Release of raw test marks allows 'tutoring to the test'!

[jamesberks]

The Information Commissioner has recently agreed with CEM’s claim to be able to provide ‘tutor proof’ tests and, to quote her directly, “disclosing raw scores will reveal too much detail about the test, and encourage tutors to attempt to calculate test difficulty in an effort to teach children to focus on achieving the required number of correct questions, rather than attempting the whole test.” (Paragraph 14, from full decision note https://ico.org.uk/media/action-weve-ta ... 661288.pdf" onclick="window.open(this.href);return false;).

I have my own views on why raw marks are withheld, which are covered in this five minute animation, https://www.youtube.com/watch?v=vgAbFZPyRJs" onclick="window.open(this.href);return false; but want to hear from tutors about what they think of the CEM's stance that releasing (anonymised) raw scores from 11+ tests could productively be used to ‘teach to the test’ through a strategy of only attempting some of the questions.

[KentMum5]

Isn't it the case that raw scores are just like :

Verbal reasoning 68, NVR 71 and Maths 53.

Would that help anyone prepare for the test? I suppose it might reveal how many questions there are. I'm not experienced enough to know if that's helpful. Medway have just switched to CEM and no one knows anything.

[streathammum]

I'm not a tutor. However, I would think an approach which relies on getting, say, 70 out of 100 answers correct and not bothering with the other 30 is not going to be very robust.

The thing about raw scores is that they are likely to change each year, depending on the cohort and the difficulty of the exam. I can't see how releasing the raw scores would be that helpful for parents, but I don't think withholding them is particularly transparent either.

[jamesberks]

... I suppose it might reveal how many questions there are ...

Yes, although the same could be achieved just by asking children who've taken the test! CEM assert that this information can be used to help with tutoring.

[BlueBerry]

I'm not a tutor. However, I would think an approach which relies on getting, say, 70 out of 100 answers correct and not bothering with the other 30 is not going to be very robust.

The thing about raw scores is that they are likely to change each year, depending on the cohort and the difficulty of the exam. I can't see how releasing the raw scores would be that helpful for parents, but I don't think withholding them is particularly transparent either.

^^ that, what Streatham mum said.

[Tinkers]

... I suppose it might reveal how many questions there are ...

Yes, although the same could be achieved just by asking children who've taken the test! CEM assert that this information can be used to help with tutoring.

Given they can vary this from year to year, I doubt it helps any way.

[Daogroupie]

HBS and previously Latymer did not release the CEM scores, CEM did.

The school withheld them but CEM were happy to sell them to any parent for £10 each.

I don't think the scores help any tutor. They only relate to that cohort and that set of questions. Nobody knows which topics are going to come up year after year so it is not like CGSE or A level where you can decide not to cover a topic and hope it does not come up.

I can't imagine any tutor would attempt to predict the questions based on the score from last year. How would knowing the score breakdown between VR English and Maths from HBS last year help you in preparation for this year? It doesn't. It may show you that more students did well in the English VR than in the Maths in that year group so perhaps you should focus on the Maths but nothing more specific than that. DG

This is not the issue, the issue is the same papers being used weeks apart by schools. So in the Herts region you have students in the same class, some of whom take the DAO CEM one week and some of whom take the same CEM paper a week apart in the SW Herts test.

That is far more of an issue than hearing about a Maths/English score breakdown from last year.