<p>Accurate measurement of household composition is fundamental for research on social indicators of well-being, including happiness, work–life balance, and the gender pay gap. In large-scale cross-national surveys, such as the International Social Survey Programme (ISSP), household composition is typically measured using a group design, where respondents distribute household members across three age categories—toddlers (0–5 years), children (6–17 years), and adults (18 + years). When implemented without interviewer assistance, this group design has produced item nonresponse rates as high as 35%. Some ISSP members have adopted an alternative list design, which requires respondents to list each household member sequentially and record their age directly. This study examines how both group and list designs affect the quality of household composition data as the ISSP transitions from interviewer-administered to self-administered modes. Drawing on the question-and-answer process model, cognitive load theory, and satisficing theory, we theorize how the simplified list design improves respondent engagement and reporting consistency. Study 1 uses propensity-score matching to compare Germany’s list design with group designs from Sweden, Switzerland, Australia, Japan, and Norway, all fielded in self-administered paper-and-pencil mode in 2020. Study 2 analyzes three German ISSP waves (2019–2023) to assess under-reporting. Across both studies, the list design substantially reduced item nonresponse and improved reporting consistency. These findings demonstrate that even modest refinements in question design can yield meaningful improvements in data quality and comparability of a key social indicator, supporting the adaptation of long-standing cross-national surveys to self-administered modes, within the ISSP context and beyond.</p>