Burnout as a Path Between Decent Work and Turnover Intention: The Buffering Effect of Calling

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

The manuscript addresses a timely and relevant topic by examining burnout as a mediating mechanism between decent work and turnover intention, with career calling as a buffering moderator. The theoretical framework is appropriate, the hypotheses are clearly formulated, and the proposed model is coherent with the JD-R and Psychology of Working Theory perspectives. Overall, the study is well structured and aligned with the aims of the Special Issue.

Nevertheless, the manuscript would benefit from a clearer articulation of its contribution relative to recent closely related studies already cited, particularly those examining career calling in stress and burnout models. More explicitly distinguishing how the present conditional process model extends beyond this literature would strengthen its originality.

Methodologically, the use of validated measures, PROCESS models, and bootstrapping is appropriate, and internal consistencies are satisfactory. However, reliance on convenience and snowball sampling should be more explicitly acknowledged as a limitation with implications for generalizability.

In addition, while common method variance is addressed via Harman’s single-factor test, this approach is limited; given that a CFA was conducted, the manuscript would benefit from clarifying the criteria used to establish discriminant validity. Some professional field control categories have very small subsample sizes (e.g., IT; lines; 155-156), which may affect estimate stability and should be interpreted cautiously.

The main mediation results are clearly presented, showing a significant indirect effect via burnout and a non-significant direct effect of decent work on turnover intention. However, the moderated mediation section includes internal inconsistencies that require careful verification. In particular, the reported sign of the t value for the effect of decent work on burnout appears inconsistent with the beta coefficient (lines 337-338), several effects described as "trends" have p values below .05 and are reported as significant in Table 3 (lines 344-348; 390-392), and the non-significant effect of burnout on turnover intention in the moderated model (lines 367-368) contrasts with its strong effect in the mediation-only model (lines 310-311).

Finally, although the analyses support the proposed conditional indirect effects (lines 350-354; 376-381), the manuscript frequently uses causal or interventive language that is not warranted by the cross-sectional design, particularly in the Abstract (lines 10-16) and in the Conclusions (lines 467-473). Revising these formulations to emphasize associations rather than causality would align interpretation with the study design and with the author's own acknowledgment of causal limitations (lines 451-453).

Addressing these points would further strengthen the manuscript's rigor, clarity, and overall contribution.

Best regards

Author Response

We greatly appreciate the positive comments and detailed evaluation of our manuscript.

 

Comment 1: The manuscript would benefit from a clearer articulation of its contribution relative to recent closely related studies already cited, particularly those examining career calling in stress and burnout models. More explicitly distinguishing how the present conditional process model extends beyond this literature would strengthen its originality.

We thank the reviewer for this valuable suggestion. In response (in red on lines 35–55), we have clarified the study’s contribution relative to recent literature on career calling and stress/burnout models. Specifically, the introduction has been revised to explicitly state that while previous studies have often focused on specific professional contexts or examined career calling independently, the present study integrates both organizational (decent work) and personal (career calling) resources within a single conditional process model across diverse occupational sectors. This change highlights the originality of our approach and specifies how the study extends beyond prior work by examining the circumstances under which career calling amplifies the protective effect of decent work on burnout and turnover intention.

 

Comments 2: Methodologically, the use of validated measures, PROCESS models, and bootstrapping is appropriate, and internal consistencies are satisfactory. However, reliance on convenience and snowball sampling should be more explicitly acknowledged as a limitation with implications for generalizability.

Response 2: We agree that convenience and snowball sampling are significant limitations. In the manuscript (line 534-536), we explicitly address this issue in the specifications section, noting that the sampling method may limit the representativeness of the sample and the generalizability of the results to the broader population of workers in Portugal. Furthermore, we emphasize the need for future studies with more diverse and representative samples to validate and expand these findings (line 537).

 

Comments 3: In addition, while common method variance is addressed via Harman’s single-factor test, this approach is limited; given that a CFA was conducted, the manuscript would benefit from clarifying the criteria used to establish discriminant validity. Some professional field control categories have very small subsample sizes (e.g., IT; lines; 155-156), which may affect estimate stability and should be interpreted cautiously.

