Rubbish, trash, detritus and journalism

I was sitting in the waiting room. The dentist would see me next.

Another man was waiting. He was reading The Daily Rubbish. (I know, it's juvenile. But I've stopped giving links to lousy news outlets.)

All of a sudden - and I'm not sure if he was addressing me, the receptionist or himself - he declared that using a mobile phone for more than [who cares?] minutes per day can triple the risk of cancer.

Triple! That's bad - right?

Maybe.

I asked for the numbers. I confess, I engaged.

He didn't have them.

So we don't know the odds. Say the odds are one in a billion, give or take, er, one. Three times the risk would be three in a billion give or take three. I don't see the trebling as a problem.

If the odds were, say, one in ten - becoming three in ten - then I'd be concerned. But if that were genuinely the case, I'd expect to have seen more people keeling over by now.

Without the actual odds - with just the number the odds have been multiplied by - we have learned NOTHING. Sorry for shouting.

And, unless the survey included everyone who's ever lived on the face of the planet (which I think we can assume it didn't), then the numbers were calculated by choosing a group of people and assuming everyone's pretty much like them, on the whole. More or less.

Which means that there's some wiggle room in how accurate the numbers are. If the wiggle room is bigger than the number, then the number is MEANINGLESS. Sorry for shouting.

And if you don't know the numbers, the size of the wiggle room or the size of the group of people then all you know is how much bigger one MEANINGLESS number is than another MEANINGLESS number.

That's probably just going to give you a third meaningless number. You know, like that the risk of getting cancer from something has trebled.

Unemployment

There's a headline to excite people. (What follows is an example. These are not real numbers but they are broadly plausible and, with any luck, might be instrumental in making an interesting point.)

A survey has told us the percentage of people who work part-time but aren't looking for any more working hours. It's 68%. Is that a good statistic?

In my example scenario, one million people work part-time. And 68% of them don't want to work any more hours. Would you be pleased about that 68% figure? It's probably a high enough score to get a first-class honours degree.

But it would mean there are 320,000 people who are short of money because they don't have enough work.

Still think it's okay?

There's a time and a place for percentages. But when we're talking about suffering or financial instability for people, it's good to know just how many of them are being whacked round the head (figuratively, naturally).

End of maths lesson

But keep it in mind. Next time you see statistics, percentages and big numbers being wafted about as if they're important, try to work out what they actually mean. Often the answer is very little.