A Levels have been in the news this year, due to how the government mismanaged the process after the exams were cancelled as a result of Covid-19.

At first, the grades were assigned by an algorithm, until the government backed down after a huge outcry over unfairness, and allowed the teacher predicted grades to stand. (A lot has been written about this; Hannah Fry has some good links here.)

This week’s visualization shows the grade proportions over the years:

A Level grades

(The interactive version allows you to hover to see the percentage for a grade in a given year.)

Up until 1986 grades were norm-referenced, so the top 10% were awarded an A, the next 15% a B, etc. Then from 1987 grades were awarded according to specific criteria, so the proportion of students being awarded a grade could change from year to year.

In 2002 the N grade (“nearly passed”) was retired, and in 2010 the A* grade was introduced.

The chart has two entries for 2020: one for the algorithm and one for teacher grades. This shows that while the algorithmically assigned grades may have been consistent with previous years in a statistical sense, it disguises the unfairness at an individual level.

Visualization type: stacked bar chart

Data source: Student Performance Analysis by Brian Stubbs (fixed width format, 2.2 KB)

Technical notes: generated using d3; code

See also: GCE Advanced Level (United Kingdom), Wikipedia