SPSS data Analysis-factor analysis

We know that the principal component analysis is a dimensionality reduction method, but it is essentially a matrix transformation process, the extracted principal component does not all have the actual meaning, and this meaning is often what we need, the next factor analysis can solve the problem

Factor analysis can be regarded as the generalization of principal component analysis, and principal component analysis can be regarded as a method of extracting factors in factor analysis, which are very similar in many cases, but there are some differences, the most direct difference between the two is that the result of principal component analysis is unique, but factor analysis is not, so we can rotate by factor, The further simplification of the factor load makes each factor have definite practical meaning, which is also the key to evaluate the success of the factor analysis.

Principal component analysis does not require much data, and the conditions for factor analysis are:
1. Sample size should not be too small
Factor analysis is to consider the internal correlation structure between variables, so the sample size should be as sufficient as possible

2. There should be correlations between the variables
The Bartlett test is required, and if the variables are independent from each other, the factors cannot be extracted and factor analysis cannot be used

3. The partial correlation between variables should be stronger
Can be measured by the KMO value, the larger the value, the more biased correlation between the variables, the better the factor analysis effect

We still use the example of principal component analysis to do factor analysis to see the similarities and differences between the two.

Analysis-dimensionality Reduction-factor analysis
















In this, we get the final factor score, and save as a variable, you can further calculate the sum factor score, in order to make a comprehensive ranking of the city.

SPSS data Analysis-factor analysis