Pathway to Quiet Quitting: Role of Sense of Community and Role Stress Between Engaging Leadership and Course Tutors’ Burnout

Descriptive Statistics

The demographic and professional characteristics of the study participants are presents in Table 1. A total of 923 respondents participated in the study. The majority were male (n = 768, 83.2%), while females constituted 16.8% (n = 155). Most participants fell within the 41–50 years age bracket (45.2%) whiles a small proportion aged 21–30 years (2.4%). Regarding teaching experience, most participants (34.6%) had 6–10 years of experience with a smaller percentage having over 20 years (0.7%) of teaching experience. In terms of academic qualification, an overwhelming majority held a Master’s degree (95.8%), with only 4.2% possessing a PhD. The data were adequate in employing structural equation modeling.

Evaluation of the Measurement Model

The initial phase of assessing the outer model entails scrutinizing the measurement model using PLS CB-SEM through the statistical software Smart PLS 4. All criteria are presented in detailed below.

Assessing the Model Fit Indices

The study’s goodness of fit was assessed based on the recommendations of Henseler et al. (2016) and this is presented in Table 2. Although results in Table 2 showed that the model nearly fitted the data, some of the AVEs (Figure 2) of the measured constructs did not meet the accepted the threshold of > .5 recommended by Hair et al. (2017). A second order model was therefore tested after few items were deleted (Figure 3) and this improved the model fit indices as presented in Table 2.

Figure 2

Figure 3

Assessing the Construct Reliability and Validity

The indicator loadings in this study were assessed using PLS-SEM, with detailed values presented in Table 3. Although a few indicators still did not meet the recommended threshold of > .708 (Mukminin et al., 2020) after the initial deletion, they were retained because of content validity and removing them wouldn’t have improved the AVE values (Hair et al., 2017). As a result, twenty-two indicators proceeded to the next stage of the PLS-SEM analysis (Figure 3).

Further, internal consistency reliability, which evaluates the statistical coherence among indicators, was examined via Cronbach’s alpha (α) and Composite Reliability (CR) following the guidelines of Hair et al. (2019), who recommend α > .700 and CR > .708. Table 3 detailed these measures: Cronbach’s alpha values ranged from .763 to .927, and Composite Reliability values ranged from .762 to .928, indicating that all constructs demonstrated satisfactory internal consistency. In addition, the study’s convergent validity was assessed. To evaluate this, we reported the AVE for each construct, calculated via the PLS-SEM algorithm in SmartPLS. An AVE of at least 0.50 is recommended, indicating that a construct explains 50% or more of its items’ variance. In the present study, every construct achieved an AVE above 0.50 with the exception of Quiet Quitting. This was maintained because both Cronbach alpha and composite reliability of the construct were achieved.

Discriminant Validity

Discriminant validity issues arise when HTMT values exceed 0.900, as this suggests that constructs may not be sufficiently distinct. An HTMT value above 0.900 indicates a lack of discriminant validity (Hair et al., 2019). As reported in Table 4, all HTMT values fell below this threshold, and each was significantly less than 1.

Structural Model Path Coefficient

The structural model’s path coefficients were examined through bootstrapping with 5,000 subsamples and two-tailed testing at the 0.05 significance level. Table 5 summarizes the path coefficients, bootstrapped means, standard deviations, t-statistics, 95% bias-corrected confidence intervals, and significance decisions of hypothesis 1–6. The results showed that all the 6 hypothesized relationship achieved statistical significance. Specifically, Engaging Leadership was found to significantly reduce Role Stress (β = –0.203, t = 4.652, 95% CI [–0.289, –0.116]) while simultaneously enhancing Sense of Community (β = 0.332, t = 8.248, 95% CI [0.255, 0.414]) among course tutors. In turn, higher Role Stress was associated with greater Burnout (β = 0.498, t = 5.697, 95% CI [0.322, 0.660]), whereas a stronger Sense of Community corresponded to lower Burnout (β = –0.163, t = 2.469, 95% CI [–0.298, –0.042]). Beyond these direct pathways, Engaging Leadership also exerted a direct negative effect on Burnout (β = –0.137, t = 3.008, 95% CI [–0.232, –0.054]). Finally, Burnout emerged as a powerful antecedent of Quiet Quitting, with a highly significant positive relationship (β = 0.672, t = 11.534, 95% CI [0.553, 0.776]). All confidence intervals exclude zero, and each path achieves significance at p < .05, thus rejecting the six null hypotheses and affirming the proposed structural model. The result is further depicted in Figure 4. Furthermore, the effect size (f²), which quantifies the strength of a predictor’s influence on an outcome construct was further examined. Six inter-construct effect sizes were computed. The path from Sense of Community to Burnout exhibited the smallest effect (f² = 0.021), whereas the path from Burnout to Quiet Quitting demonstrated the largest effect (f² = 0.415). Notably, the direct relationship between Engaging Leadership and Burnout did not yield a calculable f².

