Beyond Mean Differences: A Three-Tier Approach to Level and Structural Heterogeneity in Green HRM Research

Authors

  • Angelino dos Santos Universidade Nacional Timor Lorosa'e, Dili
  • Otto Randa Payangan Universitas Hasanuddin, Makassar
  • Indrianty Sudirman Universitas Hasanuddin, Makassar

DOI:

https://doi.org/10.58631/ajemb.v5i11.576

Keywords:

green human resource management, green strategic HRM, level heterogeneity, structural heterogeneity, ANOVA

Abstract

Green Human Resource Management (GHRM) research has expanded rapidly; however, empirical studies frequently treat organizations as homogeneous, conflating differences in construct levels with differences in structural relationships. This study addresses this methodological gap by proposing and empirically illustrating a three-tier approach to examining level and structural heterogeneity in GHRM research within a decentralized higher education setting. The research was conducted at Universidade Nacional Timor Lorosa'e (UNTL), Timor-Leste, involving 150 lecturers and administrative staff members equally distributed across ten faculties. Five constructs were examined: Leadership Integrity (LI), Personal Planning (PP), Employee Commitment (EC), Green Strategic Human Resource Management (GSHRM), and Sustainability Performance (SP). Each construct was measured using ten reflective indicators rated on a five-point Likert scale, and the data were analyzed using partial least squares structural equation modeling (PLS-SEM). The analytical procedure comprised three tiers: descriptive analysis of faculty-level means, one-way analysis of variance (ANOVA) to assess level heterogeneity, and Chow tests to evaluate the overall equality of structural coefficients across faculties. The results revealed statistically significant differences in construct means across faculties for all five constructs, with GSHRM exhibiting the largest F-statistic (F = 33.984, p < .001). The Chow tests further indicated the rejection of the null hypothesis of complete coefficient equality for the EC, GSHRM, and SP equations, providing evidence of structural heterogeneity. Notably, the largest ANOVA F-statistic indicates greater differentiation in construct levels rather than stronger heterogeneity in the underlying structural relationships. The study concludes that level heterogeneity and structural heterogeneity represent distinct analytical dimensions and that differences in construct means should not be interpreted as evidence of differences in causal mechanisms. A staged analytical approach, progressing from omnibus structural tests to path-specific multigroup analysis, is recommended for future research.

Downloads

Published

2026-11-16