MISSING Subcommand (CORRELATIONS command)

MISSING controls the treatment of missing values.

• The PAIRWISE and LISTWISE keywords are alternatives; however, each can be specified with INCLUDE or EXCLUDE.
• The default is PAIRWISE and EXCLUDE.

PAIRWISE . Exclude missing values pairwise. Cases that have missing values for one or both of a pair of variables for a specific correlation coefficient are excluded from the computation of that coefficient. Since each coefficient is based on all cases that have valid values for that particular pair of variables, this can result in a set of coefficients based on a varying number of cases. The valid number of cases is displayed in the Correlations table. This is the default.

LISTWISE . Exclude missing values listwise. Cases that have missing values for any variable named on any VARIABLES list are excluded from the computation of all coefficients across lists. The valid number of cases is the same for all analyses and is displayed in a single annotation.

INCLUDE . Include user-missing values. User-missing values are included in the analysis.

EXCLUDE . Exclude all missing values. Both user- and system-missing values are excluded from the analysis.