EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Explaining the Links Between School Administrator Leadership, Job Satisfaction, and Participatory School Climate: A Machine Learning-Enhanced Multilevel Analysis of TALIS 2024 School Administrator Data

Dönüş Şengür iD

Behavioral Sciences · 26.06.2026 · 16 (7) · 1062

DOI: 10.3390/bs16071062 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Behavioral Sciences, online erschienen am 26.06.2026. Verfasst von Dönüş Şengür. Beteiligte Einrichtung: Educational Science Department, Education Faculty, Firat University, Elazig 23200, Türkiye. Die Arbeit nennt 64 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Fırat University, Firat University, Scientific Research Project Committee.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

A participatory school climate refers to the involvement of school administrators, teachers, and other school members in decision-making processes, their sharing of responsibility, and their collaborative work for school improvement. Since this climate can be related to individual, organizational, and contextual factors such as leadership, job satisfaction, diversity beliefs, workload, well-being, and national context, identifying the key variables that support a participatory school environment is important. This study used TALIS 2024 school administrator data to identify the main predictors of participatory school climate and examined the mediating role of school administrator job satisfaction in the relationship between school administrator leadership, used here in line with school principal leadership, and participatory school climate. The research is based on a two-stage analytical framework. In the first stage, explanatory machine learning analysis was conducted by comparing Elastic Net, Random Forest, and XGBoost models; the relative significance levels of the variables were evaluated using permutation importance and SHAP methods. In the second stage, mediation analysis was performed using multi-level linear mixed models, considering clustering at the national level; the indirect association was evaluated using bootstrap confidence intervals. The analyses were conducted using data from 16,335 school administrators. The findings showed that the highest prediction performance was produced by the XGBoost model and that model performance improved with the inclusion of the country variable. Explainability analyses indicated that school administrator leadership was the strongest predictor of participatory school climate, followed by job satisfaction and diversity beliefs. Multilevel models suggested that the association between school administrator leadership and participatory school climate was consistent, with an indirect pathway through school administrator job satisfaction; bootstrap findings also supported the statistical stability of this indirect association. These findings suggest that a participatory school climate is associated not only with individual perceptions but also with multifaceted conditions such as leadership, job satisfaction, inclusivity, and country context. By combining explanatory machine learning with multilevel statistical modeling, this study identifies variables associated with participatory school climate and examines an indirect association among leadership, job satisfaction, and participatory climate. Because TALIS survey weights and the full complex sampling design were not incorporated, the findings should be interpreted as associations observed in the pooled analytical sample rather than as population-representative estimates for participating education systems.

Schlagwörter

Stichwörter der Autor:innen

TALIS 2024explainable machine learningjob satisfactionmultilevel mediation analysisparticipatory school climateschool administrator leadership

Autor:innen und Institutionen

  • Educational Science Department, Education Faculty, Firat University, Elazig 23200, Türkiye.

ORCID-Profile: Dönüş Şengür

Förderung

  • Fırat University
  • Firat University, Scientific Research Project Committee

Zitieren

APA

Şengür, D. (2026). Explaining the Links Between School Administrator Leadership, Job Satisfaction, and Participatory School Climate: A Machine Learning-Enhanced Multilevel Analysis of TALIS 2024 School Administrator Data. Behavioral Sciences, 16(7), 1062. https://doi.org/10.3390/bs16071062

Vancouver

Şengür D. Explaining the Links Between School Administrator Leadership, Job Satisfaction, and Participatory School Climate: A Machine Learning-Enhanced Multilevel Analysis of TALIS 2024 School Administrator Data. Behavioral Sciences. 2026;16(7):1062. doi:10.3390/bs16071062

BibTeX
@article{engr2026,
  author = {Şengür, Dönüş},
  title = {Explaining the Links Between School Administrator Leadership, Job Satisfaction, and Participatory School Climate: A Machine Learning-Enhanced Multilevel Analysis of TALIS 2024 School Administrator Data},
  journal = {Behavioral Sciences},
  year = {2026},
  volume = {16},
  number = {7},
  pages = {1062},
  doi = {10.3390/bs16071062}
}
Kostenlos starten →