The Effect of Safety Attitudes on Coal Miners’ Human Errors: A Moderated Mediation Model

Round 1

Reviewer 1 Report

Dear authors,

Congratulations for your interesting paper. However, and although its interest some improvements are needed, namely:

- What type of sampling did you used? Please explain.

- Discussion section needs to be completed and improved. Authors should discuss their results in the light of the literature review developed. They should compare their results with those  obtained in other studies. 

Author Response

Thank you for your comments concerning our manuscript entitled “The Effect of Safety Attitudes on Coal Miners’ Human Errors: A Moderated Mediation Model.” (Manuscript ID：sustainability-1852651). Those comments are all valuable and helpful for revising and improving our paper, as well as the important guiding significance to our research. We have studied comments carefully and have made corrections which we hope meet with approval. Revised portions are marked in yellow in the manuscript. The main corrections in the paper and the responses to the reviewer’s comments are as follows:

 

Point 1: What type of sampling did you used? Please explain.

Response 1: Thank you for your comments. The article sample was randomly sampled, thus ensuring sample representativeness and reducing sampling error.

Point 2: Discussion section needs to be completed and improved. Authors should discuss their results in the light of the literature review developed. They should compare their results with those obtained in other studies.

Response 2: Thank you for your comments. We have revised and improved the discussion section of the article. In addition, the following is a quick summary of what we have added.

 

First, the results of the article are summarized by a comparative analysis with the literature. The contents are as follows: The findings show that safety attitudes can reduce human error; situational awareness partially mediates the effect of safety attitudes on human errors. This result is supported by the literature, which shows that safety attitudes can significantly reduce human error. In addition, task complexity positively moderated the relationship between safety attitudes and situational awareness and positively moderated the indirect effect of safety attitudes on human errors through situational awareness. Previous literature on task complexity remains controversial, and this paper shows the positive moderating effect of task complexity through an empirical study, which is the same as Zhang's findings.

Second, the article is deeply discussed through three aspects: theoretical significance, practical significance, and research deficiencies. Specific revisions have been marked in yellow in the manuscript.

 

 

Reviewer 2 Report

My comments are provided in the attached file.

The research study is overall good in terms of quality, although the application of theory tends to be a bit tortuous in attempting to explain the hypotheses, such as the use of trait-activation theory.

The major issue with the manuscript that must be dealt with is that the manuscript refers to study results and especially conclusions through the lens of "normal operations" and "emergency operations."  These were not study variables and should be completely eliminated from the text (e.g., abstract, conclusions, etc.) even though the authors could discuss the implications of their findings on emergency versus normal operations in the discussion section of the manuscript.  However, the authors portray these "dichotomy" between normal and emergency states as if they researched the topic in distinct population groups when they had not.

Also a more thorough discussion of what is meant by smart mining operations would be beneficial to the reader.

 

 

 

Comments for author File: Comments.pdf

Author Response

Thank you for your comments concerning our manuscript entitled “The Effect of Safety Attitudes on Coal Miners’ Human Errors: A Moderated Mediation Model.” (Manuscript ID：sustainability-1852651). Those comments are all valuable and helpful for revising and improving our paper, as well as the important guiding significance to our research. We have studied comments carefully and have made corrections which we hope meet with approval. Revised portions are marked in red in the manuscript. 

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This paper establishes a hypothetical model of safety attitudes on human errors under routine and emergency conditions based on the attitude-behavior model. And combined with trait activation theory, we introduce moderating variables of task complexity.

 

 

1. It is recommended to strengthen the significance of the research in the introduction, e.g., what kind of complex problems are oriented.

2. Try to explain the purpose of conducting two surveys, it is easier and faster to complete the survey at once

3. Explain whether the sample size meets the requirements.

4. Validating factor analysis can also be said to validate the structural validity of the study. Some models in the subsequent analysis did not apply all potential variables, however, all variables were considered in the validating factor analysis in the text, whether another validating factor analysis is needed

5. Exploratory factor analysis is also necessary to test the scale reliability and validity and is recommended to be supplemented with exploratory factor analysis.

