[ONLINE PRESENTATION] Measuring Infrastructural Deprivation from Big Geodata: On the Inferential Validity of OpenStreetMap Proxies

Citizens' political attitudes are shaped by the local infrastructures and the built environment they encounter in everyday life – but do we always proxy their availability and accessibility correctly? In this talk, I present joint work with Daria Dementeva (Luxembourg Institute of Socio-Economic Research, LISER), in which we take a look at OpenStreetMap (OSM), a large geodata source that provides high-resolution built-environment data at no cost. Because it is crowdsourced, however, its quality for substantive research remains insufficiently validated, and establishing the validity of these measures has become an important methodological challenge. We present a hands-on tool and a validation framework to integrate OSM in your research,  OpenStreetMap-based measures of the availability and accessibility of hospitals, schools, and supermarkets with official points-of-interest data, using Germany as a test case. We examine both municipality-level measures of availability and individual-level measures of proximity, and evaluate whether differences between geodata sources affect substantive conclusions about citizens' perceptions of infrastructural deprivation, drawing on georeferenced survey data from the 2021 German Longitudinal Election Study.