Just because your PROC
TABULATE code runs without errors and your table looks fine, do not
think that there are no problems lurking beneath the surface. Checking
your table for missing data problems is an important step in the table
When you have missing
data for some or all of the class variables used in your table, PROC
TABULATE does not give you any warning messages about the problem.
Many SAS procedures produce messages such as:
312 observations have missing data, only 57 observations will be used
for analysis." PROC TABULATE, however, does not
warn you about missing data.
When producing a table, PROC TABULATE first checks every observation ...