Redfish
Diamond Member
- Thread starter
- #61
A childs view of statistical sampling
Nowhere is an absolute 5% given for all samples. In fact, the percentage drops significantly as your sample size increases.
Are you actually that statistically ignorant to believe that a sample of the US population needs to be over 15 million to be "statistically relevant" ?
no
wrong, are you claiming that 5% is needed for a population of 20 but not for a population of 300 million?since 5% of 20 is 1, does that give you a valid sample? of course not, the bigger the sample the more accurate the result.
yes, the math says that a 5% sample is the minimum for a statistically relevant result. the % does not go down as the population goes up.
The final poll average, for the presidential election, is almost always accurate to within a few percentage points.
That average is composed of a sample of about 10,000 likely voters. What percent of the actual number of voters is that?
It sure as hell isn't 5%
the pollsters have been pretty good with their guesses recently. the question is whether that is due to accurate polling or is due to people being influenced by "polling results".
its a game, why is that so hard for you to grasp?
since 5% of 20 is 1, does that give you a valid sample? of course not, the bigger the sample the more accurate the result.