Bush numbers rising

Kathianne the math doesn't matter in polling. What matters is the phrasing of the questions. You can get a poll to say almost anything you want if ask the questions the right way. That's how news organizations get the results they want. The 7 in 10 is a classic example. If you ask the question did Saddam plan the 9/11 attacks? The answer is probably no. But if you ask Do you think there is any possibility Saddam had a major role in the 9/11 attacks? You'll get a yes from a lot of people. Then you can have your big headling 7 in 10 Americans think Saddam played major role in 9/11.

It isn't hard.

acludem
 
acludem, I'm well aware of what you're saying. One of my degrees is in sociology which including polling. I agree, I've yet to see a totally 'unbiased' poll. BUT, if the questions are given to a sample the same way, every time, the results, while perhaps skewed, (ie. conducted in a reliable manner), are reliable mathematically. That was the point.

With your take, no polls are worth anything. Which is probably your take on them. Truth is, some are very good indicators, though because of caller ID and cells, the phone polls are less reliable than in the past.
 

Forum List

Back
Top