MISSING Subcommand (KNN command)

The MISSING subcommand is used to control whether user-missing values for categorical variables are treated as valid values. By default, user-missing values for categorical variables are treated as invalid.

• The MISSING subcommand defines categorical variables as dependent variables with MLEVEL specification O or N, dependent variables with measurement level set at Ordinal or Nominal in the data dictionary and no MLEVEL specification, and factors. Scale variables are dependent variables with MLEVEL specification S, dependent variables with Scale measurement level and no MLEVEL specification, and covariates.
• User-missing values for scale variables are always treated as invalid.
• System-missing values for any variables are always treated as invalid.

USERMISSING=EXCLUDE. User-missing values for categorical variables are treated as invalid. This is the default.

USERMISSING=INCLUDE. User-missing values for categorical variables are treated as valid values.