Response 3: Thank you for the comment. We acknowledge that the single-factor Harman test is limited in detecting common method variance. To address this, we conducted CFAs for each construct separately to assess convergent validity, and a full measurement model including all constructs simultaneously to examine discriminant validity. Discriminant validity was established following Fornell and Larcker (1981), as the AVE for each construct exceeded the squared correlations with other construct. While some professional field subgroups were underrepresented (e.g., Information Technology), the CFA was conducted on the full sample (N = 225), supporting the stability of factor estimates, although caution is warranted when generalizing to smaller subgroups (see lines 288-295).

 

Comments 4: The main mediation results are clearly presented, showing a significant indirect effect via burnout and a non-significant direct effect of decent work on turnover intention. However, the moderated mediation section includes internal inconsistencies that require careful verification. In particular, the reported sign of the t value for the effect of decent work on burnout appears inconsistent with the beta coefficient (lines 337-338), several effects described as "trends" have p values below .05 and are reported as significant in Table 3 (lines 344-348; 390-392), and the non-significant effect of burnout on turnover intention in the moderated model (lines 367-368) contrasts with its strong effect in the mediation-only model (lines 310-311).

Response 4: We thank the reviewer for carefully reading the results of the moderated mediation and for highlighting several important issues related to the presentation and interpretation of the analyses.

First, we corrected a typographical inconsistency in the presentation of the association between decent work and burnout in the moderated model. The sign of the t-value was changed to correctly reflect the negative regression coefficient (β = −0.573, t = −1.87, p = 0.064) (line 378).

Second, we revised the description of several occupational field effects that were previously referred to as “trends.” As some of these effects reached statistical significance (p < 0.05) but had very small magnitudes and are not theoretically central to the hypotheses, we now present them descriptively, without interpretative emphasis, and avoid inconsistent terminology (line 384).

Third, regarding the non-significant effect of burnout on turnover intention in the moderated mediation model, we clarify this point in the Results section. Although burnout showed a strong and significant association with turnover intention in the simple mediation model, this effect did not reach statistical significance after the inclusion of professional vocation and interaction terms. We now explicitly note that this pattern is consistent with models that include moderation, in which shared variance and interaction effects can attenuate direct pathways (lines 342-344).

 

Comments 5: Finally, although the analyses support the proposed conditional indirect effects (lines 350-354; 376-381), the manuscript frequently uses causal or interventive language that is not warranted by the cross-sectional design, particularly in the Abstract (lines 10-16) and in the Conclusions (lines 467-473). Revising these formulations to emphasize associations rather than causality would align interpretation with the study design and with the author's own acknowledgment of causal limitations (lines 451-453).

Response 5: We thank the reviewer for this important and constructive comment. We fully agree that the cross-sectional design of the study does not allow for causal or interventional interpretations. In response, we carefully revised the entire manuscript to systematically replace causal and interventional language with associative and correlational language. Specifically, expressions that implied causality were reformulated to emphasize associations, indirect relationships, and statistical patterns (e.g., “was associated with,” “was indirectly related to,” “was statistically explained”).

Reviewer 2 Report

Comments and Suggestions for Authors

I appreciate the opportunity to review this manuscript. The topic is timely and relevant, and the manuscript is generally well written and clearly organized. I have several minor comments that I hope will help further strengthen the paper.

