I have a data base from 2011.

I have a 3 fixed factors model as this:

Y=m + Zo + Si(Zo) + Depth*Si(Zo) + e

Where:

m: mean

Zo: Zones in a coral reef (5 of them)

Si: Sites contained (nested) in the Zones (6 of them)

Depth: As it self. There are 2 levels for this factor -Shallow and Deep-.

3 sites of a Zone were supposed to be Shallow and the another 3 Deep.

But now I figured out that only 2 zones are balanced (3 deep sites and 3 shallow sites), and the remaining 3 zones are like this:

4 shallow and 2 deep

2 shallow and 4 deep and finally

0 deep and 6 shallow.

Finally (Thank you very much if you arrived here) the question is:

Which new hypothesis can I generate?

What would be the new experimental design?

If you think more information is needed let me know at 

[email protected]