Science

Something I find fascinating is that being consistent (and trustworthy) is less effective than being 80/20% consistent (classical vs operant conditioning) at training dogs where there are contextual/environmental cues at play. It's personally counter-intuitive, but I've seen it work and am convinced (I attribute it to evolutionary mechanisms, my goto in biology).

I'm wondering what other psychology as a science results have solid statistics behind them that I'm unaware of (I'm compsci with a physics/maths background, so it's probably most), and are interesting...