Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Comments

General Comments
The manuscript presents an interesting topic that explores the effects of ESG-related advertising on consumer attitudes and intentions. The research question is relevant to the current discourse on sustainable marketing and corporate communication. The paper is generally well organized and shows effort in both theoretical development and empirical testing. However, several critical issues regarding construct validity, sampling, and methodological rigor need to be addressed before the manuscript can be considered for publication. My detailed comments are provided below.

 

1) Construct validity: mismatch between measurement items and conceptual definitions

1. The author uses ease of understanding (L279 Table 2, Information INF3: “This advertisement offers a convenient way to understand the company’s ESG goals.”) as an indicator for the information construct. However, the information scale typically emphasizes aspects such as quantity/adequacy, specificity, relevance, and timeliness of information.
2. The author interprets the semantic meaning of self-efficacy (L279 Table 2, Flow Experience FE4: “While watching this advertisement, I feel capable of addressing the ESG challenges raised by the company.”) as a reflective indicator of flow experience (immersion, concentration, loss of time awareness, attraction, and self-forgetfulness).

Such practices result in construct contamination and measurement misplacement. Measurement indicators are expected to share the same latent dimension and be consistently driven by it; mixing variations from different conceptual sources introduces systematic variance unrelated to the target construct, leading to:
① noticeably low factor loadings (INF3=0.53, FE4=0.66);
② diluted discriminant validity among adjacent constructs;
③ biased and unstable estimation of structural paths and mediation effects.
It is recommended to refine the semantic definitions or provide additional evidence of discriminant validity (e.g., HTMT) to strengthen measurement persuasiveness.

 

2) Sample selection bias and limited external validity
The study sample primarily consists of respondents aged 20–25 (L497: “The respondents were young (20–25)…”), situated within Taiwanese media and cultural contexts (L502: “…were highlighted by the media and cultural precincts of Taiwan…”), and pre-screened to include only users familiar with sustainability reports (L245: “The formalized questionnaire was conducted in a Survey Cake online format; the target consumers were users who had consumed corporate sustainability reports.”).

Although this purposive sampling enhances internal consistency, it limits external validity. The conclusions may not generalize to other age groups, regions, cultural contexts, or general consumers unfamiliar with sustainability reports. Moreover, this familiarity-based screening may influence cognitive processing and attitudinal responses toward ESG advertisements, introducing selection bias. The paper lacks evidence of multi-group invariance or hierarchical robustness tests, making it difficult to judge whether the findings hold across broader populations.

 

3) Single-brand stimulus causing stimulus-specific effects and internal confounding
The experimental material includes only one ESG image advertisement (L259: “We embedded an example of the ESG image ad of Chunghwa Telecom in the questionnaire”), which introduces the following problems:
① Limited external validity—using a single brand cannot represent the general case of ESG advertising, making generalization difficult;
② Potential internal confounding—brand equity, audience familiarity, and ad quality may simultaneously affect constructs such as “entertainment, credibility, and informativeness,” meaning that the observed mediating effects could actually be driven by brand effects.
It is recommended to include additional brands for robustness checks.

 

4) Insufficient theoretical discussion for unsupported hypotheses; lack of alternative explanations and sensitivity analyses
After adopting strict thresholds (p < 0.01, |t| ≥ 2.58, and 95% CI not crossing zero), several hypotheses were not supported. However, the paper provides insufficient theoretical reasoning for these nonsignificant results, lacks discussion of contextual mechanisms and boundary conditions, and omits alternative explanations and robustness or sensitivity tests. The discrepancies may arise from theoretical limitations, measurement or model specification issues, or the sample and stimulus specificity. Further explanation and discussion are recommended.

 

5) Model fit reporting and interpretation irregularities
Appendix B Table 2 (“Offending Estimates Test”) contains three issues:
① The thresholds and evaluation criteria were not stated in advance, with vague expressions such as “satisfy/approach,” reducing transparency and reproducibility.
② Misinterpretation of χ²: in SEM, χ² is almost always significant (p≈0) for large samples or moderately complex models; therefore, “p > 0.05” should not be treated as a hard criterion, and marking “satisfy” when p = 0.000 is logically inconsistent.
③ When fit indices diverge, the author does not specify the evaluation hierarchy or interpretive framework. The current reporting appears as a descriptive list lacking standardized justification and theoretical grounding.

