|
Edited on Mon Apr-14-08 11:23 AM by Demeter
So the way to attack a poll is to question:
1) the questions--leading, biased, push-polling, lack of suitable responses in choices
2) those questioned-- size of sample, randomness of sample, selection process bias
3) the interpretation--all of the above, and including lying about the results of the poll
4) the timing--a poll is only good until something changes, and change comes very fast these days, while some polls span days or even weeks
So, the way to attack a "scientific tool" is no different than in any other scientific controversy. Wonks needed! Bring your slide rules.
And then you need publicity--to shout good polls from the housetops, and wither bad ones with very public scorn and debunking. It's what the GOP has done for years--if only with their own agenda always in mind. That's where the biggest problem lies--the corruption of the news media.
|