Figure 4

Coefficient of Determination

The coefficient of determination (R²) represents the proportion of variance in an endogenous construct that is accounted for by its exogenous predictors. Serving as an index of a model’s explanatory and predictive capability, R² is calculated as the squared correlation between an endogenous variable and its predictor set. Its values range from 0 to 1, with higher figures indicating stronger predictive accuracy: values of 0.75, 0.50, and 0.25 are typically interpreted as substantial, moderate, and weak levels of explanatory power, respectively (Hair et al., 2019). From the result in Table 6, 29.3% of the variance in quiet quitting can be explained by its predictors and this indicates a moderate explanatory power. However, 24.3%, 3.0% and 9.7% of the variances in burnout, role stress and sense of community are accounted for by its predictors respectively- indicating low predictive powers.

Predictive Relevance (Q2)

The out-of-sample predictive performance of the model was evaluated using Q²predict for each endogenous construct (Table 8). This is assessed via the Stone–Geisser criterion and reflects a model’s ability to accurately predict the observed data points (Hair et al., 2019). Following Hair et al. (2019), Q² benchmarks of 0.02, 0.15, and 0.35 correspond to small, medium, and large predictive relevance, respectively. In Table 7, all Q²predict values exceed zero, confirming that the model holds predictive relevance for all outcomes. Specifically, Sense of Community achieved the highest Q²predict (0.092), indicating modest medium-level relevance, followed by Burnout (0.048) and Quiet Quitting (0.047), both reflecting small predictive relevance. Role Stress demonstrated the lowest yet still positive predictive relevance (0.025).

Mediation Analysis

Mediation analysis was performed to test hypothesis 7 and 8. The results demonstrated that Engaging Leadership (EL) exerts both direct and indirect influences on Burnout (BNT). Specifically, EL retained a significant direct effect on BNT (β = –.137, SE = .045, t = 3.01, 95% CI [-.232, –.054]), indicating that higher levels of engaging leadership are associated with lower burnout even when mediators are accounted for. Two specific indirect effects were also significant: EL → Sense of Community (SoC) → BNT (β = –.054, SE = .023, t = 2.35, 95% CI [-.105, –.014]) and EL → Role Stress (RS) → BNT (β = –.101, SE = .032, t = 3.21, 95% CI [-.170, –.048]). The total effect of EL on BNT was β = –.292, SE = .045, t = 6.56, 95% CI [-.380, –.208]. Because the direct path from EL to BNT remains significant alongside significant indirect paths through both SoC and RS, the pattern of results reflects partial mediation. In other words, engaging leadership reduces burnout both directly and indirectly by strengthening sense of community and by alleviating role stress but does not do so solely through these mediators. Both hypothesis 7 and 8 were therefore rejected for the null hypothesis. The result is summarised on Table 8.

Engaging Leadership, role stress, sense of community and burnout among course tutors

The findings of this study demonstrated that engaging leadership has a significant direct effect on reducing role stress and burnout among course tutors, while also positively enhancing their sense of community, supporting hypothesis 1, 2 and 3. Research consistently indicates that engaging leadership is negatively associated with role stress, particularly in educational settings (Harms et al., 2017; Diebig et al., 2016; Tims et al., 2011) while other empirical evidence also links engaging leadership to a stronger sense of community (Firouznia et al., 2021, Saari et al., 2018). Similarly, plethora of study have also confirmed that effective leadership can reduce burnout of teachers (Dishari & AlAfnan, 2024; Mitchell, 2021; Perrone et al., 2019). As the current findings resonate with empirical evidence, engaging leadership of study centre coordinators within distance education must be emphasised and extensively practice to ensure esprit de corps and reduced burnout. These results further align with the JD-R model which posits that job resources, such as supportive and engaging leadership, play a crucial role in mitigating the negative impact of job demands and stressors. In this case, engaging leadership serves as a vital resource that helps tutors manage their workload and role-related pressures more effectively, thereby lowering the risk of burnout. Our findings, however, come with a crucial caveat: the advantages of engaging leadership are not universal. In certain contexts, such as high power-distance cultures or those strained by limited resources, attempts to empower staff can paradoxically fail. Employees may perceive it as managerial withdrawal, leading to greater role ambiguity and strain. This is particularly relevant in Ghana’s distance-education context, where precarious work contracts and intensive weekend travel are the norm. Here, the aspects of leadership that foster community and define responsibilities are likely more protective against burnout than autonomy alone, which is a testable boundary for future research.

