A recent examination board survey found that ‘most A level predicted grades were wrong’ with ‘only’ 48% (pretty much half?) being accurate. http://www.bbc.co.uk/news/education-24625972 Delving a little deeper into the research, however, reveals that 90% were either correct or out by one grade and that forecasts of the highest A and A* grades, crucial for competitive university entry were more accurate than the averages presented above (presumably forecasts of the lower grades were less accurate). The survey found that independent schools were most likely to be accurate, perhaps because the generally smaller environments allow the teachers to know the students extremely well.

The predictions discussed here are made for the examination boards in February before the summer examinations in May and June rather than the predictions made as part of the UCAS university process. The latter are likely to be less accurate as they are made the previous September and most teachers would agree that these predictions are on the optimistic side with students considered to be on the A/B borderline likely to be ‘rounded up’ rather than down in order to encourage the universities to make offers and to support ambitious students (and their parents). No teacher will make a wildly optimistic prediction, however, as this might lead the student to accept university offers that require grades that are simply beyond their reach, leading to disappointment.

It is quite difficult to predict grades. Students mature through the A level course and in many skills-based subjects such as English or History this can make a considerable difference to the outcome. Some students are capable of making super-human efforts in revision and produce examination performances which are well above the level that would reasonably be expected from their work during lessons; equally a few students misread questions or allocate their time to different questions in an unbalanced fashion and frustratingly underperform in the examination. Any Head Teacher will also tell you that examination boards are not always accurate in their marking with some bewildering anomalies uncovered and challenged, not always successfully, each year. Mock examinations appear a good way of estimating grades but if the teachers use actual past—papers then both the papers and the mark schemes are widely available on the internet and can therefore be accessed in advance of the examination by the ambitious or merely curious student in advance of the supposed ‘mock’, inflating their scores. If the teachers use papers constructed for the purpose no examination board mark schemes are available and no published grade boundaries exist to indicate the likely grade. Predicting A* grades is particularly difficult with a wide variation in the percentage of these grades awarded in different subjects and, in some of the humanities, pedestrian and prescriptive mark schemes which reward the meticulous rather than the brilliant.

Statistics like this rekindle the debate over whether students would be better served by applying to university after they have received their results rather than before. This argument has run for the whole of my professional career with the most recent attempt to implement a post-qualification application system running into the sand in 2012 http://www.theguardian.com/higher-education-network/blog/2011/sep/21/why-post-qualification-applications-unlikely
To succeed, the timetable for A levels would need to be altered with the examinations taken in March, allowing the results to be available for university application in June; such a major upheaval for schools is unlikely to be implemented given that the content of A levels would have to be reduced to allow teaching to be completed in a shorter time. The universities also value the current, extended application period of almost a year which allows them ample time to develop contacts with their prospective students and ensure that they have the right number of students to fill their vacancies. To try to help the small number of students who over perform compared to their predictions enter the best universities the government has allowed the universities to recruit extra students who achieve at least ABB. A process known as ‘adjustment’ enables students to re-apply to different universities upon receipt of better than expected A level results in August.

The government is currently consulting on the content and structure of new A levels which will have less, if any, coursework and all of the examinations at the end of the U6 year. In most subjects these will be taught from September 2015 although the introduction of the new courses in Mathematics and Modern Languages will be delayed until September 2016. These changes are unlikely to improve the efficiency of matching student to university course. Firstly, it will take teachers several years to understand the new, and supposedly higher, standards and gain confidence in predicting A level grades. Secondly, the likely absence of an AS grade, as a staging post half way through the courses, is also likely to diminish teachers’ accuracy and reduces the information available to universities. Thirdly, the government is already committed to introducing new GCSE courses, beginning with English and Mathematics in 2015 and continuing with most other subjects in 2016. These new courses will culminate in the award of a number grade in the range 1-9 with 9 being a sort of superior A* grade and 1 a lowly G grade equivalent. Again the uncertainty over the status of these new numbers will lead universities to place more reliance on the A level forecast grades for several years to come.

Presumably, if teachers become really accurate at predicting grades they could simply grade the students themselves, rendering the whole examination industry obsolete and reducing the stress and cost to all concerned.