How to reduce the factor loadings of a rotated component matrix?

By Priya Chetty & Guest on August 3, 2022

Sometimes, after you run the PCA test of Factor analysis in SPSS, the rotated components matrix shows the grouping of some statements into two components. This happens when a statement is represented by two components. In that case, you may want to reduce it to one component only.

EXAMPLE

In the image below, the statements CS6, CS7 and CS8 show factor loadings in two components; 3 and 8.

Case example showing the statements CS6, CS7 and CS8 showing factor loadings in two components
Figure 1: Case example showing the statements CS6, CS7 and CS8 showing factor loadings in two components

The factor loadings value of CS6, CS7 and CS8 in component 8 are between 0.5 and 0.6. To show factor loadings only in component 3, change the ‘absolute value below’ option in the factor analysis test to .600.

Changing the 'absolute value below' criteria in factor analysis (PCA) in SPSS
Figure 2: Changing the ‘absolute value below’ criteria in factor analysis (PCA) in SPSS

This will eliminate the statements with factor loadings values below 0.600. Therefore, now the table will show CS6, CS7 and CS8 only in component 3.

Reduced number of components for the statements CS6, CS7 and CS8 after changing the 'absolute value below' criteria
Figure 3: Reduced number of components for the statements CS6, CS7 and CS8 after changing the ‘absolute value below’ criteria

Discuss