1. Demographic information: The demographic information is presented clearly and in appropriate detail. However, I wondered whether variables such as level of education and income were assessed. These factors may be relevant to the variables of interest. Additionally, career calling may differ by professional field, and some discussion or clarification regarding this point may be helpful.
2. Use of subscales: In the Measures section, the authors describe the subscales of the Decent Work Scale, burnout, and career calling. However, it is unclear whether these subscales were actually used in the analyses. If subscales were not used, it would be helpful for the authors to explain the rationale for using total scores, particularly given that the original scale developers proposed meaningful subscale structures.
3. Survey completion time and missing data: The authors note that the average survey completion time was approximately 20–30 minutes. Given that participation was voluntary and uncompensated, this is quite notable. I wondered whether a substantial number of participants did not complete the survey due to its length. If so, it would be helpful for the authors to provide additional information on how incomplete responses or missing data were handled in the analyses.
4. Model fit indices: The manuscript does not clearly specify the criteria used to evaluate model fit. In particular, the reported IFI and CFI values appear relatively low given the sample size and commonly accepted benchmarks. Additional explanation or justification for the interpretation of these fit indices would strengthen this section.
5. Confirmatory factor analysis (CFA): I may have overlooked this detail, but it was unclear how the authors approached the CFA. Given that multiple scales were used, I would expect either separate CFAs for each scale or a clearly specified overall measurement model. Clarification regarding whether a full measurement model was tested would be helpful.
6. Table 1: Table 1 appears to be cut off, which makes it difficult to interpret. Additionally, although readers may infer what variables 1, 2, 3, etc. represent, clarity would be improved by explicitly labeling them in the table (e.g., 1. Decent Work, 2. Turnover Intention) in the left column.
7. Use of total versus average scores:  The authors used total scores for all scales, despite differences in scale length. While this choice does not affect correlations or model estimates, it places the constructs on different metrics and results in standard deviations that vary as a function of the number of items. Using average scores (or standardized scores) would improve interpretability and comparability across constructs. Clarification of the rationale for this decision would be helpful.

Author Response

We greatly appreciate the positive comments and detailed evaluation of our manuscript.

Comments 1: The demographic information is presented clearly and in appropriate detail. However, I wondered whether variables such as level of education and income were assessed. These factors may be relevant to the variables of interest. Additionally, career calling may differ by professional field, and some discussion or clarification regarding this point may be helpful.

Response 1: We appreciate the positive feedback on the presentation of demographic information. We agree that variables such as education level and income may be relevant to variables of interest, particularly in the context of perceptions of decent work, burnout, and turnover intention. However, these data were not collected in this study, which represents a limitation that we have noted in the limitations section (lines 538-539). We highlight that future research could include these variables to explore potential additional effects or confounding factors. Additionally, we acknowledge that career calling can vary across different fields of work. To address this, we briefly discussed in the manuscript that the effect of calling may interact with specific occupational characteristics and suggested that future studies examine how different work areas influence the perception of calling and its relationship with burnout and turnover intention (lines 476-481).

 

Comments 2: In the Measures section, the authors describe the subscales of the Decent Work Scale, burnout, and career calling. However, it is unclear whether these subscales were actually used in the analyses. If subscales were not used, it would be helpful for the authors to explain the rationale for using total scores, particularly given that the original scale developers proposed meaningful subscale structures.

Response 2: We appreciate the detailed comments on the measures section. Although the original versions of the scales used include subscales, our study used only the total scores of each scale in the analyses. This decision reflects the central objective of the study, which was to examine global associations between decent work, burnout, career calling, and turnover intention, including modeling mediation and moderated mediation using PROCESS. The use of total scores is justified because PROCESS focuses on aggregate effects rather than specific subscales, which simplifies interpretation and avoids unnecessary complexity in the models.

Additionally, we have revised the measures section to make it clearer and more closely aligned with the reported results. We have included only representative examples of items from each scale, without describing all subscales in detail, to ensure that readers understand the nature of the measures and that the description matches what was analyzed (see lines 194-219). We believe these changes resolve the confusion identified and maintain consistency between the measures and analyses.

 

Comments 3: The authors note that the average survey completion time was approximately 20–30 minutes. Given that participation was voluntary and uncompensated, this is quite notable. I wondered whether a substantial number of participants did not complete the survey due to its length. If so, it would be helpful for the authors to provide additional information on how incomplete responses or missing data were handled in the analyses.

Response3: We appreciate the reviewer's feedback. We added to the procedures section that all items in the questionnaire were mandatory, so all 225 participants completed every section; no missing data occurred (lines 245-247).

 

Comments 4: The manuscript does not clearly specify the criteria used to evaluate model fit. In particular, the reported IFI and CFI values appear relatively low given the sample size and commonly accepted benchmarks. Additional explanation or justification for the interpretation of these fit indices would strengthen this section.

