Something really really bad has happened All that

Something really, really bad has happened.

All that is necessary for the triumph of evil is that good men do nothing. Edmund Burke What I should have done is shout, “The Emperor is start bollock naked!” But, of course it would have made no difference at all (and in fact, I did shout quite loudly!)

The background The previous GCSE was 40% course work (20% written, 20% Speaking and Listening) People thought, rightly, that this 40% was corrupt. So. . .

New spec. • 40% exam • 60% ‘controlled assessment’ – tasks set by exam boards – tasks done under ‘controlled conditions’ (quasi-exam conditions) in schools, marked by teachers and moderated by boards • at least 40% of marks must be banked at the end of the course (i. e. , this year, in June 2012) • other ‘banking’ opportunities in June 2011 and Jan 2012

e. g. AQA English Language • One exam proper (could be taken in June ’ 11, Jan ’ 12 or June ’ 12) • Speaking and listening ‘controlled assessment’ (marks could have been banked in June ’ 11, Jan ’ 12 or June ’ 12) • Three written controlled assessment units (marks could have been banked in June ’ 11, Jan ’ 12 or June ’ 12)

So. . . Candidates and schools might well have 60% of the qualification done, dusted and in the bank before the Y 11 June exams. And. . . Resits Any unit (C/A or exam) could be re-taken once The better mark counts (but the 40% terminal rule applies)

And moreover. . . • All four exam boards sold their new specifications as being more likely to result in better outcomes for candidates (by, for example) encouraging teachers to ‘chunk-up’ the controlled assessment tasks and teach to each chunk immediately before the candidates tackle each chunk. Candidates have never been better ‘prepared’ for their GCSE.

Pause for questions • new specification • modular • resit opportunities • controlled assessment with lots of teacher input • resulted in higher performance • schools and candidates had ‘banked’ higher performance before June ’ 12

So what happened? • grade boundaries moved substantially upwards between January and June 2012 (i. e. harder to get higher grades) • for one particular AQA unit, percentages of C+: June 2011 26. 7% Jan 2012 37. 0% June 2012 10. 2% • England Wales overall C+ in English / English Language fell by 1. 5% • some (good and outstanding) schools’ results ‘fell off a cliff’ (down 20%+). Some in Norfolk • ASCL say some 143 secondaries pushed below floor targets (40% 5+ C+ inc. En and Ma) by the change. Some in Norfolk.

How did this happen? • GCEs were norm-referenced (i. e. a fixed %age of the entry achieved each grade. ‘Grade inflation’ impossible • Sir Keith Joseph’s GCSEs introduced as ‘criteria -referenced’ (i. e. you achieve a given grade if you can do what the exam specification says is required to achieve that grade. ) Improved performance is reflected in improved exam statistics – ‘grade inflation’.

How did this happen? cont. • Ofqual took the criteria-referenced GCSE and applied a ‘comparable outcomes’ rule which moves the exam back towards normreferencing • comparable outcomes uses KS 2 results to ‘predict’ what the entry cohort ‘should’ achieve and adjusts grade boundaries to bring about that result

How did this happen? cont. 2 • Ofqual say they didn’t have the KS 2 data to hand in June 2011 or Jan 2012 and that the number of entries (14 K +!) was too small a sample for them to realise that ‘too many’ candidates were going to get higher grades until about two weeks before the results were to be published • they ordered boards to raise thresholds in order to ‘claw back’ the overall comparable outcome • tens of thousands (50 K+? ) candidates downgraded

Ofqual says it did nothing wrong • January grades were ‘too generous’ • they couldn’t have done anything at that time • overall results were right (so logically June ’ 12 must be too harsh!) • it’s a problem with modular exams with lots of controlled assessment • it’s a problem with English • teachers overestimated candidates’ performance (but no board informed any schools that this was happening at the time!)

And. . . • there has been no explanation of why the picture of results is so massively variable between schools. (‘We’re trying to understand that. ’) • current Y 11 following an identical assessment process • current Y 10 following a ‘linear’ process – no resits

TES leading article today. . . the August debacle. Exam boards were cowed or coopted into stringent grading of June's English papers to compensate for Ofqual's neglect. Thousands of pupils were denied predicted grades while schools crashed through floor targets. When pressed by MPs this week on why her watchdog failed to bark earlier, the chief regulator claimed it lacked sufficient data (see pages 8 -9). Why were forecasts so wide of the mark? Well, schools are prone to exaggeration. Why were they not informed? Er, it was all terribly difficult. Why were January's marks so generous? Good Lord, is that the time?

Glenys Stacey at the Select Committee “There are several unknowns. There are more unknowns than there is when we’ve got a stable qualification, or than there at the end of a qualification. So the unknowns are these, for example: the strength of the correlation between the units, which strongly influences, actually, how unit outcomes aggregate to these subject outcomes…”