Hi,
I just registered to find some answers to probably very simple questions.. I hope those are not answered too often in the forum, but my computer gives me an error when I try to use the search function (maybe due to corporate internet restrictions).

Starting situation 1:

4 characteristics (A-D) shall be tested on 9 samples (1-9).
Each of the 9 samples represents a group category within a larger category (like: tree species -- forest).

It is logical, that characteristic A,B and C could correlate with characteristic D.

Questions:
Q1-1. How can I tell if my sample is normally distributed (I have picked the samples; figuratively: samples where taken not from the forest but each from a different tree species)? Do I have to use Spearman´s correlation because of that (-->I want also to determine the significance of the correlation). How important is the small "n" for any significance consideration?

Q1-2. Is any regression analysis not possible due to the fact that I have no independet variable? Or can I just call e.g. characteristic D my independent variable?

Q1-3. What else do I have to take into consideration?




Situation 2:

Characteristic D is tested in 5 different laboratories on 5 of the above samples.

Questions:

Q2-1. The goal ist to make a more or less robust statement about the reproducibility. Which method shall I use? Analysis of variance? Inter-rater reliability? Interclass correlation with a transformed D variable (ordinal)?

Q2-2. Is a statistical robust statement possible with only that amount of laboratories and samples? How many would you suggest?

Q2-3. Can I somehow use the fact that characteristic D is the result of a test where 20 test executions are use to generate a mean value (= characteristic D)

I hope you can help me with at least some of my questions.

Thank you in advance very very much!!!