Introduction

Burnout is a chronic psychological condition resulting from prolonged and intense occupational stress. It is primarily characterized by emotional exhaustion, depersonalization, and a reduced sense of personal accomplishment1. The World Health Organization (WHO) has classified burnout as an occupational phenomenon in the 11th Revision of the International Classification of Diseases(ICD-11), underscoring its public health implications2. Nurses, who serve on the frontline of patient care and face complex clinical scenarios, are especially vulnerable to burnout due to multiple stressors, including high emotional labor demands, organizational performance pressures, and interpersonal conflicts. Evidence suggests that the prevalence of moderate to severe burnout among clinical nurses is as high as 70%3,4,5. Burnout not only compromises nurses’ mental health and job satisfaction but also threatens care quality, increases the likelihood of medical errors, and contributes to staff turnover. Accordingly, clarifying the patterns of association linked with nurse burnout remains important from both theoretical and practical standpoints.

Workplace violence and burnout

WPV is a negative social event that encompasses verbal abuse, physical assault, threats, and sexual harassment. It has emerged as a critical external factor contributing to nurse burnout6.Previous studies have shown that approximately 71% of nurses in China have experienced WPV7,8.Repeated exposure to violence can erode individuals’ sense of control and self-worth, resulting in emotional exhaustion, helplessness, and ultimately, burnout.However, prior research has mostly emphasized the direct association between WPV and burnout, with comparatively less attention to theory-consistent psychological pathways and contextual moderators.

According to Conservation of Resources Theory (COR)9,10,individuals tend to preserve and acquire critical resources (such as resilience and emotional stability) when facing stress.The depletion of these resources triggers stress responses and may eventually lead to health deterioration. In clinical settings, WPV is linked to emotional and psychological resource depletion among nurses and has been associated with higher levels of burnout.Therefore, this study conceptualizes WPV as a context of resource loss and aims to examine the pathways through which WPV contributes to nurse burnout.

Mediating mechanisms: resilience and emotional labor

Resilience refers to an individual’s capacity to maintain psychological stability and adapt positively when facing adversity or stress.It is recognized as a core internal psychological resource that helps prevent resource depletion11.In the context of WPV, resilience enables individuals to mobilize and restore internal resources, thereby improving coping efficacy and recovery12.According to the Conservation of Resources (COR) theory13,14,individuals with high resilience are more capable of regulating negative emotions, conserving resources, and attenuating pathways associated with burnout.In contrast, those with low resilience are more vulnerable to entering a vicious cycle characterized by resource exhaustion, emotional depletion, and functional impairment. Previous studies have shown that resilience not only buffers the psychological impact of occupational stress but also mediates the relationship between violence and burnout by promoting emotional regulation and behavioral adaptation among nurses15.

Emotional labor theory suggests that individuals perform emotional regulation to meet organizational expectations and professional norms through two main strategies: surface acting and deep acting16.This regulatory process consistently consumes emotional and psychological resources. In high-stress situations such as WPV, nurses are required to suppress genuine emotional responses and maintain professional demeanor, which increases emotional labor demands.As a result, emotional resources are depleted, leading to elevated exhaustion and, ultimately, burnout17,18.Existing research indicates that individuals with lower levels of resilience are more likely to rely on surface acting strategies, which has been associated with greater emotional dysregulation and exhaustion19,20.Therefore, resilience and emotional labor may jointly constitute a chain-like mediation pathway that reflects the progressive depletion of emotional resources, and are implicated in burnout.

The moderating role of organizational support

According to COR theory, external resources not only replenish internal resource reserves but also moderate the progression of resource depletion.Perceived organizational support (POS) is defined as employees’ perception of the extent to which their organization values their contributions and cares about their well-being21. Higher POS has been associated with weaker adverse associations of WPV with psychological and emotional regulation and with lower burnout levels4,22.

When nurses experience high emotional labor demands, adequate organizational support may help them attain psychological compensation and value recognition, thereby reducing the emotional costs associated with emotional labor and alleviating emotional exhaustion and burnout. Conversely, low organizational support can impair this buffering effect and exacerbate the risk associated with high emotional labor and burnout22,23.Moreover, perceived organizational support may attenuate the link between emotional labor and burnout, thereby conditioning the overall sequential indirect pathway from resilience through emotional labor to burnout.Thus, higher POS was associated with attenuation of associations and conditioned the strength of the sequential indirect pathway involving resilience and emotional labor.