Role stress, Sense of community and burnout among course tutors

The study further highlights the essential role of role stress and sense of community in influencing burnout among course tutors. Firstly, the study showed that role stress positively predicted burnout, supporting hypothesis 4. This implies that elevated levels of role stress among course tutors will result in burnout. This finding resonates with the Job Demands-Resources (JD-R) model, which suggests that high job demands such as excessive workload or unclear responsibilities deplete individuals’ energy, leading to strain and eventually burnout (Demerouti et al., 2001). The result is also consistent with Shukla & Trivedi (2008), who found that burnout among professionals such as teachers can result from excessive demands on their energy, strength, and resources. In tandem with this, other empirical studies have shown that elevated stress and burnout among educators undermine their overall job performance, diminish productivity, and erode their capacity to foster effective professional relationships (Agyapong et al., 2022; Cancio et al., 2018; Antoniou et al., 2013). Furthermore, the study found that sense of community negatively impacted burnout among course tutors, supporting hypothesis 5. Within the JD-R framework, sense of community functions as a valuable social resource that provides emotional support, belonging, and collective efficacy, thereby counteracting the negative effects of job demands. This finding is consistent with previous research that showed that a negative relationship exists between sense of community and burnout (Aryee, 2021; Fang, 2025). For instance, Fang (2025) indicated that strong emotional connectedness may reduce tutor burnout since it serves as a resource that buffers conflict-related stress. While consistent with global literature, the potent protective effect of sense of community in our study suggests that in environments with precarious work contracts and travel-intensive workloads, the community-building function of leadership may be a more critical buffer than other resources, such as autonomy-granting. This highlights the JD-R model by indicating that not all resources are equally effective across all contexts.

Burnout and quiet quitting among course tutors

The study also revealed the critical role of burnout in influencing quiet quitting among course tutors. Specifically, burnout exerted a significant positive effect on quiet quitting, thereby validating Hypothesis 6. From the perspective of the Job Demands–Resources (JD-R) model, prolonged exposure to excessive job demands without sufficient resources depletes individuals’ psychological reserves and precipitates burnout, which in turn motivates energy-conserving coping strategies such as psychological withdrawal and reduction to minimum-role prescriptions. In this context, burnout drives tutors to limit their efforts to the minimum required tasks, withdraw from extra-role responsibilities, and reduce overall commitment rather than pursue immediate turnover. This interpretation is supported by Thu Trang et al. (2024), who reported that up to 80% of those who quietly quit experienced job burnout, and aligns with conceptualizations of quiet quitting as a coping response to chronic exhaustion that is nevertheless associated with wider disengagement and counterproductive outcomes (von Hippel et al., 2019). Complementary empirical evidence from Galanis et al. (2023), Lu et al. (2023), and Xueyun et al. (2023) further corroborates the strong link between burnout and the adoption of quiet-quitting behaviours.

Mediating roles of sense of community and role stress on engaging leadership and burnout

The study found that both role stress and sense of community partially mediate the relationship between engaging leadership and burnout among course tutors, subsequently contributing to a reduction in quiet quitting and thereby supporting Hypotheses 7 and 8. Concretely, engaging leaders who clarify expectations, distribute tasks fairly, and provide timely feedback reduce role ambiguity and overload, which in turn diminishes the chronic strain that fosters emotional exhaustion and depersonalization (Maslach & Leiter, 2016). Conversely, when role stress remains high, the salutary effect of engaging leadership on burnout (and downstream quiet-quitting behaviours) is attenuated. In contrast, an enhanced sense of community functions synergistically with engaging leadership, thus, leaders who model open communication, encourage peer interaction, and facilitate collective problem-solving help build social cohesion and collective efficacy, thereby amplifying leadership’s buffering effect against burnout and behavioral withdrawal. These findings echo Cemaloğlu (2011), who showed that certain leadership behaviours can increase teacher burnout under intense stress, and Schaufeli (2015), who found that leaders who model open communication and encourage peer interaction strengthen social bonds and collective efficacy, reducing burnout. Anchored in the JD–R model, engaging leadership is a key job resource that replenishes tutors’ psychological reserves and indirectly lowers job demands especially role stress, thereby reducing burnout. Sense of community is an additional social resource that amplifies leadership’s protective effects, while persistent role stress erodes resources and increases strain.

