I have a mixed model that includes three different variables: Treatment, Stimulation, and DRG, as well as a random effect with DRG nested under a group ((1|Group/DRG)). I am looking to assess how each variable and their interactions affect Response Amplitude. Here is my model:
RespAmp ~ Treatment * Stim * DRG + (1|Group/DRG)
To note:
Treatment has two levels, indicated as 0 and 1 (
0 = control,1 = treated), and I needed to make these categorical.Stimulation has three levels, indicated as 1, 2, and 3, which were also made categorical.
DRG has four levels, indicated as 3, 4, 5, and 6, also made as categorical.
Response Amplitudes are continuous and made as numeric (see code below).
##Need to change characters into numeric
musc$RespAmp <- as.numeric(musc$RespAmp)
##Make sure numbers are categorical
musc$Treatment = as.factor(musc$Treatment)
musc$Stim = as.factor(musc$Stim)
#Use L6 as our 'reference DRG' --> change order of the levels and also convert to factor
musc$DRG <- factor(as.character(musc$DRG), levels = c("6", "5", "4", "3"))
Is it possible to create an estimation plot to visualize comparisons between Response Amplitudes from multiple variables?
For example, if I want to compare Response Amplitudes from Treatment = 0, Stim = 1, DRG = 3 to Treatment = 0, Stim = 3, DRG = 3 ....is it possible to visualize this as an estimation plot?
Or do I need to create a separate dataset with each combination of Treatment+Stim+DRG in order to compare another separate dataset combination of Treatment+Stim+DRG?
My understanding of estimation plots is that you are only visualizing two variables and a response, either as paired or unpaired. Essentially, you are visualizing the difference between means of the two variables, the effect size.
I was just hoping to get some clarification on how to best approach creating estimation plots for several variables, and perhaps my understanding of estimation plots and their use.
