Non Qualification Selection Review adv-Inconsistent Rounding

[abzee]

Hello all

DD got 120. chances of appeal. ??


DD's
Standardised score are
Verbal 124.79
Maths 137.14
Non Verbal 87.29

Weighted score
Verbal 62.39 ( 50 %)
Maths 41.14 (30 %)
Non Verbal 17.46 (20 %)
========================
Total 120.995
========================

STTS rounded by Bucks as 120.

DD has a good academic record in school.Based on the advice on this forum we are going prepare the evidence for Selection review.

What do you think are chances of being successful when requesting a Selection Review?

[Sally-Anne]

The success rates for Reviews on scores of 120 for the last 3 years are as follows (by year of admission):

2013: 47/119 (61%)
2014: 48/52 (94%)
2015: 42/49 (85%)

I hope that helps, and my advice is that you should definitely submit a case for Review, rather than for Appeal.

Edited to add: Excellent scores in curriculum subjects ... not so hot on NVR, but that is not a curriculum subject!

[Etienne]

As you can see from Sally-Anne's figures, the percentage success rate has jumped around, but you should stand a reasonable chance.

I would wait and see what the headteacher's summary sheet is like.
It is thought up till now that review panel members like to see a summary sheet with steady progression from year 2 (level 3s) to year 6 (sound level 5s).

[abzee]

Thank you Sally-Anne , yes we will.

I was wondering if there is a way away from Selection review like recalculating the scores with the decimals and then rounding it off 120.995 to 121

[Guest55]

KS2 predictions won't be levels ... there are new tests next year.

I think review panels will have fun comparing the information they get from schools - is this the nearest score we've had to 121? How close is that to 121?!

[Marylou]

I think it's really unfair that the children themselves are supposed to follow the rules of rounding if it comes up during the test, but these same rules don't seem to apply when it comes to scoring!

[abzee]

**** EDITED ******

RE - Calculation to prove inconsistent rounding.

DD's
Standardised score are
Verbal 124.79
Maths 137.14
Non Verbal 87.29



.............Weighted..........BucksRounding.....SelfRounding
Verbal........50% of 124.79.......62.39...........62.395 ** NOT Correctly Rounded to 62.40 **
Maths.........30% of 137.14.......41.14...........41.142
NVR............20% of 87.29........17.46...........17.458 ** Correctly Rounded to 17.46 **
===================================================
TOTAL.................................120.99..........121.00 ( 62.40 + 41.14 + 17.46 )
===================================================

BUCKS STTS is 120 as per their rounding technique.

PS. : Is there a way to communicate above calculation to BUCKS without the need of
Selection Review process. ???

Regards

[Sally-Anne]

There was a similar case to this last year. The child's score was 120.98, and again, the rounding was everything. (The Review was successful.)

CEM's policy* is to round everything down, regardless of usual mathematical practice. The error with these scores is therefore the rounding up to 17.46 of the NVR score. If that had been rounded in line with CEM policy, your DD's score would actually be 120.98, rounded down to a final STTS of 120.

It's a nonsensical approach (edit by S-A: mathematically speaking), and nigh on impossible to justify to a 10/11 year old child who has been taught the correct approach, but I fear a Review is the only option.

* Bucks CC is the testing administrator on behalf of the GS Heads, who in turn commission the test from CEM. The policy on rounding is therefore not of BCC's making.

[anotherdad]

I can see why CEM choose to round down as a convention and I don't think it's nonsensical or unfair, nor do I think it's comparable to what children are expected to do in the tests themselves, they are entirely different and valid methods of rounding for entirely different purposes. Rounding down from any decimal value crops up in other walks of life, so it's not unheard of.

The CEM method gives a very clear and definite position whereby you have to get "over the line" to achieve the qualifying score, not just close to it. It's like age/birthday convention. No-one claims to be a year older when they are x years and 6 months old. Your stated, legal age increases, in completed years, when you reach your next birthday. Not the day before.

That's not what you want to hear, abzee, but it seems to be the case. As Sally-Anne has pointed out, if there is an error in CEM's rounding, it appears to be in the NVR score where a recalculation would not be in your favour. I would focus on gathering the evidence you will need for a review rather than trying to convince Bucks/CEM that their method of calculation is incorrect. There's loads of help on here and statistically, a score of 120 stands a good chance of success at review if the academic evidence is there to support it.

[Sally-Anne]

anotherdad, I've made a slight edit to my post to reflect the fact that I agree with your statement that it is a nonsensical approach mathematically speaking.

As you say, a line has to be drawn somewhere and that is at 121.00, no matter how painful it is to have to wait until February in the hope of retrieving 0.01 (or 0.02) of a mark.