I determined the medians of a biomarker across 2 categories: I.e
- based on 4 study groups (ie. non-pregnant, pregnant, PE1, PE2 and ranked as 0, 1, 2, 3),
- and also by categorising BMI (eg: normal, overweight, obese, etc into 0, 1, 2 &3).
Can I do a dunns test across these 2 categories? Since I have medians, I need to determine the p values