Philippe Bocquier (DEMO, UCLouvain)
with Ashira Menashe-Oren (DEMO, UCLouvain)
and Carren Ginsburg (University of the Witwatersrand),
Integrating Sequence Analysis into Event History Analysis:
An Application to Migration Histories

The study examines how to integrate sequence analysis in an event history framework by introducing time-varying life histories as a covariate. We suggest the use of time-varying sequence analysis for researchers lacking prior hypotheses, on the condition that it is applied to the entire population at risk. We propose a new methodology to apply sequence analysis to a large population and apply it to Agincourt (South Africa) HDSS longitudinal data. The manual clustering of migration histories based on criteria like duration in each state yields meaningful results. While the second-order classic method suffices for most analyses, it underestimates the recent migrant and adaptation effects and overestimates the return migration effect, because it lacks a circular migrant category. However, the three approaches identify the recent migrant effect, the adaptation effect, and the higher mortality of return migrants. Our empirical study emphasizes the similarity of results from sequence analysis, manual grouping, and the classic second-order approach.

Les présentations auront lieu en présentiel.

Pour participer en ligne (via Zoom), veuillez cliquer ici.