I am conducting a two-way mixed ANOVA on intervention data from a randomized controlled trial (i.e., my within-subject factor is time, and my between-subject factor is treatment condition). My dependent variable is a standardized measure of reading that I collected at pre- and post.
At time 1 (pretest), my data was positively skewed, so I performed a sqrt transformation on that variable. However, my data at time 2 (posttest) is normally distributed-- do I have to perform the transformation on that variable as well before conducting my analyses?
I appreciate any help I can get!