Data, statistics and true lies.

Been seeing a lot about school data and statistics recently, none of which seems to be too helpful to getting children to learn. Two things in particular caught my eye; the impending change in the “league tables”, where the concept of CVA is likely to be dumped, and the imposition of the English Bacc on schools (I use the word imposition not in a pejorative sense, just in the sense that there was no wide spread clamour for the EBacc).

EBacc first. A lot has been said about the EBacc over the past year, almost to the extent of it now becoming part of the furniture. A wide range of reasons have been given for the Coalition bringing this measure in, ranging from it being one man wanting to hark back to the delights of his own eduction, to, to, well, to not very many other reasons really. I’ve always thought that the reason for the measure was quite obvious. Its the same reason that most politicians bring in new metrics – they think that the use of that particular measure will benefit them. And if you consider the EBacc you will understand how spectacularly successful it is going to make the Coalition look when it comes to education.

This current national achievement rate for the EBacc is 15%. This means that to improve this rate by 1% to 16% about 7,500 additional students will have had to get their EBacc, which is less than an additonal 2 students per secondary school and would amount to an almost 7% increase in attainment (did you see what I did there). Without access to the pupil level data it is impossible to say for certain, but it is a pretty good guess that this will require no more from those 7,500 students than to move from a D to a C in one or two subjects. It is also a pretty good guess that the improvement this year will be more than 7,500 students. So, we can read the press release now, about how the Coalition policies have improved education overnight, without spending more. And given the initial low level of the statistic this will be able to go on for years. Before the end of this parliament the Coalition will probably be able to claim that they have improved (according to their measure) every single secondary school in the land.

Of course, this (mis)use of statistics in education is not new, nor is it a dastardly Coalition invention. The same trick was pulled with Maths and English KS2 Level 4+ figures back in the nineties. These percentages were in the mid 40’s then, providing plenty of scope for increase. Interesting, the Science figures at the same time were 70% – this did not fit the requirement so Maths and English became our national priority. Which led to a fall off in science outcomes at this level requiring its own intervention later on.

The CVA issue is much more complex and the discussions around it have become polarised and political. My view on this is a mathematical one. The factors that lead a particular student to a particular grade in a particular subject are incredibly complex. Some of these factors can be objectively measured, but the majority either are not measured or defy objective measurement (for that individual student). Turning all this data into a single statistic renders any information contained in it quite useless. So two schools have the same CVA score. What does this tell anyone about the schools? The very complexity of the CVA measure allows it to be ignored.

The problem is one of mathematics. There is no single statistic that will provide a true measure of a school or the students in it. This desire for one number, which is a desire of politicians, is the real problem. We have allowed our politicians, all of them (and the media) to dumb us down, not because we won’t understand the complexity, but because they don’t.

Advertisements