I have data. Soil samples were taken from meadows and adjacent arable fields (see picture):
There are always three samples per each opposing site. Since those 6 spatially conjoined sampling points are supposedly related, I need a nested design.
Now, I want to do a t-test. I.e., I would like to compare all samples beneath meadows and arable fields where the subgroups are the 3 conjoined points for each type of vegetation.
Since I cannot find a way I comprehend in the internet literature, can someone help me out here?