Modify variable categories' name

[Darina Proshenska]

Hello,

I would like to ask if I could change variable categories' name.

I have this variable - Marital status with possible responses: 1="married-first and only marriage", 2="remarried-second or later marriage", 3="separated", 4="divorced", 5="widowed" and 6="have never married". I've already regrouped the first two (2=1) and the other ones into one single category. Then, I've created a dummy variable with 1=married and 0=single. But, I couldn't change the category's name. When I write: tab var or tab var, nol I can see that STATA automatically put the first two labels - 1="married-first and only marriage" and 2="remarried-second or later marriage" and it seems that I didn't create the dummy variable.

So my question is how to rename the first two categories and why I can't see the dummy variable created.

Please find herewith the code that I used to regroup and create a dummy variable. Also, I have tried to rename the categories, but it didn't work at all.

*Recode Marital status
tab e_mar_stat
tab e_mar_stat, nol
recode e_mar_stat (2 = 1)
recode e_mar_stat (6 = 3) (5 = 3) (4 = 3)
recode e_mar_stat (3 = 2)
tab e_mar_stat, nol
*Dummy for marital status
generate married=.
replace married=1 if e_mar_stat==1
generate single=.
replace single=0 if e_mar_stat==2
label variable e_mar_stat "Marital status"
label define mscats 1 "Maried" 2 "Single", modify
label values e_mar_stat mscats

Cheers,
Darina

[Raffaele Grotti]

Hi Darina,
I can't see why you labeling doesn't work - seeing what error you got from Stata would help.

Also, you haven't created one dummy variable that assumes value 1 if married and value 0 if single, but two non-dummy variables: the fist equal to 1 if married and missing otherwise; and the second equal to 0 if single and missing otherwise.

This should work for you.

recode e_mar_stat (1 2 = 1 "Married") (3/6 = 2 "Single"), gen(e_mar_stat_new)
label variable e_mar_stat_new "Marital status"

gen married = e_mar_stat_new==1

Raffaele

[Darina Proshenska]

Hi Raffaele,

Thank you very much for your response. I have no error message, but I know that it is not working because the dummy variable has not been created and STATA automatically put the first labels that are showed in the table. I have tried your code, but now I have this error message :

recode e_mar_stat (1 2 = 1 "Married")
Rules defining value labels not allowed when overwriting a variable

I have no idea why it is not working.

Cheers,
Darina

[Marcos Almeida]

Instead of regrouping, say, the levels 2,3,4 of a 4-level variable, you may just - gen mydummy = oldvar > 1.

That said, since this is just your second post in the forum, I strongly suggest to read the FAQ, particularly the recommendation to share command/output/data by using code delimiters or under - dataex - command (or SSC, for older versions of Stata).

[Darina Proshenska]

Thank you very much, Marcos, for this information. Consider it done.

Cheers,
Darina