Research objectives and hypotheses

In summary, this study conceptualizes WPV as a stressor variable and nurse burnout as the outcome variable. Resilience and emotional labor were treated as components of a sequential indirect pathway, and POS was specified as a moderator. Guided by this framework, we estimated a moderated sequential indirect association model to examine theory-consistent pathways related to nurse burnout and to inform future intervention planning.

Grounded in the Conservation of Resources (COR) theory, the following hypotheses are proposed to test the relationships among the study variables.The conceptual model is illustrated in Fig. 1:

  • H1: WPV is positively associated with nurse burnout.

  • H2: The association between WPV and burnout includes an indirect path via resilience.

  • H3: The association between WPV and burnout includes an indirect path via emotional labor.

  • H4: The association between WPV and burnout includes a sequential indirect path via resilience and then emotional labor.

  • H5: POS moderates the association between emotional labor and burnout.

  • H6: POS moderates the indirect associations of WPV with burnout via emotional labor and via the sequential path.

Fig. 1
figure 1

Theoretical model diagram.

Full size image

Methods

Participants

A cross-sectional survey design was employed in this study. From March to May 2025, convenience sampling was used to recruit currently employed clinical nurses from eight tertiary hospitals in Sichuan Province, China. The inclusion criteria were: (1) age ≥ 18 years with a registered nursing license; (2) at least one year of clinical nursing experience; and (3) voluntary participation with informed consent. The exclusion criteria were: (1) nurses working in internship, training, or administrative roles; (2) those on long-term medical or maternity leave; and (3) individuals who experienced major psychological stress events during the survey period. Because participants were recruited by convenience sampling from multiple hospitals, selection concerns are possible and generalizability may be limited. Nurses were nested within hospitals and units, which introduces potential within-cluster dependence.

Sample size

Considering the model’s complexity, which included multiple mediators and interaction terms for moderation analysis, the required sample size was estimated using G*Power 3.1.With an effect size of f² = 0.05, a significance level of α = 0.05, statistical power of 1 − β = 0.95, and 10 predictors, a minimum of 312 participants was deemed necessary.To ensure data quality and account for potential invalid responses, 600 questionnaires were distributed. After removing 37 responses due to patterned answering or logical inconsistencies, 549 valid responses were retained, resulting in a response rate of 91.5%, which met the required sample size.

Data collection procedure

Data were collected online via the Wenjuanxing platform from March to May 2025 across eight tertiary hospitals. Recruitment was coordinated with the nursing departments and head nurses at participating sites. Uniformly trained research assistants distributed the survey links and provided standardized completion guidance. Participation was anonymous and voluntary, with electronic informed consent obtained prior to data collection. Nurses completed the questionnaire independently during work breaks.

To ensure data integrity, multiple quality-control procedures were implemented. Mandatory response settings and logical skip rules were applied to minimize missingness and logical errors. A minimum completion time of five minutes was required; submissions below this threshold were excluded as abnormally fast. Platform-based checks flagged potential duplicate entries using IP addresses, device identifiers, and submission-pattern logs. No IP or device identifiers were downloaded or stored by the research team; flagged cases were reviewed on-platform and excluded when appropriate. In addition, the research team manually screened all questionnaires and excluded responses showing high similarity, straight-line or patterned responding, or internal inconsistencies. After quality control, 549 valid responses were retained for analysis. Because items were required, item-level missingness was negligible and analyses were conducted on complete cases. We did not apply automated winsorization or statistical outlier deletion; rather, implausible or duplicate entries identified by the predefined quality-control rules were removed prior to analysis.

Instruments

  1. (1)

    Demographic information: Participants’ basic characteristics were collected through a self-administered questionnaire, including gender, age, marital status, education level, type of appointment, professional title, years of clinical experience, frequency of night shifts, salary satisfaction and number of children.These demographic variables were pre-specified by the research team following a focused literature review and consensus discussions.

  2. (2)

    Workplace violence (WPV): The Chinese version of the WPV Scale (WVS)24 was used to assess the frequency and types of violence nurses experienced in the past year.The scale consists of five items covering physical assault, verbal abuse, threats, verbal sexual harassment, and physical sexual harassment.Each item is rated on a 4-point scale (0 = none, 1 = once, 2 = 2–3 times, 3 = 4 or more times), with total scores ranging from 0 to 15.Higher scores indicate more frequent exposure to WPV. The Cronbach’s alpha in this study was 0.857.

