Review Reports - Cloud Noticing and Forest-Therapy Elements in a Coastal Resort: An Exploratory Randomised Field Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall Recommendation: Major Revision

This is a clearly written manuscript that fits well within Tourism and Hospitality. Its main strength is that it treats nature based wellness programming as a matter of hospitality experience design rather than as a medical or therapeutic intervention. This is clear in the introduction, where the authors argue that scenery becomes a service resource only when guests are given “a comprehensible and safe way to engage” with it through staff practice, timing, and guest attention . The manuscript also makes a useful contribution by comparing two different forms of directed attention: sky-focused cloud observation and forest-based multisensory engagement. The study design is interesting. The authors report a randomized parent experiment with 120 adult resort guests allocated to cloud observation, forest-therapy elements, active control, or ordinary stay, while the present article analyzes 84 complete paired cases from the three organized activity arms . The active-control condition is a particular strength because it helps distinguish nature-directed content from the general effects of facilitator contact, quiet time, group participation, and an organized session.

However, I recommend major revision before publication. The manuscript needs stronger transparency about its relationship to the previously published article, clearer framing as an exploratory randomized field study rather than a confirmatory trial, more detail and caution around adapted or study-specific measures, and stronger statistical reporting. The exclusion of the standard-stay group is reasonable, but the implications of that decision should be explained more directly. Finally, the physiological findings should be further downplayed because biometrics were collected as basic field observations, not as clinical or biological endpoints.

Major Comments

1. The relationship to the earlier publication needs to be made more clear

The manuscript explains that it comes from the same parent experiment as an earlier publication. This is stated in the introduction, where the authors write that the earlier article evaluated a “22-item service-performance module” covering experience quality, satisfaction, perceived innovativeness, added value, and recommendation intention, while the present paper focuses on psychological, experiential, and physiological responses.

The authors repeat this distinction in the methods section, explaining that the earlier companion paper used the final service-performance module in the full four-condition sample, whereas the present article uses paired pre-session and post-session observations from the three organized activity arms . They also state again in the discussion that the previous report used 114 final assessments from all four conditions, while the present report uses 84 paired observations from the three facilitated conditions .

This transparency is good, but because both papers come from the same parent experiment, I think the distinction should be made even clearer. I recommend adding a table that directly compares:

the shared parent protocol;

the sample used in each paper;

the outcomes used in each paper;

the overlapping factual elements;

the non-overlapping analyses and tables;

why the standard-stay arm was used in the earlier paper but excluded here.

*Perhaps add this to the supplementary materials. This would reduce any possible concern about redundant publication or unnecessary splitting of the same dataset.

2. The study should be reframed as exploratory rather than confirmatory

The randomized design is a strength. The methods section explains that cards were mixed and selected until each of the four conditions contained 30 participants, producing equal 1:1:1:1 allocation in the parent study . The participant-flow section is also clear: 120 participants were randomized, and the present paired analysis includes 29 cloud-observation participants, 28 forest-therapy participants, and 27 active-control participants .

However, the authors also state that no formal prospective power calculation was performed and that neither the study nor the hypotheses were registered in advance . This matters because the manuscript includes four explicit hypotheses . The hypotheses are theoretically reasonable, but the study should not be framed as a fully confirmatory trial.

I recommend describing the manuscript as an *exploratory randomized field experiment or an estimation-focused companion analysis. The hypotheses can remain, but they should be described as theory-guided expectations rather than preregistered confirmatory hypotheses.

*This would make the claims more credible and better aligned with the actual study design.

3. The measurement strategy needs more caution and details

The manuscript uses several established measures, including STAI-S, PANAS, PSS, and Connectedness to Nature . This is a strength.

However, the manuscript also relies on adapted or newly created measures. The authors state that the subjective restoration module was a brief contextual module and “is not presented as equivalent to the complete Perceived Restorativeness Scale” . They also explain that the six experiential indicators—awe, mental spaciousness, perspective distancing, sensory immersion, bodily presence, and grounding—were created for this program and are not validated diagnostic or transportable instruments .

These measures are interesting, but the manuscript sometimes places substantial interpretive weight on them. For example, the results section states that cloud observation produced the highest ratings for awe, mental spaciousness, and perspective distancing, while forest-therapy elements produced the highest ratings for sensory immersion, bodily presence, and grounding . This is useful, but it should be framed as *descriptive rather than as evidence of validated experiential “profiles” or “signatures.”

