What hope do the children have?
In the UK we suffer a very bad obsession with league tables in which the performance of every public body is graded (on an arbitrary scale) and compared against others – rarely in a like for like but that is another matter.
As you can imagine, schools bear the brunt of this. Parents are understandably determined to get their children the best education possible, often moving across the country to be in the catchment area of their chosen school. Most of this one-upmanship is derived from the school tables, helpfully published on the BBC website.
All the trust has to be placed in what ever body is responsible for collecting these numbers. Are they up to the task?
Idly surfing the web, I came upon this educational report on the BBC. It is the stats for an infants/primary school (ages 3 – 11) in Shrewsbury. All normal. Have a look at the stats and it seems like its a reasonably good school – it performs above the average for its educational authority, which is also above the average nationally.
Then have a look at this:
PERFORMANCE
37 eligible, 18.9% of whom had special educational needs
At first glance it seems normal and slightly low compared to some schools.
Then look at the numbers.
18.9% of 37 is 6.993.
This means that 6.993 students have special educational needs. How is that possible? Is this just a rounding problem? No, because 7 pupils would be 19% in any normal formulation.
Either the organisation who collates the stats is mathematically challenged, or they have massaged the numbers to make it look lower than it is (and are ethically challenged).
Whichever it is, how much faith can you have in this system?
Maybe there’s one “special needs” child who’s missing a few teeth or other minor body parts. That would make him or her the .993 person.
To be fair, they’ve followed rounding rules quite correctly — 18.9189189…% rounds to 18.9%. But seriously, three significant figures from 37 people? Shouldn’t be using percentages at all at that level of accuracy.
Discarding the meaningless digits at the end will make it impossible to rank the schools of course, but that should be a clue that you’re ranking things by statistical noise.
Andrew – good point, however saying 19% would have made a lot more sense.
I think the missing .007 is significant and this is really a school for spies….