Why does stress make more empathetic people less kind, but has little effect on others? We believe this is due to a set of linked phenomena. The first is publis

Why does stress make more empathetic people less kind, but has little effect on others? We believe this is due to a set of linked phenomena. The first is publish or perish. When a researcher needs a publication, his or her cortisol levels increase. This reduces the researcher's ability to empathize with the victims of useless scientific results, or their own future self when their work turns out to be a dead end. A second cause is the garden of forking paths. When a main effect fails to reach significance, it is all too easy to find an interaction or sample subset where it is significant. For example, if stress didn't affect charitable donations, you can always check the result among high-empathy people, or women, or nursing students. These root causes come together in p hacking, where desperate researchers kid themselves that they always meant to run that particular interaction, and absurdly pretend that their p values and careers have some meaning... [Not TFA, obviously, but I wouldn't be surprised.]