If the p value is greater than 0.05, the original hypothesis of satisfactory fitting should be accepted and the fitting should be considered satisfactory.

Only for HL verification, P is just greater than 0.05 and less than 0. 1. This is because you take 5% for A. If A takes 10%, P=0.074 will reject the original assumption.

Therefore, in actual combat, when "p value is between 0.05-0. 1", we should be especially careful not to simply "accept" the original hypothesis, and generally we should expand the sample size or do more experiments to verify it.

Reference Books: Logistic Regression Model-Methods and Applications (Wang Jichuan, Higher Education Press, 1st Edition, September 1)

Goodness of fit test: Yang Zhenhai, to be published by Science Press, with the publication time of 2011-03-01.