  3. (3)

    Resilience: The Connor-Davidson Resilience Scale (CD-RISC), developed by Connor et al.25 and revised by Yu and Zhang26, was used to measure psychological resilience.The scale includes 25 items across three dimensions: tenacity (13 items), strength (8 items), and optimism (4 items).Each item is rated on a 5-point Likert scale (0 = never, 4 = always), with a total score ranging from 0 to 100.Higher scores reflect higher levels of resilience.In the present study, the Cronbach’s alpha for the total scale was 0.926, and for the subscales of tenacity, strength, and optimism were 0.921, 0.877, and 0.871, respectively.

  4. (4)

    Emotional labor: Emotional labor was assessed using the Emotional Labor Scale (ELS), developed by Diefendorff27 and translated by Bai28.The scale contains 14 items with three dimensions: surface acting (SA), deep acting (DA), and natural acting (NA).Each item is rated on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree).All items are positively scored, with higher scores indicating greater use of the corresponding strategy.In this study, the Cronbach’s alpha for the total scale was 0.900,and for the subscales of SA, DA, and NA were 0.891, 0.810, and 0.815, respectively.

  5. (5)

    Burnout: The Maslach Burnout Inventory–Human Services Survey (MBI-HSS), developed by Maslach et al.29 and revised by Feng et al.30,was used to assess burnout levels among nurses.The inventory comprises 22 items covering three dimensions: emotional exhaustion, depersonalization, and reduced personal accomplishment.Items are rated on a 7-point Likert scale (0 = never, 6 = every day). The personal accomplishment subscale is reverse-scored.Total scores range from 0 to 132, with higher scores indicating greater burnout severity. In this study, the Cronbach’s alpha for the overall scale was 0.906, and for emotional exhaustion, depersonalization, and reduced personal accomplishment were 0.898, 0.852, and 0.911, respectively.

  6. (6)

    Perceived organizational support (POS): POS was measured using the scale developed by Eisenberger et al.21 and revised by Ling et al.31.The scale consists of 9 items rated on a 7-point Likert scale (1 = strongly disagree,7 = strongly agree).Total scores range from 9 to 63,with higher scores indicating a greater perception of organizational support. The Cronbach’s alpha in this study was 0.909.

Statistical analysis

All data were analyzed using IBM SPSS Statistics, version 26.0. Continuous variables were summarized as means and standard deviations (mean ± SD), and categorical variables were summarized as frequencies and percentages (n, %). Independent-samples t tests or one-way analysis of variance (ANOVA) were used for group comparisons of continuous variables, and chi-square (χ²) tests were used for categorical variables. Pearson correlation analysis was conducted to assess linear associations among WPV, resilience, emotional labor, POS, and burnout.

Hierarchical regression analysis was performed to examine the association between WPV and burnout while adjusting for covariates selected a priori based on the nursing and occupational-stress literature (gender, age, marital status, employment type, professional title, years of clinical experience, night-shift frequency, salary satisfaction, and number of children). Indirect and moderated-indirect associations were estimated using the PROCESS macro (version 4.1) for SPSS. Model 6 was used to examine sequential indirect pathways involving resilience and emotional labor in the association between WPV and burnout. The bootstrap method with 5,000 resamples was used to estimate 95% confidence intervals (CIs) for indirect associations; an indirect association was inferred when the CI did not include zero. Model 86 was used to examine whether POS moderated the association between emotional labor and burnout and the strength of the indirect pathways. All statistical tests were two-sided, and a p value < 0.05 was considered statistically significant.

All analyses treated observations as independent. Hospital-or unit-level clustering was not modeled in this exploratory analysis, which may render standard errors anti-conservative and inflate statistical significance. Findings are interpreted as associations. Future work should account for the data hierarchy using cluster-robust standard errors or multilevel (mixed-effects) models.

Results

Test for common method bias

To examine potential common method bias, Harman’s single-factor test was conducted using exploratory factor analysis on all measured items.The results indicated that nine factors with eigenvalues greater than one were extracted, accounting for 60.534% of the total variance.The first factor explained 22.326% of the variance, which is well below the critical threshold of 40%, suggesting that common method bias was not a serious concern in this study.However, the Harman single-factor test was used only as a descriptive check and is recognized as an insensitive diagnostic; therefore, residual common method variance cannot be ruled out.

Participant characteristics

Significant differences in burnout scores were observed based on participants’ demographic characteristics, including gender, age, marital status, employment type, professional title, years of clinical experience, number of night shifts per month, salary satisfaction, and number of children (p < 0.05).Higher levels of burnout were found among female nurses, those in middle age, nurses with intermediate titles, contract-based employees, and those with more frequent night shifts.Details are presented in Table 1.

