Hi there.

I'm not a statistician, so just doing my best to understand and describe my challenges here, so I hope you'll bear with me. I'm currently working on a project in mice where I have 3 chronic treatments/factors with two levels to each (treatment or not). This of course makes it a 2x2x2, which means that most of my data analyses can be performed by 3way ANOVA. I do, however, have a a few sets of data where there is an additional dimension of either time or a separate acute treatment. In these cases I would like to perform repeated measures ANOVA since I have multiple measurements for the same mouse. This is not possible in Sigmaplot because it doesn't support that many factors, so I've turned to JMP, but run into some unforeseen challenges:

I have an experiment where two of my chronic treatments (between-subject factors) interact when looking at the "sum" analysis - so ignoring the distinction between acutely treated/untreated. There is also an overall effect of the acute treatment (the within-subject factor), but it doesn't interact with any between-subject factor.

I'm unsure about how to proceed with the analysis. Does it make sense to report the interaction of two between-subject factors from the sum-analysis, when it basically means that each mouse is used twice in the same analysis? - doesn't seem right to me. But when I don't have an interaction with the within-subject factor I also don't think there is basis for testing the acutely treated/untreated separately. The easiest way would of course be to use the RM ANOVA only to look at the contrasts portion of the analysis and then follow up with two 3way ANOVAs - one for each setting acutely treated>