Two-way mixed repeated measures MANOVA in Python or R

Question

I compare three groups:

• A control group
• Two different clinical populations

They are subjected to 6 different types of stimuli, several times each (repeated measures), and 5 dependent variables are collected on each stimulus exposition.

Therefore, I have two independent variables (IV), being the group (between-subject) and the stimulus type (within-subject): this is a mixed design approach. I also have 5 dependent variables (DV): the analysis will be multivariate.

I would like to evaluate the interaction "group" x "stimulus type", and the difference in the groups' performance in general, but also with respect to the different stimulus types (two-way analysis).

This brings me to a two-way mixed repeated measures MANOVA.

I would like to perform this analysis in Python (preferred) or in R, but I might be open to open-source solutions. Can you help me identify:

1. Is this analysis feasible in either Python or R?
2. If so, which package (and function) would be appropriate for this analysis?

I already identified:

• Python's statsmodels.multivariate.manova, but comments at the top of the MANOVA.mv_test() method lead me to think this is not for two-way analyses.

• R's Stats package comes with a manova() method, but I am having a hard time figuring the assumptions.

• R's MANOVA.RM package, with the multRM() method, mentioned in this question, but I'm not convinced it's appropriate for mixed design yet.

Answer 1

R's MANOVA.RM package provides indeed the multRM() method. The MANOVA.RM reference manual actually mentions between and within-subject variables in page 13, which comforts the fact that the package is indeed appropriate for mixed design.

Python's rpy2 should allow the use of Python and minimizing the use of R.