I don't know about you, but I hang up on pollsters. (Since almost all are robo-calls, such rudeness pricks my conscience not one bit.) Reporters use poll results; I wonder how much they trust them? Is there (pardon my skepticism) a bias towards using polling firms that produce dramatic results; not just by candidates but by media outlets who think a horse race makes good copy?
Saturday, January 16, 2016
Who is conducting the dance?
For months now we've endured sound and fury , and claims of momentum, and accusations slung freely around. The pundits that know all and predict all (in hindsight) profess astonishment with the results, and the press has already fallen in love with candidates that make reliably eye-catching headlines. But not one citizen has yet voted.
Subscribe to:
Post Comments (Atom)
2 comments:
I'm starting to get the same feeling about polls that I've had about temperature analysis since AGW became a thing, based on reading various articles on the problems pollsters admit to having. One of the biggest problems they won't even admit to having, in fact that they claim is an integral part of the process, is the adjustment of the raw poll results so that the poll sample reflects the actual population. I'll concede that some adjustments have to be done but adjustments to 'fine-tune' the results are going to be measured against the pollster's view of what the 'real population' looks like. If that view is biased then no amount of adjustment is going to make the poll reflect reality in the same way a bias towards warming produces endless claims that "last year was the hottest recorded!" every year.
Correcting for that kind of bias is very very hard. A physics research team has to crosscheck the corrections, and in the experiments I've been on, do a blind analysis. I don't see anything remotely like that in political polling.
Post a Comment