Repeated measures ANOVA but non parametric & missing values

#1
I am trying to compare a matched cohort of repeated measures (continuous variables) - within each treatment arm with 5 time points comparing to baseline. The normality test is negative ie data is non parametric but I have some missing data in some time points. I think Friedman is the most appropriate test but I cannot do this as I have missing data. What is the best alternative? I am using GraphPad but have access to SPSS
 

hlsmith

Omega Contributor
#4
normality test on residuals? Multilevel models (values clustered in subjects) runs with missing data. Not sure if you will have statistical power concerns based on missingness?


How much data is missing?