The GEE Procedure (Experimental)
The MISSMODEL statement requests a weighted GEE analysis. It specifies a logistic regression that is used to estimate the weights under the MAR assumption. If the pattern of missing data is intermittent (not dropout), the GEE procedure terminates and does not perform an analysis.
You can use the same effects or different effects in the MODEL and MISSMODEL statements. Explanatory variables can be continuous or classification variables. Classification variables can be character or numeric. Explanatory variables that represent nominal (classification) data must be declared in a CLASS statement. Interactions between variables can also be included as effects. Columns of the design matrix are automatically generated for classification variables and interactions. The syntax for effects is the same as for the GLM procedure. For more information, see the section Specification of Effects in Chapter 45: The GLM Procedure.
You can specify the following options after a slash (/).
- MAXWEIGHT=number
-
truncates the predicted weights from the missingness model if they are are larger than number, where number
1.
- TYPE=OBSLEVEL | SUBLEVEL
-
specifies the type of weighted GEE method. You can specify the following values:
-
OBSLEVEL specifies the observation-level weighted GEE method.
-
SUBLEVEL specifies the subject-level weighted GEE method.
By default, TYPE=OBSLEVEL.