2 ms·
I'm sure I read an article or book chapter that went through the astounding number of confounding variables experimenters discovered when running rat experiment
by ZeroGravitas 4y ago
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/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8073298/
Things like whether the mouse had more sisters than brothers can have an impact.
- lamename 4y agoAnd it's not just rats. Anything with multiple senses, a circadian rhythm, developmental variables, individual genetic differences can be subject to this variability. These are all factors that people should be taking into account, and yet unfortunately they do not. Some of it is practicality, it's incredibly challenging and expensive to control for everything, so you randomize certain factors instead. Confounding effects can sometimes be small enough to be washed out anyway. Nonetheless, it is known by psychophysicists (people who do their best to measure behavior rigorously) that all of these factors are important. Good behavioral research is incredibly challenging to get right, the best people know this, but most published behavioral research is quite sloppy.