The CALIS Procedure
-
Overview
-
Getting Started
-
SyntaxClasses of Statements in PROC CALISSingle-Group Analysis SyntaxMultiple-Group Multiple-Model Analysis SyntaxPROC CALIS StatementBOUNDS StatementBY StatementCOSAN StatementCOV StatementDETERM StatementEFFPART StatementFACTOR StatementFITINDEX StatementFREQ StatementGROUP StatementLINCON StatementLINEQS StatementLISMOD StatementLMTESTS StatementMATRIX StatementMEAN StatementMODEL StatementMSTRUCT StatementNLINCON StatementNLOPTIONS StatementOUTFILES StatementPARAMETERS StatementPARTIAL StatementPATH StatementPCOV StatementPVAR StatementRAM StatementREFMODEL StatementRENAMEPARM StatementSAS Programming StatementsSIMTESTS StatementSTD StatementSTRUCTEQ StatementTESTFUNC StatementVAR StatementVARIANCE StatementVARNAMES StatementWEIGHT Statement
-
DetailsInput Data SetsOutput Data SetsDefault Analysis Type and Default ParameterizationThe COSAN ModelThe FACTOR ModelThe LINEQS ModelThe LISMOD Model and SubmodelsThe MSTRUCT ModelThe PATH ModelThe RAM ModelNaming Variables and ParametersSetting Constraints on ParametersAutomatic Variable SelectionEstimation CriteriaRelationships among Estimation CriteriaGradient, Hessian, Information Matrix, and Approximate Standard ErrorsCounting the Degrees of FreedomAssessment of FitCase-Level Residuals, Outliers, Leverage Observations, and Residual DiagnosticsTotal, Direct, and Indirect EffectsStandardized SolutionsModification IndicesMissing Values and the Analysis of Missing PatternsMeasures of Multivariate KurtosisInitial EstimatesUse of Optimization TechniquesComputational ProblemsDisplayed OutputODS Table NamesODS Graphics
-
ExamplesEstimating Covariances and CorrelationsEstimating Covariances and Means SimultaneouslyTesting Uncorrelatedness of VariablesTesting Covariance PatternsTesting Some Standard Covariance Pattern HypothesesLinear Regression ModelMultivariate Regression ModelsMeasurement Error ModelsTesting Specific Measurement Error ModelsMeasurement Error Models with Multiple PredictorsMeasurement Error Models Specified As Linear EquationsConfirmatory Factor ModelsConfirmatory Factor Models: Some VariationsResidual Diagnostics and Robust EstimationThe Full Information Maximum Likelihood MethodComparing the ML and FIML EstimationPath Analysis: Stability of AlienationSimultaneous Equations with Mean Structures and Reciprocal PathsFitting Direct Covariance StructuresConfirmatory Factor Analysis: Cognitive AbilitiesTesting Equality of Two Covariance Matrices Using a Multiple-Group AnalysisTesting Equality of Covariance and Mean Matrices between Independent GroupsIllustrating Various General Modeling LanguagesTesting Competing Path Models for the Career Aspiration DataFitting a Latent Growth Curve ModelHigher-Order and Hierarchical Factor ModelsLinear Relations among Factor LoadingsMultiple-Group Model for Purchasing BehaviorFitting the RAM and EQS Models by the COSAN Modeling LanguageSecond-Order Confirmatory Factor AnalysisLinear Relations among Factor Loadings: COSAN Model SpecificationOrdinal Relations among Factor LoadingsLongitudinal Factor Analysis
- References
Changes and Enhancements
The following sections describe the new features of this version of PROC CALIS.
Case-Level Residual Diagnostics
Residual analysis at the case level or observation level is available when you provide raw data input. The main features include graphical and nongraphical residual diagnostic tools for identifying outliers and leverage observations, for detecting departures from the theoretical residual distribution, and for studying the distribution of residuals over the range of predicted values of the dependent observed variables. Use the RESIDUAL option to request nongraphical output for residual diagnostics. Use the PLOTS= option to request graphical plots for residual diagnostics. See the section Case-Level Residuals, Outliers, Leverage Observations, and Residual Diagnostics for details about these new features. See Example 27.14 for an example that illustrates the use of the residual diagnostic tools provided by PROC CALIS.
Robust Estimation
Robust estimation is available with the ROBUST option. Robust estimation methods downweight the model outliers during the estimation process so that estimates are not contaminated by outliers. Two robust methods are implemented in PROC CALIS. The direct robust method applies individual (casewise) weights to observations during the model estimation. This is the default robust method that you get when you specify the ROBUST option. The two-stage robust method estimates robust covariance and mean matrices in the first stage, and then it does an ML estimation of the model by inputting the robust covariance and mean matrices in the second stage. You can specify this method with the ROBUST=SAT option. See the ROBUST option for details about variants of the robust methods. See the section Robust Estimation for technical details of the robust estimation implemented in PROC CALIS.
Inputting Customized Baseline Model Fit
A class of fit indices, called incremental fit indices (see the section Incremental Indices for details), measures model fit by comparing the relative fit of the target model to that of a baseline model. Traditionally, this baseline model is the uncorrelatedness model, where the covariances of observed variables are all zero. PROC CALIS follows this tradition and uses the uncorrelatedness model as the baseline model when computing the incremental fit indices. However, if you have a situation where the uncorrelatedness model is not desirable, you can now define and fit your own baseline model in a separate analysis. PROC CALIS provides tools for you to input the fit information of your customized baseline model. If you use PROC CALIS to fit your customized baseline model, you can save the fit information in an OUTFIT= data set and then specify the data set as a BASEFIT= data set when fitting your target model. PROC CALIS can extract the fit information of the customized baseline model for the computations of incremental fit indices (and a few other indices that assume the degrees of freedom of a baseline model). Alternatively, you can directly input the fit function value and the degrees of freedom of your customized baseline model by using the BASEFUNC= option. See the BASEFIT= and the BASEFUNC= options for details.