This study seeks to address the existing gap by constructing a comprehensive structural model grounded in Bandura’s Social Cognitive Theory (SCT) and the Job Demands–Resources (JD-R) Model. Despite its prominence in international scholarship, the application of the JD-R framework remains limited within the Arab academic context. Published reviews indicate that most Arab research output continues to focus on conventional educational topics, while systematic implementation of the JD-R model in higher education research is largely absent.

Moreover, the few studies that have adopted the model in Arab settings often examined only direct relationships among variables, overlooking the dynamic mechanisms integral to the framework—such as mediating and moderating processes that explain how demands, resources, and outcomes interact. These studies frequently neglect to explore gender effects or interactive paths within their structural designs. Such limitations highlight the need for analytically advanced research models that capture the complexity and specificity of Arab academic environments.

Develops an integrated structural model through Partial Least Squares Structural Equation Modeling (PLS-SEM) within the Arab educational setting to test direct, indirect, and interactive pathways.

Bandura’s theory (Bandura, 1986, 1997, 2001) posits that human behavior results from a triadic reciprocal interaction among cognitive, behavioral, and environmental factors, a process known as reciprocal determinism. At the core of this theory lies self-efficacy, which reflects individuals’ belief in their capacity to regulate their actions and manage challenges effectively. This belief system serves as a foundation for understanding motivational patterns and engagement behaviors in complex academic environments (Schunk and DiBenedetto, 2020).

According to the JD-R framework, occupational well-being depends on maintaining a balance between taxing job demands and available supportive resources (Bakker and Demerouti, 2017). High demands typically trigger exhaustion through the strain pathway, whereas abundant resources—such as institutional support, self-efficacy, and psychological resilience—activate the motivational pathway, leading to greater work engagement and job satisfaction (Bakker and Bal, 2010; Salanova et al 2010).

The present research adopted a quantitative, cross-sectional, correlational–explanatory design, aiming to test a structural model incorporating both mediating and moderating effects among self-efficacy, psychological resilience, occupational stress, and work engagement among university faculty in Arab countries. The Partial Least Squares Structural Equation Modeling (PLS-SEM) approach was employed to assess both the measurement and structural models, drawing on the theoretical foundations of Bandura’s Social Cognitive Theory and the Job Demands–Resources (JD-R) Model. This design is well suited for evaluating predictive structural relationships within a single time frame.

The final study sample consisted of 266 faculty members, selected through convenience sampling from seven Arab universities located in Saudi Arabia, Egypt, and Sudan. The sample was intentionally diversified in terms of academic disciplines, ranks, and demographic backgrounds to ensure adequate representation of the target population. The demographic and professional characteristics of the participants were as follows:

Data were collected over a period of 4 months (from October 10, 2024, to February 10, 2025) through both electronic and paper-based questionnaires distributed to university faculty members. Institutional and ethical approvals were obtained in advance, and participants were assured of confidentiality and that their responses would be used for research purposes only. The questionnaire comprised two parts: demographic information (gender, age, academic rank, institution, college, economic level, marital status, and place of residence), and four validated psychometric instruments measuring self-efficacy, psychological resilience, occupational stress, and work engagement. All scales underwent a back-translation process and expert review to ensure cultural appropriateness for Arab academic contexts. On average, participants completed the survey in 10–15 min.

Data analysis was conducted in two major phases.

The applied methodology provided a statistically rigorous framework to examine the hypothesized relationships (H1–H9) within an integrative theoretical model combining the principles of Social Cognitive Theory and the Job Demands–Resources (JD-R) Model. This design facilitates a deeper understanding of the psychological and organizational mechanisms underlying work engagement in the context of Arab higher education.

Using Smart PLS v.4, a mediated–moderated structural model was developed incorporating all four main constructs. Initial analyses evaluated the relationships between 57 observed items and four latent variables, confirming indicator reliability at the criterion level of 0.70 or higher (Sarstedt et al., 2021). Items below this threshold were removed, and structural paths were subsequently estimated to confirm the model’s overall fit.

Reliability was assessed using multiple indices—Cronbach’s alpha, composite reliability, and average variance extracted (AVE)—as summarized in Table 3, alongside variance inflation factor (VIF) values, t-statistics, and significance levels.

To examine both direct and indirect effects, and to explore mediating and moderating mechanisms within the proposed model, a series of multi-group PLS-SEM analyses were performed. These analyses were conducted on the total sample, as well as separately for male and female faculty subgroups. Prior to testing structural relationships, the model’s measurement invariance across genders was established to ensure that constructs were interpreted equivalently by both male and female participants.

Figure 2 presents the structural model for the overall sample, while Table 5 summarizes the direct and indirect path coefficients for the total sample, male group, and female group, highlighting the differential patterns of mediation and moderation within the Mediated–Moderated Model framework.

