If a variable for which statistics are calculated has a missing value, that value is ignored in the calculation of statistics, and the missing values are tabulated separately. A missing value for one such variable does not affect the treatment of other variables in the same observation.
If the WEIGHT variable has a missing value, the observation is excluded from the analysis[18]. If the FREQ variable has a missing value, the observation is excluded from the analysis. If a variable in a BY or ID statement has a missing value, the procedure treats it as it would treat any other value of a BY or ID variable.
[18] In SAS 6.12 and earlier releases, the observation was included in the analysis, and the missing WEIGHT variable value was taken to be zero.