NOTE: This page has been delinked.  It is no longer being maintained, and information on this page may be out of date.

How can I count the number of missing values for a character variable?

We use the following little data set to illustrate how to count up the number of missing values for character variables with SPSS, SAS and Stata.

           id    female      race       ses    schtype       prg 
  1.        1         1         1         3        pub         1  
  2.        2         0         1         2        pub         2  
  3.        3         0         3         2                    3  
  4.        4         0         .         2        pub         .  
  5.        5         0         2         2        pub         2  
  6.        6         1         2         1        pub         2  
  7.        7         0         .         .                    .  
  8.        8         1         1         2        pub         1  
  9.        9         1         .         .        pub         1  
 10.       10         0         1         2        pub         1  
 11.       11         1         1         1                    1  
 12.       12         0         1         2        pri         1  
 13.       13         0         1         .        pub         1  
 14.       14         0         1         1                    .  
 15.       15         1         .         2        pub         1  
 16.       16         1         1         3        pub         1  

SPSS

In SPSS it is easy to request the number of missing and non-missing values for character variables.  We can use the frequencies command to request frequencies for numeric and character variables and use the /format=notable subcommand to suppress the display of the frequency tables, leaving us with a concise report of the number of missing and non-missing values for each variable (see below).

FREQUENCIES VARIABLES=RACE SES SCHTYPE PRG 
  /FORMAT=NOTABLE
  /ORDER= ANALYSIS .

SAS

In SAS, we have to go to a little extra effort to get the number of missing and non-missing values for character variables.  We can use proc format to make a format  for character variable to be either "nomissing" or "missing" and then use that format with proc freq as illustrated below.  We then get a concise table showing us the number of missing and nonmissing for the variable schtype.

proc format;
 value $miss " "="missing"
 other="nomissing";
run;

proc freq data=temp;
  tables schtype / missing;
  format schtype $miss.;
run; 

Here is the output.
 

SCHTYPE	Frequency	Percent	Cumulative Frequency	Cumulative Percent
missing	4	25.00	4	25.00
nomissing	12	75.00	16	100.00

Stata

We have created a small Stata program called tabmiss that counts the number of missing values in both numeric and character variables. You can download tabmiss by typing findit tabmiss (see How can I use the findit command to search for programs and get additional help? for more information about using findit).

Then you can run tabmiss for one or more variables as illustrated below.

tabmiss  schtype 

    schtype |      Freq.     Percent        Cum.
------------+-----------------------------------
  nomissing |         12       75.00       75.00
    missing |          4       25.00      100.00
------------+-----------------------------------
      Total |         16      100.00