Theoretical Implications

The findings provide important theoretical contributions by extending the Job Demands–Resources (JD-R) model to the context of course tutors. The results confirm that engaging leadership functions as a critical job resource that not only exerts a direct protective effect on burnout but also operates indirectly through the reduction of role stress and the enhancement of a sense of community. By demonstrating that burnout mediates the link between work experiences and quiet quitting, the study highlights burnout as a pivotal health-impairment mechanism connecting demands and resources to behavioral withdrawal. These results enrich the JD-R framework by illustrating the dual pathways, that is demand reduction and resource enhancement through which engaging leadership influences both psychological outcomes and work-related behaviors, thereby offering a more vital understanding of how leadership contributes to employee well-being and engagement.

Practical Implications

From a practical perspective, the study shows the importance of cultivating engaging leadership within educational institutions as a means of supporting tutors’ well-being and reducing disengagement behaviors such as quiet quitting. Accordingly, teacher training should embed engaging-leadership micro-skills in both pre-service and in-service programmes, covering role-clarity routines, autonomy-supportive communication, and community-building practices and these should be reinforced through coaching and mentoring. The results suggest that leadership practices that emphasize clarity of roles, regular feedback, and participatory decision-making can significantly lower role stress, while behaviors that promote communication, collaboration, and peer support foster a stronger sense of community. Together, these leadership strategies help minimize burnout and sustain tutors’ engagement by balancing job demands with adequate resources. Institutions that intentionally develop and support engaging leadership practices are therefore more likely to create work environments where tutors experience reduced stress, stronger social bonds, and greater resilience against burnout-driven withdrawals. Also, institutions may also consider implementing organizational policies that proactively monitor and address role stress while simultaneously investing in initiatives that build social connectedness among tutors.

At the policy level, inclusive education frameworks should codify standardised tutor role descriptions, set fair workload and travel caps for distance delivery, stabilise contracting and remuneration timelines, and guarantee access to psychosocial supports and reasonable accommodations. Such measures will not only help to mitigate burnout but will also reduce the risk of quiet quitting by promoting sustained commitment and engagement. Together, these teacher-training and inclusive-policy create enabling conditions that balance demands with resources, thereby strengthening tutor well-being, instructional continuity, and student learning outcomes. Future work could adopt mixed-methods designs to explain mechanisms behind the quantitative paths. For instance, a sequential explanatory design where interviews/focus groups with tutors and coordinators contextualise why engaging leadership lowers role stress and burnout, followed by joint displays that integrate effect sizes with themes.

Strengths and limitations

This study integrates engaging leadership, role stress, sense of community, burnout, and quiet quitting into a single framework, clarifying how leadership, psychological strain, and behavioral withdrawal interact. The application of the Job Demands–Resources (JD-R) model further strengthens the theoretical contribution by demonstrating how leadership functions both as a resource that mitigates role stress and as a driver of social connectedness. Methodologically, the use of Covariance-Based Structural Equation Modeling (CB-SEM) represents another important strength, as it enabled rigorous testing of complex direct, indirect, and mediation relationships simultaneously, thereby enhancing the robustness and validity of the findings. Despite these contributions, some limitations should be acknowledged. First, reliance on self-reported data raises the potential for common method bias and social desirability effects. second, the study focused on course tutors within a specific educational setting, which may limit the generalizability of the findings to other professional groups or organizational contexts. Further, although most fit indices met conventional guidelines, RMSEA was borderline, which cautions against over-interpretation and suggests scope for model refinement (e.g., revisiting item wording, adding theoretically justified indicators, and testing alternative specifications) in future research. Despite these limitations, they do not invalidate the findings of the study but rather provide avenues for future research to build upon and extend its contributions.