Statistically speaking

My maths is no good – and I’m not proud of it. I failed at O level (twice) and continue to fail to understand mathematical concepts. I scraped a pass in my economics papers and had to be gently taken through my statistics by a patient and gifted medical statistician who I was working with who probably found it light relief after crunching big numbers and many variables in proper, serious work.

However, despite my admitted weaknesses on the maths front, the combined efforts of my tutor, my colleague and my professorial statistician friend (who has breathing difficulties when coming down to my level of ineptness) have meant that I do, at least, take an enquiring approach to statistics. Journalists are often very bad at stats – or possibly willfully negligent. Given how often statistical findings are used as the centre of a story it really isn’t good enough. And anyway there are plenty of people around to help out and to interpret. If a researcher says that the results aren’t statistically significant, really that should mean the story doesn’t run – or at least not as a stats based story.

In 1995 a report on deep vein thrombosis risk associated with the contraceptive pill indictated that there was twice the risk with a certain pill. Which there was. But as the risk was still lower than deep vein thrombosis in pregnancy, the scare that led to unwanted pregnancies and an increase in abortion, actually also increased the risk to women – to say nothing of the long-term effects of pregnancy!

There was once an advertisement for Goodyear tyres which claimed, if I recall (but it was aeons ago) 4 times the holding power and 6 times the strength – a claim that led the front man, a former chief constable to claim “I’m convinced they’re a major contribution to road safety” A comic in a TV programme of the time said that he’d found a tyre that was 145 times stronger and had 750 times the holding power but that as his test piece was a bit of banana skin he doubted the relevance. A pre-stats lesson to ask “times more than what.”

I’m only posting this because I have just found an excellent piece in the BBC magazine, part of a series of six. In an effort to continue to approach news stories with the right level of enquiry I shall certainly be reading all the series.