Top tips for spotting stats porkies
14 Jun 2012
On the Today programme yesterday morning, Housing Minister Grant Shapps was questioned about his claim that there has been a ‘dramatic increase in affordable homes’. Shelter pointed out that in fact the number of new affordable homes being started has fallen by 68% since the previous year – from around 50,000 to around 16,000.*
So how do you know who’s right and who’s wrong when it comes to stats? As the debate about the number of new affordable homes being built hots up, we thought we should set down some guidelines to help you spot statistical truths from numerical howlers.
Statistics can mislead, even if they are not inaccurate
The golden rule. You can be misled with statistics without a single lie having been told.
Stats can be twisted, spun, manipulated and turned upside-down to suit your message. They can be badly collected too, particularly when it comes to public surveys.
This doesn’t mean that all statistics are always useless, just that they should rarely be taken at face value.
Ask yourself: what’s the context?
Without context, all stats are meaningless. If I said that my team will win the league because they’re scoring 40% more goals, then you should raise a collective eyebrow (not least because my team are Spurs). The key questions here are ‘how does this compare with the other teams’ and ‘since when’?
Compare like with like
A very common statistical howler is to mislead your audience by throwing out not-quite relevant numbers like chaff from a plane. This changes the terms of the debate and distracts from the question at hand.
For instance, if you asked me: what percentage of your time do you spend at the gym? I might answer: well I’m going running 12% more than a month ago. That’s great. But it’s clearly not answering your question. You should equally watch out for slippery definitions, such as ‘affordable’ rent.
Not all statistics are equal
Finally, don’t assume that just because some statistics are misleading or distorted that all must be.
This seems a reasonable response when so many numbers are banded about, but if you use the rules above you will be able to come to your own conclusions about whether you are being informed or taken for a ride. Good luck!
Top links for budding statisticians
- Shelter’s Housing Databank, a fantastic source of up to date facts and figures.
- More or Less, a BBC Radio 4 programme which sheds lights on public stats.
- The House of Commons guide to correct use of statistics.
*Mr Shapps was only able to make his ‘dramatic increase’ claim by comparing the number started in the final six months of this year (15,269) with the historically incredibly low figure for the first six months of the year (429).