Your browser does not support Javascript

New data reveals that 84% of A level predicted grades are incorrect

8th December 2016 1:00
By Blue Tutors

Analysis of the past three year’s data from UCAS, using the results of 1.3 million A level students who applied for university courses, has shown that only one in six students’ predicted marks are accurate. The research, commissioned by the University and College Union (UCU), showed that three quarters of the predicted marks were overly optimistic, forecasting a higher mark than was achieved by the student, whereas only 9% did better than expected.

Predicted marks are assigned by teachers at the beginning of A level year (after AS), and submitted in order to be used for university admissions and in case of disruption during the examination period. The results are not disclosed to students and are meant to be an honest reflection of their work so far and potential in the upcoming exams. While there will always be some error in predicting this far in advance, the staggering skew towards overprediction has led to speculation that much of the overprediction is intentional, done to increase the likelihood of students gaining a good university offer. Some teachers have admitted that they do tend to overpredict in the favour of the student as predicted marks are arguably becoming more important than the actual results. Since the cap on numbers was lifted, in the current climate where universities are actually competing to attract students, "the number who are being accepted with quite significant discounts on their offers and their predicted grades has grown quite a lot." This means increasing numbers of students secure offers on inflated grades that they are then not obliged to live up to in final examinations.

Looking more closely at the data reveals an alarming trend. Despite the overall tendency to be too optimistic about results, students from lower income backgrounds are 20% more likely to be under-underestimated, with almost one in four (24%) achieving a higher result than predicted in any given subject, as opposed to 20% of peers from wealthier families. The problem is particularly apparent in higher scoring students, achieving AAB or above, where it arguably makes the most difference in highly competitive admissions.

The research is a fundamental blow to the reliability of the current university admissions process, administered through UCAS, which relies heavily on predicted grades of half a million UK students a year when assigning offers. The UCU is demanding an overhaul of the process, saying it is time for a system which allows students to apply with real results and not “poor guestimates”. Post-qualification admissions (PQA) would eliminate uncertainty for students and universities alike; with the current conditional offer system, students are left in limbo until results day, or longer for those who miss their offers and end up in clearing, and universities have less control over student numbers, having to gamble on how many offers to make to fill places but not end up over capacity. Despite a survey which suggested 70% of university staff are in favour of PQA, UCAS argues that this would leave insufficient time to give students the proper consideration, and for successful applicants to look into accommodation and finance in time for starting their course.