Re-coding 2 race/ethnicity variables

[Andrew Steward]

Hello,
I have a survey with 2 variables related to race/ethnicity I need to re-code.

The first (labeled crl5cgrace) asks: "What race do you consider yourself to be? Code all that apply" with response options:
1) White, 2) Black/African American, 3) American Indian/Alaska Native, 4) Asian, and 5) Native Hawaiian/Pacific Islander.
The second question (labeled crl5hisplatno) asks: "Do you consider yourself Hispanic or Latino?" with response options:
1) Yes, 2) No.

I have learned that best practice (at least in my area) is those who identify as White AND non-Hispanic go into one category, those who identify as Black are coded as Black (regardless of a yes or no on Hispanic), and those who are Hispanic AND not Black/African American are coded as Hispanic. Often Asian, Native American, Pacific Islander, etc., are grouped into one “other category.”

I am wondering how I can recode these 2 variables based on the best practice described above? Is this possible since these are 2 separate variables? For example, I am not sure how I would know that the same respondent identifies as both White and non-Hispanic.

Thank you so much!

[Andrew Musau]

For example, I am not sure how I would know that the same respondent identifies as both White and non-Hispanic.

browse if  crl5cgrace==1 & crl5hisplatn==2

Assumes that White is coded 1 and non-Hispanic is coded 2. To see the coding, e.g., for the first variable

tab crl5cgrace
tab crl5cgrace, nolab

[Bruce Weaver]

Andrew Steward , you have a check (or code) all that apply item with 5 options. Is it not stored as 5 indicator (1=Yes, 0=No) variables in your dataset? If not, how is it stored? Please -dataex- to provide a small sample of (the relevant variables in) your dataset. (See the FAQ for more info about -dataex-.)

Thanks.