General Discriminant Analysis Button

Click the button to display the General Discriminant Analysis (GDA) Startup Panel. The General Discriminant Analysis (GDA) module is called a "general" discriminant analysis program because it applies the methods of the general linear model to the discriminant function analysis problem. In short, the discriminant function analysis problem is "recast" as a general multivariate linear model, where the dependent variables of interest are (dummy) coded vectors that reflect the group membership of each case.

In addition to traditional (standard) stepwise discriminant analysis (as available in the Discriminant Function Analysis module), the GDA module provides support for continuous and categorical predictors. You can specify simple and complex ANOVA/ANCOVA-like designs, e.g., mixtures of continuous and categorical predictors, polynomial (response surface) designs, factorial designs, nested designs, etc. GDA also supports multiple degree of freedom effects in stepwise selection, as well as best-subset selection of predictor effects. GDA allows you to perform model building (selection of predictor effects) not only based on traditional criteria (e.g., p-to-enter/remove; Wilks' Lambda), but also based on misclassification rates; in other words STATISTICA will select those predictor effects that maximize the accuracy of classification, either for those cases from which the parameter estimates were computed, or for a cross-validation sample (to guard against over fitting). STATISTICA computes detailed results and diagnostic statistics and plots; GDA provides a large number of auxiliary information to help you judge the adequacy of the chosen discriminant analysis model (descriptive statistics and graphs, Mahalanobis distances, Cook distances, and leverages for predictors, etc.). GDA also includes an adaptation of the general GLM (GRM) response profiler; these options allow you to quickly determine the values (or levels) of the predictor variables that maximize the posterior classification probability for a single group, or for a set of groups in the analyses; in a sense, you can quickly determine the typical profiles of values of the predictors (or levels of categorical predictors) that identify a group (or set of groups) in the analysis.

For traditional and stepwise discriminant analysis, see also the Discriminant Analysis module. Categorical dependent (criterion) variables can also be analyzed via the Log-Linear Analysis module, as well as the Generalized Linear/Nonlinear Models (GLZ) module.