Altrincham pass mark vs Sale (and Urmston)
Moderators: Section Moderators, Forum Moderators
Altrincham pass mark vs Sale (and Urmston)
.
Last edited by Elphaba on Fri Jul 06, 2012 11:57 pm, edited 1 time in total.
-
- Posts: 593
- Joined: Thu Feb 22, 2007 3:29 pm
- Location: Trafford
Re: Altrincham pass mark vs Sale (and Urmston)
I think it is because UGS and SGS standardise against a national cohort, whereas the Altrincham Grammars standardise against the actual cohort that takes the AGS exam.
So, in order to achieve an average 100 for the UGS/SGS exams you will need a lower raw score than for the AGS exam. Hence the lower standardised pass mark. Hope that makes sense!
So, in order to achieve an average 100 for the UGS/SGS exams you will need a lower raw score than for the AGS exam. Hence the lower standardised pass mark. Hope that makes sense!
Re: Altrincham pass mark vs Sale (and Urmston)
.
Last edited by Elphaba on Fri Jul 06, 2012 11:58 pm, edited 1 time in total.
-
- Posts: 593
- Joined: Thu Feb 22, 2007 3:29 pm
- Location: Trafford
Re: Altrincham pass mark vs Sale (and Urmston)
The answer is that I don't really know.
However, I have spent some time looking at this in past years and I think I concluded that the more academic a school the more likely they are to standardise against the cohort taking the exam (as opposed to nationally) because that gives you a much wider list of scores. The theory is that a wider breakdown enables you to pick a point at which to draw the line that lessens the risk that you are drawing it at the wrong point.
I don't quite see it myself since I presume a bell curve is a bell curve even if one is a flatter bell curve but I am no statistician!
Try not to worry about it too much. (Easier said than done, I know.)
However, I have spent some time looking at this in past years and I think I concluded that the more academic a school the more likely they are to standardise against the cohort taking the exam (as opposed to nationally) because that gives you a much wider list of scores. The theory is that a wider breakdown enables you to pick a point at which to draw the line that lessens the risk that you are drawing it at the wrong point.
I don't quite see it myself since I presume a bell curve is a bell curve even if one is a flatter bell curve but I am no statistician!
Try not to worry about it too much. (Easier said than done, I know.)
Re: Altrincham pass mark vs Sale (and Urmston)
.
Last edited by Elphaba on Fri Jul 06, 2012 11:58 pm, edited 1 time in total.
Re: Altrincham pass mark vs Sale (and Urmston)
I suspect the lower standardised score pass for AGBS reflects that the cohort of children entering that exam can be expected to be significantly higher than average attainers. So a score of 100 standardised against the cohort for AGBS would reflect a higher raw score average than if standardised against the general population. Assuming that to be the case, a mean average score of 100 based on the cohort rather than national attainment could allow the school better diagnostics at the higher end. If they standardised against the national cohort, i suspect that they would get quite a lot of children getting the cut off score of 141. this way they would be able to see the spread of the results at the higher end better. I am just surmising here mind, but i can see a kind of logic in it
So if the difference between AGBS and SGS/UGS is about 20 points, assuming there is not real difference between the cohorts taking each of the trafford Exams, then this suggests that on average the children taking these exams are achieving on average about 6 or 7 points per test more than could be expected in the general population eligible to take the 11 plus nationally. Of course there may be historical differences between the cohorts and so all such claims need a note of caution!
So if the difference between AGBS and SGS/UGS is about 20 points, assuming there is not real difference between the cohorts taking each of the trafford Exams, then this suggests that on average the children taking these exams are achieving on average about 6 or 7 points per test more than could be expected in the general population eligible to take the 11 plus nationally. Of course there may be historical differences between the cohorts and so all such claims need a note of caution!