What Matters in Differences Between Life Trajectories: A Comparative Review of Sequence Dissimilarity Measures

Abstract

Summary. This is a comparative study of the multiple ways of measuring dissimilarities between state sequences. The originality of the study is the focus put on the differences between sequences that are sociologically important when studying life courses such as family life trajectories or professional careers. These differences essentially concern the sequencing (the order in which successive states appear), the timing and the duration of the spells in successive states. The study examines the sensitivity of the measures to these three aspects analytically and empirically by means of simulations. Even if some distance measures underperform, the study shows that there is no universally optimal distance index, and that the choice of a measure depends on which aspect we want to focus on. From the review and simulation results, the paper derives guidelines to help the end user to choose the right dissimilarity measure for her or his research objectives. This study also introduces novel ways of measuring dissimilarities that overcome some flaws in existing measures.

Comparing methods of classifying life courses: sequence analysis and latent class analysis

“…These metrics measure the distance between sequences by counting the number of (weighted) common subsequences. For a detailed review of distance metrics for SA, we refer to Robette and Bry (2012), Studer (2012) and Studer and Ritschard (2016). In our illustration, we only present the SA approach with an OM-metric.…”

The diversity in longitudinal partnership trajectories during the transition to adulthood: How is it related to individual characteristics and regional living conditions?

“…Optimal matching (OM) is chosen because this sequence dissimilarity measure provides a flexible alternative compared to, e.g., the two extreme cases of the Levenshtein II and the Hamming distance. While the Hamming distance applies only to pairs of sequences of the same length and is very sensitive to timing mismatches (Studer and Ritschard 2016), OM makes it possible to compare sequences that are partly similar but shifted (Scherer and Bruederl 2006). However, the Levenshtein distance is sensitive to spell durations and sequencing, as it can be represented as the count of the elements in each sequence that is not involved in the longest common subsequence (ibid.).…”

“…However, the Levenshtein distance is sensitive to spell durations and sequencing, as it can be represented as the count of the elements in each sequence that is not involved in the longest common subsequence (ibid.). OM is thus a very flexible dissimilarity measure that can cope with many situations (Studer and Ritschard 2016).…”

Research on Students' Performance in Higher Education through Sequence Analysis

“…Even if their approach of this individual is different, as we have already seen, they can be considered as two complementary ways to study him. In addition some more recent papers, as those presented in this conference by Studer et al (2016) or Rossignon et al (2016), combines the advantages of the two approaches modeling "the relationships between time varying covariates and trajectories specified as processes outcomes that unfold over time". The definition given by Courgeau and Lelièvre (1997) of event history analysis appears to be also valid for sequence analysis: "Throughout his or her life an individual follows a complex itinerary, which at a given moment depends on the life course to date and on the information acquired in the past".…”