Promise and perils of population-scale social network analysis

On Thursday 26 October, POPNET PI Frank Takes will present his work online for the IAS lecture series of the Institute for Analytical Sociology in Norrköping, Sweden.

Abstract

A relatively recently emerged line of research is dedicated to harnessing large-scale population register data to address enduring questions within the realm of social science. In this presentation, we will specifically delve into the network dimension of such data, with a focus on data from the POPNET project, covering over 17 million people (i.e., the entirety of the Netherlands) and their ca. 800M family, household , school, work and next-door neighbor connections. We highlight the potential inherently present in this type of curated and comprehensive social network data through illustrative examples of results related to topics such as social capital, segregation, and migration. Furthermore, we will explore several methodological considerations and challenges related to under- and oversampling of individual connections through opportunity structures, including findings on the veracity of skewed degree distributions in the real world. We also highlight network-structure related concerns w.r.t. privacy, in particular disclosure risk and more in general the anonymization of network data for scientific research.