Burnout Among Psychologists: Direct Effects of Work Engagement and the Absence of Mediation by Areas of Worklife

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript addresses a critical topic in occupational health: burnout among psychologists, integrating work engagement and areas of worklife. Is well-structured, with an abstract that adequately covers the objectives, methodology, and main findings. It focuses on psychologists as a vulnerable group due to high emotional and relational demands.

The theoretical background is well‑structured, drawing on classical and contemporary authors However it can be more focus on burnout, engagement, and areas of worklife with some detailed information about early-career psychologists.

The article is technically sound and theoretically well-grounded, though it has statistical and sampling limitations, the sample is a reasonable size, the predominance of females (88.3%), but is mostly young adults (47.8% aged 24-29) limit the generalizability of the results.

The manuscript reports the absence of mediation but does not sufficiently explore theoretical explanations or alternative models. For example: “No evidence was found to support the mediating role…”. Consider discussing moderation, non‑linear effects, or contextual moderators.

The discussion should focus more on the specific challenges of early-career psychologists.

The title should be revised: The title claims to study psychologists in general, the results are actually representative of early‑career of psychologists.

Author Response

We would like to thank Reviewer 1 for the careful reading of our manuscript and for the constructive comments, which have contributed substantially to improving the quality of the work.

Comment 1: The theoretical background can be more focused on burnout, engagement, and areas of worklife, with more detailed information about early-career psychologists.

We thank the reviewer for this observation. The Introduction was substantially revised to improve the focus on the three core constructs and their interrelationships. A synthesis explicitly connecting burnout, areas of worklife, and engagement was added to make the theoretical framework more cohesive and directly tied to the research problem. Regarding early-career psychologists, the following sentence was added to the Introduction: "Burnout is particularly prevalent among health professionals, especially in countries with fewer resources, and is more common among younger and early-career professionals, who tend to face higher workloads, fewer resources, and greater vulnerability to occupational exhaustion compared to more experienced colleagues" (Wright et al., 2022).

Comment 2: The manuscript reports the absence of mediation but does not sufficiently explore theoretical explanations or alternative models. Consider discussing moderation, non-linear effects, or contextual moderators.

We agree with this observation and revised the Discussion accordingly. The section addressing H4 was expanded to include a more substantive theoretical interpretation of the non-significant mediation finding. Specifically, we discuss the possibility that engagement operates more proximally on burnout than structural work conditions, and that individual factors such as personality traits, coping strategies, and social support may play a more central mediating role than areas of worklife. We also acknowledge that moderation effects and non-linear relationships between these constructs remain theoretically plausible and represent relevant directions for future research, as suggested by the reviewer.

Comment 3: The discussion should focus more on the specific challenges of early-career psychologists.

We acknowledge this observation and revised the Discussion accordingly. In the section addressing H1, the interpretation of the non-significant finding was expanded to include a more explicit discussion of the challenges faced by early-career psychologists. Specifically, we added that early-career psychologists may be particularly susceptible to worklife incongruences given their limited professional experience, reduced autonomy, and fewer established coping resources — factors that may interact differently with engagement and burnout compared to more experienced professionals (Dorociak et al., 2017). The predominance of young participants in the sample is also discussed in the Limitations section as a factor that may limit the generalizability of the findings.

Comment 4: The title should be revised to reflect that the results are actually representative of early-career psychologists.

We appreciate this suggestion. However, the study was originally designed to investigate burnout among psychologists in general, including both self-employed and institutionally employed professionals. The predominance of young participants reflects a characteristic of the convenience sample rather than an intentional focus on early-career psychologists. Restricting the title to this subgroup would misrepresent the scope of the study. The limitation related to the sample's age profile is explicitly acknowledged and discussed in the Limitations section. The title was nonetheless revised to better reflect the study's most relevant finding — the non-mediating role of areas of worklife — as follows: "Burnout among psychologists: work engagement as a direct predictor and the non-mediating role of areas of worklife."

Reviewer 2 Report

Comments and Suggestions for Authors

 

The manuscript has clear potential and a solid empirical base, but it requires revision before it can meet publication standards. The main issue is not the data; it is the conceptual and analytical coherence of the paper.

