Even clinical psychology seems to me a very dubious area. In many cases it is not linear to evaluate success when the main identification of the problem is base

Even clinical psychology seems to me a very dubious area. In many cases it is not linear to evaluate success when the main identification of the problem is based on necessarily limited, indirect and biased data (patient testimony) and the result of treatment is behavioural change. Who controls for potential unintended damaging consequences of behavioural change? Fixing one issue can't be the definition of success when dealing with such a complex system. The approach seems too simplistic. I have seen people coming out of "treatments" with their initial problems apparently fixed but with new perspectives on life that proved to be just as damaging. Some level of scepticism is probably a good thing.