|Falsification: experimental or logical refutation of a general or a specific statement. Depending on the nature of the statement, there are differences in terms of whether a single counterexample is sufficient for a falsification, or a certain ratio of positive and negative cases is crucial. See also verification, verificationism, confirmation, Bayesianism, probability, hypotheses, theories.|
Books on Amazon
Verification/Confirmation/Examination/Duhem: If the announced fact does not arise, the theorem is falsified. In the examination one applies a whole group of theories (according to which the instruments are built and without which they cannot be read). The occurrence or non-occurrence of the phenomenon does not result from the contentious theorem alone, but from the connection with the whole group. The failing experiment merely teaches that among all the theorems which have served to predict or to state the phenomenon, at least one must be false. If the experimenter declares that the error lies precisely in the proposition to be tested, he presupposes that all others are true.
Confidence in the other sentences (for example, according to which the instruments are constructed and according to which they are read) does not occur with logical necessity.
Physics is not a machine that can be dismantled. It is a system when a disturbance occurs, it has indeed been evoked by the whole system. (> System). The physicist must find the organ without being able to isolate it, because then the system does no longer work.
Ziel und Struktur der physikalischen Theorien Hamburg 1998