Suppose I want to find out the number of herbivores in different kinds of trees, and also compare the numbers between day and night. How would this kind of study be run in practice? It easy to think of several possible confounding factors, such as the health of the trees, weather, etc.
If the trees are big, it is also difficult to find every living thing from that. One could of course then count the number of herbivores per defined area but again, picking the area should be carefully controlled.
From a statistical perspective, I can just make an estimate of expected effect size, and calculate the sample size that is about enough for detecting possible differences in the above mentioned conditions. However, I find these other aspects very difficult to do... All help would be appreciated :)