You can use the Output Variables tab to add analysis variables to the data table. (See Figure 24.24.) If you request a plot that uses one of the output variables, then that variable is automatically created even if you did not explicitly select the variable on the Output Variables tab.
For a multinomial response, residuals and influence diagnostics are not available.
The following list describes each output variable and indicates how the output variable is named. represents the name of the response variable. If you use events/trials syntax, then represents the name of the events variable.
adds a variable named Proportion_
, where is the name of the events variable and is the name of the trials variable. The value of the variable is the ratio . This variable is added only when you use events/trials syntax.
adds predicted values. The variable is named GenP_
.
adds 95% confidence limits for the predicted values. The variables are named GenLclm_
and GenUclm_
.
adds the linear predictor values. The variable is named GenXBeta_
.
adds residuals, which are calculated as observed values minus predicted values. The variable is named GenR_
.
adds the Pearson chi-square residuals. The variable is named GenChiSqR_
.
adds the deviance residuals. The variable is named GenDevR_
.
adds the likelihood residuals. The variable is named GenLikR_
.
adds Cook’s influence statistic. The variable is named GenCooksD_
.
adds the leverage statistic. The variable is named GenH_
.
adds variables, where is the number of parameters in the model. A classification variable with levels counts as parameters. The variables are scaled measures of the change in each parameter estimate and are calculated by deleting the
th observation. Large values of DFBETAS indicate observations that are influential in estimating a given parameter. Belsley,
Kuh, and Welsch (1980) recommend as a size-adjusted cutoff. The variables are named DFBeta
, for .