Response 4: Thank you for your comment. We have added an explanation of the model fit assessment criteria. Although ideal CFI and IFI values are often considered to be ≥ 0.95, values close to 0.90 are also accepted as indicative of reasonable fit, especially in complex models or with moderately sized samples (Kline, 2016; Hair et al., 2019) (lines 298-300).

In addition, RMSEA = 0.075 indicates adequate fit (≤ 0.08), and the PNFI and PGFI indices suggest that the model demonstrates good parsimony. All factor loadings were statistically significant and greater than 0.50, confirming convergent validity. Discriminant validity was also supported by the results. Thus, we consider the use of composite scores in subsequent analyses to be justified (lines 301-304).

Kline, R.B. (2016). Principles and Practice of Structural Equation Modeling (4th ed.). Guilford Press.

Hair, J.F., Risher, J.J., Sarstedt, M., & Ringle, C.M. (2019). When to use and how to report PLS-SEM results. European Business Review, 31, 2-24. https://doi.org/10.1108/EBR-11-2018-0203

 

Comments 5: I may have overlooked this detail, but it was unclear how the authors approached the CFA. Given that multiple scales were used, I would expect either separate CFAs for each scale or a clearly specified overall measurement model. Clarification regarding whether a full measurement model was tested would be helpful.

Response 5: Thank you for the comment. We appreciate the opportunity to clarify the CFA procedures (see lines 288-303). Given the multiple scales used in the study, we first conducted separate CFAs for each construct (Decent Work, Burnout, Career Calling, and Turnover Intention) to evaluate convergent validity, ensuring that all items loaded significantly on their respective factors (all λ > .50, p < .001).

Subsequently, we tested a full measurement model including all constructs simultaneously to examine discriminant validity. Fit indices for the full model indicated acceptable fit (χ² = 2359.159, χ²/df = 2.595, RMSEA = .075, CFI = .901, IFI = .902, PNFI = .780, PGFI = .630), and discriminant validity was supported, as the AVE for each construct exceeded the squared correlations with other constructs (Fornell & Larcker, 1981).

These results justified the use of composite scores for subsequent analyses, while acknowledging that some professional field subgroups (e.g., Information Technology) were underrepresented in the sample, which may limit generalizability across specific occupational groups.

 

Comments 6: Table 1 appears to be cut off, which makes it difficult to interpret. Additionally, although readers may infer what variables 1, 2, 3, etc. represent, clarity would be improved by explicitly labeling them in the table (e.g., 1. Decent Work, 2. Turnover Intention) in the left column.

Response 6: Thank you for your feedback. We followed the recommendation and revised Table 1 (line 331-332) to improve clarity and ease of interpretation. All variables are now explicitly labeled in the left-hand column, including both primary variables, control variables, and industry/occupation dummies. We also adjusted the numbering of the variables in the correlation matrix to accurately reflect all rows in the table and avoid confusion.

 

Comments 7: The authors used total scores for all scales, despite differences in scale length. While this choice does not affect correlations or model estimates, it places the constructs on different metrics and results in standard deviations that vary as a function of the number of items. Using average scores (or standardized scores) would improve interpretability and comparability across constructs. Clarification of the rationale for this decision would be helpful.

Response 7: Thank you for your comment regarding the use of total scores versus mean scores. We standardized all variables, except demographic ones, prior to the moderated mediation analysis to account for differences in the number of items across constructs. This ensures comparability of the scales and proper interpretation of the coefficients. We have clarified this point in the Data Analysis section (see lines 266-268).

Reviewer 3 Report

Comments and Suggestions for Authors

The hypothesis structure and results write-up need clearer alignment, especially for H1. You state H1 as a direct relationship between Decent Work and Turnover Intention, but in the Results you primarily conclude “full mediation” without explicitly stating whether H1 was supported or not. Please clarify whether H1 refers to the total effect (Decent Work, Turnover Intention without the mediator) or the direct effect. If H1 is intended as a direct-effect hypothesis, then the non-significant direct path should be explicitly reported as “H1 not supported.” If H1 is intended as a total-effect expectation, then you need to report the total effect separately and distinguish it from the direct effect in the mediation model. In the Results section, you should explicitly state which hypotheses are supported or not supported to avoid ambiguity.

