I’m revisiting the IPAM data from the Chapter 2 stress-hardening experiment, and I’m having a really hard time wrapping my head around it.

I think I’ve finally come up with a workflow for using GLMMs to test significance of percentage/proportion (0<x<1 bound data): see here. I basically follow Kevin Wong’s GLM code.

My major confusions right now are:

1. When I calculate normalized CBASS Fv/Fm values based on the final Fv/Fm measurement at the end of the treatment period, I get some negative numbers and numbers > 1. There are too many of them to filter out, so I don’t know how to statistically test these.
2. I want to test the significance of the dose-response curves (DRC) for the CBASS values, but instead of ED50 I want to test the slopes of the curves to see if those are significant. In the drc model code for R, you can specify parameters (i.e. ll.3 = “hill”, “max”, “ED50”) and I believe “hill” is the “hill slope”. I can extract those values from the model, but idk what they mean and how to statistically test them.