So, based on the first 302 respondents to the General Election 2005 Estimation Quiz who have expressed a voting intention, here are the results so far:
These results should be taken with a pinch of salt, because of the very small sample size and the self-selecting nature of the sample. But, that aside, we discover that:
- Conservative supporters (all 34 of them) are the best informed of all (must be the expensive schools they all go to, I suppose);
- Lib Dems are, by a small margin, the worst informed -- but, on the internets at least, the most numerous;
- Labour supporters are slightly better-informed than average.
(In the plot above, I've obtained the `all respondents' curve by weighting each party's supporters to reflect the average of recent opinion polls; I won't claim that that's likely to be a very effective way to produce a balanced sample of the population.)
Still, at this stage the numbers playing are far too small to make any very accurate claims about the population at large. So come on, people! Do your teams proud!
Comments
Posted by Gareth Williams, Thursday, 28 April 2005 22:50 (link):
Hello
The Estimation Quiz
You will gather, as I'm getting picky, that I came out as a bit of a dunce, but in Q10, your answer of 35 pence is the cost of the raw ingredients. A true cost of the dinners would include the cost of labour, facilities, and so on.
Regards Gareth Williams
Posted by Chris Lightfoot, Friday, 29 April 2005 00:51 (link):
Hmm. I see your point, but... the question asks for,
The other costs you quote are basically fixed costs; so far as I understand it 35p is a fair estimate of the marginal budget per child per meal, which is how I'd expect the question to be interpreted. I agree that it probably could be phrased better.Posted by dsquared, Friday, 29 April 2005 07:59 (link):
I think that the "1000 dead of MRSA" stat is also controversial, to say the least. The death certificate count is a materially lower (http://bmj.bmjjournals.com/cgi/content/full/325/7377/1390/Fu1), and I'm having trouble sourcing the numbers in your BBC reference; the "5,000/year from hospital acquired infections" is from a National Audit Office report in 1995, but the reason for believing that 20% of them are MRSA is proving elusive. The trouble is, of course, that most people who get MRSA in hospitals were already pretty ill to start with.
The ONS has this "nugget" which gives 1400 mentions of staphylococcus on death certificates, about 400 of which mention that it was a resistant bug. Mentions on death certificates isn't necessarily the same as causes, and the trend in the reporting data is quite likely to be influenced by reporting rates because the increase in lab reports is much lower. I'm not sure that the estimation quiz framework works all that well when there is considerable uncertainty about the actual number which might or might not be correlated with someone's individual estimation uncertainty.
Posted by Chris Lightfoot, Saturday, 30 April 2005 01:01 (link):
Noted. 'Course, having launched the bloody thing I can't really significantly change the questions or the answers without provoking more whinging. That's my excuse, and I'm sticking to it.
Oh well, that's the last time I believe [in] the BBC.
Post a new comment.
Comments copyright (c) contributors and available under a Creative Commons License. See also the comments policy.