Explain why demonstrating that the null hypothesis is false does not make the opposite true

Question

What will be an ideal response

Answer 1

Normally, science proceeds by ruling things out. The scientific process incrementally refines our understanding of the universe. By disconfirming the null hypothesis ("A is not related to B.") and by assuming that every other theory, measurement, contextual variable, and precondition is true and accurate, we infer that there is some relationship between A and B. The next step, normally, is to refine our understandings of A and/or B so that we can make another null hypothesis involving the refined concept(s) and see if we can then experimentally disconfirm that one. But look at all the things scientists must assume are just fine: data generation and data gathering conditions, theories about what might possibly be relevant to observe, various measurement strategies, the quality of the measurement instrumentation, and the statistical procedures for handling and for analyzing data once they are collected. Scientists do trust those things. But they also know that each one of them was itself the object of empirical investigation at one time or another. Because they understand all these factors must be assumed to be true in order to test a given hypothesis, careful scientists avoid saying "X is certainly true." They might say, "X is true to a scientific certainty," but that means that we should hold an extremely high level of confidence in X. There should be no reasonable doubt about X. But, as scientists using empirical reasoning we are not justified in saying "Ah look, we disconfirmed the null hypothesis. So its opposite must be true."