HinduKush- <br /> Been too long for me to recall 1way repeated measures ANOVA. Wondering if differing drop-outs between baseline and 48 weeks for the EPA and "Control" groups can undermine the stat findings. The N for "Control" is 72 at baseline, 69 at week 12, 66 at week 24, 63 at week 48. While for EPA the Ns are 69, 63, 59 and 55 respectively. <br /> <br /> Do you (or anyone) know if the 1way repeated measure ANOVA looks at all 4 periods or just baseline and week 48? Do the drop-outs invalidate this approach? Would the p value be the same if only subjects present at week 48 were compared to those same patients at baseline?