Handling unlabeled categorical values
The case data for a categorical variable might include values for which no corresponding category exists. This can happen in the following situations:
▪When accessing the case data with an MDM document, and when the corresponding MDM variable does not include a complete set of categories.
▪When generating an MDM document from the .sav file or accessing the case data with no metadata, and when the IBM SPSS Statistics variable has been only partially coded, and when the ScanForCategories setting described above has been changed from its default value (which is to scan all cases for unlabeled values and generate extra MDM categories for them).
When accessing the case data with an MDM document, the SPSS Statistics SAV DSC matches the MDM variable's Other category, if it has one, to any data value for which there is no corresponding MDM category. If the Other category has a helper field, and that field is of the appropriate type, the SPSS Statistics SAV DSC stores the actual data value in that field. If the MDM variable has more than one Other category, the SPSS Statistics SAV DSC uses the first one with a helper field, or the first Other category if none has a helper field. However, if the MDM variable does not contain an Other category, the SPSS Statistics SAV DSC silently ignores any data values that do not have a corresponding MDM category.
When generating an MDM document from the .sav file or accessing the case data with no metadata, the SPSS Statistics SAV DSC silently ignores all data values for which it has not generated a corresponding MDM category.
See also