# 180

Luiz Fernando Kowalskia, Érico Masieroa, António Manuel Saraiva Lopes, Marcos dos Santos, Carlos Francisco Simões Gomes and Dragiša Stanujkić

Designing streets for people: A multicriteria decision-making study

My comments:

1- I find this article very attractive, novel (at least for me) and addressing a valuable problem, related to urban planning

However, in my opinion, it lacks clarity when addressing techniques as AHP-Gauss, not regrading AHP or Gauss, which are very-well known, but in Gauss normalization, a vital step. I believe that it needs clarification with a brief concise explanation, for instance, saying that it consists in normalizing to 1, or making that the sum of all values beneath the Gauss bell is 1. In consequence, all values will be between 0 and 1.

As I understand it, you do not determine quantitative weights, but instead use a continuous of values derived from statistics that without a doubt eliminates the pair-wise comparison and the fabricated ratio between criteria. Thus, you are able to determine the value of each criterion, which is the product of the number of criteria and the probability of an integer value for each one, formula derived for the binomial distribution.

From there you compute the standard deviation, which determines the importance of the criterion. Probably I simplified this too much, and perhaps it would be necessary to expand the explanation, IF it is correct.

It is interesting the approach because it assumes that the DM probabilities are at random, which justifies the use of the Gauss distribution.

2- Page 2 you say “WISP allows for assigning weights to criterion robustness, computational robustness,computational effectiveness, and transparency”

I do not believe that the weighted sum method allows weights for criteria, rather, it uses them.

It is very simple and rational; however, its main drawback is the use of subjective weights, and the lack on interrelationships between criteria. This last concept is also applied to the AHP-Gauss, since it considers each criterion separably and not jointly. Why is this concept primordial

? Because a criterion may influence the content of another, remember that in this field not always the sum of parts is equal to the sum of the parts.

3- Page 9. “ Step 5. Calculate the global utility (ui)) for each alternative”

How? You don’t say it. By a subjective appraisal?

4- Pag. 11, Table 6. You do not explain utilities., which are gains on something that produces satisfaction, safety, comfort, etc.; they refer to sidewalks and to pavement, but what is the ‘d’? and why the utility for sidewalks is negative? What does it mean?

But answering my own question, the utilities can be negative, providing that we are talking about marginal utilities. In my opinion, to express that a utility as negative does not have too much sense, since the word becomes being an oxymoron, because is in itself contradictory.

Since how, a utility that is considered a benefit or positive, can be negative?

As you can understand, it is very difficult to follow the process due to the absences of meanings and explanations.

5- Page 12 Step 5 “Calculate the Gaussian factor for each criterion”

You did not explain what a Gaussian factor is, and now the step is calculate it?

Honesty, Table 9 means nothing to the reader. Why the mean is the same for all criteria?

6- Page 14 “It should be noted that in this study, stakeholders only assigned weights to the criteria and did not suggest them directly.

Normally stakeholders, representing different fields, are unable to assign weights because it would mean that each stakeholder must know in detail the areas of the others, and this is not realistic, let alone to discuss with each stakeholder. I do not see the common sense and feasibility of this approach. For instance, do you think that the guy in cost can discuss with the guy in carbon foot print, when both are technically speaking in different languages. Most probably, he does not even know what it is….or what it measures, and vice versa.

7- Page 14 “The sensitivity factor comes from the mean and standard deviation of the data in the matrix”

Do you think that this ‘explanation’ sheds any light on the procedure?

8 -Page 14 “Regarding the eleven criteria, both methods (AHP-Gaussian and WISP) proved to be satisfactory, in selecting the most suitable urban road system’

And how do you know that they found the most suitable solution? For that you would need a real value to compare to, and you do not have it.

These are some of my comments. Hope they may be useful to you

Nolberto Munier