Thursday, March 17, 2011

Surveys

I'm often torn about how much faith to place on surveys. Obviously, of course, the ones I agree with are the ones that are valid.



(H/T: Le·gal In·sur·rec·tion)
Posted by at
Labels: , ,

1 comments:

Stephanie said...

Excellent clip! Even wording changes in the question can make a difference. This is why Democrats say "Public option" and Republicans say "Government takeover of health care". They poll completely differently.

"Homosexual" vs. "gays and lesbians" too. Apparently the US is slightly against homosexuals openly serving in the military, but a decisive majority is in favor of gays and lesbians serving openly in the military. Go figure.

I like checking out pollingreport.com. They list several different surveys there, so you can see the difference between polls by different organizations over time. Also listed is the number of people polled and the margin of error.

They do print the original question which is helpful in finding these differences in wording. However, they don't show all of the questions of the instrument. You can dig them up sometimes by searching on the Internet.

Good polling organization try not to give leading questions though. They also randomize the order of the questions so they can avoid the type of bias in the clip.

I'm a nerd for this stuff.

3/18/2011 11:57 AM

Post a Comment

Subscribe to: Post Comments (Atom)