The focus on psychologists, including self-employed professionals, is a genuine contribution, given the scarcity of literature in this subgroup.

The title is adequate but slightly generic. It could be sharpened to reflect the non-confirmation of mediation, which is in fact one of your most interesting results.

The abstract is generally clear, but it suffers from two problems. First, it is overly descriptive and repeats obvious statements (“unexpected results and limitations”). Second, it lacks precision regarding the analytical strategy; SEM is central to your study but only implicitly mentioned. Third, the phrase “quantitative and qualitative elements” is misleading, since no qualitative data are actually presented. This should be corrected.

The introduction is one of the strongest sections conceptually, but also one of the most uneven. On the positive side, the integration of psychodynamics of work (Dejours) with neoliberal/performance society (Han) gives depth and originality. However, the section becomes overextended and insufficiently focused on your research problem. The discussion of macro-social transformations is interesting but not sufficiently tied to the operational variables (AWS, UWES, BAT).

There are also some stylistic and structural issues. The transition from philosophical framing to empirical constructs is abrupt. Some statements are too general (e.g., “studies shows…”also grammatically incorrect). The literature review becomes list-like in the middle section (pages 2–3), where constructs are described rather than critically synthesized.

Most importantly, the research gap is not sharply articulated. You mention that few studies examine these variables among psychologists, but you do not clearly state what is theoretically unresolved: is it the mediation? the role of autonomy? differences between self-employed vs. employed?

The hypotheses are clear and logically derived, but there is a conceptual inconsistency that later becomes visible in the results:. You propose mediation (H4), but in the Results section you explicitly test AWS as a moderator (“to investigate whether AWS acts as a moderator in the relationship between”,  p. 10). This is not a minor wording issue, it reflects confusion in the analytical model. Mediation and moderation are fundamentally different. This must be corrected consistently across:

• Methods
• Results
• Figure 1
• Discussion

At present, the reader cannot be fully confident about what was actually tested.

The methods section is generally adequate, but several points require clarification or correction:

First, the design is described as “quantitative with qualitative elements,” which is inaccurate. No qualitative component is reported. This should be removed. Second, the sampling strategy (WhatsApp, Instagram) is clearly convenience-based, but this is under-problematized. Given the high proportion of young participants (almost half aged 24–29), the implications for external validity should be more explicitly acknowledged already here, not only in limitations.

Third, the justification for sample size (n = 180) based on instrument licensing is understandable, but scientifically weak. You should complement this with a statement about statistical adequacy for SEM.

Fourth, the SEM description is too brief. Key elements are missing: Model fit indices (CFI, TLI, RMSEA, SRMR); Measurement vs structural model distinction;  Handling of non-normality (you report non-Gaussian distributions later)

Finally, the AI disclosure is appropriate, but it slightly disrupts the flow. It could be moved to a more standard position (e.g., end of Methods or separate note).

The Results section is detailed but overly descriptive and sometimes redundant. You spend excessive space reporting basic descriptive statistics (min, max, quartiles), which are not analytically central. At the same time, key SEM outputs are underreported.  The correlation section is acceptable, but the interpretation becomes repetitive and mechanically follows the table.  The SEM results are the most important part, yet:  The R² = 96.29% is suspiciously high and not critically discussed. This raises concerns about shared method variance or construct overlap.  You report coefficients and p-values, but no model fit indices. The mediation test is described, but without a clear statistical framework (e.g., bootstrapping).

Overall, the section needs compression in descriptive parts and expansion in analytical interpretation.

The discussion is solid in structure (organized by hypotheses), but it tends to be defensive rather than analytical, especially regarding H1 and H4.

For H1 (non-significant AWS → burnout), you essentially argue that the literature is still correct. This weakens your contribution. Instead, you should explore why your data diverge: Sample characteristics (young, early-career);  Measurement issues (AWS reliability only marginal: α = 0.6767);  Possible suppression effects in SEM.

For H4, the explanation remains too generic (“complexity of the relationship”). This needs to be sharpened. The absence of mediation could be theoretically meaningful, e.g., engagement may act more proximally than structural work conditions.

The strongest part of the discussion is the emphasis on job resources and engagement, but even here the argument could be more concise.

