Hi guys,

I have a question that is not regarding to Stata particually but to statistics in general. So I have a dataset where I want to regress the ln(wage) on different education levels and in class we wanted to find out wether the wage of bachelors is significantly lower than for those with postgraduate degrees and we did the command "test" for this. I think that means it is an F-Test and our lecturer said that we can not use a t-test as bachelor is not independent from postgraduate or the other way around and the t-test assumes independency. But why is this not given? Because people with the highest degree of postgrad do not have the highest degree bachelor so these two are different samples. Also if I would use bachelor as a reference category then the t-test would say if these are sig. different or not?.
I hope you can help me, when do I use f oder ttest.

Best regards,
John