New research in the education sector, utilising data from one of the country’s largest examination boards, suggests that more than half of predicted A-level grades are incorrect; pupils in comprehensive schools are the worst affected by this issue. The research suggested that only 48% of the estimated grades supplied to pupils by teachers earlier in the year were correct following exams; with almost a tenth of predictions being out by more than one grade – i.e. a student predicted a grade B may actually have achieved a grade D. Teachers tended to over-predict the ability of their students – suggesting they would achieve higher grades than were eventually awarded.

The implications of this research are huge in terms of the university admissions process following the announcement by the government that AS levels are to be done away with and replaced by a single qualification at the end of two years.

Currently, admissions to higher education are made on the basis of a student’s GCSEs, AS-level results and teachers’ predictions about their likely attainment in the final year of Sixth Form – i.e. their A2 levels.

A-Level Predictions Often Incorrect; May Impact University Admissions Systems
[Image: Tulane Public Relations via Wikimedia Commons]

However, from 2016, the government has announced plans to scrap examinations in the first year of sixth form and move all exams to the final year, suggesting that Universities will be forced to place much more reliance on GCSE results and teachers predictions. It may also mean that more universities start to issue their own entrance exams, although this would likely create logistical and timing issues.

The research suggests that teachers are much better at predicting grades at the higher end of the spectrum – with 64.4% of A* grades and 62.9% of A grade predictions being accurate, compared to only 32% of D grades and just over a quarter (26.9%) of E grade predictions. This data is useful in explaining why comprehensive schools, which often cater for a wider range of abilities than private schools, appear to be less effective at predicting grades overall than those not in the state sector. The rate of successful predictions dropped to 45.5% in some state comprehensives, and as low as 43% at further education colleges.

The study was carried out using Cambridge Assessment data gathered from 190,000 examinations taken through the OCR exam board.