To impute missing data, click the Multivariate red triangle and select Impute Missing Data. A new data table is created. This data table is a duplicate of your data table except that the missing values are replaced with estimated values.
Imputed values are expectations conditional on the nonmissing values for each row. The mean and covariance matrix, estimated by the method chosen in the launch window, are used for the imputation calculation. All multivariate tests and options are then available for the imputed data set.