Post hoc test

I want to run a post hoc test along with my MANOVA, but I don't know which post hoc test to run. I am looking at the difference in prey preference (soft vs. hard) between two different color morphs of the same species of salamander, over four seasons. This is data that I plan to publish and use for my thesis and from my understanding scientific journals do not like it when you use LSD. I think I would use Tukey, but I don't know why. I know that I won't use Bonferroni cause it is to conservative. Help!:confused:


TS Contributor
LSD is simply the t-test so its not the best alternative, unless you want to divide the
a=0.05 by the number of comparisons you will conduct and use that new alpha as significancve level. Now, Bonferroni dos that correction directly. Tukey is good because it is of the same family as Bonferroni's, use it, it is family errorwise, means it keeps the overall a at 0.05 (like bonferroni correction does). If it is not clear tell me.