Dear Mr. Mr. Zeljko Stevic and coauthors.

I have read the article “Multi-Criteria Decision-Making Model for Evaluating Safety of Road Sections”

and here is my feedback.

It is really a very interested project and, in my opinion, well-written. There are some aspects that called my attention and this is the reason of this letter.

I understand that the problem consists in determining which is the road with minimum risk of accidents. It is a very important issue and I have not seen in the literature this subject before.

1- You say “Based on defining a list of criteria according to their importance, determining comparative values using the scale mentioned above, it was obtained the weight values of the criteria:

Where this scale come from? It appears arbitrary.

2- Using data from you Table 4, I solved the problem using the SIMUS method, that does not need weights, and obtained exactly the same raking as you, i.e. A6 – A 5 – A3 – A4 – A1 – A2.

Sensitivity analysis shows that A6 depends on the available variation range of all criteria, except C3.

All of these criteria have a very small range of allowable variation, about 0.01, therefore, the best selection (A6) is risky, meaning that a very small variation of any of the mentioned criteria will probably produce that the best alternative is not A6, but A5, the second best.

If we now analyze A5 the situation improves, since it depends of C1 with a range of 0.06. Therefore, it seems that A5 is a better selection than A6, since it holds its position for variations of some criteria.

3- In page 83 you say “The simulation through 70 scenarios was performed by reducing the weights of each criterion in a range of 5- 95%, while the values of the other criteria were proportionally increased depending on each scenario individually”

Regarding this paragraph it appears that you worked with the OAT scenario (One a At a Time), meaning that you select a criterion, increase it, and assume that the balance of criteria increase (shouldn’t it be decrease?), proportionally.

Why proportionally? Most probably there are criteria that have a non-linear relationship.

In my opinion, this procedure is incorrect, because you are increasing the weights of each criterion assuming that all of them allow for a such large variation, and this is not normally true, since each criterion has its own range of variation, and also due to the fact that every increase affects the balance of the range.

You assume for this particular criterium that you can increase it from 5 % to 95%. Normally this very rarely happens.

There is something that I don’t understand. Out of 7 criteria, all, except criterion C3 calls for minimization.

Consequently, if A6 is considered the best section, it is assumed that it is the section with the minimum risk of accidents. However, in the conclusion, you say that A 6 is the riskiest section.

4- As a final point, again, in my opinion, you are not performing a sensitivity analysis here. You only determine which is the risk in each section. Perhaps I am mistaken, considering that I don’t know the scenario as you do. In this case, please illustrate me about this confusion.

Hope it helps.

Nolberto Munier