Thread by @BarryBlackNE, The problem is that if your 'national standards' already produce results that [...]

The problem is that if your 'national standards' already produce results that are grotesquely unequal, a statistical model designed to maintain those standards will also be grotesquely unequal. #SQAResults

Unless I am reading it wrong, historical school data is the ONLY thing that has been used to adjust grades.

This was never set out as the plan.

https://www.sqa.org.uk/sqa/files_ccc/SQAAwardingMethodology2020Report.pdf

The SQA used a twenty-three year old study as its only(!?) evidence for the rank ordering system it used.

A system where teachers had to predict student performance to within two percentage points.

Standardised Assessments were - in part - used to determine students in 4th years prior performance. @IainGrayMSP @Ross_Greer

Here is the scandal.

If your school had no historical data, your teachers' estimates were just accepted. Astonishing.

@Emma_Seith - who you should be following - got there first. Sorry Emma! https://twitter.com/Emma_Seith/status/1290579228577202176?s=20

https://twitter.com/Emma_Seith/status/1290579228577202176?s=20

The SQA told @SP_EduSkills in May that they would 'undertake a moderation exercise nationally, using a range of data, discussion and review.'

But all they have used is historical attainment data.

Here it is.

Poorest pupils Higher pass rate reduced by 15.2% between teacher estimate and statistical moderation.

Richest pupils rate reduced by just 6.9%.

Poorest young people downgraded from a *pass* to a *fail* at Higher at over twice the rate of their most affluent peers.

Predominantly based on their schools' prior performance.

Source is the SQA's Equality Impact Assessment.
https://www.sqa.org.uk/sqa/files_ccc/2020-sqa-alternative-certification-model-equality-impact-assessment.pdf

https://twitter.com/mgshanks/status/1290607333136097280?s=20