new function efa() for exploratory factor analysis

[Yves Rosseel]

Dear all,

The github version of lavaan (see
https://lavaan.ugent.be/development.html how to install) contains a new
function efa() which can be used for exploratory factor analysis.

To illustrate how it works, this is a typical example:

fit <- efa(data = HolzingerSwineford1939,
ov.names = paste("x", 1:9, sep = ""),
nfactors = 1:3,
rotation = "geomin",
rotation.args = list(geomin.epsilon = 0.01, rstarts = 1))
summary(fit, nd = 3L, cutoff = 0.2, dot.cutoff = 0.05)

The output is fairly traditional (similar to factanal()), although it
also gives an indication of the standard errors (of the standardized
factor loadings).

The function is a wrapper around lavaan(), and most lavaan options can
be used, included ordered=, missing=, and so on. The function also works
for twolevel data, but not for multiple-group data (as the latter
setting has a different focus and requires a different output style).

Comments and suggestions are welcome!

Yves.