I recommend that the authors add the exact item wording, preferably in an appendix. They should also explain why single-item indicators were used, whether any reliability checks were possible, and how these measures should be interpreted. The results can still be valuable, but the language should remain cautious.

4. The statistical reporting should be strengthened

The analysis needs more explanation. Since the study uses a randomized repeated-measures design, the authors should consider whether baseline-adjusted models, such as ANCOVA or a linear mixed model, would be more appropriate. The current difference-in-change approach is understandable, but baseline-adjusted estimates could improve interpretability, especially with small group sizes.

The authors should also clarify the following in Chapter 3.6:

How Hedges g was calculated;

How confidence intervals were calculated;

Whether statistical assumptions were checked;

Whether multiple comparisons were adjusted or intentionally left unadjusted;

Whether results are robust to alternative analytic approaches;

How complete-case analysis may affect the findings.

This is especially important because the paper presents several outcomes and contrasts. The authors should help readers understand the overall pattern without overstating individual comparisons.

5. The exclusion of the standard group is reasonable but should be explained more strongly

The authors explain that standard-stay participants did not receive an organized session and therefore did not have an equivalent pre-post program interval . This is a reasonable basis for excluding them from the present paired session-level analysis.

However, because the parent experiment had four randomized arms, readers may still wonder why the standard-stay arm disappears from the main analysis. *The authors should make this point earlier and more directly. The key clarification is that the present manuscript is not testing whether cloud observation or forest therapy outperforms an ordinary resort stay. It is testing whether nature-directed sessions differ from a structurally matched active-control session.

The authors acknowledge this limitation in chapter 5.4, stating that the standard-stay arm could not contribute to the session-level paired analysis and that the study does not estimate ordinary change during the stay . This statement should be moved earlier or repeated more prominently in the methods and discussion.

6. The physiological findings should be downplayed a bit more

The authors are already cautious about the physiological measures. They state that biometrics were included as pragmatic indicators of short-term activation in the field, not as clinical endpoints or tests of a biological mechanism . They also note that cardiovascular changes were small and interpreted only as supportive field observations .This caution is appropriate, but I think the manuscript should go even further.

Pulse and blood pressure in a resort field setting can be influenced by posture, timing, weather, expectation, measurement error, sitting quietly, and regression to the mean. Because there was no blinding and no clinical physiological protocol, these findings should not be presented as *evidence of biological restoration.In the results, the authors write that pulse and blood-pressure measures decreased in each condition, with somewhat larger reductions in the two nature-focused arms . *This should remain descriptive only. Perhaps using the phrase “physiological field descriptors” rather than “physiological indicators” or “physiological outcomes.”

7. Minor Issues

I suggest shortening the title, as the current version is somewhat long.

I also recommend using one consistent term for the sky-based activity. The manuscript alternates between “cloud observation,” “cloud noticing,” and “sky-focused observation.” The authors should choose one term and use it consistently throughout the paper.

The supplementary file confirms that the FoRRestMind Institute Ethics Committee approved the parent protocol and reviewed the protocol, participant information, consent form, questionnaires, anonymization plan, and outdoor-safety procedures. This is helpful. However, because some authors are affiliated with the FoRRestMind Institute, the manuscript should still clarify whether the ethics committee was independent from the investigators and whether any conflict-management process was used.

Author Response

Dear Academic Reviewers,

Thank you for the careful and constructive evaluations. We have undertaken a comprehensive major revision. All substantive changes in the manuscript are shown in blue. The revision strengthens transparency concerning the companion publication, reframes the work as exploratory, expands the theoretical foundation without proposing a cloud typology, replaces the principal change-score presentation with linear mixed-effects group-by-time models, adds multiplicity correction, reports the item wording and anchors, downplays the physiological observations, and adds two figures and two appendix tables.

Comment R1.1: Clarify the relationship to the earlier publication and directly compare the shared protocol, samples, outcomes, overlap, non-overlap, and treatment of the standard-stay arm.

