In research, the process of trying to invalidate a proposition. Causal relationships can never be proved; there may always be some as yet untested circumstance under which the relationship does not hold. With each successful test, the relation is said to have escaped disconfirmation.