I am interested to know the minimum sample size to apply Friedman analysis. May I apply this to 5 participants on 4 levels of treatments (Pre, mid, post and follow up)?
I think you can apply the Friedman analysis with only 5 participants. But, the less participants the less likely to have significant results (even if there are true differences between levels of treatments). The 'aligned' Friedman test has somewhat more power. In attachment an R script for such an aligned Friedman test with post hoc comparisons.