 

6) Minor errors

1. L291–292: “CR<0.70, AVE>0.50; ranges CR=0.829–0.909…”—the correct standards are CR ≥ 0.70 and AVE ≥ 0.50. The text shows a contradictory statement.
2. Table 4: “0769**” should read “0.769**”; the decimal point is missing.
3. Appendix B Table 1 “Results of Normality Test Statistics”: the abbreviation “INT” should be “INF” (for Information).
4. Section 5.2.2 “Practice inspiration” (L468–483): the paragraph beginning “As they sustain this bridge over a reasonable period…” is repetitive and should be merged for better readability.
5. Section 2.5 “The Mediating Role of Attitude Toward Advertising” (L191–192): The paper presents hypothesis H5, but later in Section 4.2 “Discussion of Findings” (L371–373) it refers to H5a and H5b, causing inconsistency and confusion in hypothesis numbering.
6. Several grammatical errors are present throughout the manuscript.

 

7) Insufficient recent literature
Among the 70 cited references, only 13 are from the past five years. It is suggested to include more recent studies to better reflect current academic developments.

Comments for author File: Comments.pdf

Author Response

{More detail please see the attachment}
We would like to express our sincere gratitude to Reviewer for an exceptionally careful and constructive review. In response to your comments, we refined the measurement of informativeness and flow experience (including HTMT evidence), clarified the implications of our purposive sampling and single-brand stimulus in Section 5.3, expanded the discussion of partially supported and unsupported hypotheses, standardized model fit and offending estimate reporting, corrected the noted errors, and incorporated more recent literature. We are genuinely grateful for your detailed and insightful suggestions, which have substantially improved the rigor and clarity of this manuscript.

 

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This is a very interesting research and the topic represents an important step towards increasing the importance of the role of ESG reporting in the customer's informed decision to purchase a certain product in the context of sustainability.

All elements of data collection and analysis were applied in the paper when a survey questionnaire was used for data collection. Objectives and hypotheses are clearly defined, materials and methods are clearly described, and results are clearly presented.

I have only 2 minor suggestions:

Maybe consider reducing the number of hypotheses (especially hypothesis 2 and its subvariants)

Describe the term flow experience in a little more detail and support it with literature.

Author Response

More detail please see the attachment

Thank you sincerely for your thoughtful and encouraging review. It was a privilege to have you evaluate our work. Your comments not only affirmed the potential of the study but also guided meaningful refinements. In fact, your suggestion helped us land on a clearer, more compelling title—'Feel the Flow, See the Value’ An S–O–R Model of Consumer Responses to ESG Advertising. Acting on your two core points, we made the following revisions: (1) Hypotheses streamlining. We reduced the number of headline hypotheses by consolidating element-level paths into two core statements, while relocating the detailed element-specific contrasts to Appendix C (Table C1) for transparency. This preserves theoretical precision without overloading the main text.; (2) Flow experience—depth and support. We substantially expanded §2.2–§2.4 to define flow more precisely (conditions such as focused attention and challenge–skill balance) and to connect it explicitly to informativeness, entertainment, credibility, and message relevance, incorporating both classic and recent literature. We are grateful for your guidance; it materially strengthened both the clarity and the scholarly rigor of the manuscript.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Paper goal is interesting although the title is too much, and quite confusing.

About the title:

• I think that it is too much of strong concepts
• maybe refine it and leave all these combinations for the general scope of the research or to objectives
• formulate a clear and simple Title, maybe concentrating on the MODEL

An S–O–R Model Explaining Consumer Responses to ESG Advertising

Are all terms necessary in the title ?

 

LR is scarce in some aspects. Authors bring 1, or 2 sources to argue a complex concept. The same with hypothesis.

Some of the references are not highlighted in the text  subchapters 2.1 and 2.2.  And many more regarding the hypothesis.

 

Model proposition and hypothesis are ambitious and intriguing.

 

The hypothesis building and arguing are rather subjective as the authors try to prove a new perspective on the concepts relations. But arguing is subjective and lacks fundamental sources.

 

Statistical methodology is well-done.

 

Discussions are well stating the positive and the negative results, and connecting these to literature.

 

After Results and Discussions, it would be interesting to see / visualize how the Model transforms according to the supported and not supported hypothesis.

 

The list of references needs some adjustments/technical corrections: too many sources from information science domain … improve with Economics & Sustainability domains.

 

Congratulation for your work,

Author Response

More detail please the attachment!
Thank you very much for the careful review and for the creative push you gave us. Your suggestion to visualize how the model “morphs” after hypothesis testing genuinely changed how we communicate the results—what began as a tricky request led us to develop a compact, reader-friendly total-effects figure that clarifies the story briefly. We appreciate your time and the constructive spirit of your comments.

Author Response File: Author Response.docx