This section about limitations is well-developed and more mature than the discussion itself. H however, some parts read more like policy discourse than scientific limitation analysis. You can shorten the prescriptive tone and keep the focus on methodological implications.

The English is generally good, but there are recurring issues: Occasional grammatical errors (“Studies shows…”); Repetition of sentence structures (“This suggests that…” appears very frequently); Slightly inflated phrasing in places (e.g., “highly significant,” “strongly reinforces”) without nuance. I think the text would benefit from tightening and variation in syntax. You should also address the excessive fragmentation of the text, particularly in the latter part of the Introduction (pp. 2–3) and again in the early Discussion section, where multiple paragraphs consist of only one or two sentences; this creates a fragmented reading experience and weakens the argumentative coherence, so I recommend consolidating these very short paragraphs into more developed units that better integrate related ideas and sustain a continuous academic narrative.

I will  summarize here  the essential revisions:

-Clarify the analytical model (mediation vs moderation) consistently.

-Strengthen SEM reporting (fit indices, assumptions, interpretation).

-Refocus the introduction toward the research problem.

-Reduce descriptive overload in Results and enhance analytical depth.

-Interpret non-significant findings more critically, not defensively.

-Correct methodological inconsistencies (no qualitative component).

-Refine language for conciseness and precision.

Author Response

We would like to thank Reviewer 2 for the thorough and detailed review of our manuscript. The comments were highly constructive and led to substantial improvements across all sections of the paper. We address each point below.

Comment 1: The title is adequate but slightly generic. It could be sharpened to reflect the non-confirmation of mediation.

We agree with this observation. The title was revised to better reflect the study's most relevant finding, as follows: "Burnout among psychologists: work engagement as a direct predictor and the non-mediating role of areas of worklife."

Comment 2: The abstract suffers from three problems: it repeats obvious statements, lacks precision regarding the analytical strategy (SEM is only implicitly mentioned), and the phrase "quantitative and qualitative elements" is misleading since no qualitative data are presented.

All three issues were addressed. The phrase "quantitative and qualitative elements" was removed and replaced with "cross-sectional quantitative design." Structural Equation Modeling (SEM) is now explicitly mentioned as the analytical strategy used to test the hypothesized relationships. The vague closing sentence was removed and replaced with a more precise conclusion summarizing the complexity of the findings and their directions for future research.

Comment 3: The introduction is overextended and insufficiently focused on the research problem. The transition from philosophical framing to empirical constructs is abrupt. The literature review becomes list-like in the middle section. The research gap is not sharply articulated.

The Introduction was substantially revised. The macro-social theoretical framework (Dejours, Han) was preserved but more explicitly connected to the operational variables of the study. A transitional sentence was added to bridge the philosophical framing and the empirical constructs, signaling how the societal conditions described relate directly to burnout, areas of worklife, and engagement. The middle section was rewritten as critical synthesis rather than sequential description, with an explicit connecting sentence showing how the three constructs interact. The research gap was reformulated to specify what remains theoretically unresolved: whether areas of worklife mediate the relationship between engagement and burnout, and whether engagement is negatively associated with burnout in this population.

Comment 4: There is a conceptual inconsistency between mediation (H4) and moderation (Results section). This must be corrected consistently across Methods, Results, Figure 1, and Discussion.

We thank the reviewer for identifying this inconsistency. Upon review, we confirmed that the analytical model tested was indeed a mediation model — areas of worklife (AWS) were included as a mediating variable between work engagement (UWES) and burnout (BAT). The term "moderator" that appeared in the Results section was an error introduced during manuscript preparation. All occurrences have been corrected to "mediator" and "mediation" consistently across Methods, Results, Figure 1, and Discussion.

Comment 5: The methods section describes the design as "quantitative with qualitative elements," which is inaccurate. The sampling strategy is under-problematized. The justification for n=180 is scientifically weak. The SEM description is too brief — key elements are missing. The AI disclosure disrupts the flow.

