I'm not sure that there's necessarily a problem with psychology itself, but just that the number of participants in studies is consistently too small. Just like

I'm not sure that there's necessarily a problem with psychology itself, but just that the number of participants in studies is consistently too small. Just like how physicists increasingly have to spend more money to get more accurate results (LIGO, LHC, redefinition of the kg, etc.), psychology needs to realize that 1000+ study participants is necessary to get valid results. Unlike physics, though, where it's easy to show that results are bad when the equipment is bad, the nature of psychology makes it easy to publish spurious results without enough support. When the psych community starts to demand higher N values, you expect fewer, lengthier, more costly, but more accurate results.