HI everyone,

I am looking to analyse data for my final year project and have come across some stumbling blocks.

I am looking at metal concentrations at various locations and at various depths of the soil profile.

I was looking to run a two way ANOVA, however after running a Shapiro Wilk test my data is not normal on all my metal concentrations (thank you environmental data!).

I have read online that 2 way ANOVA is robust enough to deal with non normal data as long as it is approximately normal - what would that look like?

If not what other non parametric test would I run? is there an equivalent to a Krukal Wallis test?

I have 3 locations, with 3 or 4 varying depths at each location and have 3 results for each depth .

Many thanks for your time

Jay Fowler Holmes