Just wondering how the standardised scores are worked out.

Is it when you are born or is it how old you are on the day you take the test.

I always thought it was when your birthday was and with so little children who normally have to take the test at a later date it wouldn't matter.

However this year there is so many children who have had to take the test a month or so later, it really does matter.

if its when you are born it wouldn't be fare as they would be a month or so older which would make the standardisation flawed.

So is it how old they are on the day or when they are born ?

I believe it is age on the test date.

Yes, standardised scores e.g. for reading tests etc are worked out using age in years and completed months on the day the test was taken.

If you were born on 26th Feb 2010 & were tested on 25th Feb 2020 you would be 9yrs 11 months. On 26th Feb 2020, 10 yrs 0 months, 25th March 2020 10yrs 0 months and so on.

Last year in Kent we reckoned it's not completed months for 11+ but month of birth, eg Sep or Oct. My late September born son had the identical raw score on a paper to another poster's son who was born in October, but my son's standardised score was one point lower. They were both 10 y 11 mo at the time of sitting the test.

When you think about it, that's the most sensible way of doing it. If you go by completed months, and the test is say 15 Sep, then you have 11 full month cohorts, but half a cohort aged 11 y 0 mo and another half cohort who are 10 y 0 mo. Makes standardisation less accurate for those oldest and youngest groups.

If its done by the month they are born how woukd that work this year

So if there was 2 kids were born november 20th
One of those kids was able to sit the test 15th october but the other was isolating due to covid and couldn't take the test till november 15th so a whole month later.

How would they compare them and their scores - as they are born the same day but 1 was a month older when they took the test.

Fair point - I really don't know. If there are raw scores threads this year, they might give pointers? But being one month older probably isn't going to give much of an advantage if you're sitting late because of self-isolation disruption and perhaps illness.

Makes sense, I have no idea how Kent test is calculated tbh. Just going from how most norm-referenced tests e.g. Salford Reading Test are done.

Yes, think it will just be due to the particular circs of 11 plus. With most age standardised things you're not looking to distinguish between one clearly identified year cohort.

For CEM exams it's age of the (in years and month blocks/segments) on the date they sat the exam. The segment granularity is typically 1 month.

Interesting - do they count the 11 y 0 month half cohort in with the 10 y 11 mo, and 10 y 0 mo in with 10 y 1 mo, to avoid the situation where you're standardising the results of a much smaller group? With a test in early Sep it could be a very small group of just 11 year olds.