The path analysis results for the total sample (Table 5) revealed that all direct effects were statistically significant except for the path between self-efficacy and occupational stress (p = 0.16), and the interaction effects of occupational stress × psychological resilience on work engagement (p = 0.48) and occupational stress × self-efficacy on work engagement (p = 0.07).

The indirect effect between psychological resilience and work engagement was statistically significant (β = −0.08, p < 0.01), whereas the indirect path between self-efficacy and work engagement was not significant (p = 0.25).

These findings indicate that self-efficacy exerts a significant direct effect on work engagement, but not an indirect one, while psychological resilience exhibits both a positive direct effect and a statistically significant indirect effect on engagement. This highlights the critical roles of self-efficacy and psychological resilience in promoting academic work engagement.

Moreover, occupational stress partially mediated the relationship between psychological resilience and engagement but did not moderate the relationships between self-efficacy, resilience, and work engagement. This underscores the challenging nature of the academic work environment, where resilience functions as a partial buffer against stress and disengagement.

Among female faculty members, all direct paths were statistically significant except for the path between psychological resilience and work engagement (p = 0.08), and the interaction term occupational stress × psychological resilience on engagement (p = 0.37).

The indirect effect of self-efficacy on work engagement was not statistically significant (p = 0.06), suggesting that self-efficacy exerts a positive direct effect but no significant indirect effect on work engagement. Conversely, psychological resilience showed a positive but nonsignificant direct effect, yet a significant indirect effect on engagement.

These results imply that female faculty members rely more heavily on psychological resilience than on self-efficacy to maintain their engagement despite high job demands. Resilience serves as a partial mediator between self-efficacy and engagement, and a full mediator between resilience and engagement. Because self-efficacy plays a comparatively smaller role, occupational stress acts as a significant moderator in the relationship between self-efficacy and work engagement among women.

For male participants, all direct paths were statistically significant except for the path between self-efficacy and occupational stress (p = 0.55) and the interaction effects involving occupational stress × resilience on engagement (p = 0.38) and occupational stress × self-efficacy on engagement (p = 0.90).

Neither the indirect path between resilience and engagement (p = 0.10) nor that between self-efficacy and engagement (p = 0.63) reached significance. These findings indicate that both self-efficacy and psychological resilience exert positive direct effects on work engagement without significant mediation via stress.

Male faculty appear to utilize both psychological resources—self-efficacy and resilience—jointly to cope with occupational stress. Accordingly, occupational stress serves as a partial mediator only in the resilience–engagement relationship but does not moderate any of the structural paths for male academics.

The analysis of moderation and gender-based differences revealed significant variations in the structural relationships among self-efficacy, resilience, occupational stress, and engagement across genders. Specifically, occupational stress moderated the relationship between self-efficacy and work engagement for female participants only.

Figure 3 illustrates this moderation effect: self-efficacy is mapped on the x-axis, work engagement on the y-axis, and the three lines correspond to high (+1 SD), medium, and low (−1 SD) levels of academic stress. The figure shows that as occupational stress increases, work engagement decreases across different levels of self-efficacy. The moderation effect of academic stress was 0.11, which is relatively small compared with the direct effect of self-efficacy on engagement (0.29).

This suggests that for every one standard deviation increase in academic stress, the strength of the relationship between self-efficacy and engagement decreases by 0.11. Conversely, each one-unit increase in self-efficacy enhances work engagement by 0.29 units, underscoring the protective yet limited buffering role of self-efficacy under high stress conditions.

The findings demonstrate that self-efficacy and psychological resilience are key personal resources that substantially contribute to work engagement among university faculty members in Arab institutions. Both constructs exhibited significant positive direct effects on engagement, underscoring their joint role as protective and motivational factors in academic work contexts.

These results are consistent with a large body of prior research. For instance, Xanthopoulou et al. (2007) found that self-efficacy positively correlates with work engagement, while Bakker and Demerouti (2008) highlighted the function of personal resources in enhancing engagement as part of the Job Demands–Resources (JD-R) model. Similarly, Luthans et al. (2007) demonstrated that positive psychological resources—particularly self-efficacy and resilience—jointly foster performance and well-being at work. Studies by Rothmann and Jordaan (2006) and Mache et al. (2014) further confirmed positive associations between these constructs and engagement, while Britt et al. (2016) emphasized resilience as a key predictor of sustained engagement under occupational stress.

However, a few studies have reported divergent findings. For example, Skaalvik and Skaalvik (2014) observed that self-efficacy exerted only a limited influence on teachers’ engagement, and Simbula et al. (2011) suggested that this relationship is context-dependent and mediated by situational factors.

These discrepancies underscore the relevance of cultural context when interpreting such relationships. Given that the present study was conducted in Arab higher education settings, factors such as collectivist values, social norms, and institutional environments may shape how personal resources relate to engagement. Hu et al. (2011) noted that cross-cultural variations in the effect of personal resources on engagement may arise from differences in both conceptual frameworks and measurement tools.

