 |
|
| The TCALIS Procedure |
| PVAR Statement |
The PVAR statement defines the variance or partial variance parameters in FACTOR and PATH models. In each assignment in the PVAR statement, you specify variables on the left-hand side and the parameter-spec on the right-hand side of an equal sign. The assignments in the PVAR statement must be separated by commas. The variables listed on the left-hand side of the equal sign are any variables defined in the model. If a variable in the list is an exogenous variable, then the corresponding parameter in the parameter-spec list is a variance parameter of the variable. If a variable in the list is an endogenous variable, then the corresponding parameter in the parameter-spec is a partial variance parameter of the variable. In most cases, the partial variance of an endogenous variable can be interpreted as the error variance for the variable.
The syntax of the PVAR statement is exactly the same as that of the STD statement. As in the STD statement, in the PVAR statement you can specify free or constrained parameters with or without starting or fixed values. You can also use prefixes to generate parameter names. Refer to the STD statement for details and examples.
The difference between the PVAR and the STD statements is conceptual in nature. In the STD statement, you can specify only variance parameters for exogenous variables. However, in the PVAR statement, both exogenous and endogenous variables can be put on the variables list. This treatment is needed in the FACTOR and PATH modeling languages because error variables in these models are not explicitly named.
Automatic Variance and Partial Variance Parameters among Variables
If the name of a variable (manifest or latent) in the FACTOR or PATH model does not appear in the PVAR statement, the variance or partial variance of the variable is assumed to be a parameter to estimate. These parameters will be named automatically with the prefix _Add and appended with integer suffixes. Therefore, the entire PVAR statement is optional in the sense that all variance and partial variance parameters will be generated automatically for the FACTOR or PATH model in PROC TCALIS. However, if you need to constrain these parameters, or assign fixed values, the PVAR statement specification must be used.
Modifying a Parameter Specification from a Reference Model
If you define a new FACTOR or PATH model by using a reference (or base) model in the REFMODEL statement, in some situations you might want to modify some parameter specifications from the PVAR statement of the reference model before transferring the specifications to the new model. If you want to change a particular PVAR parameter specification from the reference model, you can simply put the corresponding parameter location with a new parameter specification in the new model. If you want to delete a particular parameter location from the reference model, you can put that parameter location with the missing value '.' in the new model. For example, suppose you are defining a new PATH model by using the REFMODEL statement and the partial variance of variable V2 is defined as a fixed or free parameter in the reference model. If you do not want this parameter specification to be copied into your new model, you can use the following specification in the new model:
pvar V2 = .;
Note that the missing value syntax is valid only when you use the REFMODEL statement. See the section Modifying a FACTOR Model from a Reference Model for a more detailed example of FACTOR model respecification. See the section Modifying a PATH Model from a Reference Model for a more detailed example of PATH model respecification.
In confirmatory factor or PATH models, all variances or partial variances among variables are free parameters to estimate unless you specify otherwise in the PVAR statement (see the section Automatic Variance and Partial Variance Parameters among Variables). This is also true for confirmatory factor and LINEQS models specified under the REFMODEL statement. In the reference model, (partial) variance parameters generated by PROC TCALIS are named with the _Add prefix and appended with integer suffixes. These parameter specifications in the reference model do not transfer to the new model. Only after the new model is resolved from the reference model, the REFMODEL statement options, the RENAMEPARM statement options, and model respecifications do the remaining unspecified (partial) variances of variables generate additional (partial) variance parameters with the _Add prefix and integer suffixes. In this way, the (partial) variance parameters in the new model are not constrained to be the same as the corresponding parameters in the reference model. If you want any of these (partial) variance parameters to be constrained across the models, you must specify them explicitly in the PVAR statement of the reference model so that the (partial) variance specification is transferred to the new model.
Note: This procedure is experimental.
|
|
Copyright © 2009 by SAS Institute Inc., Cary, NC, USA. All rights reserved.