2 sample t test is used when the iv is nominal with 2 levels dv is interval or ratio (must be able to compute mean). The null hypothesis is that the group means are =: the iv and dv are not related. The alternative hypothesis is that the 2 group means are different: iv and dv are related. Cases have been independently + randomly sampled from the 2 pop. The 2 population variances are=: assumption of the homogeneity of variance. Robustness: violations of the last 2 assumptions do not affect if: Sample sized are similar, less than 1. 5x the # in 1 group as the other. : w/ small samples or rlly unequal gr. sized, violation of assumptions should be tested. Pooled variance formula: to test the diff. btwn means for 2 unrelated gr. ****if f from levene(cid:495)s test is significant don(cid:495)t use. Separate variance formula differences in variance use svf.