The CFA dialog is used to create and fit a confirmatory factor analysis model via the
cfa and sem functions in the sem package.
John Fox jfox@mcmaster.ca
Select two or more variables for each factor by Control-clicking on their names in the variable-list box. Optionally give the factor a name; this must be a valid R name. Then press the Define factor button.
Continue in this manner until all factors are specified.
Note that if there are not at least two unique variables selected for each factor, the
model will probably be underidentified, causing sem to fail.
The radio buttons at the top of the dialog may be used to analyze either the correlation matrix or covariance matrix of the observed variables; to specify either correlated or orthogonal factors; and to identify the model either by setting the factor variance to 1 or by setting the first loading for each factor to 1 (establishing a ``reference indicator'' for the factor). A check box is provided to compute robust standard errors and tests.
sem, cfa