Cruise Scientific                Visual Statistics Studio                Visual Statistics Studio Guide

Related Measures Designs

Classic Related Measures Design

Classic related measures design (Designs, Designs of Experiments, Classic Designs, Related Measures)

can be analyzed by clicking (Analysis II, Classic Related Measures Design)

by selecting locations of the control and experimental groups and clicking the Accept command, as

Invoking the t-distribution with symmetric critical areas (Tables, t Distribution with Symmetric Critical Areas),

dragging-in the obtained t-value with its associated degrees of freedom, and clicking on the Accept command, displays the probability that can be dragged-out into the table of results.

Coded Related Measures Design

Coded related measures design (Designs, Designs of Experiments, Coded Designs, Related Measures)

 

contain aside the predictor and criterion variables X and Y also subject binding codes. Before analysis, the data matrix must be orthogonalized. Select (Analysis II, Orthogonalization by Residualization)

and follow instructions on the Information panel.

 

Optionally, select (Analysis I, Correlation Matrices)

click the Select All command and the Correlations command under the Superimpose Correlation Matrix on a Display Panel heading.

Select (Analysis II, Coded Related Measures Design)

and click on the Accept command. Click on Descriptive Statistics, mark the Standard Variance option box, check the Select the Standardizing Variable Y box, and click on the Accept command. 

Compare standard variances on the above table with the Coded Regression Analysis output panel.

Note that Coefficient of Multiple Determination (R Square) is the standard variance of the predicted variable Y'. Coefficient of Multiple Alienation is the standard variance of the error variable Y^. Also note that the F ratio for the parent vector X (Component 1) is 3.00, equal to the t^2 ratio (1.73^2) obtained by the Classic Related Measures analysis.