Specify and estimate a factor analysis model.
Syntax
factest(method=arg, options) x1 [x2 x3...] [@partial z1 z2 z3...]
factest(method=arg, options) matrix_name [[obs] [conditioning]] [@ name1 name2 name3...]
where:
method=arg (default= “ml”) | Factor estimation method: “ml” (maximum likelihood), “gls” (generalized least squares), “ipf” (iterated principal factors), “pace” (non-iterative partitioned covariance matrix estimation), “pf” (principal factors), “uls” (unweighted least squares) |
and the available options are specific to the factor estimation method (see
“Factor Methods”).
The factest command allows you to estimated a factor analysis model without first declaring a factor object and then applying an estimation method. It provides a convenient method of interactively estimating transitory specifications that are not to be named and saved with the workfile.
Estimation of a factor analysis specification using
factest only differs from estimation using a named factor and a factor estimation procedure (e.g.,
Factor::ipf) in the use of the “method=” option and in the fact that the command results in an unnamed factor object.
Examples
The command:
factest(method=gls) g1
estimates a factor analysis model for the series in G1 using GLS. The result is an unnamed factor object. (Almost) equivalently, we may declaring and estimate the factor analysis object using the
Factor::gls estimation method procedure
factor f1.gls g1
which differs only in the fact that the former yields an unnamed factor object and the latter saves the object F1 in the workfile.
The command:
factest(method=ml) group01 @partial ser1 ser2
estimates the factor model using the partial correlation for the series in GROUP01, conditional on the series SER1 and SER2. The command is equivalent to:
factor f2.ml group01 @partial ser1 ser2
except the latter names the factor object F2.
Cross-references
See
“Factor Analysis” for a general discussion of factor analysis. The various estimation methods are described in
“Estimation Methods”.