[QUOTE="GreySeal9"]
[QUOTE="surrealnumber5"] the methods used, low sample size, a far from normal age distribution among the sample. if they were at all credible they would have tossed out that data redone it. any highschool stats student would know that. i would wager the poll was engineered not random.
sonicare
What methods do you find objectionable? Please cite from the study.
That sample size really doesn't matter that much unless it isinsufficiently low given the population parameters. Since their population parameters areRepublican primary voters in a single state, a 400 sample size is fine if the sample is representative. Do you have any evidence that this sample is not representative?
Where does the study give any information about the age distribution within the PPP release? If merely lays out what results they got from different age groups. It is interesting to note that very young people in 18-29 range was pretty close to old people as far as the percentage of them that thought it shoud be illegal. For 18-29 YO's, 54% thought it should be illegal while 56% of 65+ up thought it should be illegal. So it's not just old people influencing the base percentages.
The study set out to get the opinions of Republican primary voters in MS and that's what it did. I fail to see the problem with it.
That entirely depends on how they achieved the sample. It could have been a well done randomized study or it could have only targetted only a specific subgroup. We don't know. All that is stated in that link was that they sampled 400 people. I did not see how they chose these people and from where. So the study could have been well done or poorly done. We don't know from what was provided. For instance, if you're going to do a poll on how many citizens know advanced calculus, it makes a big difference if the 400 people you poll are all from MIT.That is true, but surreal was accusing them of having low integrity, and there's no evidence of such in the study and there are not very many studies that actually reveal their sampling method.
PPP is pretty well-respected as far as I know, so I give them the benefit of the doubt if I don't have any evidence that the sample is not representative.
Log in to comment