Identifying spouses and other family members in Finnish MTUS Sample 2009

TIME USE
1

Hello,

I was interested in a project combining MTUS data from several countries. Concerning the Finnish Sample (2009), I was curious why it’s not possible to identify people beyond hldid (household level). I’ve seen papers reference spouses within households with this data so I was wondering why that isn’t possible via IPUMS and if applying for the original data via Eurostat was necessary to do that.

Thank you so much for your help and time!

2

In the Finland 2009 sample, all individuals age 10 and older within a sampled household are asked to complete two time diaries. Information about each diarist is available, including their age, sex, couple status, cohabitation status, etc. The variable HLDID uniquely identifies each household within a sample. Within each household (HLDID value within a sample), the variable PERSID uniquely identifies each person. Each person has two observations (rows) in a data extract—one for each time diary.

Family interrelationship pointer variables are not available in this sample. There are no directly reported variables providing this information, and IPUMS has not created family interrelationship variables for MTUS data. I would guess that if you see other research referring to spouses, children, parents, etc. using this sample data, the researchers have used respondent age, sex, couple status, cohabitation status, and the household roster to make assumptions about family interrelationships.

Note: The current documentation for the variable SERIAL states that it is meant to uniquely identify each respondent within a sample. In samples where respondents have more than one time diary (such as the Finland 2009 sample), SERIAL does not uniquely identify respondents; instead, it uniquely identifies time diaries. For Finland 2009, and any other sample that contains more than one diary per person, you should use HLDID and PERSID to uniquely identify persons. We are reviewing SERIAL and other identifiers in multi-diary day samples to ensure they are clearly documented to facilitate their appropriate use. Your question prompted us to look into how SERIAL works in these types of samples, and we would like to offer you an IPUMS mug as a small token of our appreciation for helping us improve IPUMS data. You can send us an email at ipums@umn.edu with your mailing address.

3

Thank you! That was my original assumption going in, so it’s good to get that confirmed.