Future research should therefore explore cross-contextual dynamics, including the temporal and situational variability of these relationships, as well as potential interactive mechanisms between self-efficacy and resilience. These insights could deepen theoretical understanding and inform targeted interventions to enhance faculty engagement.

The observed positive effects can be interpreted through Social Cognitive Theory (Bandura, 1986), Conservation of Resources Theory (Hobfoll, 2002), and the JD-R Model (Bakker and Demerouti, 2017). Self-efficacy enhances confidence in one’s ability to meet professional demands, boosting motivation and persistence. Resilience, in turn, helps individuals effectively adapt to stressors and maintain high engagement levels even under challenging academic conditions.

Results revealed that academic occupational stress partially mediated the relationship between resilience and engagement among the overall and male samples and fully mediated this relationship among females. It also partially mediated the link between self-efficacy and engagement among women. Additionally, occupational stress exerted strong negative direct effects on engagement, indicating the substantial pressures inherent in academic environments across Arab universities.

These stressors likely derive from multiple sources, such as teaching overloads, research and publication demands, administrative responsibilities, community service, and quality assurance duties. Importantly, the relationship between stress and engagement may not be linear: moderate stress can sometimes enhance engagement by promoting challenge appraisals, whereas excessive stress undermines motivation and performance.

Gender differences in stress perception and coping also emerged. This aligns with Tytherleigh et al. (2007), who found gender-based discrepancies in identifying and managing occupational stressors, and with Mache et al. (2014), who reported that women tend to rely more on resilience-oriented coping strategies. The current results thus support the idea that males and females employ different personal resources to navigate academic demands.

Consistent with earlier work by Kinman and Jones (2008), Barkhuizen and Rothmann (2008), and Watts and Robertson (2011), the current study found negative correlations between academic stress and engagement. Yet earlier research such as Abouserie (1996) and Gillespie et al. (2001) suggest that for some academics, stress can stimulate achievement, with effects varying by career stage and experience.

From a theoretical standpoint, the mediating role of academic stress can be explained within the JD-R framework (Bakker and Demerouti, 2017): stress represents a job demand that depletes personal resources, thereby reducing engagement. However, self-efficacy and resilience operate as protective personal resources, mitigating these adverse effects.

For female participants, academic stress moderated the relationship between self-efficacy and engagement—indicating that the strength of this relationship depends on the level of stress experienced. This result mirrors findings by Xanthopoulou et al. (2007) and Bakker et al. (2010), which confirmed stress’s moderating role between personal resources and organizational outcomes. Similarly, Barkhuizen and Rothmann (2006) found that the positive effects of personal resources on well-being were most pronounced under high-stress conditions.

By contrast, other studies such as Simbula et al. (2011) and Schaufeli and Taris (2014) did not support a significant moderating influence, suggesting the relationship’s sensitivity to context and occupational setting.

This moderating effect can be interpreted through Cognitive Appraisal Theory (Lazarus and Folkman, 1984), which posits that stress effects depend on individuals’ perception and evaluation of stressors. Female academics may perceive self-efficacy as a particularly critical personal resource to maintain engagement despite elevated stress levels. Within the JD-R model, self-efficacy thus buffers stress’s negative influences, functioning as a key motivational resource that sustains engagement among women in academia.

The results revealed notable gender differences in how self-efficacy and resilience predict engagement. While resilience played a stronger role for females, self-efficacy had a greater influence among males. These distinctions suggest that men and women utilize different coping mechanisms to deal with academic demands, supporting findings from Mache et al. (2014).

This distinction aligns with Huang (2013), who found that men display higher self-efficacy levels in certain academic domains—especially in science and technology—yet differs from Schaufeli et al. (2006), who reported no significant gender differences in engagement.

Such inconsistencies may stem from multiple factors, including biological differences, social role expectations, and institutional structures. For instance, Taylor et al. (2000) found biological variations between genders in stress responses, affecting how personal resources are utilized. From a socio-cultural perspective, Eagly and Wood (2012) argue that cultural role expectations predispose men to demonstrate competence and confidence, while women are socially encouraged toward flexibility and adaptability.

In Arab societies, these social expectations may be reinforced by cultural norms and constraints, such as limited work–life balance and societal pressures regarding women’s roles. Studies like O’Laughlin and Bischoff (2005) have shown that such pressures compel women to rely more heavily on resilience as a survival and adaptation mechanism. Moreover, social learning processes (Bussey and Bandura, 1999) influence how gendered behaviors and coping strategies develop over time.

Institutional contexts also matter. Some Arab universities maintain gender-segregated education systems (e.g., Saudi Arabia), while others adopt coeducational models, potentially shaping professional experiences and engagement patterns differently across genders.

Overall, these findings emphasize the importance of contextual sensitivity and gender awareness in examining how personal resources, stress, and engagement interact within Arab higher education environments.