Human subjects are crazy to work with for a few reasons
- People don’t follow instructions perfectly
- Research subjects often don’t take the research project very seriously.
- It’s not uncommon to have dropouts, thus you either have to find more subjects or have less data.
- It’s impossible to know what the subjects are doing to cause data variability (diet, vices, etc)
- You can’t lock subjects in a room and force them to eat and drink the same food every day.
- There’s a financial (time) penalty to many research studies that can get in the way of enthusiastic participation.
Laboratory mice literally live 5 to a cage with almost no diet variability, in a controlled environment. Yes shit does happen with research mice, but it’s something that is easy to control overall.