I have collected a nested data measuring cell growth which has two levels (3 patients with disease A, another 3 patients with disease B, another 3 patients with disease C). For each patient, cell growth was measured repeatedly to minimise random error in the experiment.
I would like to access whether cell growth differs by disease groups.
However, considering the small sample size and non-normality of the data, I think that using mixed model may not be appropriate: Is there nonparametric alternative of mixed model where I may analyse non-normally distributed data?
