Popper made this argument in his book The Logic of Scientific Discovery (1934), which became one of the most influential works in the philosophy of science. In it, he proposed three criteria for a scientific hypothesis:
The hypothesis must be falsifiable: all possible observations must be capable of contradicting it. If no observation can ever contradict a theory, then it cannot be tested and so is not scientific.
The hypothesis must be testable: there should exist at least one possible experiment which could prove the theory wrong (if it were false). This experiment must not be too difficult to perform or else there would be no point in doing so because we would already know that the theory was true anyway!
The experiment must be repeatable: if someone else tries to repeat your experiment and gets different results from you then this does not necessarily mean that your result was wrong but could just mean that there was some error in the way you performed your experiment.
