Tuesday 30 November 2010

Murder rates worldwide

Always an interesting statistic.

Data such as these have to be taken with a large dose of salt. But it's intriguing to see England and Wales at 43rd out of 198, with a murder rate of 1.6 per 100,000 (where one is the lowest rate in the world, on this account at least; Luxembourg at 0.4). Exactly equal with France and slightly worse than Canada and New Zealand. Scotland, as ever, has a higher murder rate, coming in 72nd with 2.6 per 100,000.

Plainly, the UK isn't the violent hellhole the tabloids and some government ministers would like us to think it is. But it does have a slightly higher murder rate than equivalent countries - and in Scotland the rate could be characterised as substantially higher. So some mixed messages here, and it isn't easy to disentangle them. Are we to conclude that the UK has something of a problem, compared with some of its European neighbours (Germany, 1.0 per 100,000; Italy and Spain 1.2)? Or that we're about on par with other similar countries in terms of size and economy (France) or 'culture' (Canada, New Zealand)?

My money is on the later, actually, but that's personal preference, rather than thought out position, at this stage.

Saturday 20 November 2010

'Nuff said

By the always excellent Ben Goldacre, picking up a piece from here.

One interesting note is the CiF comments, which split almost precisely between "I support nuclear power, what's your problem, people aren't that stupid"; and "I don't support nuclear power, this is terrible manipulation and typical of the nuclear industry".

It is an emotive issue, of course, but it always a shame when people can't (at least make an effort to) separate their personal preferences from a methodology in this way. Whether you support nuclear power or not you should be able to make a reasoned judgement about the way in which the poll was conducted.

I suspect this links into widespread mistrust of opinion polling and statistical techniques. Many people simply don't believe these can provide truly useful information and are by definition to be mistrusted (being linked to spin, focus groups etc.). So the idea that (for example), the survey might be wrong but the idea of a station still a good idea is a bit too challenging.

The preference for nuclear (or not) comes first, and this informs judgement of the survey, whereas in reality this are two quite distinct things.

Thursday 11 November 2010

Students riot!

Whoda thunk it?

What's interesting is the speed with which Cameron et al get on the cop's backs when something like this happens - as I write he's on the news saying 'the police have admitted they got it wrong' - yet when protesters are injured they are much less ready to say the police mishandled things. It's almost as if the police should be there to protect property, not citizens, if those citizens have the temerity to protest. Whoda thunk that, either. (Note that this doesn't justify the 'rioters', of course, especially the idiot who threw the fire extinguisher.)

I suppose the wider point is that these events weren't the police's fault. The blame lies with the protesters involved. So this shouldn't be used as a way to lever in heavier police presences than there are already at many other demos. Police should be there to facilitate protest, not treat protesters as potential terrorists as they did at the G20. The fact that sometimes the demonstrators break the law doesn't alter this, just as the fact that some members of the public break the law doesn't mean that police should treat everyone has a potential criminal.