Quote from shortie:
are you suggesting that my ~30 weekly polls are meaningless?
none got 100 votes so far, ~25 is typical
It depends on how you look at the Poll. Suppose you want to assess ET's opinion (which is larger in size than the readers of this thread). Then what determines the margin of error of your poll is the ratio of responds to the total population on ET, in addition to the size of respondents . So the first thing to do is to assess who the population is and its size, and work backwards. If r is the percentage of responds to size of population, and n is the number of respondents, the formula I would use is something like: ((r*(1-r))/n)^0.5, times 1.96 (for 95 percent margin). So you get a conclusion like Poll margin of error is 10% with a probability of 95%.
Notice that the poll test does not say anything about what is needed to draw individual conclusions. You still need a size like what you stated above to draw individual conclusions.
The interesting other part of your polls is that you can observe the arrival of responses, while typical polls are looked at at the end of polls. I think you understand this part because I read something from you that proves that you do.

The readers now have the proof why they should vote.
NOTE: I renew my call to readers of Shortie's threads to vote immediately. Help our friend shortie so he can help us. Please vote now. Do not delay it. a minimum of 100 voters are needed for this to be meaningful. [/B][/QUOTE]