My design is quite complicated - I need to analyze a 2 x 2 x 2 design (all independent variables)
but I don't have enough number of trials to have SOA for each condition.
Some conditions are missing a SOA for each run and some have only one value - what would be the best way to approach this issue?
I'm wondering if I should combine all three runs, or do ANOVA with beta values later on.
Thanks,
Catherine
|