Integrating Conventional Population Surveys with Individual Device Data

Develop robust, generalizable methodologies to integrate data from conventional population surveys with anonymized individual-level device datasets (for example, mobile phone GPS traces) in a manner that accommodates variability across information types and preserves privacy, enabling reliable joint analysis of behavioral and socioeconomic indicators.

Background

The paper highlights the growing use of large-scale behavioral data (e.g., mobile phone GPS) alongside traditional population surveys to study social and economic dynamics. While each source offers complementary strengths, combining them is challenging due to differences in granularity, privacy constraints, and the types of information they contain.

Within this context, the authors explicitly note that integrating data collected via conventional methods (such as surveys) with individual device data remains an open challenge, particularly because the integration requirements vary depending on the information categories being merged. This underscores the need for methodological advances that can bridge these data sources consistently and responsibly.

References

However, integrating data from conventional data collection methods (e.g., population surveys) with data gathered from individual devices remains an open challenge, the nature of which varies depending on the types of information being integrated.

— Job loss disrupts individuals' mobility and their exploratory patterns  (2403.10276 - Centellegher et al., 2024) in Introduction