Modeling Rollover Crash Risks: The Influence of Road Infrastructure and Traffic Stream Characteristics
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
The manuscript entitled "Modeling Rollover Crash Risks: The Influence of Road Infrastructure and Traffic Stream Characteristics" examines a significant concern pertaining to transportation safety. By introducing a Crash Index and investigating inadequately explored variables such as ramp density, U-turn maneuvers, and violations of speed regulations, the research addresses substantial gaps in the existing literature. The employed methodology is thorough, encompassing a comparison of multiple models and the application of rigorous statistical techniques. Nonetheless, certain elements of data interpretation, the breadth of analysis, and the manner of presentation could be refined to augment the overall clarity and efficacy of the manuscript. Here are a few suggestions/shortcomings for improvement of this manuscript.
1. The abstract needs to be improved and re-structured especially the problem statement/statement about the current research gaps being addressed in the paper and the conclusion part. See line 20, “This paper seeks to fill research gaps.” And line 29-30, “which showed how to reduce the number and severity of rollover accidents.”
2. Introduction section is too short, it can be enhanced further with addition of relevant facts and figures about accidents in Iran and why only one Province is selected in this study. Provide some evidence from previous research studies on the topic in particular to Iran.
3. Research objectives are not clearly defined, even after reading the literature review section and introduction as well. Define the research objectives clearly so that the readers can understand.
4. Line 57-62, The authors have mentioned six sections but in actual there are only five sections in paper. Also, “ It is worth mentioning that Iran crash data employed in this paper.” not appropriate at this point.
5. The literature review section should be clearer and properly organized. The authors have made a good attempt but it is unclear what they are trying to convey to the readers. Presently, it is difficult to understand what the authors are conveying, the writing style needs to be more academic and thought provoking. (For example, line 65-75 can be rewritten as, “A study on rollover crashes on mountainous roads indicated that adverse weather and topography can pose risks. The study employed variables including speed limit, road median conditions, road surface type, season in which crash occured, day type, time of day, weather conditions, and driver characteristics like speed and age group. Three modeling approaches were evaluated: conventional logit, non-correlated random parameter logit, and correlated random parameter logit. Findings revealed the correlated random parameter logit model's superiority over the others. Furthermore, an investigation was conducted into the interrelations among the variables. Nonetheless, all variables were incorporated into the analysis. Additionally, variables such as roadside slope and other road characteristics were omitted from this analysis.”)
6. Line 171-172, literature review section’s conclusion is not appropriate, needs further improvement read some academic articles to get a better idea.
7. Figure no 2, the information being shown is not clear in the figure also typographical mistakes in figure. It is suggested to include graphical charts to show the intensity or number of accidents on the selected road segments for the particular year.
8. Line 200-203, “In addition, coefficient A indicates the performance coefficient of the desired segment, which is equal to 8 for the highway, 6 for the main arterial, 4 for the intra-arterial road, and 2 for the secondary roads.” How these values are defined give references.
9. Line 233-235 Provide appropriate reference and reason as well.
10. Line 253-254 and the relevant text in table 2 are not similar.
11. What are the reasons apart from the one mentioned in manuscript of forming the compound variables (references from previous research works).
12. Table 9 and 10 are too dense with large amount of numerical data that are difficult to interpret. It is suggested to summarize the key findings in the text and add visual aids to further improve the understanding of readers. Also include large tables in appendix along the main text.
13. The authors have used 10 models but no details about these models have been included in the manuscript. It is suggested to include the literature about these models and also mention the software tools that were used to run these models.
14. The authors examine ten models, yet some results lack comprehensive discussion, particularly the Poisson versus Negative Binomial models.
15. Marginal effects are computed but inadequately contextualized, necessitating theoretical justification for the unexpected link between increase in road shoulder width and rollover crash risks.
16. The result discussion section lacks references to support the findings of the research work.
17. I would recommend improving the quality of references in the literature review and introduction. Higher-quality, recent, and more relevant sources would strengthen the foundation of the study and provide a clearer context for the research questions and findings. This improvement could also help support the assumptions and interpretations made in the study more robustly.
