Using Sequence Analysis to Uncover Patterns of Social Trajectories

Using Sequence Analysis to Uncover Patterns of Social Trajectories
Speaker: Jacques-Antoine Gauthier # Université de Lausanne / UoE
Hosted by
Introduced by
Date and Time
25th Jan 2018 13:00 - 25th Jan 2018 14:15
Practice Suite 1.12, Chrystal Macmillan Building

In social sciences, the increasing need for quantitative longitudinal data during the last decades is associated with the development of new analytical strategies to capture time related phenomenon. In this seminar, Jacques-Antoine Gauthier will walk us through one of these methods, sequence comparison.

It is an approach that offers powerful tools to adequately perform pattern searches in these newly available data. In social sciences, as there is no such strong theoretical model as the evolutionary one, the variability of individual careers is huge, making the search for typical patterns more relevant than that of developmental laws. Interdependence, enchainment, order and convergence are here key issues to understand how social phenomenon are shaped over time. Methodologically, the standard approach consist in comparing strings of symbols using (pairwise) alignment techniques, followed by cluster analysis to build up typologies of (possibly multidimensional) sequences of states in the spheres of family, occupation, place of residence, health, ... Eventually, the types are contextualized using logistic regression and or multiple correspondence analyses techniques. Recent methodological framework aim at estimating the effect of past trajectories on the chances of experiencing a life course event. 

This event will take place on Thursday 25th January from 1pm-2.15pm in the 1st floor Practice Suite of the Chrystal Macmillan Building.

Three students looking at laptops