SW Herts standardized scores

[WP]

The consortium uses papers in Maths and VR, that each apparently come with age standardization tables based on some standard sample. So I would expect the scores to have an average of 200, or a bit higher, due to preparation and more distant children applying on academic criteria. It recently occurred to me that I could check that against the bands used by Bushey Meads School. They divide all those that take the consortium tests in a particular year into 4 equally sized bands. Thanks to the super-efficient BM admissions secretary, here are the lowest scores in bands A, B and C for the last 6 years:

Code: Select all

2005 | 2006 | 2007 | 2008 | 2009 | 2010 |
-----+------+------+------+------+------+-----------------
 204 |  212 |  209 |  213 |  215 |  213 | Band A (top 25%)
 185 |  194 |  190 |  198 |  196 |  194 | Band B (median)
 168 |  176 |  173 |  178 |  178 |  176 | Band C (top 75%)

Clearly something funny happened in 2005 (was that when they switched VR supplier?) but the long term trend is that these are lower than I'd have expected (220, 200 and 180 based on the standard normal distribution).

[AlisonR]

I wonder if the discrepency is because the bands A, B, C and D are only for the children that applied for Bushey Meads school each year (about 700 in each of the past 2 years) and not for the whole of the 2500 or so children sitting the SW herts consortium tests.

I suspect that a lot of the children getting higher scores are those only applying for the harder to get in schools which pushes the scores for BMS applicants slighty below the average scores.

Alison

[WP]

It shouldn't be. Their admissions arrangements say "top 25% taking the tests", etc.

[AlisonR]

I stand corrected, just looked back at their admissions rules and you are quite right. That completely passed me by last year when we applied.

I do like the way that BMS do their allocations, it seems a really fair way.

Alison

[WP]

Doing it the way they say they do would be sensible strategy in the presence of selective schools.

And yet your theory would partially explain the low scores. It's a puzzle.