Table 1 Group differences in nurse burnout scores across demographic characteristics (N = 549).
Full size table

Correlation analysis

Pearson correlation analysis revealed that WPV and emotional labor were significantly and positively correlated with nurse burnout (r = 0.325,0.416, p < 0.01).In contrast, resilience and POS were significantly and negatively correlated with burnout (r = − 0.453, − 0.372, p < 0.01). Detailed results are shown in Table 2.

Table 2 Correlation matrix of key study variables (N = 549).
Full size table

Main association testing

After controlling for demographic variables such as gender, age, and professional title, hierarchical regression analysis showed that WPV was significantly and positively associated with nurse burnout (β = 0.269, p < 0.001), with the model explaining 21% of the variance. This indicates that higher levels of exposure to WPV are associated with higher levels of burnout among nurses, thus supporting Hypothesis 1. Detailed results are presented in Table 3.

Table 3 Hierarchical regression analysis of WPV and nurse burnout.
Full size table

Analysis of indirect associations

The bootstrap method in PROCESS v4.1 was used to test the mediating roles of resilience and emotional labor in the relationship between WPV and burnout. The results for each mediation path and the distribution of indirect effects are presented in Tables 4 and 5.

  1. (1)

    Indirect path via resilience.

    Controlling for demographic covariates, WPV was negatively associated with resilience (β = −0.335, p < 0.001), and resilience was negatively associated with burnout (β = −0.273, p < 0.001). After introducing resilience into the model, the direct path coefficient decreased(β = 0.269→β = 0.122), indicating a partial indirect association. These findings support Hypothesis 2.

  2. (2)

    Indirect path via emotional labor.

    Further analysis showed that WPV was positively associated with emotional labor (β = 0.143, p < 0.001), and emotional labor was positively associated with higher burnout levels (β = 0.216, p < 0.001). After including emotional labor in the model, the direct effect of WPV on burnout decreased, indicating a partial indirect association. This supports Hypothesis 3.

  3. (3)

    Sequential indirect pathway via resilience and emotional labor.

    The results showed that resilience was significantly and negatively associated with emotional labor (β = −0.335, p < 0.001). The total effect of WPV on burnout was further reduced after accounting for the sequential mediators. Bootstrap analysis indicated a significant sequential indirect pathway, with an indirect effect of 0.024 (95% CI: [0.014, 0.036]). These findings support Hypothesis 4.

Table 4 Hierarchical regression analysis of WPV, resilience, emotional labor and burnout.
Full size table
Table 5 Sequential indirect pathway analysis of WPV and burnout.
Full size table

Moderation analysis

Moderated indirect associations of POS

When emotional labor, POS, and their interaction term were entered into the model, the interaction between emotional labor and organizational support was significantly and negatively associated with burnout(β = −0.114, p < 0.01). As illustrated in Fig. 2, emotional labor had a stronger positive effect on burnout under conditions of low organizational support. These findings provide empirical support for Hypothesis 5. Detailed results are presented in Table 6.

Table 6 Hierarchical regression analysis testing the moderating effect of POS.
Full size table
Fig. 2
figure 2

Interaction of emotional labor and POS in relation to burnout.

Full size image

Moderated indirect associations of POS

Further moderated mediation analysis revealed that POS significantly moderated both the indirect effect of WPV on burnout through emotional labor and the sequential pathway involving resilience and emotional labor. The differences in indirect effects under high versus low levels of organizational support were statistically significant, supporting Hypothesis 6. See Table 7 for details.

Table 7 Conditional indirect effects at levels of POS.
Full size table

Discussion

This study, grounded in COR theory, examined theory-consistent associations between WPV and burnout through resilience and emotional labor, and assessed whether POS conditioned these associations.The results showed that WPV associated with burnout, with statistical indirect paths via resilience and emotional labor; POS statistically moderated these associations. These findings are consistent with the proposed framework and offer new theoretical insights into the formation of nurse burnout.

Association between WPV and burnout

This study demonstrated that WPV was significantly and positively associated with nurse burnout even after controlling for demographic variables, supporting Hypothesis 1. As a high-intensity external stressor, WPV has been linked to emotional exhaustion and reduce professional identity independently of mediating psychological mechanisms such as resilience or emotional regulation, highlighting its fundamental role in initiating burnout10,32.

