I think they are both quite reliable when their conclusions are properly understood. Both physical and social scientists tend to be quite clear in their published papers about the degree to which their theory is supposed to be probabilistic, the amount of support they’ve gathered, and what the possible sources of error.
It’s when we ignore those qualifications and start treating the conclusions presented as absolute that the science may look unreliable. But the fact that one type of science may have more trouble controlling its sources of error does not make it less reliable so long as we understand the conclusions in light of that fact.
@YARNLADY “Human science” is another term for “social science” (e.g., psychology, sociology, anthropology).