The logic and computational details of this procedure are described in Chapter 9 of Concepts and Applications. |
The difference between the means of two samples, A and B, both randomly drawn from the same normally distributed source population, belongs to a normally distributed sampling distribution whose overall mean is equal to zero and whose standard deviation ("standard error") is equal to
square.root[(sd^{2}/n_{a}) + (sd^{2}/n_{b})]where
sd^{2} = the variance of the source population (i.e., the square of the standard deviation);-1sd | mean | +1sd | ||
<== | source population | |||
<== | sampling distribution | |||
sd of source population | ||
Home | Click this link only if you did not arrive here via the VassarStats main page. |