4 ms·

You can, but why trash information that is present when you can leverage it with a different approach?

• MichailP 8y ago
  Like PCA? But that way you loose physical meaning of the predictors.
  • closed 8y ago
    PCA is a special case of factor analysis, so you are representing them as observations of a latent variable (which is often a narrative people use when explaining why two x variables are correlated)