Crossing Levels: The Potential for Numerical Taxonomy and Fuzzy Set Approaches to Study Multi-Level Longitudinal Change

Abstract

This article demonstrates how we might explore trajectories of complex systems through time when we are dealing with multiple levels of systems with intersecting causal propensities. Using ‘social exclusion’ as an illustrative example, the objective is to develop a methodological approach, using available secondary data sources, that transcends the linear and uni-directional faults of conventional multi-level modelling. This alternative approach consists of a sequential combination of cluster analysis, tallying, documentary analysis and qualitative comparative analysis (QCA). The authors argue that this alternative methodology allows for both a descriptive and causal explanation of change and continuity within and between multiple levels of observation, whilst also stressing the importance of context and the possibility of all configurations of change and continuity within and between the different levels to be considered.