Falsifiability

Falsifiability is the property of a claim that allows some possible observation to count against it. It distinguishes testable claims from those compatible with every possible outcome.

Karl Popper: A philosopher of science who made falsifiability a central criterion for distinguishing empirical science from non-science. He framed scientific theories as claims that must risk conflict with possible observations.
Empirical testability: The capacity of a claim to be assessed through observations or experiments. It is the practical condition that lets observations bear on a claim's truth.
Verificationism: The view that a statement's meaning or significance depends on how it can be empirically verified. It emphasizes possible confirmation, whereas falsifiability emphasizes possible refutation.
Psychoanalysis: A family of theories and therapeutic practices founded by Sigmund Freud that emphasize unconscious mental processes. Popper used some psychoanalytic explanations as examples he considered compatible with too many possible outcomes.
Imre Lakatos: A philosopher of science who assessed research programmes by their development of novel predictions and explanatory progress. He shifted attention from isolated theories to whether a sequence of theories makes progress or merely patches failures.
Scientific prediction: A statement about an observable outcome derived from a scientific model or theory. A theory becomes vulnerable when it predicts an outcome that could fail to occur.
Logical possibility: The property of being consistent with the rules of logic and therefore not self-contradictory. A falsifiable claim must have logically possible outcomes that would contradict it.
Confirmation theory: The study of how evidence supports hypotheses and theories. Evidence can support a theory without making it falsifiable or conclusively confirming it.
General relativity: Einstein's theory of gravitation, in which gravity arises from the curvature of spacetime. Its prediction of starlight bending near the Sun became a celebrated testable consequence.
Bayesian epistemology: An approach to reasoning that represents uncertainty with probabilities and updates beliefs in light of evidence. It models degrees of evidential support rather than treating tests as simple refutations.

Linked from 57 pages