According to COR theory13,33,34, when individuals experience resource loss without adequate compensation mechanisms, they are more likely to develop defensive psychological responses such as emotional numbness or disengagement, which ultimately lead to burnout. In the present study, the mean MBI-HSS burnout score among 549 nurses was 55.45 ± 21.31, indicating a substantial psychological burden in this occupational group.These findings emphasize the need for organizational interventions targeting the source of stress.Integrating violence prevention into institutional health strategies may not only reduce individual resource depletion but also contribute to a supportive and health-promoting work environment.

Indirect and sequential indirect associations via resilience and emotional labor

The results confirmed significant indirect associations via resilience and via emotional labor, and a sequential indirect pathway, supporting Hypotheses 2, 3, and 4. This finding highlights a progressive process, wherein nurses exposed to violence undergo psychological resource depletion, behavioral dysregulation, and eventual emotional exhaustion, which has been associated with higher levels.

Resilience, as an internal psychological resource, serves as a key protective buffer in stressful environments35.Violent incidents diminish nurses’ emotional recovery and stress adaptation capacities, making it more difficult for them to engage in cognitive reappraisal and emotional regulation. This impaired ability to mobilize psychological resources increases their vulnerability to prolonged emotional fatigue and cognitive depletion14,36.COR theory posits that individuals in a state of resource imbalance may activate maladaptive emotional defense mechanisms, thereby increasing the risk of burnout13,34.

Emotional labor, as an outward behavioral response, functions as a compensatory mechanism when psychological resilience is insufficient37.Nurses with lower resilience are more likely to rely on surface acting to meet professional display rules38,39.Although this strategy may offer short-term adjustment, its long-term use is associated with increased emotional and cognitive load, contributing to the progression of burnout.

Importantly, the sequential mediation path from resilience to emotional labor to burnout reflects a structured chain of resource depletion.Specifically, WPV may be linked to lower resilienceand greater reliance on surface acting, and these patterns are associated with heightened burnout18,40. Although initially adaptive, this strategy entails substantial emotional energy consumption and cognitive burden, eventually resulting in heightened burnout41.This sequential pathway reflects a pattern of associations linking internal psychological resource depletion, behavioral strain, and emotional exhaustion.The findings reinforce the central role of resilience in the resource system and extend our understanding of the internal dynamics underpinning burnout, providing empirical support for the application of COR theory in high-stress occupational settings.

Moderating role of POS

This study indicated that POS moderated the association between emotional labor and burnout and conditioned the strength of the sequential indirect pathway, supporting Hypotheses 5 and 6. These findings underscore the potential role of external resources in attenuating the chain of associations in resource depletion.

POS statistically moderated the association between emotional labor and burnout. Under low POS, the positive association between emotional labor and burnout was stronger40,42, whereas under high POS this association was weaker and in some cases not statistically significant21,43, consistent with a resource-substitution account. Given that models were estimated on observed composite scores without formal diagnostics or explicit mean-centering, and in light of possible common method bias, the interaction and conditional-indirect associations should be interpreted as exploratory and associational.

POS also conditioned the indirect pathways involving emotional labor, including the sequential indirect pathway. Under low support, WPV showed a stronger positive association with emotional-labor demands, whereas under high support this association appeared weaker44,45. These patterns provide evidence consistent with the resource-buffering hypothesis within COR theory, suggesting that external resources may attenuate the chain of associations along resource-related pathways16,46.

In summary, POS functions as a vital external protective resource that buffers the emotional demands associated with emotional labor and structurally moderates the process of resource depletion. These findings expand the boundary conditions of COR theory by highlighting its regulatory mechanisms and offer practical implications for healthcare organizations to strengthen support systems and build protective work environments through institutional design.

Conclusions and recommendations

Study conclusions

Grounded in Conservation of Resources theory, we examined theory-consistent associations between workplace violence and nurse burnout using a moderated sequential indirect pathway model.WPV was positively associated with burnout and showed indirect associations via resilience and emotional labor, separately and sequentially; higher POS was associated with weaker positive associations between emotional labor and burnout and statistically conditioned the sequential indirect pathway.Furthermore, higher POS was associated with weaker positive associations between emotional labor and burnout and statistically moderated the strength of the sequential indirect pathway.

