Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers
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 in...
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Artículo |
| Lenguaje: | inglés |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | Enlace permanente (DOI o Handle) Ver en el sistema de origen |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1881700533645869056 |
|---|---|
| author | Amanda Patricia Martínez-Yacelga Alexander Fernando Haro-Sarango Andy Rick Sánchez‐Villena Victor Manuel Siesquen-Custodio Erika Roxana Cercado-Vásquez |
| author_browse | Alexander Fernando Haro-Sarango Amanda Patricia Martínez-Yacelga Andy Rick Sánchez‐Villena Erika Roxana Cercado-Vásquez Victor Manuel Siesquen-Custodio |
| author_facet | Amanda Patricia Martínez-Yacelga Alexander Fernando Haro-Sarango Andy Rick Sánchez‐Villena Victor Manuel Siesquen-Custodio Erika Roxana Cercado-Vásquez |
| author_sort | Amanda Patricia Martínez-Yacelga |
| collection | Producción científica |
| description | 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. |
| format | Article |
| id | openalex:W7225111092 |
| institution | UPeU |
| language | English |
| publishDate | 2026 |
| publishDateRange | 2026 |
| publishDateSort | 2026 |
| record_format | dspace |
| spelling | openalex:W7225111092Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workersAmanda Patricia Martínez-YacelgaAlexander Fernando Haro-SarangoAndy Rick Sánchez‐VillenaVictor Manuel Siesquen-CustodioErika Roxana Cercado-Vásquezorganizational trusttrust in supervisororganizational cynicismorganizational climatejob satisfactionemployee characteristicsexplainable machine learningCatBoostrandom forestXGBoostLightGBMlogistic regressionnested cross-validationprofile analysisJob Satisfaction and Organizational BehaviorEmployee Performance and ManagementPsychometric Methodologies and TestingOrganizational 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.Articlehttps://doi.org/10.3389/fpsyg.2026.1910559https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2026.1910559/pdfW7225111092enFrontiers in Psychology1664-10782026-11-02 |
| spellingShingle | organizational trust trust in supervisor organizational cynicism organizational climate job satisfaction employee characteristics explainable machine learning CatBoost random forest XGBoost LightGBM logistic regression nested cross-validation profile analysis Job Satisfaction and Organizational Behavior Employee Performance and Management Psychometric Methodologies and Testing Amanda Patricia Martínez-Yacelga Alexander Fernando Haro-Sarango Andy Rick Sánchez‐Villena Victor Manuel Siesquen-Custodio Erika Roxana Cercado-Vásquez Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers |
| title | Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers |
| title_full | Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers |
| title_fullStr | Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers |
| title_full_unstemmed | Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers |
| title_short | Organizational trust in the workplace: explainable machine learning evidence from Ecuadorian workers |
| title_sort | organizational trust in the workplace explainable machine learning evidence from ecuadorian workers |
| topic | organizational trust trust in supervisor organizational cynicism organizational climate job satisfaction employee characteristics explainable machine learning CatBoost random forest XGBoost LightGBM logistic regression nested cross-validation profile analysis Job Satisfaction and Organizational Behavior Employee Performance and Management Psychometric Methodologies and Testing |
| topic_facet | organizational trust trust in supervisor organizational cynicism organizational climate job satisfaction employee characteristics explainable machine learning CatBoost random forest XGBoost LightGBM logistic regression nested cross-validation profile analysis Job Satisfaction and Organizational Behavior Employee Performance and Management Psychometric Methodologies and Testing |
| url | https://doi.org/10.3389/fpsyg.2026.1910559 https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2026.1910559/pdf |
| work_keys_str_mv | AT amandapatriciamartinezyacelga organizationaltrustintheworkplaceexplainablemachinelearningevidencefromecuadorianworkers AT alexanderfernandoharosarango organizationaltrustintheworkplaceexplainablemachinelearningevidencefromecuadorianworkers AT andyricksanchezvillena organizationaltrustintheworkplaceexplainablemachinelearningevidencefromecuadorianworkers AT victormanuelsiesquencustodio organizationaltrustintheworkplaceexplainablemachinelearningevidencefromecuadorianworkers AT erikaroxanacercadovasquez organizationaltrustintheworkplaceexplainablemachinelearningevidencefromecuadorianworkers |