Generating a new variable (gen)

[Gabi Costa]

Hello friends,

I am attempting to generate and regroup a variable "jbstat" into three new variables (employed, unemployed, and neither). Unfortunately, the values do not fall consecutively, so I cannot recode through the greater than, less than or equal to function. Any help will be greatly appreciated (I have attached the code I have written below - the green text represents the labels that belong in each section)!

Regards,
Gabi

[Daniel Schaefer]

Hi Gabi, I may be missing something, but I am not seeing any attachment. can you please provide a data example using -dataex-? It also helps if you include your code in the body of a comment rather than an attached document. Remember to surround it in CODE tags.

[William Lisowski]

Expanding on Daniel's advice, to present code and results readably, please copy them from the Results window and paste them into a code block in the Forum editor using code delimiters [CODE] and [/CODE], as explained in the Statalist FAQ linked to at the top of the page. For example, the following:

[CODE]
// sample code
sysuse auto, clear
describe
[/CODE]

will be presented in the post as the following:

// sample code
sysuse auto, clear
describe

[Gabi Costa]

gen employed = .
replace employed = 1 if jbstat>= & jbstat<
replace employed = 0 if jbstat< | jbstat>=



gen unemployed = .
replace unemployed = 1 if jbstat>=3 & jbstat<4
replace unemployed = 0 if jbstat<3 | jbstat>=4


gen neither = .
replace neither = 1 if jbstat>= & jbstat<
replace neither = 0 if jbstat< | jbstat>=


*EMPLOYED/// 1, 2, 5, 6, 9, 10, 11
*UNEMPLOYED/// 3
*NEITHER/// 4, 7, 8, 97

[Daniel Schaefer]

Hi Gabi,

If I am interpreting this correctly, I believe you want something like this:

gen employed = 0
replace employed = 1 if jbstat == 1 | jbstat == 2 | jbstat == 5 | jbstat == 6 | jbstat == 9 | jbstat == 10 | jbstat == 11

gen unemployed = 0
replace unemployed = 1 if jbstat == 3

gen neither = 0
replace neither = 1 if jbstat == 4 | jbstat == 7 | jbstat == 8 | jbstat == 97

EDIT: Just to clarify, do you mean that the comment lines at the bottom represent the *values* of jbstat that should correspond to one for each of these dummy variables?

[Daniel Schaefer]

I am also guessing that when jbstat equals 97, it is missing. Based on your post, I think it is clear you want missing values assigned to "neither". You could of course alternatively preserve missingness on jbstat like so:

gen employed = 0
replace employed = 1 if jbstat == 1 | jbstat == 2 | jbstat == 5 | jbstat == 6 | jbstat == 9 | jbstat == 10 | jbstat == 11
replace employed = . if jbstat == 97

gen unemployed = 0
replace unemployed = 1 if jbstat == 3
replace unemployed = . if jbstat == 97

gen neither = 0
replace neither = 1 if jbstat == 4 | jbstat == 7 | jbstat == 8
replace neither = . if jbstat == 97

Naturally, the best solution depends on a clear understanding of why the value is missing. If this were survey data and missingness indicated that the person had not responded, I might prefer to impute the missing values rather than assigning them to 'neither'. This does, of course, depend on your particular research question.

Regardless, you might also want to use -checkvar- to confirm that the new variables match your expectations.

checkvar jbstat employed unemployed neither

[Gabi Costa]

Hi Daniel,

The first one was just what I needed - thank you very much!

I have already dropped the missing, inapplicable, proxy, refused and 'don't know' observations, the 97 variable represented 'doing something else'.

Best,
Gabi

[William Lisowski]

Note that a little typing can be saved using the inlist() function.

replace employed = 1 if jbstat == 1 | jbstat == 2 | jbstat == 5 | jbstat == 6 | jbstat == 9 | jbstat == 10 | jbstat == 11

can be shortened to

replace employed = 1 if inlist(jbstat, 1, 2, 5, 6, 9, 10, 11)

and since inlist() returns a 0/1 value

gen employed = 0
replace employed = 1 if jbstat == 1 | jbstat == 2 | jbstat == 5 | jbstat == 6 | jbstat == 9 | jbstat == 10 | jbstat == 11

can be shortened to

generate employed = inlist(jbstat, 1, 2, 5, 6, 9, 10, 11)

[Daniel Schaefer]

Awesome, thanks!