2008-08-15

Defensive Statistics

The results are in, so now is the time for all conscientious Physics teachers to analyse their A Level results.

The newspapers are always first off the mark with ranked tables of gross percentages, showing what proportion of each grammar or independent school gained A or B grades. The tables are then dissected and the top schools pronounced. That they are always selective, either academically or socially, will not be dwelt upon.

Within schools, the Physics teachers will have a couple of weeks to analyse the results of their own students. There are two pressing reasons: to properly assess you own performance, and to have a ready defence against naive assessments by managers.

There will be decisions made using unjustified or unreliable comparisons between very different subjects and with small sample problems. (Why on Earth do class stats get reported to three significant figures?) There will be a need to explain that Johnny did not necessarily underperform in Physics just because he got higher grades in Computing and English.

I have had a cross Head of Sixth Form who presented data showing that half the Physics class had Physics as their worst grade at AS Level. Was I really suggesting that Physics was harder than other subjects? - he asked me. Well, er, yes.

It would be funny if important decisions weren't being made based on these amateur stats and analysis.

No comments: