Search This Blog

Sunday, 7 April 2013

Trusting statistics, part 1

I have, I must confess, fallen somewhat silent, the reason being pressure of work as I check page-proofs of a new book (it's a massive Australian history for younger readers, and there's a bit of information here).  Once, at this stage in the book-cycle, I would have a mountain of expensive colour photocopies, but now, it's just PDFs that are placed in a secure location for me to download.  I attach comments, place my file on the cloud, and back it goes,

Still, I would like to maintain a sort of presence here, and I was inspired by this hilarious piece of false statistics that came my way yesterday.

But rather than write something fresh (my brain is too clogged, right now), I pulled out a radio talk that I presented some years ago, which is basically about putting your trust in statistics.

Remember what they say: figures don't lie, but liars can figure, though often they don'y bother: they just say "statistics prove my point" and assume that there is no need to do or say any more.

* * * * * *

I'm a reformed smoker.  I gave up multiple decades ago, but I did it for political reasons.  I just wasn't prepared to support the government of the day by paying tax on both liquor and tobacco, so I gave up the demon weed.  Soon after that, the government changed, but I didn't like the other mob very much either, so I kept on not smoking.

Now it goes without saying that reformed smokers are tiresome people.  At least if you're a smoker they are.  They will keep on at you, trying to get you to stop as well, so they can hang another scalp on their thoroughly smug and sanctimonious belts.

To all non-smokers, those who still puff smoke are tiresome people, who can't see the carcinoma for the smoke clouds.  Stupid fools who deny any possibility of any link between smoking and anything.

Like the tobacco pushers, the smokers dismiss the figures contemptuously as "only statistics".  The really tiresome smoker will even say a few unkind things about the statisticians who are behind the figures.  Or about the statisticians who lie behind the figures.

But the smokers don't just settle for simple attacks, like alleging the statisticians are secret non-smokers.  It's all very much nastier than that.  So much nastier, there must be a deeply ingrained cultural hatred of statisticians in our society, and probably an equal contempt or detestation for statistics as well.

At one stage in my infamous career, I freely confess to having quite a lot to do with gathering statistics, and messing about with numbers, an honourable and harmless activity, I would have thought.  But it was then I discovered that such people, while sometimes accused gently of being "mathematicians", more often suffer the far heavier opprobrium of being called "statisticians".

I certainly encountered this all-too-human tendency, and all too often at that.  The things people used to say about statistics and the users of statistics offended me greatly.

If you've never suffered from being called a statistician, you may think it's a minor inconvenience to suffer.  That's only because you've never heard the jokes which go with the label: there are more jokes about statistics than I know about the dismal science of economics, even if you let me throw in all of the many jokes I know about the economists as well.

Take, for example, the definition of a statistician as "somebody who's rather good around figures, but who lacks the personality to be an accountant".  Or the story about the statistician who drowned in a lake with an average depth of six inches.

These days of course, we really ought to say fifteen centimetres, rather than six inches, but I'd like to stress the hoary agedness of so many of these witticisms about statistics: my reasons will become clear soon enough.

Then there are other clever-clogses who earnestly assure and advise us that a statistician collects data and draws confusions.  We are also told statisticians are people who draw mathematically precise lines from an unwarranted assumption to a foregone conclusion.

With this sort of bias floating around, it's little wonder the great physicist, Lord Rutherford, once harumphed, "If your experiment needs statistics, then you ought to have done a better experiment".

An example of the blogger's illustrative imperative?
No, rather it serves as a reminder that statistics and
wooden horses have something in common. This
one is real enough, located just outside the site
where Troy once flourished. All right, it really
was the imperative working.  So what? Do you
want pretty pictures, or don't you?
Then there was the cruel and cutting comment, attributed to various witty people, and said to be about various people, that "X uses statistics much as a drunkard uses a lamp-post: rather more for support than for illumination".

On a slightly different tack, but still a debunking one, people will sometimes assert that statistics show how the vast majority of people have more than the average number of legs.  Which is a bit like the common discovery, popular with conservatives, that tests reveal half our nation's school leavers to be below average.

Or the mildly sexist one-liner that statistics are like bikinis: what they reveal is interesting, but what they conceal is vital.  And politicians like to get into the act as well, so we find Fiorello La Guardia, one-time mayor of New York saying statistics are like psychiatrists — alienists he called them — statistics are like psychiatrists because they'll testify for either side.

Finally, there's the grand-daddy of them all, the famous line about "Lies, damned lies, and statistics".  Now quickly answer out loud, so there's no cheating: who was it who first said that?

The odds are if there are two or three know-alls in your house, you'll now be locked in bitter dispute.  At least, I hope you are.  It may be hard on you, but it will help me prove my point.

The official version is that this line was first voiced by Mr. Disraeli, the well-known politician, but many quite reputable and reliable reference books attribute it to no less a personage than the author Mark Twain.

Now you can see why I expect to have started a few arguments by asking you to say your answers out loud.  Even the experts can't agree on who it was said it!  So if the authorities can't answer with one clear voice, how could you and your neighbours?

Well, the true facts of the case are fairly simple.  That catchy snippet about "Lies, Damned Lies" et cetera was first published by Mark Twain all right, this can be proven to anybody's satisfaction, but Twain attributed the line to Disraeli.

The only problem is this: search as hard as you like, you won't find the story in any earlier publication than Twain's Autobiography.  In short, Mark Twain made the whole thing up!  Disraeli never spoke those words: Twain invented them all, but he wanted the joke to have a greater force, and so gave the credit to an English politician.

Twain wasn't only well-known for his admiration of a good "Stretcher" (of the truth, that is), he even lied when he was talking about lies, and his name wasn't even Mark Twain, but Samuel Clemens!  Now would you buy a used statistic from this man?

Come to think of it, the yarn's pedigree should have been enough in itself to cast doubt on the its veracity, with an arch-liar like Twain quoting, of all things, a politician!  Yes, sad but true, there are more jokes about politicians than there are about statisticians, but only by a short head.  It must be because so many politicians are trained originally as economists.

When you look to the background of the "Damned Lies" story, something that I did recently, there's an even stronger link between statistics and politics.  Last century, when Disraeli is supposed to have made the remark, statistics were just numbers about the State.  The state of the State, all summed up in a few simple numbers, as it were.

That's enough for a bit.  This continues in Part 2 and Part 3.

No comments:

Post a Comment