Practical recommendations

Establish WPV prevention and response systems

Hospitals should implement comprehensive mechanisms for identifying and reporting WPV risks. For instance, violence incident report cards can be placed at nurse stations, and anonymous online reporting platforms can be established to facilitate early warning, timely intervention, and post-incident review. The hospital safety management department should conduct regular risk assessments and prioritize monitoring in high-risk departments. It is also recommended to form “violence response teams” to provide timely conflict mediation and emotional support services.

Enhance nurses’ psychological resilience

Resilience-building modules should be incorporated into pre-employment training programs. These may include mindfulness practices (e.g., breath awareness, emotion labeling), cognitive restructuring techniques, and stress management strategies. Annual psychological health evaluations should include resilience assessments, and nurses with low resilience scores should be offered individual counseling or group-based interventions to support resource restoration.

Implement structured emotional labor management training

Regular workshops focused on emotional expression and regulation skills should be conducted to help nurses understand the differences and implications of surface and deep acting. Training should also include practical communication strategies such as empathy and non-violent communication techniques tailored to clinical scenarios. In departments with high emotional workload, the implementation of a “rotational emotional labor break system” can be considered, allowing nurses to periodically rotate out of emotionally demanding roles to reduce cumulative emotional strain.

Build a multi-level organizational support system

A multi-faceted emotional support network should be developed: First, supervisors should conduct regular one-on-one emotional check-ins with subordinate nurses.Second, the position of “nurse psychological support officer” can be created, filled by experienced senior nurses to provide peer support and coping guidance.Third, an anonymous peer-support platform can be established within the hospital to facilitate informal emotional exchanges among nurses.Organizational support perception should be included as a domain in staff satisfaction surveys, and evaluation results should inform ongoing improvements in support strategies and management mechanisms.

Study limitations and future directions

This study has several limitations that should be considered when interpreting the findings. First, the single-wave, cross-sectional design does not permit causal or temporal inferences. Temporal ordering among WPV, resilience, emotional labor, POS, and burnout is unobserved, and reverse or reciprocal directions are plausible. For example, higher burnout may relate to greater perceived WPV or to different patterns of emotional labor, and POS may covary with burnout in both directions. Accordingly, all conclusions are associational.

Second, participants were recruited by convenience sampling from eight tertiary hospitals in one province. This approach may introduce selection concerns and limits the generalizability of the findings beyond similar settings. Nurses were nested within hospitals and units, but clustering was not modeled; therefore, standard errors may be optimistic and statistical significance may be inflated. Future studies should account for the data hierarchy using cluster-robust standard errors or multilevel (mixed-effects) models.

Third, all variables were collected via single-source self-report at one time point, which raises the possibility of common method bias. The Harman single-factor test was used only as a descriptive check and is known to be insensitive; residual common method variance cannot be ruled out. Future studies should incorporate procedural remedies (for example, temporal or psychological separation of measures) and statistical remedies (for example, a marker variable or an unmeasured latent method factor) to assess and mitigate CMB.

Fourth, Analyses were conducted on observed composite scores. Although internal consistency was acceptable across scales, we did not conduct confirmatory factor analyses or measurement-invariance tests in this sample; accordingly, latent factor structures and comparability across key groups remain to be established. Emotional labor was summarized as a single overall index rather than modeling surface acting, deep acting, and natural expression separately, even though heterogeneous associations are plausible, particularly for moderation. Burnout (MBI-HSS) was treated primarily as a continuous score, and categorical severity thresholds were not used in inferential analyses because validated cut-points for the Chinese adaptation vary across studies. The WPV measure summarized frequency across event types and did not encode event severity or recency. Future research should employ latent-variable models, analyze emotional-labor and burnout subdimensions explicitly, evaluate measurement invariance, and consider multi-facet or weighted WPV instruments that capture severity and timing.

Fifth, PROCESS paths were estimated on observed scores without modeling measurement error, and we did not report formal diagnostics for linearity, influential observations, heteroskedasticity, or multicollinearity; explicit mean-centering for interaction terms was not implemented. Some plausible contextual covariates (such as unit type, workload/staffing/acuity, job control/autonomy, and leadership climate) were unavailable, so omitted variable bias cannot be ruled out. Future work should incorporate latent-variable and multilevel models, pre-specified covariate sets that include contextual factors, and comprehensive diagnostic and sensitivity analyses.

To address these limitations, future studies should adopt longitudinal or multi-wave designs (for example, cross-lagged panels), link self-reports with external sources such as supervisor ratings or incident reports, and use latent and multilevel structural equation models. These steps would better address temporal ordering, measurement error, clustering, and common method bias, and would allow a more rigorous examination of theory-consistent pathways.