When using spss for multi-scale correspondence analysis, the following results are obtained. What does this chart mean?

It can be considered that the greater the correlation, but I think this is flawed. First of all, dimensions 1 and 2 can be understood as two principal factors obtained by principal component analysis, so the meanings of these two dimensions need to be explained with reference to the dimension score, which is the abscissa and ordinate of the above two-dimensional image. If the coordinate value (point in the figure) is a variable, the farther away from the origin, it means that the corresponding principal component is more influenced by this variable, or the variance of the principal component is more influenced by the variance of this variable. The variance in statistics can be understood as the content of information, so this dimension contains more information about this variable. Of course, the closer the points in the image are, the stronger their correlation is.

Let me explain your question 1: Why do I think the smaller the included angle, the greater the correlation? This explanation is flawed. For an extreme example, if there is a variable at the starting point of all rays, then its angle with all variables is 0. Obviously, this variable cannot have a strong correlation with all variables.

Question 2: If we have to explain the meaning of included angle, we can only explain it with the explanatory dimension 2. For example, q 1, 4 and q 1,1are very close together, and the included angle is very small, which shows that these two variables have a strong correlation. If dimension 2 makes economic sense, then the included angle is small, indicating that their contribution to dimension 2 is not much different.

Question 3: If you ask how to explain dimensions 1 and 2, you must give the principal component specific economic significance through the results of principal component analysis. For example, the absolute values of the coefficients of variables x 1 and x2 in the 1 dimension are relatively large, indicating that the1dimension mainly explains the information of X 1 and x2. If x 1 is the housing area and x2 is the residential property fee, then the dimension 1 mainly measures the housing level of the residential area, so it can be given such economic significance. After giving different dimensions certain economic meanings, it can be explained.

Limited to my level, these are the main explanations I can give you at present, and I hope I can report them to you. If you still don't understand, you can find a multivariate statistical textbook and look at the three chapters of principal component analysis, factor analysis and correspondence analysis, which will certainly inspire you. . Thank you.