Experimental Design Builder Results - Design Summary Tab

Select the Design summary tab in the Experimental Design Builder Results dialog box to access the options described here.

Efficiency measures. Click this button to display the D, A, and G efficiency measures of the current design.  

Variances of coefficients. Click this button to display the relative variance of each coefficient in the design. The variances of the estimated ordinary least squares estimates are given as the diagonal elements of the matrix:

When designing the experiment, an estimate of the variance is usually not known in advance, so the diagonal elements of the matrix:

are displayed and are referred to as relative variances.

Percent of design space. Click this button to display the fraction of design space graph. This graph displays the fraction or percent of the design that has prediction variance at or below a specific value. This graph is useful in the comparison of competing designs whose goal is precise prediction. STATISTICA takes a large random sample, N, of points in the design space and computes the relative variance of each sampled point. Those relative variances are then sorted from smallest to largest and the coordinate pair (100.0 i/(N+1), relative variance) is plotted for each sampled point. For example, if the pair (25, 3.5) is plotted, then 25% of the data has a relative variance less than or equal to 3.5.

Alias matrix. Click this button to display the alias matrix.

Moments matrix. Click this button to display the moments matrix associated with the current design.

Information matrix. Click this button to display the information matrix associated with the current design. The information matrix is the inverse of the relative covariance matrix. The determinant of this matrix is maximized for D-optimal designs.

GLM syntax. Click this button to display a report that contains the GLM syntax of the optimal design.