Linking Burnout Emotional Exhaustion and Assessment Related Stress to Mental Health Outcomes Among Early Career Teachers: Counselling Support as a Dual Mediating and Moderating Factor

Abstract The study revisited the silent mental health crisis among early-career teachers in Ghana by examining how burnout and assessment-related stress predict mental health, with counselling support analyzed as both a mediating and moderating factor. Employing a cross-sectional correlational design grounded in the positivist paradigm, the study surveyed a nationally representative sample of 790 early-career teachers using validated psychometric instruments, including the General Health Questionnaire (GHQ-12). Multiple regression analysis revealed that emotional exhaustion (β = 0.43, p < .001), depersonalization (β = 0.18, p = .002), and role conflict (β = 0.22, p < .001) were significant predictors of psychological distress, while reduced personal accomplishment had a negative association (β = -0.14, p = .003). Assessment-related stressors such as test pressure (β = 0.25, p < .001), accountability load (β = 0.28, p < .001), and curriculum inflexibility (β = 0.20, p < .001) also significantly predicted poor mental health. Mediation analysis using PROCESS Model 4 confirmed that counselling support partially mediated the relationship between burnout components and mental health (indirect effect for emotional exhaustion: B = 0.15, 95% CI 0.10, 0.22; Sobel z = 4.50, p < .001). Moderation analysis using PROCESS Model 1 further indicated that counselling significantly buffered the negative impact of assessment stress on mental health (interaction effect: β = -0.23, p < .001), with the strongest protection observed under high test pressure (β = -0.21, p = .001) and fear of evaluation (β = -0.22, p < .001). Multigroup SEM analyses revealed that the impact of burnout on mental health was stronger among female teachers (β = 0.50) than males (β = 0.42), and that counselling exerted a more protective effect for females (β = -0.23) compared to males (β = -0.18). Hierarchical Linear Modeling (HLM) showed that 11.2% of the variance in mental health outcomes was attributable to between-school differences (ICC = 0.112), and that the full model, which included both teacher- and school-level predictors, explained 45% of the variance in mental health (conditional R² = 0.45). The study highlighted the importance of addressing both individual and institutional risk factors and advocates for the inclusion of mental health benchmarks in school quality assurance frameworks to ensure a psychologically sustainable teaching profession in Ghana.