AI-Driven Risk Management and Sustainable Decision-Making: Role of Perceived Environmental Responsibility

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I liked the topic and methodology used to conduct the research and test the hypothesis. However, the model components such as SDM1-3 need to be defined and explain why you construct the model in this way (reflective vs formative). After these explanations, the discussion and conclusion should be updated. Table 2 should be cited, also, Figure 1 needs to be presented in a way aligned with the text. 

Finally, I recommend that the result section be results and analysis. Also, the discussion section should be separated from the conclusion part. 

 

Comments on the Quality of English Language

It is good and needs just a minor review.

Author Response

Thank you for your valuable comments and suggestions.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The topic is interesting, but in order to be published, the authors must make a few improvements such as:

1.     In the Abstract, the authors must add a few information about the novelty of the article and the obtained results and the fulfillment or not of the research hypothesis.

2.     In the Introduction, in the final part, the authors must add a short description of each section the authors developed.

3.     At 5.1. section, is indicated for the authors to say only Implications and devide them into Theoretical Implications and Practical Implications, and these practical ones to be also devided into Implications for managers, for employees and for society, all written in italic to increase the interest for readers and to be perceived easily.

From these points of view it is proposed minor revision before the article publication.

Author Response

Thank you for your valuable comments and suggestions. 

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

IMPORTANT ISSUES

I'd like to start with the conclusions.

"This study showcases how an AI-driven risk management approach is central to
long-term thinking in the construction industry. "

it is not supported by the text

The tools of AI including predictive in- 605
sights and scenario modeling capabilities, provide the decision makers with accurate in
formation to make the right decisions that go in line with the sustainability objectives,
optimize resource allocation, and successfully mitigate risks.

This general statement is not supported by the text of the article

 

"The results show the importance of AI-driven risk management as one of the most effective tools in the hands of decision-makers in construction companies."

The other tools (not AI-driven) are not analysed (i.e. their effectiveness). The statement is not supported by the text of the article.

Furthermore, the identification of perceived environmental responsibility as a mediating factor also underlines the holistic connection
between AI-based risk management and sustainability outcomes.

The holist connection is not proved in the article. Only the existace of mediating factor is proved.

My conlusion is that conclusions if the article are not supported by the tesxt of the article.

The second issue: The article is read as a storytell. Without presenting the merit questions the respondents were asked, without presenting statistics of the answers, prooving the hypotheses looks unreliable.

The third improtant issue: In several statistic hypothesis testing I did, the p-value was compared with the assumed confidence level alpha (usually 0.05). If p>alpha e.g. null hipothesis was acepted, if not the alternative hypothesis was choosen as the valid one. Here - in the article - there are not alternative hypotheses (explicitly shown). It is not known wich are null, which are alternative. p-value = 0 every time. It is not known what is a meaning of comparison to alpha. Moreover the level of alpha is not known too.

The central figure 1. There are ortographic mistakes in the text presented in this figure. I can find many examples for other tools (than presented in this figure) that were used to build the models presented there. The figure is misleading.

Comments on the Quality of English Language

The English language of the text is good. Fig 1 suffer from mistakes.

Author Response

Thank you for your valuable comments and suggestions. 

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Thak you for considering the remarks

Author Response

Thank you for pointing out the areas that need further improvements:

Point 1: Are the research design, questions, hypotheses and methods clearly stated? Needs Improvement

Response 1: The research design has been revisited and improved. The research approach, survey technique, sampling frame and sampling technique, and data collection method are clearly stated for a better understanding for the readers. The research questions have been added in the introduction as per the objectives of the study. Section 2.6 presented hypothesis development which has been carefully examined and improved for better comprehension.  

Point 2: Are the arguments and discussion of findings coherent, balanced and compelling? Needs Improvement

Response 2: The discussion section has been revisited and further improvements have been made to enhance the clarity. The implications of the study are also carefully rechecked.

Point 3: For empirical research, are the results clearly presented? Needs Improvement

Response 3: The empirical results are presented in a structured way starting from CMB to the measurement model assessment (convergent validity and discriminant validity) and structural model assessment (hypothesis testing through path modeling). The interpretation of the results has been carefully stated for the better understanding of the readers.

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

The revisions have brought significant improvement of the clarity of the article. However, I have still some following doubts:

- there are 14 questions of the questionnaire mentioned in line 334, but Table 4 contains only 12 statements. Are the statements the questions of the questionnaire?

- if yes - please correct; if no - the questions of the questionnaire are really needed

- why the descripive (basic) statistics of the answers are not presented? (5 point scale answers are very informative if the histograms are presented)

- the meaning FL, alpha, AVE, CR, T-value, p-value, SRMR, Q2 are not explained

- there are a lot of conclusions based on the values of the abovementioned coefficients, but in fact, I don't know e.g. for Q2 calculated for SDM (there are probably 3 answers) how it was calculated (despite that Q2=0.33 is assessed as moderate; I find it low; similarily is with R2)

Author Response

Comment 1: there are 14 questions of the questionnaire mentioned in line 334, but Table 4 contains only 12 statements. Are the statements the questions of the questionnaire?

Response 1: 

Thank you for your comment

 

According to the criteria, the threshold values for all the items needed to be loaded at a value of at least 0.50 (Ali, Rasoolimanesh, et al., 2018; Dijkstra & Henseler, 2015). Two Items concerning low factor loadings (less than 0.50) were omitted.

However, the Appendix A: Questionnaire Survey is presenting all measurement items (14) at last.

Comment 2: why the descripive (basic) statistics of the answers are not presented? (5 point scale answers are very informative if the histograms are presented)

Response: 

Thank you for your valuable feedback. Histogram are generally not presented in empirical research due to few reasons. Many academic journals have strict word or page limits. Given this, researchers prioritized presenting essential descriptive statistics that efficiently communicate the central tendency and variability of the data.

Secondly, While histograms can be informative visualizations of data distribution, our primary focus in this research stage is to establish relationships between variables and assess the strength and direction of these effects. Due to this, we presented tables and figures showing the Std. Error, T-value, Beta, and P-value. to derive structural Equation Modeling (SEM) results.

Comment 3: the meaning FL, alpha, AVE, CR, T-value, p-value, SRMR, Q2 are not explained

Response: The meaning of FL, alpha, AVE, CR have been explained on page 11.

T-value, p-value, SRMR, and Q2 have been explained on page 13. 

Comment 4: for Q2 calculated for SDM (there are probably 3 answers) how it was calculated (despite that Q2=0.33 is assessed as moderate; I find it low; similarily is with R2)

Response: 

Several researchers suggest several thresholds for the assessment of predictive relevance and coefficient of variance. Henseler et al. (2009) recommends that the model is considered having predictive relevance when the Q2 value is greater than zero. Therefore, the higher the Q2 value, the higher the model‘s predictive relevance.

According to Hair et al. (2011) and Henseler et al. (2009), the value of R2 such as 0.25, 0.50, and 0.75 represent weak, moderate, and substantial level of variance respectively. Moreover, few researchers (Cohen, 1988; Chin, 1998a) suggest three ranges for structural model quality assessment such as the R2 value ranges from 0.26 to 0.67 is substantial, from 0.13 to 0.33 is moderate and 0.02 to 0.19 is regarded as weak variance level. However, Falk and Miller (1992) opine that the R2 value 0.10 (10 percent) is acceptable while the R2 value 0.015 (1.5 percent) is satisfactory.