Shrub Job Approval...what's Gallup up to now?!

[Career Prole]

The chimp's job approval in the latest Gallup/CNN poll stayed the same as their last poll (52% approve, 44% disapprove), but look at what's happened to the sample size!
Who'd they throw out?

Since the election, their sample size for every poll has ranged from 1,002 to 1,016.
This last one's sample? 909! 10% just went out the window.
:wtf:

http://www.pollingreport.com/BushJob.htm

[Selatius]

1. Sample size is irrelevant; sample make-up is what matters

Statistically, if you are over the magic number 30, you have a sample that roughly emulates actual conditions. However, the more you have in your sample, the better the match. The only thing that affects this is if you oversample Republicans, which might be what is happening.

[punpirate]

3. I agree that sample distribution...

... is more likely to affect a poll, but sample size does matter--that's what determines the margin of error. Standard statistics indicates that the smaller the sample, the larger the MOE. If it weren't that way, pollsters definitely wouldn't spend the money to call 1000 people if they could get a similar MOE with thirty people.

Cheers.

[mcscajun]

2. Maybe more people are hanging up on them as soon as they hear

the word "Bush"?

[Career Prole]

5. Or maybe Gallup hangs up when they hear the words

"fascist war-mongering lying sonuvabitch".
Hey...that would account for the missing 10% right there!

[roguevalley]

4. gallop always skews. they are fundies and worthless.