Originally posted by: Zenmervolt
Originally posted by: Luthien
Concerning evolutionary creationism versus biblical creationism. Wrong it does because it says 55% of all americans believe we were created as we are today. If 55% believe that and 71.7% all americans are christians it is very easy to see that the vast majority of those that believe we were created as we are today are christians hence the reason why I mentioned the 55% of all americans to begin with way way back.
I did miss that, you're right. However, the poll itself from CBS is rather severely flawed in the manner in which the sample is selected.
You did look at the method that CBS uses to conduct their polls, correct? The sample is not truly random. The system used first chooses an area code at random, and then an exchange within that area code at random, and then polls random numbers within that exchange by phone. The entire poll is taken from within a single exhange.
If the random first area code is somewhere in Mississippi, the results of the poll are immediately an irrevocably skewed. For example, even though the number are "random" if the area code and exchange happen to be those of Bob Jones University, the overall poll is not representative.
Even though the specific exchange is chosen randomly, the selection process fails to account for the fact that which exchange is selected will heavily bias the results of the poll even though the numbers appear to be random. In order for the poll to be truly random, the entire phone number, not just the last four digits, would need to be selected randomly before each call to include people from everywhere in the country. As it is, the poll is only valid within a 10-30 mile radius of the particular telephone exchange that was chosen.
Originally posted by: Luthien
Here is my link which says 17.2% in 2001 baptists, lol. I believe that poll is a year newer than yours?
LINK
You will need to scroll down a bit.
Oh, and another point! My survey sampled 50,000 while yours sampled 13,000...
The 2001 ARIS survey gave a result of 16.3% Baptist, not 17.2%. The Wiki page is wrong. (If you look at the link to the ARIS page on the Wiki article, you'll find that the ARIS site lists the correct 16.3% figure for the 2001 poll.)
The Harris poll also only polled eligible voters, while the ARIS has no restriction and therefore includes people below the age of 18 who are less likely to have formulated their own belief system yet and more likely to simply parrot what their parents believe, which would slightly inflate the numbers for the more conservative/fundamentalist religions. (This would further be exaggerated by the tendency of the more fundamentalist sects encouraging larger families.)
In any case, the Harris poll was chosen because, unlike the ARIS, it separates Baptists from Southern Baptists, which is an important distinction. There are some large denominational differences between Southern Baptist and more generic Baptists in general and it's simply not correct to lump the Baptists who do not identify with the Southern Baptist Convention in with the Southern Baptists. We can extrapolate from the Harris poll that roughly 55% of "Baptists" are actually "Southern Baptist", which would yield a result of approximately 9% of the US total population being Southern Baptist. Not a significant difference from the Harris poll numbers speaking from a statistical standpoint. As far as the sample size argument, once a statistically-significant sample size is reached, there is no meaningful accuracy gain from even large increases in sample size. There's no statistical accuracy difference between a sample of 13,000 and a sample of 50,000. Though it is apt to appear that way to those unfamiliar with the workings of statistical analysis.
ZV