Inspired by a conversation on OUT, I got carried away as usual...

First off: as all the teachers seem to be saying, there's a decline in standards, and I can't disagree. Having two older brothers, I did my A-levels with 8 years' worth of past papers and the decline was clear and visible (we all did physics, maths and chemistry). Chatting recently to someone who got his results today, he mentioned that his physics paper was "dumbed down" significantly from last year's, and in fact he was annoyed at how much material he'd learnt that hadn't been covered by the exam at all. I think the evidence that standards are declining is unassailable. The real question is, what's causing the decline?

The problem is grade inflation. Teachers and exam boards are *both* under constant pressure to produce better results. Students meanwhile only have a certain level of natural ability. Due to improved diet and medicical technology, their natural intelligence may be increasing slowly -- IQs are rising -- but slowly, not nearly fast enough to make a noticeable difference to exam results. And therein lies a vicious cycle, illustrated below:

1997, January. White Town is making its one and only appearance on the charts, and you're an examiner.
When setting an exam, examiners have to decide how difficult to make that exam. They have to choose questions such that a certain percentage of people will be able to pass that exam -- the *examiners* decide what percentage of people, roughly, are going to pass that exam. People don't think about it that way: the prevailing mindset seems to be that exam papers are magically set to the same difficulty every year. But how do you gauge how "difficult" a paper is? It's all relative to how much the students know. So, as an examiner under pressure to improve results this year, you will tweak your questions a bit, and leave off a little material. No harm done, right?

1997, July. Oasis are at number 1, and you're a teacher.
You saw last year's exam. It had a little less material on it. You are under pressure to produce the best results possible, so you know that covering material that isn't going to be examined, while worthy, is useless to you. So, reasonably enough, you didn't teach it, and instead taught a slightly smaller set of material, covering it more thoroughly.

But here's the problem: it doesn't matter how much you teach your students, or even how you teach it. No teacher is perfect, and no students are perfect either. Your students will always understand less than 100% of the material you have taught them, and they will go into that exam understanding less than all of it.

1998, January. All Saints are wailing on the radio, and you're Mr. Examiner again.
You're looking at last year's exam results, and despite the fact that last year's paper had less material on it, mysteriously, the students seemed to get roughly the same distribution of marks. 1% seemed to know almost everything about the paper, 1% almost nothing, and the remainder a sort of bell curve between those two extremes. But how is this? There was less material! They should have known it better! But no. So, to compensate for these mysteriously dim students, you put a little bit less on the paper this year.

And the cycle continues. Examiners examine less, so teachers teach less, so students learn less, and so just to *maintain* current pass-levels it's necessary to reduce the syllabus. To effect an *increase* in pass rates, examiners examine a *lot* less, so teachers teach even less, and the cycle accelerates. That's the cycle we're seeing, and that's why it's getting worse and worse until universities are receiving applications from thousands of students all with 5 A's at A-level, most of whom don't seem to know nearly enough to handle the course, and with no way of telling who the really excellent students are.

Grade inflation (or standards decline, whichever way you want to look at it) is clearly unavoidable in this cycle. Can we avoid this cycle? Can it maybe even work the other way? Let's do a hypothetical case:

2005, January: Busted are number one again. You're an examiner.
Under instructions from a crazy government who don't want to get elected ever again, you consciously make this year's exam *harder* than last year's. You put more stuff on it.

2005, July: Will Smith releases another cheerful summer hit. You're a teacher.
You cruise along blissfully unaware of the disastrous results awaiting you. All hell breaks loose when the pass rate plummets to 60% for the first time in decades. Parents call for for you to be sacked for incompetence.

2006, January: Britney Spears new death-metal track goes straight to number one. You're an examiner again.
You got some crappy results last year. Angry students threw rocks at you, and the government took a roasting for a "decline in educational standards". But you're instructed it's not to worry: go ahead and make the exam harder still, we don't want to get elected ever again!

2006, July: Christina Aguilera releases a 3-minute recording of herself performing oral sex. It reaches #2, and you're Mr. Teacher.
Fearing another disaster like last year, where the exam covered more than you taught, and to greater depth than you taught it, you tried hard. You taught more advanced subjects, and covered more material. Pass rates jump to 80%, an excellent increase. Parents laud your skills and fight to gain entry to your school

2007, January: George Michael "covers" Christina's summer hit. It doesn't even chart, but you like it. You're Mr. Examiner.
Brilliant teaching, says everyone, or are students just getting smarter? It's all those computers we gave them. So to compensate for your scary-smart new students, you add yet more material to the exam.

2007, July: You're tired of forced-sounding pop music jokes, and you're Mr. Teacher.
Exams have been getting harder for 2 years running. So you cover yet more material, in even further depth. Exasperatingly, you only just keep pace with the expanding syllabus, even though you worked your students half to death. Calls go out to stop the madness that's over-pressuring our students. You call for the syllabus expansion to stop, and for pass rates to be maintained at their current levels.

At which point the downward spiral sets in again, because there is never any political will to maintain standards, only to raise them -- except that in education, raising the pass rate actually decreases the standards. And this plan would never have got this far anyway, since in 2006 the pressure from angry parents at the massive decline in pass rates would have killed the political will to continue expanding the syllabus.

So there you have it. Not only is grade inflation and a decline in standards real, it's inevitable as long as teachers and parents have this bizarre idea that pass rates should increase, rather than -- as they were intended to do -- indicate the relative ability of students, to allow for selection.