the t test for the difference between the means of 2 independent populations when standard deviations are unknown assumes that the samples are randomly and independently drawn population variances are equal populations are approximately normal or sample sizes are greater than 30 all of the above