Predicting COVID-19 Vaccination Decision-Making Profiles Among Dutch Adults: From Survey to National Administrative Data
Publication date
2026-05-21
Authors
Abreu, Taymara C.
Bernardo-Garcia, Javier
de Wolf, Isabelle
Matthijssen, Mitchell
Timmermans, Danielle
Buskens, V.W.
Editors
Advisors
Supervisors
Document Type
Article
Metadata
Show full item recordCollections
License
cc_by_nc
Abstract
Individual's vaccination behaviors are influenced by factors such as values and beliefs. Applying latent class analysis (LCA) to such factors from the Longitudinal Internet Studies for the Social Sciences (LISS) panel[1], Matthijssen et al.[2] identified 12 distinct COVID-19 vaccination decision-making profiles in a sample of 2,567 Dutch adults. However, the extent to which membership in these profiles can be predicted using sociodemographic administrative data remains unknown. We assessed that by linking survey data to administrative records from the Statistics Netherlands (CBS) database and employing an Explainable Boosting Machine (EBM) model. The model showed substantial predictive performance and revealed the central role of income, interacting with other variables, in predicting profile membership.
Keywords
COVID-19, Machine Learning, Registry-based study, Biomedical Engineering, Health Informatics, Health Information Management, SDG 3 - Good Health and Well-being
Citation
Abreu, T C, Bernardo-Garcia, J, de Wolf, I, Matthijssen, M, Timmermans, D & Buskens, V 2026, 'Predicting COVID-19 Vaccination Decision-Making Profiles Among Dutch Adults : From Survey to National Administrative Data', Studies in Health Technology and Informatics, vol. 336, pp. 2107-2108. https://doi.org/10.3233/SHTI260628