jj, I agree that exit polls are accurate, in Germany for example (we don’t do them much in CH) error has always been under, say, .2% (that figure is off the cuff and leaves a wide margin..) Properly done, the error would be yes, within .1%. And I did note the peculiar and hasty addition of the word early to ‘exit poll’ as if there were early ones and late ones, different one from the other, which is nonsense – an exit poll is an exit poll and until it is completed it is meaningless. However as in the US anything to do with simple numbers seems to be hopelessly screwed up (of late), maybe the exit poll was bizarre or off as well. The other explanation bruited about – that people lie to pollsters, while true (e.g. about TV watching and smoking and sex) it is here immaterial, as there is no reason to assume a bias in responses (reducing the no. of cigs smoked daily) nor any reason to take into account massive or consistent lying. Statistically, in a case like Kerry/Bush, whatever lies appear simply cancel each other out and can be ignored. I tend to think that the exit poll was properly done and accurate, and that Kerry did win, but to be sure, you would need the examine whole procedure, all the numbers, etc.
annie, good, but I am not too optimistic about the outcome of challenges or accusations of fraud…
Talking of polls – a very curious thing. CNN’s today’s ‘the buzz’ asks: Do you believe there is a US cover-up surrounding 9/11? (answer: 90% = yes.)
CNN, top left corner
What does this mean?

Posted by: Blackie | Nov 11 2004 17:10 utc | 101