I found a significant four-way interaction in my mixed model 5-ways ANOVA (analysis of ERP data, 2*2*2*3*2).
My variables are:
- Within-subject factors: magnitude of outcome (big, small), likelihood
of outcome(high, low), valence of outcome (gain, loss), electrode
(Fz, Cz, Pz)
- Between-subject factors: group (group 1, group 2)
Interaction is magnitude * likelihood * valence * electrode
I will try to disentangle this interaction by analysing factor levels of my data set separately, similar to https://stats.stackexchange.com/a/73064/132815
I have two questions regarding this process:
I wonder, whether there is need to correct for multiple comparisons, if I do more than 2 ANOVAs? For example, Bonferonni or FDR correction?
Naïve question: is there any special name for this kind of procedure?
Thanks in advance!