View source for Statistical Significance for MDD

This formula helps you calculate the statistical significance of the difference between two sample groups. 
<pre>Mc +n.Ec = Mv - n.Ev
n(Ec+Ev)=(Mv - Mc)
n=(Mv - Mc)/(Ec+Ev)
where
M is the mean
E is the std error of the mean (ie the std dev/sqrt(sample size))
c is for the control
v is for the variant
n is the number of std error from the mean between the control and the variant.
</pre>

You should then calculate the area under the Gaussian curve for -infinity to n, which is 0.5 + area between 0 to n which is erf(n)