Response R1.1: We agree. The Introduction and Section 3.1 now state the analytical separation more directly. New Appendix Table A1 provides the requested side-by-side comparison. It identifies the shared four-arm parent protocol, distinguishes the earlier sample of 114 final service assessments from the present 84 linked pre-post observations, lists the separate outcome families and analytical purposes, specifies the unavoidable factual overlap, and explains why standard stay was suitable for the service-evaluation paper but not for the session-level paired analysis. No service-performance endpoint, hypothesis, principal table, model, or conclusion is reused.

Comment R1.2: Reframe the study as exploratory rather than confirmatory because there was no prospective power calculation or preregistration.

Response R1.2: The title, Abstract, Introduction, theoretical section, Statistical Analysis, Discussion, Limitations, and Conclusions now consistently describe an exploratory randomised field study and estimation-focused companion analysis. Section 2.4 has been renamed 'Theory-Guided Expectations' and explicitly states that H1-H4 are not preregistered confirmatory hypotheses. Interpretation prioritises effect sizes, confidence intervals, multiplicity-adjusted p values, and consistency across outcomes.

Comment R1.3: Provide exact wording for adapted/study-specific measures, justify single items, discuss reliability, and avoid treating them as validated profiles or signatures.

Response R1.3: Section 3.4 now reports the common 1-5 response format, the reason for using concise single items in an immediate field protocol, and why internal-consistency reliability is neither available nor conceptually appropriate for six distinct indicators. Appendix Table A2 gives the common prompt, full English wording, and anchors for awe, mental spaciousness, perspective distancing, sensory immersion, bodily presence, and grounding. Throughout the manuscript, 'profiles' and 'signatures' have been replaced by 'descriptive item-level patterns' or equivalent cautious wording. The Results, Discussion, Limitations, and Conclusions now state that these items do not establish latent mechanisms, psychometric profiles, stable market segments, or transportable scales.

Comment R1.4: Strengthen the statistical analysis with a repeated-measures framework, calculation details, assumptions, multiplicity control, robustness, and discussion of complete-case analysis.

Response R1.4: Section 3.6 has been fully rewritten. Linear mixed-effects models now form the principal framework for the four primary outcomes, with fixed effects for group, time, and group-by-time interaction and a random participant intercept. Table 4 now reports B, SE, 95% t-based CI, t(81), raw p, Holm-adjusted p, and Hedges g for all planned contrasts. The manuscript specifies the small-sample J correction for g and its confidence interval, identifies residual Q-Q and residual-versus-fitted plots as diagnostic checks, and compares the model-based contrasts with the original difference-in-change estimates as a design-based robustness check. Holm correction is applied to the three planned contrasts within each primary outcome. The complete-case limitation is now explicit: missing values were not imputed, withdrawal reasons were unavailable, missing completely at random cannot be established, and results are conditional on completers. We did not claim a richly adjusted covariate analysis because the available complete-case sample would not support it responsibly.

Comment R1.5: Explain the exclusion of the standard-stay arm earlier and more strongly.

Response R1.5: The Abstract, Introduction, Section 3.1, the participant-flow note, Figure 1, Appendix Table A1, Discussion, and Limitations now make the comparator logic explicit. The present question is not whether an organised nature session outperforms an ordinary resort stay; it is whether the nature-directed content differs from a structurally matched organised active-control session. Standard stay had no additional study session and therefore no matched session-level measurement interval.

Comment R1.6: Downplay pulse and blood pressure and use the term physiological field descriptors.

Response R1.6: We have adopted the reviewer's suggested terminology throughout. Pulse and blood pressure are now 'physiological field descriptors,' not indicators, outcomes, biomarkers, or evidence of biological restoration. Section 3.5 explicitly lists posture, timing, ambient conditions, expectation, quiet sitting, device error, and regression to the mean as alternative explanations. Section 4.4 and Table 7 are descriptive only, and the Discussion and Conclusions make no biological or clinical claim.

Comment R1.7: Shorten the title, use one term for the sky-based activity, and clarify ethics-committee independence and conflict management.

