Is CEM testing fair?

Eleven Plus (11+) in Birmingham, Walsall, Wolverhampton and Wrekin

Moderators: Section Moderators, Forum Moderators

Post Reply
11 Plus Platform - Online Practice Makes Perfect - Try Now
MSD
Posts: 1731
Joined: Tue Dec 16, 2008 6:08 pm

Is CEM testing fair?

Post by MSD »

We know CEM applies 50% weighting to English element for Birmingham consortium. Looking at the traditional content, I would say Synonyms/Antonyms and Cloze exercises pretty much make up to 40% of that. Is CEM right in testing potential candidates on the amount of English words they are familiar with, instead of testing their innate talent and ability.

Would it not be impossible for someone emigrating from a foreign country, where no English is spoken, to pass this test? Do we really think the potential candidates need to have a good grasp of complex English vocabulary to become part of, and be successful at a grammar school.

My boys don't really read out of choice, but they had to cram thousands of words for the purpose of this test. Is that not really cheating what the test is actually testing for? I thought CEM was trying to filter out candidates that have been keen readers from a young age.
BucksBornNBred
Posts: 1031
Joined: Thu Aug 13, 2015 4:01 pm

Re: Is CEM testing fair?

Post by BucksBornNBred »

I have to agree; the weighting for each element seems strange. NVR is supposed to be the one that tests for innate ability (IQ as it were) and yet that gets the lowest percentage. NVR does not rely on previous learning or language and can identify children with a natural ability, also know as "fluid intelligence". Why CEM rates this so low is strange imo.
MSD
Posts: 1731
Joined: Tue Dec 16, 2008 6:08 pm

Re: Is CEM testing fair?

Post by MSD »

BucksBornNBred wrote:I have to agree; the weighting for each element seems strange. NVR is supposed to be the one that tests for innate ability (IQ as it were) and yet that gets the lowest percentage. NVR does not rely on previous learning or language and can identify children with a natural ability, also know as "fluid intelligence". Why CEM rates this so low is strange i
Completely agree - command on a specific language rather than natural talent must not become the qualifying criteria.

Few years back it was even weighting across all subject areas. It would be interesting to know what made them change that? I would say the decision would be down KE foundation rather than CEM itself. No offence intended, but could the reasoning be down to potential ethnic make of Birmingham grammar schools at that point in time.
Petitpois
Posts: 1440
Joined: Thu Jul 02, 2015 7:44 am

Re: Is CEM testing fair?

Post by Petitpois »

How do you apply that weighting though, if you have just estimated raw scores to go on? First question
Do the numbers of questions approximate to the weightings or even equal the weightings???

second question

Fair to who?? A number of people we know are speakers of second languages. Some of the words that came up would have bamboozled the parent never mind the 10 year old. That must be so hard for the kids.



Kind regards
Petitpois
Posts: 1440
Joined: Thu Jul 02, 2015 7:44 am

Re: Is CEM testing fair?

Post by Petitpois »

that assertion is not suggested by the proportion of NVR question posed though
kenyancowgirl
Posts: 6738
Joined: Mon Oct 21, 2013 8:59 pm

Re: Is CEM testing fair?

Post by kenyancowgirl »

Maths and NVR have been proven to be linked - so if you are strong in maths you are likely to be strong in NVR too. And, as has been said, NVR is the one are that cuts out all language bias. However, as NVR and Maths are linked, the equal third weighting that Bham/Warks used to have, was actually favouring mathematicians. Ability in English has also been proven to link strongly to ability in all other subjects (even science/maths questions require an understanding of English to understand the question) with essays etc and all GCSE exams have the warning on the front about the need to use good English.

The change was therefore to reflect the need for a good level of English as being a better indicator of overall intelligence and ability in all subjects - nothing to do with race or ethnic mix - there are threads on this every year...
MSD
Posts: 1731
Joined: Tue Dec 16, 2008 6:08 pm

Re: Is CEM testing fair?

Post by MSD »

Petitpois wrote:that assertion is not suggested by the proportion of NVR question posed though
PP - I think there is slight confusion on weighting here. :D

Irrespective of how many question they had on NVR, it only constitutes 25% of the final mark.
MSD
Posts: 1731
Joined: Tue Dec 16, 2008 6:08 pm

Re: Is CEM testing fair?

Post by MSD »

kenyancowgirl wrote:Maths and NVR have been proven to be linked - so if you are strong in maths you are likely to be strong in NVR too. And, as has been said, NVR is the one are that cuts out all language bias. However, as NVR and Maths are linked, the equal third weighting that Bham/Warks used to have, was actually favouring mathematicians. Ability in English has also been proven to link strongly to ability in all other subjects (even science/maths questions require an understanding of English to understand the question) with essays etc and all GCSE exams have the warning on the front about the need to use good English.

The change was therefore to reflect the need for a good level of English as being a better indicator of overall intelligence and ability in all subjects - nothing to do with race or ethnic mix - there are threads on this every year...
KCG - I take your point that Maths and NVR are pretty much linked, hence the reason to weigh them on par with English. But I would argue that English is something that can quite easily be learnt over few years, but not sure if it's the same with innate ability. I hardly knew any English when I landed in this country, but it only took me few years to satisfactorily start communicating with people. Having said that my grammatical skills are still quite poor and love using an exclamation everywhere :D
Petitpois
Posts: 1440
Joined: Thu Jul 02, 2015 7:44 am

Re: Is CEM testing fair?

Post by Petitpois »

Trying to compute this, because I know the numbers of questions for each area, but in generic terms

If Area A (lets say maths) has a 25% weighting but they only ask say 10% questions (ie 20 questions), as long as you get the 20 right you are guaranteed say roughly 70 Standardised marks (ie 25% of 280).

And the reverse applies if they ask 20% of questions on maths, you have to now get 40 questions right to achieve the same standardised 70 Marks.

Okidoke I will work on the basis of

140 Marks for English
70 Marks for NVR
70 Marks for Math, see how that plays out compared to the formula
Tolstoy
Posts: 2755
Joined: Wed Aug 13, 2008 5:25 pm

Re: Is CEM testing fair?

Post by Tolstoy »

I have only had one DC's results from the CEM test. He did it with one/two weeks prep in what can only be described as very difficult circumstances. For the two years prior to the test he attended a non-english speaking school and because of his dyslexia he didn't read English only the language he was being taught in. He is the most knowledgeable of my DC and perceived by all who know him as being intelligent.

He passed both the NVR and NR sections of the test whereas in the VR his score was abysmally low and this in an area that is supposed to select the top 30% academically of its cohort. I will never be convinced that the CEM is a fair test when the weightings are biased in favour of VR. I am fairly confident however that my DC will do well regardless despite having had to give up a place at a highly selective school in the country we left.
Last edited by Tolstoy on Tue Sep 22, 2015 7:43 pm, edited 1 time in total.
Post Reply
11 Plus Mocks - Practise the real exam experience - Book Now