6. Why is the SEM model not used for the study? The differences and advantages and disadvantages of the methods used in the paper relative to SEM.

7. The results of the study are consistent with everyday understanding, the distinction and addition of the article relative to existing studies, and the suggestion to strengthen the research value of this article?

Reviewer 2 Report

Dear Authors, 

I suggest some serious corrections to be applied to the article at this stage of reviewing:

1. Improving Introduction Section, and moving the Fifure to the Methods. Also the introduction should observe the modern trends in State-of-Art of the problem. 

2. The methods used should be described more precisely. 

3. The Results must have a generalization of the fidings.

4. Conclusion section must be more comprehensive and summarize methodological results as well.

Good luck!

 

Reviewer 3 Report

- According to Fig1 this work much easier can be done through the structural equation modeling. Just using the correlations and known metrics such as SD never ever can be considered as novelty. I think the biggest potential drawback in this paper is that you wasted a lot of time on very obvious statistical method which scientifically doesn’t show any novelty, progress or any development. There is no information on acquired data!!! which type of data? categorical! Numerical! ordinal! nominal! Time dependent! Discrete! Continues! from questionnaire form? If yes how they have been analyzed? How the problem relating to dishonest answers, unanswered questions, differences in understanding and interpretation, hardness in convey feelings and emotions, hidden agenda in respondents, lack of personalization, unconscientious responses and accessibility issues have been treated? How did you assure for sufficient data, because of needing a considerable amount of clean data to generate useful results? They can deliver mixed results if the data set if it is not large enough or is not clean or is too noisy. What will happen if the statistics use the wrong use case or different categories represented in a dataset? None of these concerns neither are presented nor discussed. 

- In the case of CR, it has been approved that composite indicators provide misleading messages about quality if poorly constructed or misinterpreted. It then led to simplistic policy as you exactly mentioned in your conclusion ‘We hope that our findings will con-422 tribute to the study of the relationship between employee safety attitudes and human errors in the context of smart mines.’, where you just totally used three times the ‘smart mines’ and much more wondering after doing the work and finishing you still use present tenses. However, I strongly believe that the CR can be misused because the construction process is not transparent and lacks sound statistical or conceptual principles. 

- Another concern is referring to sensitivity analysis. You have several attributes. They must and have to be weighted for modelling. The CR for selection of metrics and weights can be challenged by other stakeholders. Which attributes are analyzed? According to which criteria have been considered? Given results only uncovers the data and won't determine what attribute have the most influence. Definitely the results can adversely be impacted by extraneous variables. You must and have to sensitivity analysis when you have different involved parameters and thus model calibration. These are some references dealing on sensitivity using different approaches https://www.sciencedirect.com/topics/medicine-and-dentistry/sensitivity-analysis; https://iwaponline.com/jh/article/22/3/562/72506/Updating-the-neural-network-sediment-load-models; …

- I cannot see any structure of IMRAD in this paper. 

- Up to the end of paper, nothing concerning the applied variables, processing, screening, scalability, unification and … for comparison can be found.

- When test items meet the assumptions of the tau-equivalent model, alpha approaches a better estimate of reliability. In practice, Cronbach's alpha is a lower-bound estimate of reliability because heterogeneous test items would violate the assumptions of the tau-equivalent model.

- scores of Cronbach’s alpha that have a low number of items associated with them tend to have lower reliability, and sample size can also influence your results for better or worse. Your discussion?

- In the concept of statistical theory, it is widely accepted that a p-value, or statistical significance, does not measure the size of an effect or the importance of a result. By itself, a p-value does not provide a good measure of evidence regarding a model or hypothesis. P-value doesn’t have reasoning capability about the probabilities of the hypothesis. It is only used as a tool for deciding whether to reject the null hypothesis. P-value whether small or big does not have any importance of the result or measure the size of an effect. A small effect can produce a small value if the sample size is high and vice versa. Then for hypothesis you relied on P-value? Why?

