Hi,
I am working on analyzing a microbiome dataset that has two factors - Sex and Treatment. Microbiome data is non-parametric, and transformation is not recommended (since important information can be lost), so as a default, I always use the Kruskal-Wallis test, Mann-Whitney test, or Spearman Correlation. In order to evaluate interaction, the only test I can perform is a two-way ANOVA, which is meant for parametric data, so this test cannot be used for the microbiome dataset. In this regard:
- Is Kruskal Wallis an acceptable test to compare 4 groups (Male_Control, Female_Control, Male_Treatment, and Female_Treatment)?
- Is there an alternative test that can be used instead of a 2 Way ANOVA that can test the interaction between sex and treatment?
Your suggestions would be very helpful for proceeding with the analysis.
Thank you,
Best,
R