Posted by cccp1014 on 3/14/2018 1:04:00 PM (view original):
Posted by bad_luck on 3/14/2018 12:38:00 PM (view original):
Posted by cccp1014 on 3/14/2018 12:10:00 PM (view original):
Posted by tangplay on 3/13/2018 11:29:00 PM (view original):
Posted by cccp1014 on 3/13/2018 10:41:00 PM (view original):
Posted by wylie715 on 3/13/2018 7:13:00 PM (view original):
Posted by cccp1014 on 3/13/2018 12:55:00 PM (view original):
About 100 give or take.
dude, in a country of over 300 million 100 people is not statistically significant.
Actually mathematically it is.
https://www.surveysystem.com/sscalc.htm#one
Actually no it isn't. You need, at the very least, 300 people to be 'significant' and even if you do, you need to cut corners.
Nope
You just need 32 persons for it to be somewhat statistically significant. Now if I want to start playing with standard deviations I need to speak to more. But 100 is a good enough sample. This is not 15 people. This is 100 more or less. If you want to extrapolate they then speak about their uncles, fathers, brothers, friends and neighbors who voted the same way for those same reasons. That equates to more or less 500+ people.
You’re literally just making **** up. There is no set number that makes something statistically significant.
Nope. Take a stats class. Actually you don't have any class. Never mind.
Um, I have taken stats classes. In this case, both 32 and 100 are way too small to rely on the result.
For example, if you have a population of 50,000 and you want a 2% margin of error and a 99% confidence level, you need over 3800 samples. Now, that's a very small margin and a high confidence level. But even if you reduced those to 5% and 95%, you still need almost 400 samples. And that's at a 50,000 person population.