Each time you choose a response and click Run in the Response Specification panel, a multivariate response report is added to the Manova Fit report. The multivariate response reports are named using the response specification names.
The red triangle menu in each response specification report contains the following option:
Custom Test
Shows the Custom Test launch control in the report. This launch control enables you to set up custom tests of effect levels. For more information about creating custom tests, see “Custom Test”.
Each model effect report contains a table of multivariate tests. For more information about these tests, see Comparison of Multivariate Tests. The red triangle menu for each effect name in the multivariate response report contains the following options to request additional information about the multivariate fit:
Test Details
Shows or hides canonical details about the test for the whole model or for the specified effect. This report contains the eigenvalues and eigenvectors of the E−1H matrix that is used to construct multivariate test statistics.
Eigenvalue
The eigenvalues of the E−1H matrix used in computing the multivariate test statistics.
Canonical Corr
The canonical correlations associated with each eigenvalue. This is the canonical correlation of the transformed responses with the effects, corrected for all other effects in the model.
Eigvec
The eigenvectors of the E−1H matrix, or equivalently of (E + H)−1H.
Centroid Plot
Shows or hides a table of centroid values and a plot of the centroids (multivariate least squares means) on the first two canonical variables formed from the test space.
The table of centroid values appears above the Canonical Centroid Plot. Click the CentroidVal disclosure icon to show the table of centroid values for up to four centroids.
In the Canonical Centroid Plot, the first canonical axis is the vertical axis. If the test space is one dimensional, the centroids align on a vertical axis. Relationships among groups of variables can be verified with biplot rays and the associated eigenvectors. See Details for Centroid Plot Option.
The Canonical Centroid Plot red triangle menu contains the following options:
Centroid Circles
Shows or hides the centroid points and circles in the Canonical Centroid Plot. The centroid points appear with a circle that corresponds to the 95% confidence region (Mardia et al. 1979). When centroid plots are created under effect tests, circles that correspond to the effect being tested appear in red. Other circles appear blue. The coordinates for the centroid points appear in the CentroidVal matrix above the Canonical Centroid Plot.
Biplot Rays
Shows or hides biplot rays on the canonical centroid plot. Biplot rays show the directions of the original response variables in the test space. The intersection point of the biplot rays is labeled Grand. The coordinates for the biplot ray intersection and endpoints appear in the CentroidVal matrix above the Canonical Centroid Plot.
Show Points
Shows or hides the individual points in the Canonical Centroid Plot.
Save Canonical Scores
Saves new formula columns to the data table. The new columns contain the canonical scores for an individual effect. The columns are named Canon[i], where i refers to the ith canonical score for the Y variables. The canonical scores are computed based on the E−1H matrix that is used to construct the multivariate test statistic. Canonical scores are saved for eigenvectors that correspond to nonzero eigenvalues. See Statistical Details for Canonical Calculations.
Tip: Canonical correlation analysis is not a specific option, but it can be performed using a sequence of options in the multivariate fitting platform. First, click the Whole Model red triangle and select Test Details. Then click the Whole Model red triangle and select Save Canonical Scores. The details list the canonical correlations (Canonical Corr) next to the eigenvalues. The saved variables are called Canon[1], Canon[2], and so on. These columns contain both the values and their formulas. To obtain the canonical variables for the X side, repeat the same steps, but interchange the X and Y variables. If you have already appended the columns Canon[n] to the data table, the new columns are called Canon[n] 2 (or another number) that makes the name unique. For an example of canonical correlation analysis, see Example of Canonical Correlation Analysis.
Contrast
Runs a customized F test for the statistical contrasts of treatment levels that you specify in the contrasts dialog.
Note: The Contrast option for a multivariate model is similar to the corresponding option for a regression model with a single response. See “LSMeans Contrast” for a description and examples of the LSMeans Contrast options.