As part of this summer’s installment of the annual media ritual of bashing Britain’s system of secondary-school qualifications, the Guardian has sent 23-year-old journalist (and maths graduate) Tom Whipple undercover to take the AS-Level sociology exam. Next week, the paper will reveal how he did.

As someone with a postgraduate degree in this maligned discipline, I find this tale of what’s going on at the secondary-school level rather depressing.

The article does include some funny stuff I can relate to. In the process of taking the necessary secondary school courses, he discovered he was carrying out a covert participant observation of a discipline with lots of jargon.

And then there’s this:

Think of sociology as the slightly hairy offspring of politics and economics. Perhaps the best claim it has to legitimacy is that, like any decent humanities subject, it has a good, honest academic schism; positivists on one side and interpretivists on the other (actually, this is one of many schisms, but let’s keep it simple). Positivists are the sort of sensible chaps you’d go to a pub with. They think people should be studied using scientific methods. Take, for example, the use of official statistics in research. Positivists would say they are extremely useful, thank you very much.

Interpretivists are the sort of people you would abandon your pint in the pub to avoid. If you asked them about official statistics, they would look at you and say: “What is a statistic? Isn’t it just a social construct that reflects the questioner’s prejudices?”

And then, as your eyes glaze over, a barely pubescent guy in the corner of the pub puts down his Socialist Worker and, attracted by the controversy, comes over to explain that you’re both wrong. Actually, statistics are “an instrument of the ruling classes used to propagate and legitimise an exploitative capitalist system”. He’s a Marxist.

We can safely assume that our maths graduate is a positivist (or “naive empericist”, some might put it). But we interpretivists who want to study the social construction of statistics (see, eg Best 2002) aren’t so scary and boring.

We help empericists understand why there are limits to their claim to carrying out objective research modeled on the natural sciences — why people can’t be studied like electrons.

What you probably don&rquo;t get in A-Level sociology is the idea that analyses of “social construction” aren’t about uncovering bias or “prejudices”.

All accounts and narratives are selective and an effort by people to assign meaning to phenomena. This doesn’t mean they are lying, or telling un-truths. It just means that truth is arrived at in a selective social process that is necessarily partial. The best description I know of is that the sociology of knowledge is the emperical cousin of epistemology.

Rather than uncovering the logical rules for deriving knowledge, sociologists who worry about this sort of thing examine what people who create narratives — natural scientists, journalists or government statisticians — actually do, what social constraints condition them to behave in that particular way, and what the wider consequences are.

The debate about statistics in sociology is actually a very good example of why all this is important. Statistics are central to the functioning of modern state and corporate bureaucracies and therefore understanding how they work is essential to understanding how contemporary governance works.

OK, I probably just proved his point.