Age Standardisation

Eleven Plus (11+) in Essex

Moderators: Section Moderators, Forum Moderators

11 Plus Mocks - Practise the real exam experience - Book Now
hessian77
Posts: 11
Joined: Thu Oct 18, 2018 1:53 pm

Re: Age Standardisation

Post by hessian77 »

There were just 72 results in that spreadsheet vs a total of 5600 for God's sake! What does this tell us statistically? Nothing. 0. Just made us congratulating a few beautiful outliers that in most cases are free of any concern for any school.
hessian77
Posts: 11
Joined: Thu Oct 18, 2018 1:53 pm

Re: Age Standardisation

Post by hessian77 »

Wontsleeptonight, are you a statistician?
Wontsleeptonight
Posts: 42
Joined: Mon Oct 15, 2018 5:28 pm

Re: Age Standardisation

Post by Wontsleeptonight »

For the sake of mine and my child's anonimity I'd rather not divulge my profession on a public forum.
Wontsleeptonight
Posts: 42
Joined: Mon Oct 15, 2018 5:28 pm

Re: Age Standardisation

Post by Wontsleeptonight »

hessian77 wrote:Wontsleeptonight, are you a statistician?

Why? Are you recruiting? ;)
hessian77
Posts: 11
Joined: Thu Oct 18, 2018 1:53 pm

Re: Age Standardisation

Post by hessian77 »

Because Maths was very difficult this year, most people have raw scores closed to the mean - e.g. there is a high density of values or a 'spike' around the mean. Which will contribute nothing to their standardised values as those will be 100 since actual value - mean will be 0 or will tend towards 0.

The SD do not describe the density of distribution, but just their dispersion, e.g. how far are values from the mean.

The whole very long discussion is because we don't have descriptive and distribution data for the standardised overall data - CSSE just offered raw data on both Maths and English, which are not very helpful in this case.
Wontsleeptonight
Posts: 42
Joined: Mon Oct 15, 2018 5:28 pm

Re: Age Standardisation

Post by Wontsleeptonight »

hessian77 wrote:Because Maths was very difficult this year, most people have raw scores closed to the mean - e.g. there is a high density of values or a 'spike' around the mean. Which will contribute nothing to their standardised values as those will be 100 since actual value - mean will be 0 or will tend towards 0.

The SD do not describe the density of distribution, but just their dispersion, e.g. how far are values from the mean.

The whole very long discussion is because we don't have descriptive and distribution data for the standardised overall data - CSSE just offered raw data on both Maths and English, which are not very helpful in this case.

Groan. The process of standardisation forces the data into a normal curve, with mean 100 and SD 15.
The curves will look the same every single time.

Except this year.

And if you look again, on the csse website, they have given the information needed to fully geek out over the standardised results.
Last edited by Wontsleeptonight on Thu Oct 18, 2018 4:33 pm, edited 1 time in total.
hessian77
Posts: 11
Joined: Thu Oct 18, 2018 1:53 pm

Re: Age Standardisation

Post by hessian77 »

Thanks. No. I am working with stats data in my job.

Anyway, thanks for contacting the CSSE statistician. Perhaps his/her answer will en-light us.
hessian77
Posts: 11
Joined: Thu Oct 18, 2018 1:53 pm

Re: Age Standardisation

Post by hessian77 »

The standardisation you keep mention is intended to normalise the raw distribution overall but not to give us identical shapes of 'bell 'curves and distribution parameters across years. The differential difficulty of exams between years is important too on the final thresholds. It isn't that simple with age-standardisation: not everything should be re-thought as adding 5.3 to the last years threshold and having this year's thresholds.

Wait for the CSSE reply on this as I said.

What I try to suggest is that people shouldn't be stopping their celebration at scores of 356 because thresholds will be increased by ...? (5, 6 or more points) and everything we received as guidance is unreliable.
Wontsleeptonight
Posts: 42
Joined: Mon Oct 15, 2018 5:28 pm

Re: Age Standardisation

Post by Wontsleeptonight »

The csse fit the distribution to SD of 15 each year.
This year they have not

The mean is 100 each year.
This year it is not


Do we agree this far?


And absolutely everyone should celebrate the results, just to be aware that a huge change has happened, that not to close any door, keep options open and take a cautious approach to completing the admission form. I just hope CSSE respond in time.
test7
Posts: 1
Joined: Thu Oct 18, 2018 4:47 pm

Re: Age Standardisation

Post by test7 »

My DC scores
M- 48
E-48
SC - 369.56
Born month-- Sep

from the https://docs.google.com/spreadsheets/d/ ... edit#gid=0" onclick="window.open(this.href);return false;
I saw one score
M-47
E-48
SC -- 376.82
Born Month July

:roll: still cannot get my head around age standardisation
Post Reply