07 August 2015
Research by exam group Cambridge Assessment shows how hard it is to predict how students will do in exams, with teachers getting more than one in two forecast grades wrong.
Figures from Cambridge Assessment’s UK exam board OCR show that in 2014 43% of A Level forecast grades were correct, compared to 48% in 2012*. Researchers Tim Gill and Dr Tom Benton say that one explanation for the fall in accuracy might be reforms which have led to the removal of most January exams, as teachers often used these to help their predictions.
The researchers point out that in the vast majority of cases, the forecast grade was either correct or within one grade. In 2014, around 88% were correct or within one grade, while nearly 12% were more than one grade out. In 2012, nearly 92% were correct or within one grade, while around 8% were more than one grade out.
Teachers are more likely to be over-optimistic in their forecasts (43%) rather than pessimistic (nearly 14%).
The data is broken down by school type and shows that independent and grammar schools were the most accurate at forecasting A Level grades. The researchers say this might be because students at these schools generally achieve higher grades, and higher grades are easier to predict.
The figures are not the predicted grades sent to UCAS in January each year, but are instead forecast grades sent to OCR in May. They are part of a series of statistics that OCR used to help set grade boundaries.
Sylke Scheiner, OCR’s Director of Assessment Standards, said the figures showed how important it is for teachers to be involved in examining and develop their assessment skills.
“This underlines the need for more teachers to become examiners so they can really understand the nuts and bolts of how their students can perform,” she said.
The research is published alongside reports on forecast grades for OCR GCSEs in 2013 and 2014, which show a fall in accuracy from nearly 47% to nearly 44%.
Statistical reports - The accuracy of forecast grades for OCR GCSEs and A Levels