Each of these points was addressed. The description "qualitative elements" was removed — the study is described exclusively as cross-sectional and quantitative. The sampling strategy is now explicitly identified as convenience-based, with a discussion of its implications for external validity, particularly given the predominance of young, early-career participants. The justification for n=180 was complemented with a statement of statistical adequacy for SEM, citing Westland (2010), who demonstrates that the ratio of indicators to latent variables in the present model substantially exceeds recommended minimum thresholds. The SEM description was expanded to include the number of latent variables, estimation method (maximum likelihood with 1,000 iterations), the mediating role of AWS, and the significance level adopted. Regarding the AI disclosure, it was already positioned at the end of the Methods section in the original manuscript and was maintained in that position, which aligns with the reviewer's suggestion.

Regarding the missing fit indices (CFI, TLI, RMSEA, SRMR), we acknowledge this as a limitation. These indices were not calculated in the original analysis, which was conducted as part of a master's dissertation with a primary descriptive and inferential focus. We recognize this as a methodological limitation and have explicitly acknowledged it in both the Methods and Limitations sections. We recommend that future replications of this model report complementary fit indices to allow for a more comprehensive evaluation of model adequacy.

Comment 6: The Results section is overly descriptive and redundant. Key SEM outputs are underreported. The R² = 96.29% is suspiciously high and not critically discussed. The mediation test lacks a clear statistical framework.

The Results section was revised to reduce descriptive overload. Tables were simplified to report median, mean, and standard deviation only, removing minimum, maximum, and quartile values that were not analytically central. The SEM results were expanded to report direct, indirect, and total effects with their respective path coefficients and p-values. The R² = 96.29% is now critically discussed, acknowledging that such a high value may reflect conceptual overlap between constructs or shared method variance given the simultaneous self-report data collection. The mediation test is described with explicit reference to the indirect effect analysis, including the effect size and significance level.

Regarding the use of bootstrapping for the mediation test, we acknowledge that this procedure was not applied in the original analysis. The mediation was assessed through the indirect effect within the SEM framework using maximum likelihood estimation. We recognize bootstrapping as a more robust approach for mediation testing and recommend it for future studies replicating this model.

Comment 7: The discussion tends to be defensive rather than analytical, especially regarding H1 and H4.

The Discussion was substantially revised. For H1, instead of simply noting that the literature supports the relevance of areas of worklife, we now offer three specific analytical explanations for the divergence between our findings and the literature: the predominance of young, early-career participants who may not yet have accumulated sufficient exposure to worklife incongruences; the marginal reliability of the AWS (α = 0.6767), which may have attenuated the detection of real associations; and possible suppression effects within the SEM, where the strong UWES-BAT relationship may have absorbed much of the explained variance. For H4, the absence of mediation is now interpreted as a theoretically meaningful finding rather than a limitation, suggesting that engagement may operate more proximally on burnout than structural work conditions, and that individual factors not included in the model may play a more central mediating role.

Comment 8: The limitations section reads more like policy discourse than scientific limitation analysis.

The Limitations section was revised to reduce prescriptive tone and refocus on methodological implications. Statements directed at policymakers and organizations were removed, and each limitation is now discussed in terms of its implications for the validity and interpretation of the findings and for future research design.

Comment 9: The English has recurring issues — grammatical errors, repetition of sentence structures, and inflated phrasing. The text would benefit from tightening and variation in syntax. Short paragraphs throughout the Introduction and Discussion should be consolidated.

The manuscript was revised throughout for grammatical accuracy, syntactic variation, and conciseness. Recurring structures such as "This suggests that..." were replaced with varied formulations. Inflated expressions such as "highly significant" and "strongly reinforces" were replaced with more precise language. Short isolated paragraphs in the Introduction and Discussion were consolidated into more developed argumentative units to improve coherence and academic narrative flow.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

After this revision, the manuscript has been strengthened and accomplished the merit to be consider for publication.

Author Response

We would like to sincerely thank Reviewer 1 for the positive evaluation of the revised manuscript and for the recognition that the revisions addressed the concerns raised in the first round of review. We are grateful for the constructive and thorough engagement throughout the review process, which contributed substantially to improving the quality of the work.

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the revised version of the manuscript and for the detailed point-by-point response to the review comments.

The manuscript has improved substantially compared to the previous version. Although the revised title is more informative than the original version, it still reads somewhat densely and mechanically due to the formulation “direct predictor” and “non-mediating role.” The title could benefit from slightly greater fluency and readability while still preserving the central contribution of the study.