Because the study is cross-sectional and based on self-report data, you need to tone down the causal wording throughout the Results and Discussion. Please replace causal verbs with association-consistent phrasing.

You report a positive main effect of calling on burnout in the moderated model (calling associated with higher burnout), while theorizing calling as protective. Please reconcile this inconsistency by acknowledging calling as potentially double-edged, and add theoretical interpretation and boundary conditions.

Author Response

We greatly appreciate the detailed evaluation of our manuscript.

Comments 1: The hypothesis structure and results write-up need clearer alignment, especially for H1. You state H1 as a direct relationship between Decent Work and Turnover Intention, but in the Results you primarily conclude “full mediation” without explicitly stating whether H1 was supported or not. Please clarify whether H1 refers to the total effect (Decent Work, Turnover Intention without the mediator) or the direct effect. If H1 is intended as a direct-effect hypothesis, then the non-significant direct path should be explicitly reported as “H1 not supported.” If H1 is intended as a total-effect expectation, then you need to report the total effect separately and distinguish it from the direct effect in the mediation model. In the Results section, you should explicitly state which hypotheses are supported or not supported to avoid ambiguity.

Response 1: We appreciate the feedback. We have revised the formulation of Hypothesis 1 (lines 92-93) and the presentation of the results for greater clarity and alignment between theory and analysis. We clarify the following points:

1. Hypothesis 1 (total effect): H1 refers to the total association between decent work and turnover intention, considering the simple correlation between the two variables without including mediators. H1 was partially supported: significant correlation, but not a significant direct effect, explained as an indirect effect via burnout.
2. Hypothesis 2, regarding burnout mediation, was confirmed, as the indirect effect of decent work on turnover intention via burnout was significant.

Hypothesis 3, regarding mediation moderated by professional calling, has been partially updated: only DW → burnout was moderated per call; burnout → turnover was not moderated.

Each hypothesis is now clearly distinguished by the type of effect expected (total, direct/indirect, and moderated) and its confirmation status, eliminating ambiguities in the interpretation of the results.

 

Comments 2: Because the study is cross-sectional and based on self-report data, you need to tone down the causal wording throughout the Results and Discussion. Please replace causal verbs with association-consistent phrasing.

Response 2: We thank the reviewer for this important and constructive comment. We fully agree that the cross-sectional design of the study does not allow for causal or interventional interpretations. In response, we carefully revised the entire manuscript to systematically replace causal and interventional language with associative and correlational language. Specifically, expressions that implied causality were reformulated to emphasize associations, indirect relationships, and statistical patterns (e.g., “was associated with,” “was indirectly related to,” “was statistically explained”).

 

Comments 3: You report a positive main effect of calling on burnout in the moderated model (calling associated with higher burnout), while theorizing calling as protective. Please reconcile this inconsistency by acknowledging calling as potentially double-edged, and add theoretical interpretation and boundary conditions.

Response 3: We thank the reviewer for their comment, which helped us clarify and strengthen the theoretical interpretation of our findings.

We thank the reviewer for their comment, which helped us clarify and strengthen the theoretical interpretation of our findings.

In response, we revised the Discussion (lines 476-480) to explicitly acknowledge the potentially ambivalent role of vocation. Specifically, we now discuss how vocation, while often functioning as a personal resource, can also be associated with greater emotional investment, higher internalized standards, and increased personal demands, which can make individuals more vulnerable to burnout when contextual resources are insufficient.

Reviewer 4 Report

Comments and Suggestions for Authors

1. 1. 1.Introduction

      Introduction is still very weak, please explain your debate with the previous study, then the contribution of your present study.

      Literature Review and Hypothesis Development

      Hypothesis H1 and H3 didn't clear the relationship ( must be clear direct, indirect( mediating), moderating)

      3. Materials and Methods

      Note 1: Please organize again the Materials and Methods - how you get the results of your survey without you design your survey (questionnaire)?,

      Note 2: When you Materials and Methods - Please clearly explain How you design, and what means and measurement your study, distribution and data collection, data analytic tool and approach.

