Why is Ofqual trying to dumb down English exams?

If you wanted a good working example of the concept of dumbing down in practice, look no further than Ofqual, the exams regulatory board, the one that covered itself in ignominy when it oversaw the exam algorithm fiasco during lockdown. Its latest idea is to get exam boards for English to replace ‘complex’ language elements, such as idiom, sarcasm and metaphor, with simpler alternatives in some assessments to make the tests more accessible for pupils. The temptation at this point to respond with sarcasm, irony, idiom and metaphor – not at all nice ones either – is almost irresistible, but let’s not go there. How, in any language, but particularly

Prepare for the next A-level fiasco

When I was at school, the best grade you could hope to achieve on your termly report card was A5, with A being the highest grade for attainment, and 5 being the lowest grade for effort. I expect there will be a lot more students hoping for, and outright expecting, their own A5s this summer.  In light of the news that GCSEs and A-levels exams will be cancelled this year, Ofqual has now confirmed that grades will be decided by teachers. Schools can use mock exams, coursework and essays, or assessments set by exam boards, but these are optional, will not be taken in exam conditions, nor decide final grades. Make no mistake:

Ofqual boss’s algorithm malfunction

Gavin Williamson has taken a lot of stick for the A-level exams debacle, but Mr Steerpike thinks we should perhaps look to Roger Taylor, the chair of Ofqual, who also happens to be head of the Centre for Data Ethics and Innovation. Not many people think that using an algorithm to decide exam results was the best option, but it becomes even more questionable when you realise that Taylor led a study last year, warning of algorithms propensity to ‘make decisions which reinforce pre-existing social inequalities’. The study states: ‘concerns are growing that without proper oversight, algorithms risk entrenching and potentially worsening bias.’ Unfortunately, concerns hadn’t grown enough to prevent