Not to criticise psychologists or psychology, but it does have one of the highest turnover rates of information of any science. Every several years, I believe, half of what was previously accepted is overturned. (I can't find my source for that, my google fu has failed me.)
It's to be expected though; humans are complicated. We can't dismiss an entire branch of science just because it's difficult.
Maybe we'll get a paradigm shift one of these days.
It's something around half, yeah, so far as I can recall, or at the least was a decade or two ago. That it's that
little is honestly kinda' impressive given how gorram young the field is, and how many different things influence it. It's not just a matter of humans being complicated, it's a matter of what that complication
means. Hard sciences should be called
easy mode sciences vis a vis experiment construction and execution, t'be frank; they're goddamn
child's play in comparison to how rough it is getting good rigorous data with any of the soft sciences. The hard sciences' experimental scopes are almost always comparatively miniscule, their variables few, their ethics barely a concern, their cross discipline influence relatively small and easy to isolate, their core precepts often fairly well established decades or centuries ago, and the time scales required to get
really good data from any one experiment generally functionally nonexistent.
None of that is true for bloody close to
anything the soft sciences do. It's amazing they're managing as well as they are given the several magnitudes more trouble involved in just
studying them, never mind coming to any conclusions as well.