PROC TABULATE by Example, Second Edition

Report the Missing Data for Analysis Variables with the NMISS Statistic

The previous examples showed how to spot missing data for class variables in your table. But the MISSING option does not work for analysis variables. To show how many observations you have with missing data for the analysis variable, you use the NMISS statistic.

You could always use PROC MEANS to see how much missing data there is for an analysis variable, but PROC MEANS will not tell you where these values are. If you are concerned about which particular categories in your table have missing data, you need to use PROC TABULATE. With PROC TABULATE and NMISS, you can produce a table that shows which table cells have missing data for your analysis variable.

We start with ...