What does the score 121 mean?

[doting_mum]

Does anyone know how the scores are calculated from the raw score? What does 121 mean in percentage? Is 121 just a cut-off mark??

[Sanna]

The raw scores go through a statistical process called standardisation which starts by looking at the average of the raw scores for this exact test. Okanagan posted a nice explanation in the Warwickshire forum here.

http://www.elevenplusexams.co.uk/forum/ ... 02#p329502" onclick="window.open(this.href);return false;

The standardisation is done by age, so older children compared against older children, younger against younger to make it fairer.

At the end of the day, all the raw scores will be translated into standardisation scores, one for verbal, one for non-verbal and one for maths. These three scores are then combined by "weighting" so that verbal will contribute to 50% of the final standardised score, non-verbal 20% and maths 30% giving an overall final STTS (standardised transfer test score). This is the key number.

This year the STTS scores for Bucks will be more spread out than in previous years, they will range from 40 to 180+, and in theory the average is 100 (though it's not always been in Bucks and figures like 110 have been mentioned). An overall combined STTS score of 121 or more will mean the child does qualify for Grammar schools. It doesn't matter if it's 121 or 180 both qualify equally.

This year hopefully the spreading out of scores will work so we should expect less people to achieve 120 and 119 etc making it easier for selection reviews and appeals to differentiate between the near misses.

[scary mum]

110 have been mentioned

I think that's for CATs - for some reason the schools often say that the average score for CATs tend to be higher in Bucks (something in the water ?). I don't understand standardisation very well, but if the cohort is just Bucks, I think the average has to be 100. Or maybe not?

[Okanagan]

The average in theory should be 100. But because the old Bucks test was so heavily tutored for, but the standardisation based on a whole population sample without tutoring, then the average score actually achieved was typically higher, especially as some who might have expected lower scores may have opted out in advance.

With CEM in other areas they have always standardised on the cohort who actually take the test - so the average will be 100, and 121 ought to only be achieved by about 9%. However Bucks want to select around 30%, meaning you'd usually be looking at a standardised score of around 108. We already know that Bucks have decided to retain 121 as the headline figure for a pass, presumably because they know that if they do anything else some people will misinterpret it changing it as making the test easier/harder to pass.

BUT ... before you all panic about this, the number of Bucks school places hasn't changed - so they still want to select around 30%, meaning that they need the same number to "pass". Therefore I suspect that there will be some adjustment of the scores to stretch the range out so that they get a suitable number of "121+" scores.

For those interested in the stats, they'd perhaps use a factor different to the educational norm of 15 to standardise to.

The standardised score (ignoring for now the age adjustment) is
S = x(b — a)/sd + 100
where S is the pupil’s standardised score, b is the pupil’s raw score, a is the average raw score of all the pupils, sd is the standard deviation of the raw scores, and x is usually 15. If they needed to get more high scores they could just change the value of x, and hey presto you'd have enough 121+ scores to be consistent with previous years.

[Sanna]

110 have been mentioned

I think that's for CATs - for some reason the schools often say that the average score for CATs tend to be higher in Bucks (something in the water ?). I don't understand standardisation very well, but if the cohort is just Bucks, I think the average has to be 100. Or maybe not?

Ah that makes more sense. Yes if children are given a CATS score on a national benchmark then the average for any specific county could be different. I'm not sure if it's the water or the house prices in Bucks !!!

However just to get a bit more technical: surely the must be some "adjusting" done in the standardising process in order to guarantee that the 121 qualifying score delivers enough qualifying children to fill the Grammar schools? In most educational tests the average is 100 and one standard deviation is set at +- 15 which would mean that only 16% of people taking the test would have a score of 115 or above, and far, far fewer with 121. So either the standard deviation is set differently or the average is adjusted or perhaps the number of out-of-county people taking the test is so huge that a 16% pass rate delivers close enough to the supposedly 30% of Bucks children who are meant to go to Grammar School? Though I've never see any real life normal Bucks State school evidence of the 30% unfortunately....

[Sanna]

Just see my message crossed with Okanagan's - as usual very well explained! Yes, I was assuming they would have to adjust the standardisation factor or the average to make it work.

[anotherdad]

surely the must be some "adjusting" done in the standardising process in order to guarantee that the 121 qualifying score delivers enough qualifying children to fill the Grammar schools?

In recent years Bucks grammar schools began their year 7 intakes with many places unfilled. In September 2011 for example, I knew of 72 unfilled year 7 places across just two of the grammar schools. A lot of those places were subsequently filled in year 8 by successful 12+ sitters in September 2012.

[Kiwimum]

My dd had a lovely small class of around 25 in year 7 last year and all the year 7 classes at her school had spaces. This year, according to dd, all those classes have 32 students except for hers which has 33 - although apparently most classes only have 32 seats! (I don't know how that works - perhaps when you have 33 pupils there's a good chance at least one of them will be away!)

[guest666]

In certain areas, e.g South Bucks, there are many children who take two tests Bucks & Slough, which often produced different outcomes for some pupils.

But the brighter do tend to pass both tests. Trying to adjust the mark so 30% got 121 could see lots of unfilled places at certain schools when those who passed opted for an OOC school.

[Tinkus-Pinkus]

There are also lots of people who put their children in for the test as a practice run for other areas. I met a number of these at the Central location. There is a boy in my DD's class (which is in a neighbouring county), who has no intention of taking up the place should he pass, but is going to an indie. His parents just felt it gave them other options, if needed.