Viewing a single comment thread. View all comments

Pr0gr3s t1_iwnxqdq wrote

45 is actually a decent N for a RCT with a within subjects design. The design increases statistical power significantly over a between subjects design, which is likely what you are picturing.

23