Major questions on scores in the three test subjects

[mystery]

Can you see your standardised score for each of the three areas on your bucks score sheet - it would be a number where 100 is average (mode) and 140 top score. Edit: looks like I made a false assumption here so scenarios below need reworking.

Your final standardised score is 0.5 x verbal plus 0.3 x maths plus 0.2 x non- verbal I understand. Then this has to add to 121 for you to pass in bucks this year. Is that right?

Some possible scenarios from this:

Child very strong in maths and nvr - if you had 140 in these two you could pass with a minimum verbal standardised score of 102 - so with very average verbal. Below 102 in verbal you could not pass with 140 in nvr and maths.

( Bucks standardisation - wOuld this be very average for the population as a whole, or very average of those sitting the bucks test? I assume it is standardised on that year's cohort sitting the test - so we are talking the average of a probably above average bunch if the bucks test is now opt in- is that correct?)

Child very strong in verbal and nvr - if you had 140 in these two you could pass with a minimum standardised score of 77 for numerical - below average. Even with an above average bunch sitting the test this is probably a poor performance. The old bucks vr gl assessment test had some maths type questions in it. I wonder if you could get away with worse maths under the new test than the old one if you have great vocabulary?

Child strong in verbal and numerical - if you had 140 in these two you could pass with a nvr standardised score of 45. Very poor score.

Are people finding that their standardised score in each of the three areas are very different or similar?

Are people finding that the areas their children had lower standardised scores on are the ones they seemed concerned about after the test?

In bucks grammars on the old test did you get many children passing who did not have level 5 maths on entry to grammar?

Did the grammar heads agree this method of deciding how to determine a pass from the individual scores in bucks or was this left to c e m? Presumably the method used is highly significant in determining the ability profile of your typical bucks grammar school entrant.

In kent under the old test you had to have a minimum of around 117 in each and every one of these three areas to pass.

I find it curious that no one discusses this area of detail.

In kent probably none of the above scenario candidates would pass the old test. In bucks, you can get in if you are a duffer in maths or nvr (way below a 100) but brilliant in the others, but you can never get in if you are even a bit mediocre in the verbal reasoning - you have to have a standardised score in that of above 100 come what may.

Is the final pass / fail arrived at by a similar method in other c e m areas?


Edit: looks like my assumptions above are wrong as okanagan says below that the bucks standardisation is different and goes up to a much higher top score - see next post in this thread from okanagan

[FLG]

I have absolutely no idea what that means lol!?

[FLG]

And I went to grammar

[Okanagan]

Can you see your standardised score for each of the three areas on your bucks score sheet - it would be a number where 100 is average (mode) and 140 top score.

You can work it out. The scores given are the weighted ones -e.g. as per the red section

VR standardised score of 55.34 x 50% = 27.67
Numerical standardised score of 97.23 x 30% = 29.17
NVR standardised score of 123.8 x 20% = 24.76

But it isn't 140 top score - Bucks have used a slightly different standardisation mechanism to that used elsewhere, to maintain the "121" pass mark they used on the old papers, so you can't directly compare Bucks scores to those elsewhere (roughly Bucks scores will be 2.5 times further away from 100 than they would be on other systems). Scores ranged from 5 to 199.

Your final standardised score is 0.5 x verbal plus 0.3 x maths plus 0.2 x non- verbal I understand. Then this has to add to 121 for you to pass in bucks this year. Is that right?

Yes

(Bucks standardisation - would this be very average for the population as a whole, or very average of those sitting the bucks test? I assume it is standardised on that year's cohort sitting the test - so we are talking the average of a probably above average bunch if the bucks test is now opt in- is that correct?)

cohort standardisation. There's opt-in and opt-in - for a test taken in schools on a school day where the number of places relative to the whole population is quite high the opt-out rate is probably a lot lower than in other areas ehere opt in means a test taken out of school on a non school day.

Some possible scenarios from this:

Child very strong in maths and nvr - if you had 140 in these two you could pass with a minimum verbal standardised score of 102 - so with very average verbal. Below 102 in verbal you could not pass with 140 in nvr and maths.

