In a phone survey, asking individuals to give data on other members of their household is a useful way to collect data on larger numbers of people or on specific populations (such as married women or students), but it requires listing household members in a roster to determine who is being described. It is uncommon to include household rosters in phone surveys because respondents typically are not willing to spend as much time on the phone as they are in face-to-face interviews. In a recent project, IPA Colombia found that household rosters can be conducted in a short amount of time: an average of less than 3.5 minutes for households that averaged 4.1 members. This provided an opportunity to target questions to and/or about specific household members in order to create a richer, more comprehensive dataset.