Creating Age Groups from Age Variable

[jamuriuk]

Hello team,

i am trying to group the variable "age" to assign a dummy with groupings 0-14 years, 15-65 years and above 65 years. My variable representing age is z2. How do I set the codes please?

James

[WarrenKuhfeld]

x1 = 0 le z2 le 14;

x2 = 15 le z2 le 65;

x3 = z2 gt 65;

[jamuriuk]

Thank you although I did not want to add extra variables x1, x2 and x3. I have apparently used

if age <= 15 then z2=1;else if 12 <= age < 65 then z2=2;else if age >=65 then z2=3; and it has worked.

Thanks so much.

James

[Jim_G]

 proc format; value  age
 
 0-14='0-14  '
 15-65='15-65'
  65-120='gt 65';
  
  
  data;   input name $ age ;
  
  format age age.;
  
  cards;
  mary 12
  joe  60
  ;
  
  proc print;  run

 proc format; value  age
 
 0-14='0-14  '
 15-65='15-65'
  65-120='gt 65';
  
  
  data;   input name $ age ;
  
  format age age.;
  
  cards;
  mary 12
  joe  60
  ;
  
  proc print;  run

 proc format; value  age
 
 0-14='0-14  '
 15-65='15-65'
  65-120='gt 65';
  
  
  data;   input name $ age ;
  
  format age age.;
  
  cards;
  mary 12
  joe  60
  ;
  
  proc print;  run

You can use proc format   like this;

 proc format; value  age
 
 0-14='0-14  '
 15-65='15-65'
  65-120='gt 65';
  
  data;   input name $ age ;
    format age age.;
   cards;
  mary 12
  joe  60
  ;
 
  proc print;  run 

Jim

[ballardw · Accepted solution]

If you use formats such as @Jim_G suggested then you can avoid adding variables to the data set for almost any analysis tasks. The groups assigned by a format will be used in almost all of the analysis procedures and have the very signficant advantage of if you want to see what the difference may be when you change the boundary for one of the categories you make a new custom format and use that in the procecure code for the age variable. No added variables and if you data is largish then you don't spend extra time with multiple data sets.