how to perform a wilcoxon test on 10 groups ?
2 views (last 30 days)
i have 10 sets of data that i want to perform wilcoxon test on to find if there a significant difference between them, and chose the best one
but i found that it is not possible to use it on 10 sets of data and i have to use anova
which anova test do i use ? 1 one /2 way ? n way i don't understand the difference ? and what the result is
( i am familar with wilcoxon on two sets but don't know anything about anova )
Scott MacKenzie on 11 Sep 2022
The Wilcoxon test is used to compare two sets of within-subjects data (aka data from correlated samples). The equivalent test where there are > 2 sets is the Friedman test, which is available using MATLAB's friedman function. That's probably the test you should use.
Both the Wilcoxon and Friedmans tests are non-parametric tests. The anova is a parametric test. So, whether to use the Friedman test or an anova test is a matter of whether the data are considered parametric or non-paramteric.