User:Jclaer/Can factor analysis yield correlated personality traits, and why do some people think it can?

Factor analysis singular-value theory algorithms yield orthogonal vector pairs. Personality theorists use factor analysis but report "the personality traits are correlated". My years of experience applying inverse theory in Geophysics suggests the personality researchers may be using factor analysis incorrectly.

Not having used singular value theory myself I plan to prepare a short tutorial of its use including a clear statement of what orthogonality results must be expected.

It's too early for me to guess where they are having difficulty, but if it's like my experience in Inverse theory, people may use it without adequate attention to important details like weights and regularization.

Let us hope we can all come to understand two things:

1. If on a personality test one question is included twice, and everyone answers it consistently, then we have learned nothing and the extra inclusion should not affect the analysis.
2. If someone answers the repeated question inconsistently, then we have learned something, but what?

Definitions

Shall I do this all in Latex and import it later? Lets test what happens to an in-line equation $0=Lm-d$, see how much effort it is to change that to $0=Lm-d$. Exactly what I feared. The math button doesn't strip the dollar signs. We get $\$0=Lm-d\$$