The reduced correlation matrix and its eigenvalues

In PCA, the starting point for analysis is the original correlation matrix of the variables in the analysis. In FA, the starting point for the analysis is the reduced correlation matrix, which consists of the correlations of the measures off the main diagonal, and communalities on the main diagonal.

The fundamental equation underlying factor analysis is as follows:

original_correlation_matrix= reduced_correlation_matrix + uniquenesses

SPSS Statistics FACTOR does not report the reduced correlation matrix, but you can obtain it via the SPSS code. In truth, the communalities are not known at the start of the analysis, nor is the uniqueness. For initial estimates of the communalities, we can ...

Get Data Analysis with IBM SPSS Statistics now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.