Question: In situations for which the parameter of interest is the difference in population means for independent samples, is it necessary for the two samples sizes n1 and n2 to be equal in order for the sampling distribution of the sample statistic to be approximately normal? If not, what condition is required for the sample sizes?