A number of government, university, and privately-sponsored surveys use within-household proxy reporting to collect information about multiple members of a household from a single household member. The benefits offered by within-household proxy reporting include reduced survey costs and effort (compared to self-response from each household member), making this type of response a potentially advantageous alternative to self response in some situations. However, there are open questions as to whether data collected from one household member answering on behalf of other household members is comparable to the data collected from each individual household member’s self-report. In this research, we interviewed multiple participants from the same household to explore whether factors such as their relationship to each other and topic area (questions on volunteering, civic engagement, Internet and device usage) can inform comparability of responses. This paper discusses the design considerations necessary in planning and executing a research study using paired cognitive interviews in an approach similar to that of Blair, Menon, and Bickart (1991) to evaluate the use of within-household proxy interviewing in surveys conducted with related and unrelated household members. Data from two recent paired cognitive interview projects are used as case studies.
Others in Series
Cognitive Testing the New York City Housing and Vacancy Survey Round 1
Testing of new and revised modules on housing quality, residential history and utilities.
Cognitive Testing the New York City Housing and Vacancy Survey Round 2
Testing questions on housing quality, residential history, and utilities.
Testing of the 2017 School Crime Supplement
Testing of new and revised questions for the School Crime Supplement, focusing on bullying.