If a researcher were studying the effects of a teaching method on patient learning outcomes, how does the research question need to be worded to use a t-test to test for statistical differences? What type of data needs to be collected and why?