Dear Nolberto ,

That is a good question. If the original problem structure has enough reliable quantitative data including numbers of criteria, alternatives ,scores under each criteria and the corresponding weights, it is more objective to apply MCDM . However, if the problem intended for solving incorporates some subjectivity in the model and you need an input from public or from experts , it is more reliable to limit that as much as possible. The powerful of some of MCDM such as AHP is that you can incorporate both quantitative and qualitative scores in the model and check the inconsistency. 

It is totally depend on the objective of the study and the DM who will design the problem and the availability of the data. From your question, in my opinion it is better to go with  run the model with the initial reliable data and apply judgment at the end. Moreover, I recommend to do a scenarios and a sensitivity analysis to consider experts inputs and their impact on the final decision to present holistic approach in your model.

All the best .

Hassan 

Dear Nolberto Munier,

Thanks for nice question. MCDM like AHP has three options approach: bottom-up approach, top-down approach and a combination of both. In your case, you can use a bottom-up approach to give judgment on each criterion but any judgment it must be constructed scientifically and empirical experience.

Model untested not necessarily unsuitable for the application, it is highly dependent on the social, culture and environment  conditions in which the model will be applied.

All the best,

Yasir Haya

by top down approach - do you mean the team deliberates on the pair-wise comparison at first before the final ranking is known?

For your description of the bottom up approach (with initial reliable data and apply judgment at the end ) - I prefer to not apply judgment instead like to do a Monte Carlo on the criteria weights (doing on pair-wise comparisons is a little too cumbersome) 

I have discussed how to resolve some of the subjectivity (in pair-wise comparisons - for your top down approach) in my paper at Intl conference for AHP, June 2014.

Dear Prof. Munier, 

Your question kept me up late last night !!

It is not clear if the following lines of thought have been explored by you.  There are some references I want to bring to your attention.

1. Hopefully there is no "moral hazard" problem (as Economists define it) - i.e., there is no mal-intent. Like making a decision first and then using data, technique etc. to justify the decision. Refer to Amartya Sen's write-up on Rational Fools.

2.  More importantly, is the information availability which has been addressed by Ragner Frisch in his 1969 nobel prize (first time for ...) in economics. He states on page 25 -

We must stress the fundamental distinction between information that is

available before the optimization of the decision model and information that only emerges after this optimization.

I think/feel that you have raised another version of this fundamental conundrum, i.e., (following is modified from Frisch)

information available before optimization  is a necessary basis, but very far from being a sufficient basis for reaching well founded decisions.

Thanks for the fun question.

Refs:

1. https://www.uclouvain.be/cps/ucl/doc/cr-cridis/documents/sen_on_TCR_rational_fools.pdf

2. https://www.nobelprize.org/nobel_prizes/economic-sciences/laureates/1969/frisch-lecture.pdf

Comment on:

Assume for instance that selecting equipment the ranking is D>A>B>C. This is a purely mathematical approach, and then the DM can say, “OK the best equipment selected by the model is D, but in my opinion I would reverse the ranking and select equipment A. Why? Because I have worked with equipment A and I know it is reliable, sturdy, and with reasonable maintenance costs, while equipment D incorporates a new technology that I don’t know if it has been sufficiently tested”.

My view of the above statement is that your MCDM should have a criteria regarding reliability/ sturdiness and maintenance costs.

Dear Chakradhar

Answer to your first letter

Thank you for responding to my question

by top down approach - do you mean the team deliberates on the pair-wise comparison at first before the final ranking is known?

Yes, that is what I am saying

For your description of the bottom up approach (with initial reliable data and apply judgment at the end ) - I prefer to not apply judgment instead like to do a Monte Carlo on the criteria weights (doing on pair-wise comparisons is a little too cumbersome) 

Interesting your proposal

I agree with you. You can apply different tools such as Monte Carlo, Fuzzy Logic or Entropy.

However, remember that for using the two first tools you need a ‘seed’ or initial value.

In case of Monte Carlo you need this initial value µ  and its standard deviation if you think that it behaves as a Gaussian distribution, or its central value and limits if you prefer for instance a triangular distribution.

 For Fuzzy Logic you also need a ‘central’ value, establish limits and also determine if you want to use triangular or trapezoidal membership functions. The problem is how you determine this seed value.

For this reason, I think that in this case AHP has an advantage because you have weight values obtained by certain reasoning, and you can use them as seed values.

I have discussed how to resolve some of the subjectivity (in pair-wise comparisons - for your top down approach) in my paper at Intl conference for AHP, June 2014

Could I get a copy of your conference presentation?

You enriched this discussion with very good points, which are worth to discuss.

I hope more colleagues will join us

Thank you

Dear Chakradhar

Answer to your second letter

Your question kept me up late last night !!

You make me feel guilty………….

It is not clear if the following lines of thought have been explored by you.  There are some references I want to bring to your attention.

1. Hopefully there is no "moral hazard" problem (as Economists define it) - i.e., there is no mal-intent. Like making a decision first and then using data, technique etc. to justify the decision.

No, obviously there is no moral hazard here because nobody is providing misleading information or hiding something; quite the opposite, I believe that researchers are doing their best in looking for better ways to manage this very difficult problem. In my opinion it is the way as it is because we have not found a better manner to do it, but we have to recognize that the approch in which MCDM is conducted nowadays is a little too flexible, if not naïve.

Naturally my position is that I am not in favor of deciding first and using data later at least in MCDM, albeit perhaps it applies to others.

I am not an economist, but remember that economists also accept as valid and acceptable some concepts when they do know that they aren’t

The GDP is an example since it does not consider society, environment and externalities, and however is normally accepted as country development index….

