The problem of missing values is well known to data analysts. And, it is always important to know beforehand the Complete-case Analysis Sample Size for variables under consideration. The SAS-code program, below, provides a quick way to calculate the latter sample size.




data IN;
input ID 2. GENDER $1. MARITAL $1.AGE 2. INCOME 3.;
cards;
01MS23123
02MM32345
03M 234
04 34
05FS
08FM34234
07F 237
08 M35
09 S 234
10MD
;
run;

data IN;
set IN;
MISSING_GENDER=MISSING(GENDER);
MISSING_MARITAL=MISSING(MARITAL);
run;

data IN;
set IN;
NO_MISSING_NUMERIC_VARS=NMISS(AGE , INCOME );
NO_MISSING_CHAR_VARS =MISSING_GENDER+MISSING_MARITAL;
NO_MISSING_VARIABLES=NO_MISSING_NUMERIC_VARS+NO_MISSING_CHAR_VARS;
proc freq data=IN;table NO_MISSING_VARIABLES;
run;

The Complete-case Analysis Sample Size is the frequency count corresponding to
NO_MISSING_VARIABLES = 0.