Nested Design ANOVA in GLM, GLZ, and PLS

Select Nested design ANOVA as the Type of analysis on the GLM Startup Panel - Quick tab to specify nested ANOVA or MANOVA designs. Nested designs are incomplete (as opposed to complete factorial designs, where all treatment combinations are observed). Often, the missing treatment combinations are specifically arranged, so that the main effects in the design are not confounded (correlated) with each other. This is, for example, the case in Latin Square designs, which are frequently used in applied research in cases when it would be impractical or too costly to observe all possible treatment combinations (see also the Latin square designs section in Experimental Design for details and examples). These types of designs are also discussed in the context of the ANOVA/MANOVA module, and the Variance Components and Mixed Model ANOVA/ANCOVA module.

In the subsequent Quick Specs dialog box, you can specify one or more categorical predictor variables. STATISTICA will construct a default simple hierarchical nested design, which can be further customized in the Nesting for between group factors dialog box or via the GLM Analysis Syntax Editor.

Note that in nested designs, nested factors never appear as main effects; see the Nested designs topic in the Introductory Overview for additional details. For more information, refer to the Introductory Overview, GLM - Index, and General ANOVA/MANOVA and GLM Notes - Nested Designs and Latin Squares.