This is my first forray into a study of trees. I’m learning and absorbing a lot, but also starting to find myself asking questions and wanting to experiment.
My question is:
When setting up a ‘test’ to study some hypothesis for a certain aspect of tree development. How do you analyze the results without hurting the tree? How do you try to attain some consistency between trees?
even if you start with two trees, they will still have slightly different starting points
→ the root/canopy can have subtle differences that might distort your interpreted results
How to approach it scientifically, so that you don’t make incorrect conclusions?
The less control you have over the experiment the more samples you need. For living organisms that will not be sacrificed at the end of the experiment it may require 10, 20, 50, etc. samples for each experimental condition. The ability to statistically differentiate between treatments scales with the square root of the number of samples in each trial. You cite that two plants will have different starting conditions. If they are not grown from cuttings (clones) they will also have slightly different genetics. I imagine that some of the studies using seedlings start with 100s of plants in each trail group.
depends entirely on the research question. but to control for differences between trees you would generally randomise trees into a treatment group and a control group, quantify the differences between them, mass, age, height, etc and then show that the two groups are similar on average across these parameters.
Good point about the randomization. Sometimes you have to do the randomization several times in order to get groups that are statistically equivalent - both average and variance ideally.