Re: Bush's poll approval rating state by state
Wolf_ODonnell wrote:
You see, 600 people is not statistically valid. The larger the sample size, the better.
As stated previously, Zogby has gotten the popular vote margin correct in the last three presidential elections using a sample of only 1,000 voters for the entire nation. If these polls average 600 people per state, that is the equivalent of 30,000 people for the nation-much, much larger than any poll I have ever seen.
Suffice it to say that the sample size in these polls is more than adequate.
Wolf_ODonnell wrote:Also, sampling has to be completely at random, ensuring a good mix of people from the countryside and the cities. If not, the study is statistically invalid.
These are all from a professional polling organizations which can be counted on to take these things into account. Granted, some polls seem to be able to nail it better than others, but in that case, the inaccuracy is as likely to favor Bush as go against him. The trend among almost all states is against Bush. Baldimo is just trying to throw up some smoke here.
States, being smaller than the nation, can swing one way or the other quite quickly. But here, we see almost all the states swinging against Bush-the trend is clear.