10 Factor Analysis
Principal component analysis in Section 4 rewrote \(p\) variables as \(p\) uncorrelated components and then discarded the smaller ones. It is a transformation of the data, and it proposes no explanation of them.
Factor analysis proposes one. It supposes that the observed variables are correlated because they are driven by a small number of unobserved quantities, and it asks what those quantities would have to look like. The two methods are often confused, produce similar-looking output, and answer different questions.
10.1 The Orthogonal Factor Model
10.2 Non-uniqueness and Rotation
10.3 Estimation
10.4 Practice Problems
10.2 Non-uniqueness and Rotation
10.3 Estimation
10.4 Practice Problems
Questions on this section
Stuck on something here? Ask below and it stays attached to this topic.