I built a linear mixed effects model to estimate the effects that different variables have on biomass. However, it tells me that it estimates them using a t-test, even if they are continuous. Therefore, I would like to ask you if the code is really doing the wrong thing, or if I'm just misunderstanding it.
To check whether patch size and connection had an effect on biomass, I built the following mixed effects model with lmerTest:
full_model = lmer(biomass ~ patch_size_ml * day + connection * day + day + (day | culture_ID), data = filtered_data, REML = FALSE, control = lmerControl(optimizer = "Nelder_Mead"))
where biomass, patch_size_ml, and, and culture_ID are continuous variables and connection is a categorical variable (connected vs isolated).
I then have gotten the statistics of the model using summary(full_model), which tells me:
Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's method [lmerModLmerTest]
Does this mean that it is uses a t-test to estimate the parameters of all the variables? That doesn't make much sense, as some are continuous. Also does it use a t-test to estimate the effects of interactions patch_size_ml : day and connection : day?
I looked the vignette of the package lmerTest but it doesn't include this information at all.