Poset sensitivity analysis reveals post 2020 changes in human development index components
This study assesses the sensitivity of the Human Development Index (HDI) using partially ordered set (poset) analysis. Unlike conventional methods, such as correlation, dominance, or simulation-based techniques, poset analysis maintains the multidimensional structure of the HDI and captures cases of...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2026 |
| País: | España |
| Institución: | Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya) |
| Repositorio: | Recercat. Dipósit de la Recerca de Catalunya |
| OAI Identifier: | oai:recercat.cat:20.500.14342/5984 |
| Acceso en línea: | https://hdl.handle.net/20.500.14342/5984 https://doi.org/10.1016/j.isci.2026.114964 |
| Access Level: | acceso abierto |
| Palabra clave: | Data analysis Social sciences Research methodology social sciences Ciències socials Conjunts parcialment ordenats Investigació--Metodologia 510 |
| Sumario: | This study assesses the sensitivity of the Human Development Index (HDI) using partially ordered set (poset) analysis. Unlike conventional methods, such as correlation, dominance, or simulation-based techniques, poset analysis maintains the multidimensional structure of the HDI and captures cases of incomparability, offering a more nuanced understanding of variable influence. Applying this method to HDI data from 1990 to 2023, the analysis shows that expected years of schooling have been the most sensitive variable historically. However, since 2020, life expectancy has gained prominence, particularly during the COVID-19 pandemic, with the greatest impact observed in very high HDI countries. This shift reveals how external shocks can alter the comparative structure of development rankings. By focusing on structural changes rather than marginal effects, poset-based sensitivity analysis provides valuable insights for refining HDI methodology and informing policy strategies. It offers a robust alternative for evaluating indicator influence without relying on arbitrary weighting assumptions. |
|---|