Pairwise comparisons after RM ANOVA
Hi,
JASP uses the pooled error term from the ANOVA. SPSS does not. I thought that the prevailing view was that in RM designs it's better not to use the pooled error term. Could you please explain?
Thanks.
Glenn Schellenberg
Comments
Hi Glenn,
I'll forward this to the team member responsible for the classical ANOVA. I recall there was a discussion, but perhaps I misremember.
E.J.
Thanks. As far as I can tell, this isn't considered to be correct by anybody. Even SPSS doesn't do this. GS