11 Plus Exams Forum
http://www2.elevenplusexams.co.uk/forum/11plus/

Age Standardisation
http://www2.elevenplusexams.co.uk/forum/11plus/viewtopic.php?f=15&t=55981
Page 19 of 19

Author:  ToadMum [ Mon Jul 01, 2019 4:49 pm ]
Post subject:  Re: Age Standardisation

DSDSDD wrote:
I used the numbers and assumptions above for simplicity, but is my interpretation correct? I.e. 15 students from each monthly cohort would be offered a place if all entrants applied for the school.

If yes, then how is this method of age-standardisation justified if it is only applied intra-month, BUT not across the year as a whole (as implemented by CSSE for 2019 entry)?

For example: you could have a child born 31Jan who was ranked 16th in his/her monthly cohort (hence not offered a place), but a child born 1Feb who ranked 15th in his/her monthly cohort (offered a place) that scored ‘materially’ lower.


No, that wouldn't be the way it was worked out. If you standardise the overall cohort in e.g. 12 birth-month sub cohorts, you would still end up with a range of scores within each that generally looks quite similar, e.g.between about 200 and about 420. Then all those groups would be merged back together and you would still have a range (in that example) from c.200 to c.420. It would probably (should) work out that the proportion from each birth month 'passing' will be more or less the same as their representation in the cohort overall.

Author:  DSDSDD [ Mon Jul 01, 2019 7:14 pm ]
Post subject:  Re: Age Standardisation

If the standardisation only occurs within each monthly cohort before being remerged with the rest of the year, how does it then cater for the perceived disadvantage that August-born children have when compared with September-born?

Author:  ToadMum [ Mon Jul 01, 2019 8:41 pm ]
Post subject:  Re: Age Standardisation

DSDSDD wrote:
If the standardisation only occurs within each monthly cohort before being remerged with the rest of the year, how does it then cater for the perceived disadvantage that August-born children have when compared with September-born?


Not that this is how the CSSE didi it anyway, but...

For each grouping, you have a range of scores, calculated from the mean raw scores and standard deviations from that group. So the mean for the September birthdays for English, say, may be 32/60 and the standard deviation 8.33 and for the August birthdays, 31/60 and 10.125. For each, you will end up with a normal distribution of standardised scores. For example, you may have a score of 345 in each range; that score is still '345' and will in both cases represent a position on the curve above a certain percentage of candidates and below a certain (much smaller) percentage. And the same for all the other months. '345' as a standardised score is '345', although the underlying raw scores may be different in each of the sub-cohorts.

Page 19 of 19 All times are UTC
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/