Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers
Frontiers in Psychology · 02.11.2026 · 17
DOI: 10.3389/fpsyg.2026.1910559 ↗Kurz gefasst
Fachartikel in der Zeitschrift Frontiers in Psychology, online erschienen am 02.11.2026. Verfasst von Amanda Patricia Martínez-Yacelga und 4 weiteren Autor:innen. Beteiligte Einrichtung: Facultad de Ciencias Económicas, Administrativas y de Negocios, Universidad Tecnológica Indoamérica, Ambato, Ecuador. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Organizational trust reflects employees' willingness to accept vulnerability on the basis of positive expectations about supervisors and the organization. This study examined whether organizational cynicism, organizational climate, job satisfaction, and worker characteristics contain reproducible information for distinguishing high organizational trust among workers in Ecuador. A quantitative, non-experimental, cross-sectional design was applied to 899 valid survey records obtained through non-probability availability sampling. The focal construct was the 12-item Organizational Trust Inventory, comprising trust in the immediate supervisor and trust in the organization. High trust was defined a priori as a mean score of at least 5.0, the first response anchor labeled “High,” yielding 555 high-trust and 344 non-high-trust cases. Ordinal confirmatory factor analysis supported the proposed measurement structures; composite reliability ranged from 0.833 to 0.965, and the four-construct measurement model showed good fit ( CFI = 0.991, TLI = 0.991, RMSEA = 0.036). Logistic regression, Random Forest, XGBoost, LightGBM, and CatBoost were compared through nested cross-validation, training-only hyperparameter tuning, a held-out test set, and bootstrap uncertainty intervals. Logistic regression achieved the highest nested cross-validated ROC-AUC (0.865), while CatBoost obtained the highest test ROC-AUC (0.855); paired bootstrap intervals indicated that these differences were not statistically distinguishable. Satisfaction, cognitive cynicism, and organizational climate were the most informative predictor groups. A stability-validated two-profile solution separated a favorable profile with 84.4% high trust from a critical profile with 34.1%. The results support cautious organizational diagnosis and hypothesis generation, not employee-level automated decisions. An incremental benchmark compared demographic, satisfaction-only, cynicism-only, climate-only, and combined predictor sets under the same repeated nested cross-validation protocol; job satisfaction alone reached a mean nested ROC-AUC of 0.843, and adding cynicism and climate raised it to 0.875, so the integrated framework yields a modest rather than transformative increment. High trust is an analytically defined category, not a diagnostic threshold, and the held-out set is drawn from the same convenience sample, so the evidence constitutes internal out-of-sample validation rather than external validation.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Facultad de Ciencias Económicas, Administrativas y de Negocios, Universidad Tecnológica Indoamérica, Ambato, Ecuador.
- Escuela Profesional de Contabilidad, Universidad Peruana Unión, Lima, Peru.
- Universidad Señor de Sipán, Chiclayo, Peru.
Zitieren
APA
Martínez-Yacelga, A. P., Haro-Sarango, A. F., Sánchez-Villena, A. R., Siesquen-Custodio, V. M., & Cercado-Vásquez, E. R. (2026). Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers. Frontiers in Psychology, 17. https://doi.org/10.3389/fpsyg.2026.1910559
Vancouver
Martínez-Yacelga AP, Haro-Sarango AF, Sánchez-Villena AR, Siesquen-Custodio VM, Cercado-Vásquez ER. Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers. Frontiers in Psychology. 2026;17. doi:10.3389/fpsyg.2026.1910559
BibTeX
@article{MartnezYacelga2026,
author = {Martínez-Yacelga, Amanda Patricia and Haro-Sarango, Alexander Fernando and Sánchez-Villena, Andy Rick and Siesquen-Custodio, Victor Manuel and Cercado-Vásquez, Erika Roxana},
title = {Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers},
journal = {Frontiers in Psychology},
year = {2026},
volume = {17},
doi = {10.3389/fpsyg.2026.1910559}
}
EUVIMED