Tuesday, 24 August 2010

Why GCSEs are Meaningless

It has been reported today that record numbers of students have passed and achieved A* grades in their GCSEs, much to the delight of the students and their parents. However, there is a slight problem with this. The value of something - anything - is determined by supply and demand. If something is easy to obtain, i.e. the supply is high, then it is likely to be cheap.

So, if 98.7% of students leaving school have GCSEs, how much are they worth?

This question is worth one mark.

Here's the answer... Nada. Niente. Nil. Nothing.

The whole point of GCSEs, indeed any formal educational qualification, is to provide an indication of the student's aptitude towards a particular subject, and on the whole, their overall intelligence.

If you compare this to another test for measuring intelligence, say an IQ test, you will notice some frightening disparities. IQ generally follows a normal distribution pattern, as illustrated below.

Normal Distribution

Most people have around average IQ, some people have above average, some below average and a few select few have exceptionally high or exceptionally low IQs. So, you would expect any qualification system that purports to measure intelligence to be reflective of this distribution. In such a distribution, the mean value - calculated by adding up all the discrete values and then dividing by the number of values - equals the median value - calculated by finding the midway point between the highest and lowest value. And it also equals the modal value - the most commonly occurring value. If this was mapped onto the GCSE grading levels, the middle grade is a C, so most students should actually get Cs, with only a very small percentage getting an A*.

Is this reflected in the GCSE system? Quite simply, no. This year, 69.1% of the grades issued were C or above. 22.6% of grades were either A or A*, and only 1.3% actually failed to register a grade at all. On a normal distribution, the number of fails should be equivalent to the number of As and A*s. They're nowhere near.

So this leaves two possibilities:
  • The student intelligence pattern does not conform to a normal distribution - highly unlikely, given that most measurable criteria conform to a normal distribution given a wide enough sample range;
  • The GCSE system is fixed to produce better results, i.e. the examinations are made easy in order to boost pass rates, thus making our (effectively nationalised) education system look good.

Therefore, as truckloads of students leave school with GCSEs, and they do not accurately measure aptitude or intelligence, they are worthless.