using a=0.05, test for mean 1 greater than mean 2.
group n mean std dev.
1 30 13.5 1.1
2 30 12.0 1.2
Calculate standard error (SE) for each mean.
SE = SD/sq. root of n -1
Get z score of difference between means divided by square root of sum of SE^2.
Look up value in bask of statistics text in table labeled something like "areas under the normal distribution."
I hope this helps. Thanks for asking.