18. Line 58, 168, 214-215, 322, 330, 337-338 silly grammatical mistakes.
19. It is also recommended to include the current study limitations and future work recommendations.
20. Minor grammatical errors and awkward phrasing are present throughout the text. For example: "The higher density of U-turns increases the index of rollover crashes." Suggested: "A higher density of U-turns is associated with an increased rollover crash risk." Proofread the manuscript for clarity and grammatical consistency
Comments on the Quality of English Language
The English language quality requires enhancement for better readability and comprehension in the manuscript.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors1. Abstract should concisely summarize and provide additional details of the study. For example, how many years of data were analyzed, what was the sample size, what modeling techniques along with FENB were utilized etc. Currently it fails to capture the gist of the work.
2. Line 38 - This is not a good example as COVID years proved to be an anomaly to the recent crash trends due to several reasons in the US as research from this period indicates. Can you provide some more compelling crash statistics/trends from recent years?
3. Line 41 - "driver vehicle"? Should not it be only driver?
4. Line 43/44 - Is it a worldwide statistics? State the region.
5. Throughout the manuscript - Authors mention several studies indicated something, but provided reference of just one study. Consider providing more in the text.
6. Line 53 - demographic characteristics of drivers or all occupants?
7. Line 64 - "regarding" or "affecting"?
8. Throughout the manuscript - Instead of crash intensity, crash severity will be more technical and appropriate.
9. Line 93 - readers might not be familiar with these severity levels. Consider explaining them first. For example, does "damage" mean no injury property damage only crashes?
10. Line 101-102 - "with random parameters" is redundant.
11. Line 107 - Shouldn't it be "weather conditions" instead of "atmospheric characteristics"?
12. Line 110 - Why and how is evaluating individual variables separately noteworthy to the authors? Currently the rationale for this argument is completely unclear.
13. Throughout the manuscript - Editorial formatting is not removed in several places (e.g. line 168).
14. Figure 2 title - Consider changing "distribution" to "spatial distribution"
15. Line 246 - Service area "density" is coded as binary - does not explain completely. How do you justify it being coded as just presence or absence of service area, where the variable itself deals with a count factor (density)?
16. Line 263 - Explain what is VIF and why is it more appropriate to use compared to mere correlation coefficients with the data in hand?
17. Line 264 - What do you mean by problematic. Please avoid these vague terms throughout the manuscript and support it with a more numerical//technical terms.
18. Line 271- What do these results mean, especially for the average VIF to be greater than 10? Authors did not do a good job in laying the background for analysis and results.
19. Line 302/303 - This phenomenon is called overdispersion where standard deviation exceeds the mean.
20. FENB is one of the most widely accepted technique. Thus, it does not add much value to the extant research by merely stating FENB was superior to the rest. Authors need to provide rationale for using each method, assuming all are appropriate for the data in hand and its distribution. Then they have to provide a probable explanation as to why FENB outperformed all other methods.
21. Line 337- 42% is not less. How do you assume the passing vehicles are only passenger cars?
22. Throughout the discussion - The explanations given are too definitive and hence erroneous or confounding. They can only provide a probable/potential reasoning and insights.
23. Conclusions - The authors need to make the case more compelling by categorically providing the contributions of this study. Moreover, the conclusions lack the limitations and future scope for research too.
Comments on the Quality of English LanguageVery poorly written article with significant grammatical errors and editorial flaws. Not suitable to qualify for a journal quality manuscript.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThere are specific areas where further improvement is needed.
Major comments:
1. The authors should explicitly elucidate the contributions of this manuscript.
2. In the literature review section, it is strongly recommended that the authors incorporate more recently published articles from reputable journals.
3. The authors outlined the drawbacks of current methods in the literature review section. To facilitate readers in evaluating the practical applicability of the research findings, it is recommended that the paper incorporates a discussion on the inherent limitations of the proposed methodology. This addition would contribute to a more comprehensive understanding of the study's outcomes.
4. The authors need to emphasize the necessity and importance of this study. The significance of this study is not sufficiently persuasive in its current form.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors addressed most of the comments satisfactorily. However, I am still doubtful about the paper's contribution to the extant body of knowledge (novelty of this work). Also, the writing still needs a lot of improvement.
Comments on the Quality of English LanguagePoor.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript has been thoroughly refined by the authors based on my feedback, and I find no requirement for further comments.