      Sample in your 3.2. Measures - your measurements are very mess-up didn't match with the results of the study.

      Please design your survey (questionnaire) in one table

      4. Results

      The data didn't analysis properly. For example: Table 2. Analysis of mediation effects. didn't match to the figure 1, also similar to another.

      The following must also revise.

      2026-1-8

Comments on the Quality of English Language

Many in diferences part of the body text expesially in the 3. Materials and Methods

Author Response

Comments 1: Introduction is still very weak, please explain your debate with the previous study, then the contribution of your present study.

Response 1: We thank the reviewer for this constructive feedback. In response, we have strengthened the introduction to clearly articulate the debate with previous studies and the contribution of the present study (lines 35-55).

Specifically, we now explicitly discuss that previous research has investigated the role of career calling in buffering stress and burnout, often within specific professional contexts such as healthcare or education. While these studies highlight important relationships, they typically examine career calling in isolation or focus on a single occupational sector, which limits the generalizability of their findings.

Our study addresses this gap by integrating both organizational (decent work) and personal (career calling) resources within a single conditional process model across diverse occupational sectors. This approach allows us to examine burnout as a mediator of the relationship between decent work and turnover intention, while testing career calling as a moderator, thereby identifying the circumstances under which career calling amplifies the protective effect of decent work.

Through these revisions, the introduction now clearly situates the study within the existing literature, articulates the debate regarding career calling and burnout models, and emphasizes the originality and theoretical/practical contributions of our work.

 

Comments 2: Hypothesis H1 and H3 didn't clear the relationship (must be clear direct, indirect( mediating), moderating).

Response 2: Thank you for your comment. For greater clarity, we have revised the hypotheses, explicitly distinguishing the types of associations tested:

Hypothesis 1 (Total association): Higher levels of decent work are associated with lower levels of turnover intention (lines 92-93)

Hypothesis 2 (mediation): Burnout mediates the association between decent work and turnover intention (lines 121-122).

Hypothesis 3 (moderate mediation): The indirect association of decent work with turnover intention via burnout is conditional on employees’ levels of career calling, such that career calling moderates both the association between decent work and burnout and the association between burnout and turnover intention (lines 153-156).

With this revision, each hypothesis independently clarifies the type of effect tested (total, direct/indirect, and moderator), aligning the theoretical formulation with the statistical analysis conducted.

 

Comments 3: Note 1: Please organize again the Materials and Methods - how you get the results of your survey without you design your survey (questionnaire)?,

Note 2: When you Materials and Methods - Please clearly explain How you design, and what means and measurement your study, distribution and data collection, data analytic tool and approach.

Sample in your 3.2. Measures - your measurements are very mess-up didn't match with the results of the study.

Response 3: We appreciate the feedback and have reorganized the “Materials and Methods” section to ensure clarity and alignment with the recommendations. The section now explicitly states:

1. The digital questionnaire consisted of two main parts: (i) sociodemographic characterization of the participants, and (ii) psychometric instruments used to assess decent work, burnout, career aptitude, and turnover intention. All items were mandatory, ensuring no missing data.
2. Although the original versions of the scales include subscales, our study used only the total scores of each scale in the analyses. This decision aligns with the central objective of the study, which was to examine global associations between decent work, burnout, career calling, and turnover intention, including the modeling of mediation and moderated mediation using PROCESS. Using total scores is justified because PROCESS focuses on aggregate effects rather than specific subscales, which simplifies interpretation and avoids unnecessary complexity in the models. Additionally, we revised the measures section to make it clearer and more consistent with the reported results. We included only representative examples of items from each scale, without detailing all subscales, to ensure that readers understand the nature of the measures and that the description matches what was actually analyzed. We believe these changes address the confusion raised by the Reviewer.
3. Participation was voluntary and unpaid. The questionnaire link was distributed through social networks (Facebook, Instagram, WhatsApp) and the researchers' professional networks, using non-probabilistic, convenience, purposive, and snowball sampling. Data collection took place between October and December 2024, with an average completion time of 20–30 minutes.
4. Analyses were conducted in SPSS 29.0 with PROCESS Macro (Hayes, 2013). Descriptive analyses, normality assessment, internal consistency (Cronbach’s α), single-method test (Harman’s Single-Factor Test), and CFA were performed to confirm construct validity. Mediation effects (burnout) and moderated mediation (vocational vocation) were then tested using 5,000 bootstrap samples with non-parametric confidence intervals. All variables, except demographic ones, were standardized for comparability.

