Shown below are examples of GDA syntax specifications for various common designs; refer to the GDA Syntax Overview for details concerning the GDA syntax. For additional examples, also refer to the GLM Syntax Examples.

In practically all cases, the most efficient way to write GDA syntax is to use the Quick-specs dialogs. Specifically, from the General Discriminant Analysis (GDA) Models Startup Panel - Quick tab, select the desired Type of analysis, and Quick specs dialog as the Specification method; next click OK on the General Discriminant Analysis (GDA) Models (Startup Panel) to display the respective Quick specs dialog. Then click the Syntax editor button, and the analysis design (as specified in the Quick-specs dialog) will be written out to the syntax editor. You can now further modify the Design statement to customize the design (e.g., to drop effects).

General Discriminant Analysis Syntax Examples. The following are examples involving GDA Syntax:

Example 1: Simple Standard Discriminant analysis

Example 2: Stepwise Discriminant Analysis with Categorical Predictors

General Linear Models (GLM) Syntax Examples. Many of the General Linear Models (GLM) Syntax Examples as well as General Regression Models (GRM) Syntax Examples can be run with minor modifications in GDA as well. However, remember that a few specialized GLM (or GRM) features and options are not available in GDA; for example:

No nested designs can be specified;

No choice of parameterization is available (sigma-restricted parameterization for categorical predictor effects is always used in GDA);

No choice of methods for partitioning the sums of squares is available; Type VI (effective hypothesis) sums of squares is always used (see also Six types of sums of squares in GLM for details);

No within-subject (repeated measures) designs are available;

No mixed-model ANOVA-like models (with random effects) are available;

Only a single categorical dependent variable can be specified;

The best subset analysis options are different than those described for GRM, and tailored to discriminant function analysis.