- This paper doesn’t have any real discussion. 

- Very short and uninformative conclusion.

 

More than the described issues the following concerns also should be addressed:

1. Title: I suggest for short but concise without question mark like a statement. Something like ‘the effectiveness of safety attitudes in reducing the human errors …’. It is not specified which area you are going to work on. When I read the Abstract then I found that you aim to focus on coal mines. This issue must be characterized in title.

2. English of the work definitely must be revised by native expert or verified proofread system. There are several linguistic flaws, word repetitions, punctuation as well as long and vague statements which should be edited. Simply can be seen from L8 in Abstract. ‘Year by year’ can be replaced with ‘annually’, L12 starting with ‘And’, then  L13, …

3. You have done the job, so it is past, why use of present tenses. To be consistent, it is highly recommended to use third passive voices.

4. L14: I cannot follow what do you mean for ‘two-wave data’?

5. I like short Abstract, but it should be concise, self-informative and to the point. From the Abstract I as a reader couldn’t find what is the novelty and applied method. If you are relying on correlation, unfortunately this analysis doesn’t sound and meet the criteria for novelty and publication. In L19, it is claimed for theoretical implication in which scientifically cannot be accepted because the concept of correlation and its application for error analysis through different metrics is widely recognized and applied. You urgently should show what is new and what progress have been done in this field.

6. Keywords should be representative and available in both Abstract and context. 

7. From the whole introduction, I couldn’t find what is the main problem and what you are pursuing? What is the limitation of applied methods and models and which gap of them is going to be filled here? With what method? What motives for? What is the novelty of model? Significant of contribution?

8. The last paragraph of the introduction should be assigned to brief summary of applied method and bolded findings.

9. How the variable reduction for nonlinear analysis has been applied? What about the uncertainty of applied inputs? How the uncertainty involved in the datasets and achieved results have been evaluated? What about the reliability-based analysis? Look at https://link.springer.com/article/10.1007/s11053-022-10051-w; https://www.tandfonline.com/doi/abs/10.1080/09535311003661226; https://www.sciencedirect.com/science/article/abs/pii/0304380095001859; https://www.tandfonline.com/doi/full/10.1080/00045608.2012.700616; 

 

Reviewer 4 Report

Review of the Manuscript ID ijerph-1790934 for the International Journal of Environmental Research and Public Health.

Title: Can Individual's Safety Attitudes Effectively Reduce Human Errors? The Role of Situation Awareness and Task Complexity.

First, I would like to thank the authors and editors for giving me the  opportunity to review the research paper.

The authors propose a a hypothetical model of miners' safety attitudes towards human errors in various conditions. They based their approach on the attitude-behavior model along with trait activation theory. Their proposal is verified on data from 246 miners gathered in two waves of questionnaires.

The study is potentially interesting and of theoretical and practical importance for reducing human errors in miners.

The proposal along with the paper seem to be prepared decently from the theoretical point of view. Nevertheless, there is missing information about theoretical models developed for similar purposes. In this regard, the literature review should be significantly extended. The authors' proposal should also be better confronted with other approaches including those that would be added to the improved literature review.   

Regarding the methodological aspects, I have significant doubts if the confirmatory factor analysis is  applied correctly here. Confirmatory factor analysis is usually applied to verify existing factor model of some theoretical construct. In this paper, it is not clear what the main, top level construct is. The confirmatory factor analysis assumes orthogonality of their factors, but the further authors' analyses revealed that there are significant correlations between factors? What is presented here rather resembles more general Structural Equations modelling, however too few data is presented to fully evaluate the proposal.

Given the serious above doubts regarding the core methodological approach I do not recommend this paper to be published in a present form.

Some exemplary minor issues:
"Task Comlexity" -> "Task Complexity", Figure 1., page 2
"seaboch" -> "Seaboch", page 5, line 232
"shakerian" -> "Shakerian", page 6, line 241
"... as "_Have you ..." unnecessary space, page 6, line 242
...