There are different kinds of scientific hypotheses. For the same problem, the development of thinking will be different because of the different facts, principles and ways of thinking. For example, different associations and analogies will lead to different guesses and assumptions, and different viewpoints will inevitably lead to different assumptions. Various hypotheses have to compete, and the same hypothesis itself will evolve. In the process of competition and evolution of hypotheses, people will constantly evaluate and choose various hypotheses. At this time, not only the internal factors of hypothesis, such as the simplicity of concept and the perfection of structure, but also various external factors, such as social and cultural background, ideological tradition and psychological habits, are playing a role. However, the test of practice plays a key role in the evaluation and selection of scientific hypotheses.

The statement that cannot be tested in principle has no scientific value, so it is not a scientific hypothesis. Testability in principle is a necessary condition of scientific hypothesis, and the most powerful support of scientific hypothesis is that the facts it predicts are confirmed by later practice. However, people's cognitive process is complex and tortuous, and the test process of hypothesis also presents complexity and stumbling. A successful forecast can't completely confirm this hypothesis, but it does prove or supplement its authenticity to some extent; Failure to predict does not necessarily overturn this assumption, because a hypothesis always needs to be combined with other preconditions (or auxiliary assumptions) to get a prediction. Even if the prediction fails completely, the problem may lie in the hypothesis itself or other conditions, and sometimes it is necessary to check the practice itself, such as experimental instruments, experimental operations and even calculation methods for errors. The history of science shows that the failed scientific hypothesis will "come back to life" with the passage of time under new conditions; Successful scientific hypotheses may be in trouble again, need to be improved, and even be replaced by new hypotheses.