Validating survey measures of acquaintance networks for political behavior research

Survey research examining social influence in political behavior has often relied on name generators to measure individuals’ social networks. As the term implies, name generators ask respondents for a list of their network ties—these batteries often focus on core networks, or the handful of intimate contacts with whom the individual discusses important matters. Recent work has broadened its focus from these small, dense, core networks to larger, but less intimate acquaintance networks; this is the broader set of people an individual knows on a first-name basis, and these contacts have been measured with aggregated relational data (ARD) batteries. In this paper we use data from the Civic Health and Institutions Project (CHIP50) and a series of Cooperative Election Study (CES) modules fielded in 2018 and 2024 to examine the reliability and validity of acquaintance network data. Our analysis pays particular attention to points of similarity and distinction between ARD measures and standard name generators, and we close by offering some suggestions to researchers studying patterns of interpersonal interaction in mass publics.