Nevermind - I'm not in the mood for the snark today.

[jillan]

Yes, I'm feeling a little more than grumpy living here in Arizona....

[TK421]

1. those 500 are probably the ones still breathing n/t

[pinto]

2. Good catch. Thanks for the post.

:thumbsup:

[Captain Hilts]

12. No, it's an ignorant post. A survey sample of 500 for a state is enough. nt

[gristy]

3. 500 people is a reasonable sample size

It gives a 95% confidence interval of +/-4.4%. Meaning that it is likely (95% likely) that between 65.6% and 74.4% of all Arizonans were in favor of this bill passing. Of course, it's also possible that it was a push poll or that the sample wasn't randomly selected. But your concern about sample size is misplaced.

calculator here: http://www.surveysystem.com/sscalc.htm

[Lyric]

7. Of course, that's only true if we also assume that the sample was truly random.

Depending on who did the polling and what method was used, that might not be the case.

[ChairmanAgnostic]

9. with a purely random sample, true. But, was this a random sample?

I love how Ras puts a computerized thumb on the scales when they "select" a pseudo random sample for their polls. I was called once by Ras, and advised them that I count myself as an independent. They ended the poll early.

[Arkana]

4. What did they poll, Maricopa County?

[TK421]

11. that there is Arpaio Country, boy!!! You damn well better like it!!!!!

or get the hell out of AMURKA!!

[Arkana]

21. That was my point.

Sheriff Joe was elected by a bunch of old white retirees who hate minorities.

[sandnsea]

13. Exactly

I've read in the past that the only thing the pollster chooses is the zip code. Well gee, you can get any result you want if you choose the zip code yourself.

[ProdigalJunkMail]

5. all polls use extrapolation

why should this one be any different? if the set used is representative of the population as a whole then i don't see the number being invalid at all.

sP

[Lyric]

6. 500 people are not enough to give a statistically valid sample.

Not even CLOSE to enough. And was it even truly random? If they didn't even bother to poll enough people to get a valid sample, they probably didn't bother to make sure the samples were totally random, either.

[kristopher]

8. Yes it is, see post #3.

[Captain Hilts]

14. Yes, it's more than enough. nt

[gristy]

15. Of course it is. The statistical "validity" of a sample is quantified by the confidence interval.

Try the link I gave above.

[Captain Hilts]

10. For a statewide survey, 500 is more than enough people. They use 2000 for a national survey. nt

[sakabatou]

16. Yeah, way to small a poll number

[Codeine]

17. Posts like this remind me of why everyone should be required

to take a course in basic statistics and applied mathematics before graduating high school. Assuming a random selection 500 is a reasonable sample size. Not always safe to assume random sampling with Rasmussen, but that is a separate issue.

[bridgit]

18. There are sample sizes down as low as 350 respondents, the key is any bias in the taking...

But allot of people have aced themselves out of being able to worry about it when they all went on the 'do not call list'

[Rex]

19. What are the chances of 500 people having the same opinion

out of 6.8 million?

[OHdem10]

20. Proof, --you can make a poll say anything you wish.