Regression analysis: study the influence relationship, whether there is any influence relationship, and how it is.

Correlation analysis is to study whether there is a relationship, and regression analysis is to study the influence relationship. Obviously, correlation analysis is the basis, and then regression analysis. First of all, we need to know whether there is any correlation; Only when there is correlation can regression affect the relationship; If there is no correlation, there should be no regression effect.

Therefore, from the analysis point of view, correlation analysis should be carried out first, and then regression analysis should be carried out after the correlation analysis is completed.

Sometimes there are strange phenomena, such as:

There is regression effect, but there is no correlation. At this time, it is suggested to conclude with' irrelevance'

Negative influence, but positive correlation. At this time, it is suggested to draw a conclusion that there is correlation but no regression influence.

Correlation and regression analysis of SPSSAU