|
|
|
relt
Notes from the GHS
Question
I am exploring the General Household Survey to look
at interethnic marriage in Britain and I've got some questions regarding some
variables in the dataset. Since I'll need to match couples in the same household,
I intend to use the variables RELTO01 - RELTO14 which are Filled relationship
grid. But I'm not quite sure whose relationship the variables refer to. For
example, in RELT001, does it refer to the relationship between person 1 and
person 2? I'm also wondering if Person Number is unique within a household.
Since I find in the dataset that sometimes individuals in the same serial
number (HSERIAL) have exactly the same person number even though they appear
to be different person. It would be very beneficial if you could suggest the
best method to match husband and wife within a household.
Answer
Relto01 is the relationship to person
number 1, Relto02 is the relationship to person number 2 and so on. Person
number is obtained from the variable 'Persno'. So for example, in household
number (hserial) 1001011 there are two people who are a married couple (persno
1 and persno 2). For persno 1, Relto01 = 0 (self) and Relto02 = 1 (spouse).
For persno 2, Relto01 = 1 (spouse) and Relto02 = 0 (self). In a larger household
this would continue in the same fashion. In relation to your question about
persno being unique within hserial, this variable should be unique. I've checked
one dataset for error and didn't find any. If you've found a duplicate, could
you let me know the hserial, persno and which file you found it in, and I'll
check it out.
|
|
|
