Hi,
I am not super good with stat and I would need some help/advice.
I estimated the seasonal abundance of a marine species population for 3 winters and 3 summers using CMR modelling. During fieldwork, I also collected data on the salinity of the water of the entire study area using a continuous data logger. So, I calculated the mean and the median of the salinity water for each season, and I would like to test wether or not the salinity is affecting the abundance estimates. I think it is a very simple test to do, but not sure which one would be best? Correlation test (Pearson)? GLM? or what will you suggest?
Does anyone know if there is an or a website where you can enter the type of data you have, what kind of relationship you want to test and the app or the website tells you what test is best or appropriate to run? That tool would be super helpful.
Thanks a lot!
I'm rambling/nitpicking here:
I think it's pretty certain that salinity has an effect. I guess the question is how precisely you can estimate that effect given the data and working model. The fact that the hypothesis of "no effect" is compatible with the data (and model) does not mean the effect is actually zero and I guess biologically it is very unlikely to be zero.
As the name suggests, GLM is a very general class of models which includes among other things anova and Pearson correlation (which is the coefficient after you scale and center the two variables to correlate). It seems to me that some form of GLM is what you want. I can't/haven't thought more exactly what though...