Clarifying a Common Mistake in Using Hypothesis-test
Zhong Xiaobo
Abstract
Zhong Xiaobo
Abstract
Whether we will accept a hypothesis should be determined by the probability that the hypothesis is right.When we obtain the observed value X of a sample from an experiment,the probability that hypothesis H is right is P(H/X).Therefore,the qualification for receiving H and thereby the method for prescribing the receiving area for H should be P(H/X)≥ e(e is a large enough probability constant).The method is true not only for alternative hypothesis but also for null hypothesis,so it can be proved that the receiving area for null hypothesis is not the surplus set of the receiving area for alternative hypothesis.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Whether we will accept a hypothesis should be determined by the probability that the hypothesis is right.When we obtain the observed value X of a sample from an experiment,the probability that hypothesis H is right is P(H/X).Therefore,the qualification for receiving H and thereby the method for prescribing the receiving area for H should be P(H/X)≥ e(e is a large enough probability constant).The method is true not only for alternative hypothesis but also for null hypothesis,so it can be proved that the receiving area for null hypothesis is not the surplus set of the receiving area for alternative hypothesis.
Key concepts: Null hypothesis, Alternative hypothesis, Statistical hypothesis testing, Mistake, p-value, Null (SQL), Statistics, Mathematics