# Should we use the same formula when using the T stat?

If we know the variance of both populations, but our sample size is small, we should use T instead of Z, but in that case, do we still use the same formula for the standard error? Usually T is paired with the sample variance S, not with sigma, but at the same time not using the population variance knowking it sounds like a waste, so what should we do?

0 answers ( 0 marked as helpful)