Querying Results

[mattsurf]

I wonder what legal basis there might be for challenging the age-related standardisation as an "easy" exam (which after all is down to the consortium to control) appears to favour a younger candidate disproportionately. There is a point where this becomes discriminatory.

So long as the paper was age standardised against the cohort who sat the same paper, I cannot see how it is discriminatory. Basically Age standardisation works so that the mean score of children born in the same month is 100.

If the mean score of children born in November was 75% and the mean score of children born in April was 72% then the standardised score would be 100 for a November born child who scores 75%

for each standard deviation above the average a child gets 15 points - so if the standard deviation for children born in November was 10%, a child who gets 85% would get score of 115. If the standard deviation of scores for children born in April was 7%, then a child born in April who gets 79% would also get a standardised score of 115

As the mean gets closer to 100%, then the standard deviation will get smaller, therefore, if the test was very easy and the mean was very high, then the standard deviation would be very small, because a student could not score above 100%, ie if the mean was 100% then the std dev = 0. As a result a small number above the mean could equal a big difference in score. Ideally the test should have a mean of between 50% to 60%, then the std deviaton would be quite big (probably in the range 15% - 20%), so a child with 60% would get a standardised score of 100 while a child getting 78% would get 115. However, if the mean is 75% then the standard deviation would probably be around 7-8% so a child with 75% would get a standardised score of 100 while a child getting 82% would get 115

I think that statistically you need at least 100 data points to be accurate; as long as the examing board has at least 100 children for each birth month who sat the test, then the standardised score should be very fair

[maddad]

I'm not sure I agree that this pattern can be explained. The common theme on this thread is "My DC's maths score was lower than his/her VR score". The scores people are referring to are normalised scores (quoted as out of 141 for the one I saw), so the average score for maths & VR should be the same (100). Some children will get better maths scores than VR and vice versa (and in roughly equal numbers because these are normalised scores). It makes no sense if lots of children are getting high VR scores (then they're not normalised).

Talking to a friend yesterday, they have exactly the same result: (a much higher normalised VR score: top 2% compared to maths score: top 30%).

[Twiga]

Etienne, very helpful post. Thank you.

Mattsurf, a fair point and I may not have grasped the maths behind the standardisation (mark me down for that!).

[hairydog]

Thanks Etienne. I think I would like to show ds his raw scores so will do a FOI request. Any idea where this should be sent? The school where the test was taken? I can't seem to find a central contact for the Consortium.

[Etienne]

Thanks, Twiga.

I wonder what legal basis there might be for challenging the age-related standardisation ....

I'm afraid I'm very doubtful. The Schools Adjudicator will normally only consider objections with regard to the lawfulness of admission arrangements between 15 April and 30 June in the year that the arrangements are being determined.

An appeal panel is required, as a matter of routine, to consider the lawfulness of the arrangements, but - even if they were to conclude that the arrangements are unlawful (which seems unlikely) - all they can do is refer the matter to the Local Authority, while the appeals proceed as normal.

I think I would like to show ds his raw scores so will do a FOI request. Any idea where this should be sent?

Thanks, hairydog, but take care not to confuse FOI (which is for non-personal information) and the DPA. See:
http://www.elevenplusexams.co.uk/appeals/general#a19" onclick="window.open(this.href);return false;

Sorry I don't know the address.

[Twiga]

FYI: we have now officially been told that the consortium will not remark or consider any appeals about the examination papers or marking and will not reveal papers to parents until after school allocations have been made next year.

[Parent68]

Standardisation is a mathematical action, not aimed at any individual specifically but has a collective effect on everyone. Why would this constitute a basis for appeal or legal recourse escapes me, unless a child's paper had been wrongly marked. It does appear that this is a remote possibility since almost everyone is saying the same thing, that their math marks are lower than expected. Therefore one has to deduce that this is representative of what the entire population of exam participants' experienced as well. Granted that there could be isolated cases of erroneously marked papers but that is normal in any examination and the usual recourse can be taken to manage this by affected parents.

[2Girlsmum]

At Latymer they set a hard maths paper which children scoring 73+/80 in the 1st round NVR often struggle with. DD sat it 2 years ago and the average score was 28/50. She was scoring 48-50/50 in normal 11+ papers and 44-46/50 in old Latymer and Habs Boys ones. She scored 100% in KS2 Maths, but managed 44/50 in the actual Latymer test (she though she'd scores 48/50). Maths is her best subject BTW. Her English score was shockingly bad (though she thought that she'd done well in that too). Some papers are different in style/much harder than ones practiced, but the outcome will be the same for all the students- lower scores!

[bluepeach]

Why is age standardisation deemed appropriate at 11 plus when it is not applied at KS1 or KS2 sats, when arguably the effect of age related performance is more marked.

[mitasol]

Age Standardised scores are calculated for KS2 exams. KS2 Age standardised scores for 2012