I'm sure I read an article or book chapter that went through the astounding number of confounding variables experimenters discovered when running rat experiments.

Things like what they could smell, what they could hear, navigating by lights in the roof, individuals interacting with them, them learning 'tells' from the way the experiment was run that leaked info.

I assume there's some textbook on this for rat experiments.

edit not the one I remember but this is an overview:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8073298/

Things like whether the mouse had more sisters than brothers can have an impact.