st: non parametric test for multiple sites

["McCollum,Kelly" <Kelly.F.McCollum@drexel.edu>]

Greetings,

Is anyone aware of a non-parametric test that would allow me to compare the median values (or categorical values) from three groups, while accounting/correcting for repeated measures? I have 50 women from which 3 samples were collected on each (a total of 150 samples) and it is not appropriate to compare the means of each sample (i.e.-I cannot do the repeated measures ANOVA). I considered using the Kruskal Wallis test; however, I am concerned that this will not account for the repeated measures nature of the data. Any suggestions?

Sincerely,

Kelly