Dear Dave

Friday 10 August 2007

Statistics

Dear Dave,

Ninety-three percent of statistics are mis-reported.

Think about it. Your average journalist is an arts graduate who hasn't studied any maths whatsoever in twenty-three and a half years. Even the four-fifths-forgotten statistics lessons they have had will have been to a very low level. The typical journalist, for instance, will only know two and a third types of average and will use the mean in eighty-nine percent of cases. Ask them to calculate the probability of three cases of a ten-thousand-to-one non-contagious illness in a village of a hundred people and they will either answer, 'Small enough to make a big headline,' or run away screaming. Neither of these is the correct answer.

The correct answer runs closer to, 'Is this the chance it will occur in a particular village chosen at random or that it will occur in some village somewhere in the United Kingdom? And besides, surely it should be the probability of three or more cases? And what are the assumptions, again? You haven't even mentioned timescales, for crying out loud!'

An article I read the other week kept going on about the effects of interest rate rises on a typical mortgage. What on earth is a 'typical' mortgage?! Was it the mean value of all mortgages they were talking about or of new mortgages? Something else reckoned that a 'typical' baby sleeps sixteen hours a day. Since I felt lucky if any of mine slept twelve hours out of twenty-four when they were small, I can only assume that some children barely open their eyes before the age of two. What?

Then there are phone-in polls. Don't get me started on those. (No, really - it won't be pretty...)

To be fair, I don't think getting mathematicians to write the newspapers would, in general, be a good idea. (Can you imagine?) But it was with some scepticism I viewed reports of recent surveys of dads. One, for instance, claimed 43% of dads surveyed had 'put their careers on hold' to spend more time with their families, passing up work worth an average of £2,800 a year. What does that really mean, though? 10% of all the dads claimed to have gone to working part-time. Surely this must account for the vast majority of the wage reduction? It's possible to conclude that over half of the dads had made no changes to their work patterns (maybe they were laid-back to start with!), one third had cut back on overtime a bit and one tenth had gone part-time (but who knows what their partners are up to?).

Does this actually tell us much?

Not really. I probably achieve equally scientific surveys by hanging around outside school at collecting time. On this basis, I'd say there are plenty of couples where both partners are heavily involved in childcare. One or both of them may be working part-time but it's just as likely they're both working strange shifts and granny is filling in the gaps in the schedule. There are housedads about but you could probably cram us all into a phone box.

Looking at parent and toddler attendance gives a different picture. Dads are rare. I am 'the man', quite often. Either the percentage of dads staying home and looking after very small children is extremely low or they're all scared. Maybe we're as endangered as pandas or female executives or crumple-horned snorkacks. Who knows? Not me.

I think all the various surveys show is that the number of housedads is on the rise but, you know, that wouldn't be hard. I don't think there will be vast numbers of us any time soon. Probably more important is that, 'on average', dads are far more involved with their kids than they were a generation ago. Scarily, this makes you and me role models for a revolution where men engage more fully with their offspring. We demonstrate that it's perfectly possible for men to look after children whatever the occasion, not just on Saturday afternoons. We can educate and encourage. We can teach other men that, 'Kids are fun once you know which way up to hold them!' (We might want to think of a better slogan that that, though...)

That may sound like a tall order but the simplest way to start is to walk along the road smiling with a stack of children in tow. If we can do it without everyone treating us like pandas, so much the better. ('Look! Look! There's one. It's a housedad! Aw, isn't he cute with all the little ones. Do you think if I poke this stick at him, he'll eat the leaves off the end?')

Probably best to the ignore the statistics and get on with it. Of course, you may be thinking that if ninety-three percent of statistics are mis-reported, what about the rest? Should we pay attention in order to learn from the other seven percent of stats we hear and read about?

I wouldn't bother, to be honest - those ones are entirely made up.

I'm off to prepare for the paintballing tomorrow. Apparently, Scary Karen has been watching Rambo III all week as training. This could get brutal...

Yours in a woman's world,

Ed.

PS I read a great book about statistics last year. It's called Freakonomics and it's by Steven D Levitt and Stephen J Dubner. Did you know that there's a correlation between having a house full of books and having children who do well at school but that there isn't a correlation between reading to children every day and them doing well?

Conclusion: Books are magic! Buy books!

Kidding. It's not what you do that makes you a good parent, it's who you are. Isn't that both reassuring and frightening...

3 comments:

Anonymous said...

78.1% of statistics are made up on the spot

Anonymous said...

This reminds me of a poster I saw a few years back claiming in large letters

40% of children will use drugs

It bothered me for weeks. Even if it's true that 40% of today's children will use drugs at least once at some point in their lives, is that so surprising?

Or did the poster mean to tell us that at any given time 40% of children were using drugs? If drug use in the under-nines is rare, that pushes the figure nearer 80% for nine to 17 year-olds.

That didn't seem likely.

It seems the matter was raised in parliament. A quick Google search found this report.

Lord Blackwell asked Her Majesty's Government:

Whether the assertion in the current poster advertisements by Action on Addiction that 40 per cent. of children will use drugs is correct.[HL3880]

and received the answer:

The 1996 National Drugs Campaign Survey by the Health Education Authority suggested that 16 per cent. of 11-14 year olds and 40 per cent. of 16-19 year olds have ever used drugs.

Would you call 16 to 19 year-olds children?

Anonymous said...

depends on the individual.

remember,you can prove anything and disprove anything with statistics