Page 1 of 2

Top aggregated marks for Tiffin Boys

Posted: Wed Mar 12, 2008 10:23 am
by STRESSED OUT
Hi,

Does anyone know what the top score was for tiffin boys for the sitting they just did? 8)

I have been checking the threads and i have not come across any links so far?

Not sure how they work it all out?

Hope someone will able to enlighten me?

Thanks
8)

Posted: Wed Mar 12, 2008 11:11 am
by allegra
The top score was 280 (140 for VR and 140 for NVR).

Posted: Wed Mar 12, 2008 11:21 am
by STRESSED OUT
allegra wrote:The top score was 280 (140 for VR and 140 for NVR).
HI Allegra,

Thanks for the information. It does show it is not very easy to get into any grammar schools.
This year the pass mark was higher than it has been over the previous two years.

Good luck to all who are going through the 11+.

Posted: Wed Mar 12, 2008 11:29 am
by allegra
No it isn't easy. I'm not sure that the slight variation in the pass mark is all that significant - it might just mean the papers are slightly easier / harder some years - although the amount of tutoring that goes on seems to have increased considerably in recent years, whether DIY at home or with a tutor, so maybe everyone's getting better at the tests!

Posted: Wed Mar 12, 2008 10:07 pm
by SunlampVexesEel
The scores are standardised so that within a month bucket of peers the top scorer will get 140 and the bottom will get 70. If a candidate was the best at both NVR and VR they would score 280.

And for those that go on about "summer babies".... this is irrespective of age. It is possible to score 280 regardless of date of birth. The "summer baby" advantage is an 11+ myth.

100 is average in the bucket with a standard deviation of 15. So if you whip out your normal distribution I think it means that if you score >135 ish then you are in the top 5% ish of the tested candidates (again IRRESPECTIVE of age).

The pass mark is set in order to make the same number of offers as there are places.

The NVR is rock hard; so a good score will need good intelligence, a good undertanding of the question types (via preparation), good technique and an element of luck. It is difficult to cover all the questions in the alloted time.

Also... remember the average 100 is based on the group being tested. It is a self-selecting group comprising many smart and extremely well prepared candidates.

Regards
SVE

Posted: Thu Mar 13, 2008 10:30 am
by STRESSED OUT
SunlampVexesEel wrote:The scores are standardised so that within a month bucket of peers the top scorer will get 140 and the bottom will get 70. If a candidate was the best at both NVR and VR they would score 280.

And for those that go on about "summer babies".... this is irrespective of age. It is possible to score 280 regardless of date of birth. The "summer baby" advantage is an 11+ myth.

100 is average in the bucket with a standard deviation of 15. So if you whip out your normal distribution I think it means that if you score >135 ish then you are in the top 5% ish of the tested candidates (again IRRESPECTIVE of age).

The pass mark is set in order to make the same number of offers as there are places.

The NVR is rock hard; so a good score will need good intelligence, a good undertanding of the question types (via preparation), good technique and an element of luck. It is difficult to cover all the questions in the alloted time.

Also... remember the average 100 is based on the group being tested. It is a self-selecting group comprising many smart and extremely well prepared candidates.

Regards

SVE

Hi Sve,

Thanks for the infromation it makes it all crystal clear now.

Yes NVRis a subject the child need to be rock hard to understand it and needs to do lots of practice to stand any chance of passing these exams, as they have to answer the questions within 30 sec, so they have to be well prepard either at home or outsidefor these exams.

More and more candidates are sitting for these exams not sure what the ratio is for getting a place in these schools 1 to 10?


:lol:

Posted: Sat Mar 15, 2008 12:31 am
by xe
This year Tiffin offered 146 places in the first round and there were 1170 applications, so it is 1 to 8 now. Hope the ratio will go down when more kids on the waiting list get the offer before Sept.30th.

Posted: Sat Mar 15, 2008 12:49 am
by Bexley Mum 2
Dear Sunlamp, I've been trying - and failing - lately to get to grips with standardisation. I think the penny almost dropped when I read your very clear explanation, but I don't know what you mean by "a standard deviation of 15". I've seen that phrase used by others, but I don't know what it means. Could you explain it for me - in simple terms!!!!

Many thanks

RE: Standardisation

Posted: Mon Mar 17, 2008 11:32 am
by oesmugso
I must say I didn't find the post particularly helpful to understand the whole business of what goes on and why. So I spent a few hours yesterday re-learning some basics and digging into the process. I located the source of the main meat of the post and in trying to understand it, I wrote myself 2-3 pages of verbose detail on it - I find it helps me to understand the issues better. Maybe I might end up posting it, much in the guise of a consultative offering, if anyone wants to help knock it into shape - our own Dumbed Down Score Standardisation. And of course, we all make mistakes and I'm sure I have. More on that later perhaps. To begin with I need to test my ground:

SD = standard deviation

SVE states something which I interpret as saying that there is no adjustment for age in the standardisation process and it's a myth. So let's assume an example where in the oldest age group the best raw marks of 80% (1st) and 75% (2nd) are found AND in the youngest age group the top raw mark is 75%. If we assume that the actual SD was exactly the Utopian one of 15 - perfectly plausible - we only have a shift of +60 to arrive at the standardised mark. So the 75 maps to 135 in this group whilst it maps to 140 in the youngest group, a very clear advantage. The effect of the standard deviation is such that if it's less than 15 the difference in the marks of 80 & 75, namely 5 here, is stretched out (increased by 15/SD) & if the actual SD is more then difference is reduced. That depends on the actual data within each 'bucket'.

What am I missing ? Have I misunderstood something ? Not diffcult to, I'll say!

RE: Standardisation

Posted: Mon Mar 17, 2008 3:02 pm
by oesmugso
Correcting my mistake, it should read:

The effect of the standard deviation is such that if it's MORE than 15, the difference in the marks of 80 & 75, namely 5 here, is stretched out (increased by 15/SD) & if the actual SD is LESS then difference is reduced.

Eg. sd = 10 ==> the difference of 5 would map to a difference of 7.5
sd = 20 ==> the difference of 5 would map to a difference of 3.75

I think it is now correct - hopefully.