Response R1.7: The title has been shortened to 'Cloud Noticing and Forest-Therapy Elements in a Coastal Resort: An Exploratory Randomised Field Study.' 'Cloud noticing' is now the consistent activity label. References to sky, horizon, openness, and atmospheric change are retained only as descriptions of the attentional field, not as alternative intervention names. Section 3.7, the Institutional Review Board Statement, and the Conflict of Interest statement now explain that the ethics review and decision were conducted separately from the research team; affiliated authors did not participate in deliberation, voting, or approval, and recusal constituted the conflict-management procedure.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear author(s),

This is an interesting manuscript. However, I have several comments that may help improve the clarity, theoretical foundation, and presentation of the manuscript:

Abstract: The first and second sentences should be revised to provide a clearer and stronger opening to the study.
Theoretical foundation: The theory underlying this study should be explained and emphasized more clearly. Please strengthen the discussion of how the selected theory supports the research framework and findings.
Cloud characteristics and mental spaciousness: Please provide justification for the relationship between cloud characteristics and the concepts of mental spaciousness and/or perspective distancing. In particular, are there specific types, forms, or characteristics of clouds that may contribute to these psychological experiences? This relationship should be supported with relevant literature.
Methodology: Please include a research design flowchart in the methodology section. A visual representation of the research procedures and flow would help readers better understand the overall research design and methodological process.
Results and findings: Since the analysis mainly involves t-tests, please consider presenting some of the key results using appropriate figures or visualisations. This would make the findings more visually appealing and easier for readers to interpret.

Author Response

Dear Academic Reviewers,

Comment R2.1: Revise the first and second sentences of the Abstract to provide a clearer and stronger opening.

Response R2.1: The Abstract now opens with the practical knowledge gap and the two programme formats: 'Wellness resorts often market natural scenery without knowing whether a structured way of attending to it changes guests' immediate experience. We examined cloud noticing and forest-therapy elements as two low-infrastructure forms of nature-directed hospitality programming.' The remainder of the Abstract has been updated to reflect the exploratory framing and model-based analysis.

Comment R2.2: Strengthen and clarify the theoretical foundation.

Response R2.2: Section 2 has been substantially expanded and retitled. The revision now integrates Attention Restoration Theory, soft fascination, Stress Reduction Theory, noticing-nature research, awe theory, perceived vastness and accommodation, expanded time perception, the small-self/decentring literature, and the hospitality concept of attentional choreography. The framework clearly separates the shared restorative rationale from the two attentional orientations: distal, open, dynamic visual attention during cloud noticing and proximal, multisensory, embodied attention during forest-therapy elements.

Comment R2.3: Justify the relation between cloud characteristics and mental spaciousness or perspective distancing, including whether specific cloud types or forms matter.

Response R2.3: We addressed the conceptual question while deliberately avoiding an unsupported cloud typology. The revised Section 2.2 states that the relevant explanatory features are an open visual field, distant focus, visual depth, horizon, low-demand slow movement, continuous transformation, changing edges, light, and absence of a classification task. These features are linked to soft fascination, widened attentional scope, awe, accommodation, expanded time perception, and temporary decentring. New references include Keltner and Haidt (2003), Rudd et al. (2012), Piff et al. (2015), Lin et al. (2023), and Tasser et al. (2023). Crucially, the manuscript now states that cloud type, coverage, morphology, motion, and brightness were neither classified nor manipulated, so the results cannot identify a specific cloud form as causal.

Comment R2.4: Add a research-design flowchart to the Methods.

Response R2.4: New Figure 1 shows recruitment, screening, randomisation, allocation to all four parent-study arms, completion in each arm, exclusion of standard stay from the session-level interval, and formation of the 84-person complete-case analytical sample.

Comment R2.5: Present key results using appropriate figures or visualisations.

Response R2.5: New Figure 2 is a forest plot of Hedges g and 95% confidence intervals for cloud noticing versus active control and forest-therapy elements versus active control across the four primary psychological outcomes. Effects are oriented consistently so that positive values favour the nature-focused condition. This figure complements, rather than replaces, the complete model estimates in revised Table 4.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I apologize for taking so long for this final review. The authors have addressed my previous concerns satisfactorily. The revised manuscript is substantially improved in terms of methodological transparency, statistical analysis, positioning of the study as exploratory, reporting of study-specific measures, clarification of the companion publication, and interpretation of the physiological observations.

In particular, the revised repeated-measures analysis, multiplicity correction, clearer explanation of the comparator structure, and more cautious interpretation of the findings have strengthened the manuscript considerably. The limitations are now appropriately acknowledged, and the conclusions are consistent with the exploratory nature of the study.