What is the role of Size Classes ? That is, you have an abundance value for each of 6 Size Classes at each Station ?

Does it make sense to just compare Size Class 1 at Station 1 to Size Class 1 at Station 2, and so on ?

Hi Sal,

fist of all thanks for your reply.

yes I am interested to check if there is any difference between size class 1 in station 1 and size class 1 in station 2, size class 2 in station 1 and size class 2 in station 2 and so on.

I had a look to possible test and i came up with Wilcox test which seems to be the most appropriate. Do you agree or you suggest other type of analysis?

thanks again

You have one observation for Size Class 1 for Station 1, and so on ?

If so, yes, you could use a Wilcoxon signed rank test, so that Station 1 is paired to Station 1, and so on.

But note that you will need to have one station with all six observations greater than the other station to get a p-value < 0.05 (for a two-sided test).

A paired t-test may also be an option. Or a sign test. Or various other tests for two paired groups.

Yes exactly, I have only one observation for each size class

Now i have a look to the other test you suggest just to check the requirement to use them.

thanks so much very helpful.