Quote:
Originally Posted by Delta I've been saying this a lot lately -- The overall population number is completely irrelevant to the accuracy of any poll (internet or not, self-selected or not). If the poll fails to be representative, it has nothing to do with the size of the fraction polled.
Statistical sampling theory in fact goes and assumes that you've got an infinite sized population, because it totally doesn't matter what that number is, and the math is simpler that way. It's when the fraction polled gets too big (over 5%) that you have to do an a bit more calculation to account for it. See: http://en.wikipedia.org/wiki/Margin_...opulation_size |
The "added math" is to account for the larger sampling having a smaller margin of error... in other words, more accurate. While obviously hyperbole, a smaller margin of error is not "completely irrelevant to the accuracy...".
Either way, taking a poll from ENWorld and applying it to D&D Fanbase as a whole is worthless.