If a researcher wants to compare the means of two different groups where the data points are paired (e.g., before and after measurements on the same subjects), which test is most suitable?