This reorganization aims to present the entire methodological procedure in a clear and sequential manner, making evident how the questionnaire was structured, which instruments were used, how data distribution and collection occurred, and which analytical methods were applied.

 

Comments 4: The data didn't analysis properly. For example: Table 2. Analysis of mediation effects. didn't match to the figure 1, also similar to another.

Response 4: We thank the reviewer for raising this important point. We would like to clarify that the data analyses were conducted correctly and in accordance with the statistical procedures described (PROCESS Models 4 and 59). The apparent inconsistency between Table 2 and Figure 1 did not result from analytical errors, but rather from an initial misalignment between the theoretical model, the formulation of the hypotheses, and their graphical representation.

Specifically, the original version of the manuscript did not clearly distinguish between the total association (Hypothesis 1), the mediation model (Hypothesis 2), and the moderated mediation model (Hypothesis 3), which may have led to confusion regarding the correspondence between Figure 1 and the mediation results reported in Table 2. In response to the reviewer’s comment, we have reformulated the hypotheses and revised the theoretical model to ensure full conceptual and methodological coherence among the study objectives, hypotheses, analytical strategy, and reported results.

We believe that these revisions resolve the inconsistency noted by the reviewer and that the manuscript now presents a coherent and transparent alignment between theory, hypotheses, analyses, and findings.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed my comments thoroughly and revised the manuscript accordingly. I have no further concerns and recommend acceptance.

Author Response

Comment 1: The authors have addressed my comments thoroughly and revised the manuscript accordingly. I have no further concerns and recommend acceptance

Response 1: We sincerely thank the reviewer for the careful reading of our manuscript and for the constructive feedback provided throughout the review process. We greatly appreciate the time and effort invested in evaluating our work. 

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript has been progress but still needs to revise, when seeing the topic and the results of the report a little bit missing 

“Burnout as a Path Between Decent Work and the Turnover 
Intention: The Buffering Effect of Calling”

In the report of the results of the study: please make sure the report of the H1, because you write mess-up with the mediating effects. That’s didn’t only confuse to the reader but also unclear the position of the report and the topic you have chosen.

Note: Please find in the page 8 and revise it.

Good, Luck 
2026/2/11

 

Comments on the Quality of English Language

Many in diferences part of the body text expesially in the 3. Materials and Methods

Author Response

Comment 1: The manuscript has been progress but still needs to revise, when seeing the topic and the results of the report a little bit missing. In the report of the results of the study: please make sure the report of the H1, because you write mess-up with the mediating effects. That’s didn’t only confuse to the reader but also unclear the position of the report and the topic you have chosen. Note: Please find in the page 8 and revise it.

Response1: Thank you for your constructive comments. We have carefully revised the Results section to improve clarity and explicitly distinguish between total, direct, and indirect effects as they relate to Hypothesis 1.

Specifically, we have added a clarifying sentence after the correlation results in Section 4.2: “It is important to note that Hypothesis 1 refers to the total association between decent work and turnover intention, whereas subsequent regression analyses distinguish between total, direct, and indirect effects as required for mediation testing.”

This ensures that readers clearly understand the role of H1 in the context of mediation.

Clarified the conditional indirect effect of burnout in the moderated mediation model (H3) in Section 4.4. We explicitly state that although burnout did not reach conventional significance in the full model predicting turnover intention, bootstrap confidence intervals confirm that the indirect effect of decent work on turnover intention via burnout is significant at moderate and high levels of career calling. This highlights that the mediation effect remains meaningful when accounting for the moderator.

 We believe these changes address your concerns and improve the interpretability of the results for readers.