Showing posts with label statistics. Show all posts
Showing posts with label statistics. Show all posts

Wednesday, 1 July 2009

City roulette

I'm always amused by those dubious surveys declaring that London (or New York or Stockholm) is the world's best city for something or other - quality of life, happiness, tourist potential etc.

The indicators used to measure these things are always quite arbitrary and seldom the ones ordinary folk like me would use.

It's all very well quoting household recycling, life expectancy and school success. All very worthy, I'm sure, but are they really the things that get you and I excited? I think not. In my case it's much more likely to be the accessibility of Belgian chocolate, a well-stocked bookshop or a flattering pair of jeans. Bugger life expectancy - what's the point of living for 80 years if you've been thoroughly miserable for half of them?

You can see how arbitrary it all is when every new survey contradicts the previous one. Tokyo's the world's greatest city? You must be joking, it's Berlin of course - it has more electric cars and hairdressers than any other city. Or was it espresso bars and cycle lanes? Good news for tourist chiefs in the chosen city, but for the losers - it probably just prompts a cynical snort of disbelief.

As for the things that can't be measured at all, like breathtaking scenery, a friendly atmosphere or an intellectual buzz - what happens to them in the midst of all these statistical calculations and pie charts? It's these intangibles the researchers overlook that make our own city such fun.

And what about the areas outside the cities that the number-crunchers don't even consider? Are they saying the Orkney Islands are a soul-destroying backwater? Or Nova Scotia is a God-forsaken wilderness? There's an underlying assumption that city living is better, even if your city's only number 21 in their roll-call. Many rural dwellers would beg to differ.

But personally, I'm absolutely certain Belfast's the greatest city on earth - and I've got the figures to prove it.

Photo: Donegall Square, Belfast
.................................................................................

Katie, the sweet tortoiseshell cat from three doors up, one of our regular visitors, has been missing since the weekend. I do hope she hasn't come to a sticky end.

Sunday, 19 October 2008

Bedtime stories

The story of the 105 year old virgin got me thinking. Just how believable are surveys on sex when we have no way of checking people's unlikely replies?

The newspapers are full of breathless headlines about how often, or how little, we're getting our end away. Or how many affairs we've had or which unconventional form of sex. But is it really true?

People are probably more likely to lie about sex than just about anything else. Of course we all want others to think we're having an ecstatic time in bed, that our partners are irresistible sex bombshells, that we can't get enough of it. Who wants to be thought of as a frustrated loner resorting to guilty fumblings in a shabby bedsit?

So when a survey says that 37 year old Jack is at it every night, has had 91 lovers and his wife is insatiable, is it any more credible than Pinnochio giving birth to triplets? Where's the evidence? Especially if he's anonymous and there's nobody out there to say "Him? Are you joking? The only thing that gets him going is the football results."

Survey after survey claims men have had a lot more affairs than women. So exactly who are they having the surplus affairs with? The dogs? A series of Romanian women? Amnesiacs? Other men?

And with sexual frequency, we're obviously relying more on personal memory and estimation than a tangible record of what happened when. How many times in the last month? Ooer, let me think. Choose a number, any number, then double it....

But there again, how many people take any notice of sex surveys, whether bogus or genuine? Sex is such a personal business and preferences so unique, why should we care what other people do? Somehow I think knowing your own loved one's hot spots is more useful than knowing the climax rate in Huddersfield.
.................................................................................

Nicole has a really perceptive and thought-provoking post about lying and deception. Go have a look.

Wednesday, 27 June 2007

Beware statistics

How sceptical we've all become about statistics. We just assume straightaway they've been massaged, fiddled and adjusted ad nauseam to give the desired result rather than the real one.

The health service is improving by leaps and bounds? Schools are doing a fantastic job? We've all got more money to spend than ever? We just don't believe the figures any more, there's a funny smell about them.

A lot of the time, they just don't chime with our personal experience, or our friends' experience. The health service has never been better? So how come Uncle Ted has been waiting six months for a heart op? How come I've been waiting nine for physiotherapy?

The fact is that too many people have a vested interest in cooking the figures to their own advantage. Politicians looking for votes. Drug firms dependent on safe medicines. Police forces expected to solve crimes. If the stats look bad, they'll give them a little tweak in the right direction.

We all know examples of people being caught with their pants down. When A&E waiting times dramatically decreased it was discovered that senior managers had stealthily improved them by falsifying admission and treatment records. The figures were worthless.

Individuals are no more reliable. If someone's asked how often they have sex, "Several times a week" is a lot more likely than "Actually, I can't get it up. I've been a washout for six months." Who freely admits their failings and weaknesses?

Statistics gathered by reputable academics with no axe to grind may be 100% legit. Unfortunately these principled souls are thin on the ground.

So those precise-looking figures may look good in a newspaper headline. They may be presented by a jolly authoritative chap in a smart suit from the Institute of Something-Or-Other. But most of us take them with a bucketload of salt and think "Sure. So who stands to gain? Where's the pay-off?"

(Bet you didn't know that 57% of estate agents are transvestites, 23% suck their thumbs and 39% are satanists? Who'd have thought it?)

NB: This post was completely rewritten on 17.10.08