------------------------------------------------------------------------------- help forconfapostestimation author: Stas Kolenikov also see: confa, bollenstine -------------------------------------------------------------------------------

Postestimation tools for confaThe following commands are available after confa:

command description -------------------------------------------------------------------------

estat fitindicesfit indicesestat aicAICestat bicBICestat correlatecorrelations of factors and measurement errorspredictfactor scores -------------------------------------------------------------------------

Special interest postestimation commandsThese commands provide some additional post-estimation output.

estatcorrelatetransforms the covariance parameters into correlations for factor covariances and measurement error covariances. The delta method standard errors are given; for correlations close to plus or minus 1, the confidence intervals may extend beyond the range of admissible values. Additional options are allowed:level(#)specifies the CI levelboundprovides an alternative CI based on Fisher'sz-transform (arctanh) of the correlation coefficient. It guarantees that the end points of the interval are in (-1,1) range, which may not produce desirable results for Heywood cases.

estat aicandestat biccompute the Akaike and Schwarz Bayesian information criteria.

estatfitindicescomputes, prints, and saves intor()results a number of traditional fit indices. The following options ofestat fitindicesrequest specific indices:option fit index -------------------------------------------------------------------------

aicAIC, Akaike information criteriabicBIC, Schwarz Bayesian information criteriarmseaRMSEA, root mean squared error of approximationrmsrRMSR, root mean square residualtliTLI, Tucker-Lewis indexcfiCFI, comparative fit index_allall of the above indices, the default -------------------------------------------------------------------------

predict[type]newvarlist[if] [in] [,scoring_method]

predictcan be used to create factor scores followingconfa. The number of variables innewvarlistmust be the same as the number of factors in the model specification; all factors are predicted at once by the relevant matrix formula, anyway. The following methods are supported:option factor scoring method -------------------------------------------------------------------------

regressionregression, or empirical Bayes, scoreempiricalbayesalias forregressionebayesalias forregressionmleMLE, or Bartlett scorebartlettMLE, or Bartlett score, alias formle-------------------------------------------------------------------------

bollenstine,reps(#)saving(filename)confaoptions(...)bootstrap_options

bollenstineperforms Bollen and Stine (1992) bootstrap. The original data are rotated to conform to the fitted structure. By default,bollenstinere-estimates the model with rotated data, and uses the estimates as starting values in each bootstrap iterations. It also rejects samples where convergence was not achieved (implemented throughreject( e(converged) == 0)option supplied tobootstrap).

The following options are supported:

reps(#)specifies the number of bootstrap replications. The default is 200.

saving(filename)specifies the file where the simulation results (the parameter estimates and the fit statistics) are to be stored. The default is a temporary file that will be deleted as soon asbollenstinefinishes.

confaoptions(...)allows to transfer options toconfa. Some bootstrap replications produce non-convergent samples that may never converge, so in order to speed up computations, it might make sense to limit the number of iterations, say withconfaoptions( iter(20) ).

All non-standard model options, like unitvar or correlated, must bespecified with bollenstine to produce correct results!All other options are assumed to be

bootstrap_optionsand passed through tobootstrap.

Example

. use http://web.missouri.edu/~kolenikovs/stata/hs-cfa.dta, clear. confa (vis: x1 x2 x3) (text: x4 x5 x6) (math: x7 x8 x9), from(iv)corr(x7:x8). estat fit. estat corr. estat corr, bound. predict fa1-fa3, reg. predict fb1-fb3, bart

Also seeOnline:

confa,bollenstine.

ReferencesBollen, K. and Stine, R. (1992) Bootstrapping Goodness of Fit Measures in Structural Equation Models.

Sociological Methods and Research,21, 205--229.

ContactStas Kolenikov, kolenikovs

atmissouri.edu