Imagine a scenario that could easily be a question in a mathematics A-level exam: If a class of 27 pupils is projected to obtain 2.3% A* grades and 2.3% U grades, how many pupils should receive each of these grades? Show your working.
Solving this problem can be approached in a few ways. For instance, if every pupil in the class is 3.7% of it, it could be inferred that not one pupil deserve to receive either an A* or a U grade. Given that the projected percentage of achieving these grades was less than one, the class size of 27 means that there is no number of pupils less than one and the answer is, therefore, none.
Alternatively, the opposite method can be used. Since 2.3% is over "half" a pupil in that class, the logical step is to round up in this case. Thus, one A* and one U would be given.
In contrast, another option is to opt for a grading system similar to the one used by the exam regulator on Thursday. This process was heavily criticised as being unjust, but it declares that no pupil should be awarded an A*, yet one pupil should receive the U. The U stands for ‘unclassified’ and is equivalent to fail.
Under the Office of Qualifications and Examinations Regulation system, not only should a student receive a U, but they must receive it even if their teacher predicted a much better grade and even if the algorithm predicted that no more than one pupil in that school would get a U. This choice of forcing grades to be lowered across the board has generated a sense of dissatisfaction among many in England on Thursday. If Ofqual had predicted any possibility of a U grade in a class, even if it was less than one pupil getting that grade, then one pupil would be given that grade, no matter how well they did up to that point.
Moreover, this unfair action had a flip side. Even if you were a good student, you could only attain an A* if the Ofqual algorithm predicted that at least one pupil in the class would get that grade.
If a class is predicted to receive slightly less than an A* and a bit over zero U grades, the logical outcome would be that no A* would be awarded while one U would be given.
Dave Thomson, who is the chief statistician at education think tank FFT, demonstrated the issue with data from a real, anonymous school. The results from 2017 to 2019 indicate that 12.5% of students achieved an A*. None of them, however, got a U grade. Considering this, what happened to the school’s exam-free results for 2020?
This data was combined with the students’ information in a process called the "prior attainment adjustment". This is the part of the method, often referred to as "the model" or "the algorithm" by officials, that utilises information about A-level pupils, including their GCSE grades, to determine the accuracy of their predicted grades.
The adjustment process for the school analysed by Thomson predicted a 2.3% chance of a U grade in 2020, despite the fact that it had none over the last three years, and the 12.5% A* achievement from the past was reduced to a 5.71% possibility of that grade this year.
After this, the rounding process made things worse. The school was given 3.7% for each one A* and one U, which seems unfair given that the prediction model suggested that less than one pupil (2.30%, when each student counts as 3.70%) in the class would achieve this grade.
However, beyond the rounding, the adjustment process is the most fundamental aspect of understanding how this year’s grades were determined. Unfortunately, this creates more issues than it solves.
Moreover, the student who was ranked last in the class by teachers, even if they performed exceptionally well, must be given the U that was only ever predicted for half a student.
An Ofqual document shows that the process, which was developed using historical data, was best at predicting grades for history A-level. It was correct slightly more than two-thirds of the time. For the worst exam, Italian, it was correct only a quarter of the time.
Rather than focusing on the accuracy of predictions for each exam, Ofqual chose to determine their accuracy in broader terms of being within a grade. In this regard, it performed better, with 98.7% of English language grades being within a grade of the truth. However, for other exams, such as further maths, the accuracy was still wrong by more than 10% of the time, which is ironically disappointing.
Your duty is to rephrase the complete passage using improved vocabulary while making it original with a more natural tone. Each segment of your output should be in the English language. Here is the initial text to rephrase:
