). The data is then normalised as follows: whereY-hat = Y, a candidate is scored at 0. Over- or under- performance is expressed
in terms of ‘standard deviations’ from the slope of the linear regression line.

Er, thanks :wink:

So was this correct? "I *think* it means the candidate gets a positive score if they are outperforming their peers in the context of their school."

Reply 8

DrAdrenaline

14

Original post by BrasenoseAdm

Contextualised GCSE A*s estimates, using leastsquares, the expected number of A*s a candidate will have using the
proportion of pupils obtaining 5 or more GCSEs at grades Cand higher at the candidates’ schools as a single predictor.

The dataset consists of all University applicants with 6 ormore GCSEs. Actual A* performance (Y) is compared with the predicted
score (Y-hat). The data is then normalised as follows: whereY-hat = Y, a candidate is scored at 0. Over- or under- performance is expressed
in terms of ‘standard deviations’ from the slope of the linear regression line.

The proportion of pupils getting 5 A* - C changes each year. When doing this do you use the average of say the last five years or the data of year the applicant sat the GCSE? how does it work?
thanks

(edited 8 years ago)

Reply 9

Pars12

OP

9

Original post by BrasenoseAdm

Contextualised GCSE A*s estimates, using least squares, the expected number of A*s a candidate will have using the
proportion of pupils obtaining 5 or more GCSEs at grades C and higher at the candidates’ schools as a single predictor.

The dataset consists of all University applicants with 6 or more GCSEs. Actual A* performance (Y) is compared with the predicted
score (Y-hat). The data is then normalised as follows: where Y-hat = Y, a candidate is scored at 0. Over- or under- performance is expressed
in terms of ‘standard deviations’ from the slope of the linear regression line.

Are you saying (broadly) that the result is positive if an applicant's number of A*'s is greater than the expected number for the applicant's school? If so, they must be interviewing an awful lot of geography applicants at Keble!

Also, from 2012 The Welsh Baccalaureate GCSE was compulsory and awarded on coursework only. I am not aware of any pupils failing it (I don't think the teachers let that happen). It counts as 3 Gcse grade A's! I quite approve of the work they do for it but it does seem that it will seriously skew your statistics. Or possibly, the Welsh Assembly has seriously shot itself in the foot.

Contextual GCSE Scores

Quick Reply

Related discussions

Latest

Trending

Trending

Articles for you