I plan to use ACS 2010 to determine the degree of exogamy/endogamy of Franco-Americans, overall and stratifying for year of birth and year of marriage. I have looked at various variables and they seem straightforward except one. It seems to me that I could identify couples with the RELATE variable, where 1 is the head of household, and 2 the spouse. What I don’t understand is the SPLOC variable.

I would exclude the households where there is more than one couple, to simplify. Do I then need to use SPLOC, or RELATE would be enough?


Jacques Pepin

RELATE captures the relationship to a single household reference person in whose name the housing unit is owned or rented. Using RELATE will omit couples who do not include the reference person (e.g., an older householder living with their adult child and child-in-law). SPLOC iterates through all the different combinations of RELATE values (with consideration for marital status, age, and sex) to identify a person’s probable coresident spouse. Using SPLOC will increase the number of couples and capture a more representative sample of couples.