Since its introduction in the social sciences by Andrew Abbott in the nineties, Sequence Analysis has been increasingly used to study longitudinally trajectories coded as sequences, i.e. using categorical states. It is even considered a key method for life-course research by many authors (e.g., Shanahan 2000; Mayer 2009; Buchmann and Kriesi 2011; Piccarreta and Studer 2019; Liefbroer 2019; Liao et al. 2022). The standard use of SA revolves around the creation of a typology of trajectories with cluster analysis. This typology allows the identification of recurrent patterns or, in other words, typical successions of states through which the trajectories run. By doing so, it provides a holistic perspective on the trajectories. This typology is then generally used in subsequent analysis, such as regression to identify profiles linked with each type of trajectory.

This symposium regroups methodological developments made within the “Strengthening Sequence Analysis” SNF-funded research project aiming to overcome some of the current limitations of sequence analysis. It includes four presentations preceded by a short introduction. The first compares different approaches to deal with missing data in sequences analysis using the SHP data. It distinguishes several missing data cases linked to longitudinal data collection processes, such as attrition for prospective panels or missing spells in retrospective interviews. The second presentation makes use of newly developed sequences indicators to analyze the evolution of life-satisfaction across the life-course using the SHP data as well. The third presentation proposes a new method to measure the relationship between trajectories and covariates of interest in sequence analysis, which is illustrated through a study of healthcare utilization patterns using the prospective cohort study CoDiab-VD of diabetes patients in the canton of Vaud, Swizterland. The last presentation reviews and compares different methods to create a typology of trajectories using sequences analyses relying this time on the TREE database.