A theoretical and empirical review of sequence dissimilarity measures

What is it about?

This article provides a comparative review of the multiple ways of measuring dissimilarities between state sequences. The study examines the sensitivity of distance measures to these three sociologically meaningful aspects analytically and empirically by means of simulations. These aspects concern the sequencing (the order in which successive states appear), the timing and the duration of the spells in successive states. 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.

Featured Image

Why is it important?

Many distance measures have been proposed in sequence analysis for the social sciences. However, the development was scattered and we lack a comprehensive review that would give arguments on how to choose a distance measure in a practical application. The article aims to fill this gap by reviewing the available measures, but also by providing practical guidelines on how to choose one among the different options.

Perspectives