GDA Models Results - Regression Residuals 1 Tab

Select the Regression residuals 1 tab of the GDA Models Results dialog to access options to display regression-like residual statistics and predicted values. As described in the Introductory Overview, the GDA module applies the general linear model to the discriminant function analysis problem, after (internally) recoding the class (group) memberships recorded in the categorical dependent variable into multiple (dummy-) coded dependent variables.

Note: the names and organization of the options on the Regression Residuals 1 tab may be different depending if you are viewing the smaller or the larger GDA Models Results dialog. The order of the options below reflect that of the smaller GDA Models Results - Regression Residuals 1 tab. The names of the options below reflect that of the smaller GDA Models Results - Regression Residuals 1 tab with the names of the larger GDA Models Results - Regression Residuals 1 tab in parentheses.

Sample. Select the respective option button in the Sample group box to specify which type of sample to base the predicted and residual values. You can display and plot predicted and residual values for all observations that were used to compute the current results (select Analysis), all observations that were not used to compute the current results, but have valid data for all predictor and dependent variables (select Cross-val (Cross-validation)), both the analysis and the cross-validation samples (select Both), or display and plot predicted values and related statistics for all cases that have valid data for the predictor variables, but missing data for at least one dependent variable (select Pred). If these options are dimmed, no valid cases in the prediction sample were found, and/or no cross-validation sample was specified on the GDA Quick Specs Dialog - Advanced tab, or via the SAMPLE keyword in the GDA Models Syntax Editor dialog.

Classes (Dependent variables). Click the Classes button to display the Classes for Residuals dialog in which you select the classes of the dependent variable to be used for the residual statistics. The predicted and residual values are computed for the dependent variable's classes (or levels).

Show predicted and residual values
Use the options in the Show predicted and residual values group box to compute the desired predicted and residual statistics, and to display them along with other variables in the current data file.
Predicted and residuals
Click the Predicted and residuals button to display a spreadsheet of the observed, predicted, and residual values for each dependent variable class; if the Cross-val option button is selected in the Sample group box then only the observations in the cross-validation sample will be shown (see above); if the Pred. option button is selected, then only the observations in the prediction sample will be shown.
Extended (Extended stats)
Select the Extended check box to display predicted and residual statistics in addition to the basic observed, predicted, and residual values. Note that the setting of the Extended check box will also determine the number of different statistics for each dependent variable that will be displayed via the Save button. If the Extended check box is selected, then in addition to the predicted, observed, and residual values, STATISTICA will compute the (default) 95% Prediction intervals and 95% Confidence limits, the Standardized predicted and Standardized residual scores, the Studentized residual scores, the Leverage values, the Deleted residuals and Studentized deleted residual scores, Mahalanobis and Cook distance scores, the DFFITS statistic, and the Standardized DFFITS statistic.

Spreadsheet for each dependent variable (Separate Spreadsheet for each dep. var). By default the predicted and residual statistics for all classes (levels) of the Classes are shown in a single spreadsheet, where consecutive classes will occupy consecutive groups of columns. Select the Spreadsheet for each dependent variable check box to display a separate spreadsheet for each class.

Sort (Sort obs by)
The Sort combo box provides a number of options for ordering the cases or runs shown in the Predicted and residual spreadsheet. From the combo box you can select options to sort the cases by Case numbers, by Observed, Predicted or Residual values, by Standard errors of predicted values, by Standardized predicted, Standardized residual or Studentized residual scores, by Leverage values, by Deleted residual, Studentized deleted residual, Mahalanobis distance, or Cook distance scores, by the DFFITS statistic, or by the Standardized DFFITS statistic. The default is to sort cases by Case numbers; note that if any other sort order is requested, then the Spreadsheet for each dependent variable check box is automatically selected.
Save
Click the Save button to display a standard variable selection window, which is used to select additional variables to be displayed together with the predicted and residual values. After you select the additional variables, a spreadsheet containing the specified variable(s) along with numerous residual and predicted values/statistics will be displayed in an individual window (regardless of the settings in the Options - Output Manager tab or the Analysis/Graph Output Manager dialog). You can, however, add the spreadsheet to a workbook or report using the or buttons, respectively. Note that in order to save the spreadsheet, you must select the spreadsheet and select Save or Save As from the File menu. This is useful if you want to use the predicted and-/or residual scores for further analyses with other STATISTICA analyses. See the description of Extended (above) for a list of the default statistics saved when the Extended check box is selected.
Plots of predicted and residual values
The options in the Plots of predicted and residual values group box allow you to produce various plots of predicted and residual values for each class of the dependent variables; if the Cross-val. option button is selected in the Sample group box then only the observations in the cross-validation sample will be shown; if the Pred. option button is selected, then only the observations in the prediction sample will be shown (see above). To produce 2D or 3D plots of predicted or residual values against the observed values of other variables (e.g., predictor variables), use the GDA Models Results - Regression Residuals 2 tab.
Pred values (Predicted values)
Click the Pred values button to display a histogram of the predicted values.
Pred. & resids.
Click the Pred. & resids button to display a 2D scatterplot of the predicted versus residual values; the specific residual values that will be computed can be chosen via the Resids for default plots group box (see below).
Residuals
Click the Residuals button to display a histogram of the residual values; the specific value that will be computed can be chosen via the Resids for default plots group box (see below).
Obs. & pred.
Click the Obs. & pred. button to display a 2D scatterplot of the observed versus predicted values.
P-plot of resids (Normal)
Click the P-plots of resids button to display normal probability plot of the residual values; the specific value that will be computed can be chosen via the Resids for default plots group box (see below).
Obs. & resids.
Click the Obs. & resids. button to display a 2D scatterplot of the observed versus residual values; the specific residual values that will be computed can be chosen via the Resids for default plots group box (see below).
Half-n (Half-normal)
Click the Half-n button to display a half-normal probability plot of the residual values; the specific value that will be computed can be chosen via the Resids for default plots group box (see below); if the  Abs (Plot absolute values) check box is selected then the respective absolute residual values will be plotted.
Res. & del. res.
res. Click the Res. & del. res. button to display a 2D scatterplot of the residual values versus the deleted residuals.
Detr. norm. plot (Detrended).
plot (Detrended). Click the Detr norm. plot button to display a detrended normal probability plot of the residual values; the specific value that will be computed can be chosen via the Resids for default plots group box (see below).
Case no. & res.
Click the Case no. & res. button to display a 2D scatterplot of the residual values versus the case numbers (of the observations in the plot). This plot is useful in order to detect any serial correlation between the residuals for consecutive observations.
Resids for default plots
Select the type of residual that is to be shown in histograms, probability plots, and scatterplots of residuals in the Resids for default plots group box. Instead of the standard Raw (residuals) (the default), you can choose Standardized residuals, Deleted residuals, Studentized residuals, or Studentized deleted residuals. Note that this group box is only available on the larger GDA Models Results dialog. If you can not see this option, select the More results button at the button of the smaller GDA Models Results dialog).

See also GDA - Index.