There's a difference between hard science and soft science. The argument over psychology is whether it's a hard science (chemistry, physics) or a soft science (economics). Most people would argue it's a soft science. Statistical inferences can be drawn from psychological experiments that may or may not hold true for an individual as determined by probability. The core problem with psychology is that our models aren't great.

Physics as experienced by 99.99% of humanity is obnoxiously rigorous. Newtonian mechanics dictates the stupendous majority of the physical interactions in most peoples' lives, and the left over bits subject to relativity (GPS, electronics, lasers) are so surface-level relativistic that they're well within the sweet spot of theory. You have to get pretty deep into quantum mechanics before the effects of assumptions predominate.