So, the annual flurry of discussion about the general and increasing uselessness of exam results as measures of [whatever it is we think we are measuring] continues. Too many people are getting A grades at A level, we need a better discriminator.
Well, far be it from me to excuse the general dropping of exam "hardness" (be quiet at the back, and don't try to tell those who sat a mixture of old and new exams (in Scotland) that standards aren't dropping), but it does seem that Universities and employers are managing to discriminate quite well, albeit to an inevitable chorus of complaint one way or another.
The BBC quotes Geoff Lucas, secretary of the Headmasters' and Headmistresses' Conference, as saying,
If no solution to the problem of discrimination at the top is found (and soon), the A-level itself and any concept of national standards will be replaced by anarchy and incoherence as universities and employers go their own way.
The implication is that we need to have a national, uniform system, based on unambiguous (ha!) numbers, through which Universities and employers can assess the quality of applicants. Well, sorry to make things difficult, but quality comes in more varieties than that which is measured, badly, by A Levels, so however good a discriminator A Levels are, they won't be anywhere near enough.
Blair with his black/white targets again, methinks. If we distill everything down to one, pointless, meaningless number, then perhaps the messy complications of the real world will go away.
Assuming we all accept an urgent need to "level the playing field", and to make "the University experience" (who said that? I can't remember, but it makes me angry every time I remember it) available to all based on ability, not background, then do we want to leave room for more discretion, or less?
On a small technical note, it is a feature of light sensing technology (the CCD sensors in digital cameras, for example, although the effect is more noticeable in satellite remote sensing) that as you increase resolution, and narrow the range of frequencies which you are detecting, you get a lower signal, and a lower signal to noise ratio. At some point you lose your signal completely in the noise. Which is why claims of military satellites which can read number plates don't excite me.
Relevance? The top 5% of an A Level now does not mean the same as the top 5% of an A Level then. When exams get easier, trying to improve discrimination by pushing pass bands up increases the noise in the signal. What you are measuring is, more than ever, the ability of the top few percent not to make the odd silly mistake in an exam.
Comments