What is the benefit of a two-sample t-test (i.e., independent and dependent)? In other words, under what circumstances would a researcher chose to design an experiment using two samples?
If you were given the opportunity to design you own experiment which of the inferential statistics we've discuss so far would you chose and why?