Angelo León
Angelo León has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Cristian Cáceres
Facultad de Ciencias de la Salud, Departamento de Psicología, Universidad Católica del Maule, Curicó, Chile
1 shared publication Luis Yáñez-Sandivari
1 shared publication- Sebastian A. Rios
University of Chile
1 shared publication
Latest Publications
- Open Access
- 17 September 2026
- Published by Elsevier in SSRN Electronic Journal
From Psychosocial Explanation to Behavioral Prediction: Anticipating Donation Frequency and Resource Type in Post-Disaster Humanitarian Logistics
Post-disaster humanitarian logistics faces major challenges related to uncertainty, volatile aid flows, and material convergence. Although donation prediction has advanced at aggregate and organizational levels, individual-level behavioral and psychosocial determinants remain insufficiently integrated into predictive frameworks. This study proposes a predictive extension of a previously validated socio-technical framework, integrating outcome-free SEM-derived latent constructs, donor profiles, and questionnaire variables into a machine learning layer to anticipate donation frequency and resource type.The empirical analysis uses 362 valid observations from adults in rural municipalities of the Maule Region, Chile. Several predictor configurations and model families were evaluated, including multinomial logistic regression, ensemble models, and benchmark classifiers, complemented by SHAP-based interpretability analysis. Results show that psychosocial variables contain meaningful predictive signal. For donation frequency, the best-performing models achieve Quadratic Weighted Kappa (QWK) values close to 0.30. For donation type, Random Forest models using classifier chains obtain the strongest performance, with a macro-averaged F1 score close to 0.59. SHAP results indicate that frequency is mainly discriminated by granular resilience and motivation items, whereas perceived constraints and donation-type dependencies help predict resource composition.These findings provide behavioral inputs for anticipating donation intensity, resource composition, and aid-flow congestion in post-disaster contexts.
- 17 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessFrom Psychosocial Explanation to Behavioral Prediction: Anticipating Donation Frequency and Resource Type in Post-Disaster Humanitarian Logistics
Post-disaster humanitarian logistics faces major challenges related to uncertainty, volatile aid flows, and material convergence. Although donation prediction has advanced at aggregate and organizational levels, individual-level behavioral and psychosocial determinants remain insufficiently integrated into predictive frameworks. This study proposes a predictive extension of a previously validated socio-technical framework, integrating outcome-free SEM-derived latent constructs, donor profiles, and questionnaire variables into a machine learning layer to anticipate donation frequency and resource type.The empirical analysis uses 362 valid observations from adults in rural municipalities of the Maule Region, Chile. Several predictor configurations and model families were evaluated, including multinomial logistic regression, ensemble models, and benchmark classifiers, complemented by SHAP-based interpretability analysis. Results show that psychosocial variables contain meaningful predictive signal. For donation frequency, the best-performing models achieve Quadratic Weighted Kappa (QWK) values close to 0.30. For donation type, Random Forest models using classifier chains obtain the strongest performance, with a macro-averaged F1 score close to 0.59. SHAP results indicate that frequency is mainly discriminated by granular resilience and motivation items, whereas perceived constraints and donation-type dependencies help predict resource composition.These findings provide behavioral inputs for anticipating donation intensity, resource composition, and aid-flow congestion in post-disaster contexts.