Shades of Spinal Tap in A-level grades: This one goes to A*

It has been about 2 weeks since students in the UK received their A-level results and there is an unlimited number of university places for students with ABB grades. But what do those grades mean? A lot depends on them.

Consider two presentations: trends in SAT scores (US system) and trends in A-level grades (UK system). Set aside any philosophical problems you might have with standardized tests; there are many reasons not to like them – SAT score is horribly dependent on ethnicity, race, and socio-economic status  and A-levels have gender issues (scroll to the bottom to the nicely presented spreadsheet) and type of school matters (if you can afford to pay your results will be better). There is much to dislike, but still...

...there is an interesting narrative embedded in the UK data courtesy of the BBC. There is a graphic telling us that from 1960 to about 1980, the proportion of A grades was flat at around 8-9%. Then things took off and by about 2008 over 25% of students were receiving A grades. This was not as helpful as it might have been, particularly in sifting out the best students, so... The A* was born. As of 2013, roughly 7% of students received A* grades. For more details about A-level grades see here. 

Contemplating this graphic and listening to the absurd commentary about the return to “gold standard” A-levels (c.f.: here) I am reminded of the famous scene in the film Spinal Tap where the musician is particularly proud of his guitar which has a volume control knob that goes to 11 (instead of 10). This is what the introduction of the A* grade is all about and my argument is the same as the interviewer. Why not have the same range covered by A-E, to which the response would be “but this one goes to A*”. SATs have shown no such “inflation.”

Let’s be charitable about what happened to A-level grades between 1985 and the present. Let’s pretend there was no grade inflation (just like we can pretend there has been no grade inflation in University degree class) and that the whole problem is due to us getting better and better at teaching and a greater and greater number of people able to teach the subject competently has been created so achievement overall is much better than it was before. Yeah right, but suspend disbelief for a moment and repeat after me “there is no grade inflation, there is no grade inflation...”

However, if eventually all students get 100% , it means they are all average. This is unhelpful.

What people want to know is how good a particular student is and how well prepared they are. This allows them to be tracked into the best place to serve their needs. When working with very large cohorts of students, the “curve” is probably the best way to go. No one really cares that a student only got 30% of the total possible IF they are the best in their cohort. 

In the US, many institutions indicate average SAT scores, but these are flexible and the more important metric from an SAT is the ranking by percentile. Getting a 600 on a particular SAT section is less meaningful than being in the 90^th percentile of students taking the test. The top 10% will the top 10% no matter what you call the grade (A, A*, or A**) and 25% of students cannot be in the top 10%. It cannot be done.

The lesson here is that countries investing in education want to evaluate achievement. If we pretend that grade inflation does not exist, long term trends toward increased grades can be likened to a car with a speedometer that only goes to 50 MPH. Once upon a time that would have been pretty fast, but it isn’t any longer. The measurement should change with the cohort. The SAT has done a much better job over time than the A-level system as long as we recognise its limitations with regard to ethinicity, gender, and socioeconomic background.