Apart from the KMO and Bartlett’s test table, the most important output while running the factor analysis test in SPSS is the rotated component matrix table. This table shows the classification of the variables into different components as per their factor loadings. The factor loadings will depend on how similar the entries of a variable in a dataset are. Therefore, variables with similar data will be grouped together in a single component, showing similar factor loadings.
For example, in a dataset of 100 consumers of a shampoo brand where the values of 20 statements are pertaining to 5 variables which are:
- smell and,
- texture of the shampoo.
It is probable that the factor loadings of all statements under ‘feel’ will be similar, therefore they can be grouped under one component, ‘feel’. Take a look at the below image. It shows the statements with similar factor loadings grouped into one component.
Extraction method: Principal component analysis.
Rotation method: Varimax with Kaiser Normalization.
a. Rotation converged in 7 iterations.
Now, to group statements EM1, EM2, EM3, EM4, EM5, EM6, EM10, EM11, and EM12 together instead of 3 different components, then this is what should be done:
- Open the ‘Data View’ option of the dataset in SPSS.
- Duplicate some entries (at least 35%) from EM1 to the rest of them.
- Run the factor analysis test again.
The statements EM1, EM2, EM3, EM4, EM5, EM6, EM10, EM11, and EM12 will now be grouped together in one component in the rotated component matrix table.