In probability theory, this is called hypothesis testing. Generally speaking, it is a course only involved in universities, but now it may be difficult to show it to high school students.

Chi-square distribution is only a distribution, and there will be acceptance domain and rejection domain, which is different from the quantile you set, but it is only possible for AB, and it is not necessarily completely irrelevant. Because after all, it is a probability distribution, not a number.

The two concepts of independence and independence can be determined by covariance, that is, Cov(A, B)=0 means that they are independent of each other. But independence doesn't mean it doesn't matter Independence only means that the change of A will not cause the change of B, but that two things are completely different. But they don't necessarily affect each other.