It's sort of a conceptual question: one of the assumptions for 2-sample t confidence interval is equality of variances in two populations, and I need to come up with a concrete example of when this is likely to fail. I sort of have an idea that it really depends how one chooses samples. For example, if one compares mean hight of males and females on a certain campus where females are predominantly asian. Then since variance in hight of asian women is much smaller than variance in hight of white males, the assumption fails. ALthough I am not sure if this is a good example because I make an explicit assumption that asian females have small variance in height. Could someone help me out a bit with an example? Thanks in advance.