Quote (EndlessSky @ Dec 3 2016 04:53pm)
Standard deviation isn't useful. We need a p-value, or the chance that being an ILLEGAL SPIC is not correlated with VOTER FRAUD.
Code
SO... Mr ICE deported 1,000 immigrants and 100 of them committed voter fraud.
We are trying to prove that being an ILLEGAL SPIC is not correlated with VOTER FRAUD.
Lets set the rate that would be observed if non-correlation was true to a 2% rate of voter fraud, to be conservative (far higher than a control population would have, but will still prove our point if the null hypothesis is proven wrong).
The chance that a 10% rate of voter fraud would be seen in a sample of 10 when 2% is the real occurrence rate is exactly 19.6%.
Increase that to 1,000 sample size and the p-value comes to a freakishly infinitesimal number...
far greater than 6 sigma... 1.027E-28
BTFO
Correct me if I'm wrong, but you're trying to prove a negative? Shouldn't your post read:
Quote
We are trying to prove that being an GEOGRAPHICALLY AND LEGALLY CHALLENGED PERSON OF CHICANO ORIGIN is most definitely correlated with VOTER FRAUD
???
This post was edited by sylvannos on Dec 4 2016 02:47am
Guests cannot view or vote in polls. Please register or login.