Humboldt-Universität zu Berlin - Ressourcenökonomie

Sustainability research often seeks to transfer insights across cases, but the heterogeneity of contexts and outcomes presents significant challenges. What works in one case may fail in another, requiring an approach that combines classification with explanation. Classifying cases into several types, each with a particular explanation, however, involves a trade-off between too broad and too fine-grained classes. Existing studies often address this trade-off in ways that are difficult to reproduce, highlighting the need for more systematic and replicable methods. To address this gap, this study develops quantitative metrics and standard procedures that are replicable across contexts. They enable identifying archetypes from binary data sets using formal concept analysis (FCA). The novel procedures are demonstrated by replicating two previously published archetype analyses on land use and climate change adaptation. We propose three core steps (formal concept analysis, concept filter, theoretical analysis) alongside two optional steps (grouping, optimal concept selection). Key metrics, including consistency, coverage, richness, size, and lift, guide these steps. We show that the procedures enhance reproducibility and speed up analysis compared with previous approaches, and help determine the appropriate number of archetypes to provide more parsimonious research findings. This study thus contributes to methodological rigor in case-based sustainability research by balancing generality and particularity of archetype analysis.

Eisenack, K., and R. Wang. 2026. Divide and explain: novel metrics and procedures for archetype analysis in case-based sustainability research. Ecology and Society 31(1):12. [online] URL: https://www.ecologyandsociety.org/vol31/iss1/art12