Household members present

Short version: Hi! I’m working with the ATUS and trying to figure out if there is a way to know the number of household members that were in the house at the time of interview. I know I can get data on the number of hh members in general, but I want the specific number during interview time.

Long version: I am interested in finding out how much time household members are in a household at the same time. For example, if x respondent spends 15 hours at home, for how many of those hours were other household members there? I don’t feel that I can answer this question using “WHO” data since this asks more specifically if an activity was done with someone. E.g., if someone was working from home in their bedroom for 3 hours, they will probably say they were alone for this activity, even if their roommate was also in the apartment but in a different bedroom.

Hope this makes sense and thank you!

I agree with your assessment that the WHO data is unlikely to capture household members who are present, but not participating in the activity with the respondent. Unfortunately, there is no other way to track how long other household members were at home during the day if they did not participate in activities with the respondent.

You might however be interested in taking a look at the National Couples’ Health and Time Study (NCHAT). One distinction that it makes is that it asks respondents both who was present during an activity and who was directly involved in it. Note that the NCHAT only asks questions from the main couple and not of any other household members.

Outside the US, there are a number of countries that have their own time use surveys, which are included as part of IPUMS MTUS. Many of these survey multiple household members. Although I’m not certain if any record time use in a way that’s helpful for your research, having multiple respondents might give you a better sense for time spent at home with others.

Hi Ivan,

Thank you so much for your response! I will take a look at NCHAT. The other idea I had was to use other ATUS respondents and, controlling for various characteristics like age, occupation, etc, figure out the average times during the day we expect the respondents to be home (e.g., males that are NILF can usually be found at home from 6 pm to 9 pm). Then, we can match household members to these averages and use that to estimate whether the main respondent’s spouse (e.g., male not in the labor force) was in the household.