Many psychological experiments involve shocking animals, playing loud sounds near them, depriving them of social contact, etc. For example, Martin Seligman's studies of learned helplessness in dogs are of this nature.
However, to my understanding, studies of learned helplessness in dogs at best provide an analogy to depression in humans, and it seems unlikely that actually shocking and trapping dogs was necessary for insight into human depression.
Similarly, Harry Harlow's studies of deprivation of maternal contact in monkeys seem to give extraordinarily obvious results, telling us what almost anyone already knows - that infants crave maternal contact. I don't think I need socially maladjusted monkeys in a laboratory in order to understand this point.
Is this wrong, or if not, what are some other counter-examples? What experiments involving inflicting physical or psychological pain on animals were necessary parts of important results in human psychology?