RM ANOVA - Estimated Marginal Means for Interaction
Hi,
I am trying to get the EMM from a 2 [within] x 2 [between] ANOVA. In the corresponding menu for RM ANOVA, there is "Marginal Means". Here, I can select both factors, but they are estimated separately/independently, see:
I would expect that there is the option to get EMMs for the interaction, i.e., to have two separate values (one per group) for each repeated measurement? I need that values (interaction EMMs + SEs and CIs) to create a custom plot elsewhere. Thank you!
EDIT: I have noticed that if I put two within-subject factors in the RM ANOVA, there is the option to insert WS-FACTOR1 * WS-FACTOR2 interaction in the Marginal Means menu.
Comments
I believe you've identified a limitation in JASP that tends not to be present in other stats programs.
R
Hi Martin,
You have indeed found a limitation of the software. It is something that the developers seem to be aware of since other users have requested this feature be added:
[Feature Request]: RM ANOVA - allow interactions between within/between factors in marginal means · Issue #2351 · jasp-stats/jasp-issues (github.com)
I have attended them to the fact that you seem to have noticed the same problem.
Oh, thank you. It seems, I am forced to use alternative free-soft for now but hope to be able to see this feature soon. JASP is just great!
Best regards,
M.
Hi @TarandeepKang and @MartinM ,
Yes, somehow this got removed at some point. I added the feature back, so it will available again in JASP 0.19 (probably released somewhere in April).
Cheers,
Johnny
Hi @JohnnyB , I did wonder if some previous versions of the software did have this functionality. :) I'm glad it was an easy fix/ repair.
Best,
Tarandeep