Child very strong in verbal and nvr - if you had 140 in these two you could pass with a minimum standardised score of 77 for numerical - below average. Even with an above average bunch sitting the test this is probably a poor performance. The old bucks vr gl assessment test had some maths type questions in it. I wonder if you could get away with worse maths under the new test than the old one if you have great vocabulary?

Child strong in verbal and numerical - if you had 140 in these two you could pass with a nvr standardised score of 45. Very poor score.

Try reworking them on up to 199 as your 140 assumption isn't correct for Bucks.

Child very strong in maths and nvr - 199 in these two you could pass with a minimum verbal standardised score of 43 - so with very low verbal.
Child very strong in verbal and either maths or nvr - if you had 199 in these two you could pass with a minimum standardised score of 0 for the third element!

Is the final pass / fail arrived at by a similar method in other c e m areas?

Not exactly. Birmingham have 2 way standardisation - verbal (50%) and numerical/NVR 50%, then the standardised scores are aggregated. Warwickshire and Walsall do 3 way with each element standardised separatley.

[southbucks3]

Gs school subjects where vr used extensively all studied in yr 7: english lit, english lang, latin, french, spanish, german, history, re, geography, biology (chemistry and dt for write ups).

Gs school subjects where maths is used extensively: maths (further maths at a level), physics, chemistry, dt.

Maths is obviously v v important in life, but we do not hold a conversation in maths, thank goodness because i would be struck dumb.

Btw, I am an engineer!

[mystery]

Thanks okanagan. Is bucks still doing its own way of standardising? In all c e m tests does verbal make up 50 % of the final score? And do they never require a minimum in individual subjects?

[2Girlsmum]

It's interesting how different areas/schools balance the results. Latymer in North London, who introduced the test this year, separate the Verbal Reasoning and Numerical Reasoning/NVR into two results with equal importance. This is a massive shift from previous tests where a NVR test was given and the top 500+ candidates invited back to sit the Maths and English tests. It's thought that they are trying to balance the boy/girl ratio this way since statistically girls do better at English (though my two daughters are stronger in Maths and NVR!)

[Sanna]

Scores ranged from 5 to 199.

I'm sure Okanagan has this correct - but the standardised scores are only being reported as 40-180+ Rather than as 5-199. I assume the usual applies that any differences at the extreme top and bottom ends are too unreliable to split out.

[KenR]

Coming from B/Ham, I find the Bucks weighting which is biased in favour of verbal reasoning very odd.

Birmingham changed a couple of years ago to standardise equally across both papers(sum of 2 standarised scores), so there is a natural weighting depending on how many questions there are of each type. However for the first 5 years of Durham CEM they weighted the Verbal, Numerical/Maths and NVR sections equally. ie the sum of 3 equally weighted standardisations.

When my son did the B/Ham KE Grammar 11+ many years ago, he got a very high score in Maths/Numerical; a low score in verbal and a well above average score in NVR. His overall score was very good - well above the last entry score for KE 5-ways (B/Ham entry is strictly on ability not distance). He did exceptionally well at KE 5-ways and came away with excellent A-levels in the end. However, translating his old 3 standardised scores using the Bucks weighting algorithm he would not have reached the magic 121 score. I find this interesting. I can't help feeling that in using such a weighting biased in favour of verbal reasoning that they are missing out on some very talented Mathematicians and Scientists.

It may be small consolation to parents of bright mathematicians who missed the 121 target, but at least they can note that if Albert Einstein at taken the Bucks 11+ aged 11 (as a German immigrant with poor english vocabularly) then he would almost certainly have failed!

[Sally-Anne]

I'm sure Okanagan has this correct - but the standardised scores are only being reported as 40-180+ Rather than as 5-199. I assume the usual applies that any differences at the extreme top and bottom ends are too unreliable to split out.

There was a last minute change to the way the scores were reported, and they are indeed being reported as 5 - 199.

This is the message that was sent to Heads on Thursday morning by the Admissions team at BCC:

It may be a small surprise to parents when they realise the following: the results range is not limited to 40-140 as we originally intended. Instead it runs between a low score of 5 to a top score achieved of 199 (qualification mark 121, mean of 100). This decision arose due to including the composite weighted marks for the three elements ( verbal , non-verbal and numerical) in the letter as well as the STTS score and we did not want parents to worry about the maths not being correct, plus if they did the maths they could find the actual score anyway.