Two possible alternatives could be: “Burnout among psychologists: the role of work engagement and areas of worklife” or “Burnout among psychologists: direct effects of work engagement and the absence of mediation by areas of worklife”

I also appreciate the consistent correction of the mediation/moderation inconsistency throughout the manuscript, as well as the clearer presentation of the SEM framework and hypotheses.

The Results section is now considerably more concise and analytically oriented. The reduction of descriptive overload improved readability, while the interpretation of the SEM findings is more balanced and methodologically reflexive. The discussion of the unusually high R² value and the acknowledgment of potential construct overlap/common method variance represent important improvements.

The Discussion section is also substantially stronger. The interpretation of the non-significant findings is now more analytical and theoretically meaningful rather than defensive, particularly regarding H1 and H4. The inclusion of possible suppression effects, measurement limitations, and sample characteristics significantly strengthened the interpretative depth of the paper.

The Limitations section has likewise improved and is now more appropriately centered on methodological implications and future research directions.

At the same time, a few methodological issues remain unresolved, although the authors now acknowledge them transparently in the manuscript:

-Standard SEM fit indices (CFI, TLI, RMSEA, SRMR) are still absent. While this limitation is explicitly recognized, it nevertheless restricts a fuller evaluation of model adequacy.

-The mediation analysis was not conducted using bootstrapping procedures, which should be acknowledged as limiting the robustness of mediation inference.

-Although substantially improved stylistically, a few sections of the manuscript could still benefit from minor tightening for conciseness and fluency.

Overall, however, the revised version represents a clear and meaningful improvement, and the authors responded constructively and professionally to the review process.

 

Author Response

We would like to thank Reviewer 2 for the careful and constructive re-evaluation of the revised manuscript. We are pleased that the revisions were well received and appreciate the continued engagement with our work. We address the remaining points below.

Comment 1: The revised title is more informative but still reads somewhat densely. Two alternative formulations were suggested.

We thank the reviewer for the suggested alternatives. We adopted the second option, which we consider more precise and informative: "Burnout among psychologists: direct effects of work engagement and the absence of mediation by areas of worklife." This formulation preserves the central contribution of the study while improving readability.

Comment 2: Standard SEM fit indices (CFI, TLI, RMSEA, SRMR) are still absent. While this limitation is explicitly recognized, it nevertheless restricts a fuller evaluation of model adequacy.

We acknowledge this limitation. As stated in the manuscript, the complementary SEM fit indices were not calculated in the original analysis, which was conducted as part of a master's dissertation. The raw data are no longer available for reanalysis, which precludes the retrospective calculation of these indices. This limitation is explicitly acknowledged in both the Methods and Limitations sections, and future studies replicating this model are encouraged to report complementary fit indices to allow for a more comprehensive evaluation of model adequacy.

Comment 3: The mediation analysis was not conducted using bootstrapping procedures, which should be acknowledged as limiting the robustness of mediation inference.

We agree with this observation. A paragraph was added to the Limitations section explicitly acknowledging that bootstrapping procedures were not employed in the mediation analysis. The paragraph recognizes that maximum likelihood estimation, while valid, provides less precise confidence intervals for mediation inference compared to bootstrapping, particularly in the presence of non-normal data distributions. 

Comment 4: A few sections of the manuscript could still benefit from minor tightening for conciseness and fluency.

The manuscript was reviewed for fluency and conciseness. The following adjustments were made: the inaccurate reference to "healthcare settings" in the Introduction was corrected to accurately reflect the original study context (Portuguese hotel employees); the heading of section 3.1.6 was updated to better reflect the analytical focus of the section; the column header "AVG" in Table 3 was replaced with "Mean" for consistency across all tables; the opening sentence of the Discussion was revised to differentiate it from the Conclusion, avoiding repetition; and minor punctuation adjustments were made throughout to improve readability.

We believe the revised manuscript now addresses all outstanding concerns and represents a substantially improved contribution to the literature on occupational health among psychology professionals. We thank the reviewer once more for the rigorous and constructive engagement throughout the review process.

Author Response File: Author Response.pdf