What are the common problems in logistic regression application?

What are the common problems in logistic regression application? in view of this problem, this article introduces the corresponding analysis and answers in detail, hoping to help more partners who want to solve this problem to find a more simple and feasible method.

In medical research, especially in epidemiological research, it is common to analyze the quantitative relationship between disease (outcome) and multiple factors (exposure). When the outcome is classified (two-classification, multi-classification) data, in order to study the quantitative influence of multiple factors and their interaction on the outcome, Logistic regression (Logistic regression) analysis can be used. Logistic regression is a multivariate analysis method of probabilistic nonlinear regression. With the development of computer technology, more and more Lo-gistic regression has been applied to medical research, and there are often problems of misuse and improper interpretation of results. This paper mainly discusses the common problems of modeling and result interpretation in the application of Logistic regression.

Binary logistic regression analysis

The significance of the ratio of advantages

Note that there are the following misunderstandings in the application and interpretation of OR

1) it is considered that OR > 1 represents a risk factor, OR1 represents a protective factor that promotes cure and survival, while OR1 represents a risk factor, OR