Another example is in classic economics is the ceteris paribus criteria, witch does not represent reality since in fact it is very strange or outlandish to think that you can play with the variation of one variable while assuming than the other are constant……

Another is the Cobb-Douglas production function…

2.  More importantly, is the information availability which has been addressed by Ragner Frisch in his 1969 nobel prize (first time for ...) in economics. He states on page 25 -

We must stress the fundamental distinction between information that is available before the optimization of the decision model and information that only emerges after this optimization.

Thank you Chakradhar, I did not know this extremely important statement, and makes me happy because it highlights by far more precisely than me, what I am suggesting. No doubt that there are to stages of information, and I sustain that from MCDM point of view we must work with information that emerges after this optimization. However, in all truth the last word used by Frisch (optimization) generally does not have a realistic meaning in MCDM since it is quite impossible to reach this optimization in real life, because amongst other things, there could be objectives totally incompatible, such as simultaneously maximize a benefit and minimize a cost, and this is the reason why emerged many heuristic models aiming instead at reaching a compromise solution.

I think/feel that you have raised another version of this fundamental conundrum, i.e., (following is modified) information available before optimization is a necessary basis, but very far from being a sufficient basis for reaching well founded decisions.

Yes, I agree, especially in the word ‘necessary’ because for me it implies a flexibilisation between the data we have and the real data (that we do not know)

Thanks for the fun question.

Thank you for your sharp and interesting comments. I believe that they really help in understanding this issue.

Nolberto

Dear  Chakradhan

I am answering your third letter

Comment on:

Assume for instance that selecting equipment the ranking is D>A>B>C. This is a purely mathematical approach, and then the DM can say, “OK the best equipment selected by the model is D, but in my opinion I would reverse the ranking and select equipment A. Why? Because I have worked with equipment A and I know it is reliable, sturdy, and with reasonable maintenance costs, while equipment D incorporates a new technology that I don’t know if it has been sufficiently tested”.

My view of the above statement is that your MCDM should have a criteria regarding reliability/ sturdiness and maintenance costs.

Exactly.

This should be the conclusion of the DM. Consequently he must go back to the initial problem and modify it adding these three criteria.

Running the model again it could be for instance that the ranking does not change and if this is the case it means that his first result agrees with his preferences. Also it could be that the scores of the alternatives changed, and if now for instance D is very close to A this is not a good signal, and a new analysis by the DM is in order.

As a bottom line the DM must act as a feedback mechanism and then modifying or not initial information according to results.

Dear  Yasir

Thank for responding to my query

Thanks for nice question. MCDM like AHP has three options approach: bottom-up approach, top-down approach and a combination of both.

As far as I know AHP has only one option which is to determine weights from the very beginning, that is, top down. It is the very essence of the method and there is no way to eliminate the pair wise comparisons

I don’t understand how you can consider the bottom up as a different approach, because it means that previously you have had a result using a top down approach. What indeed you can do, once you get the final result in a top dorm approach, is correct or adjust the initial preferences values, but in so doing you are correcting something based on previous SUBJECTIVE assumptions not on REAL values.

 In your case, you can use a bottom-up approach to give judgment on each criterion but any judgment it must be constructed scientifically and empirical experience.

You are right, you can judge a criterion weight based on scientific (like using entropy to generate weights).

However, I don’t believe that empirical experience is of much help because projects are all different , and thus,  which is OK for one of them may not be adequate for another, even similar.

Suppose you have an agriculture scenario and you have to select which crop to cultivate, all of them subject to criteria related with quality of soil, water availability, yields, size of the undertaking, weather, plagues, etc. Do you believe that you can extract values for criteria weights from similar undertakings in other areas, even using the same criteria? Not in my opinion.

Model untested not necessarily unsuitable for the application, it is highly dependent on the social, culture and environment  conditions in which the model will be applied.

What you say is true about untested models, but I don’t think you can test a model; AGAINST WHAT, if you do not know which the ‘true’ solution is? You realize of course, that if for a problem involving say three alternatives we knew which the best solution is, we wouldn’t need to use any MCDM method to select one of the three. That is more or less what happened in the past using a tool such as Cost/Benefit ratio, because the only objective was to obtain the best economic results. That changed some decades ago, when projects involved, in addition to economics, social, cultural and environmental aspects as those you mention, many more like Risk, Safety, Sustainability, etc.

Considering the above I agree with you about your second paragraph. All scenarios are subject directly or indirectly to social, cultural and environmental conditions, and for this very reason these influences must be considered in the original criteria. Some criteria values corresponding to alternatives or options are quantitative, for instance you know that there are limits for every kind of pollution in soil, air and water, which are established by international or local organizations. There are landmarks for social for instance the Human Development Index or the Gini index than can be used as references, and there are also cultural indexes from international organizations.

Best regards

Nolberto

Dear Hassan

Thank you for your contribution to this subject

That is a good question. If the original problem structure has enough reliable quantitative data including numbers of criteria, alternatives ,scores under each criteria and the corresponding weights, it is more objective to apply MCDM .

Sorry, I don’t understand, since whatever the procedure we are always applying MCDM, not Cost/Benefit or IRR or NPV, because we have more objectives than Economy as before.

But even in the case you mention most models apply subjectivity from the very beginning since they assign weights to criteria, from experts, from AHP of from Delphi. I wonder based on what, since there are no identical projects from where information can be collected, and normally with different conditions of any type.

However, if the problem intended for solving incorporates some subjectivity in the model and you need an input from public or from experts , it is more reliable to limit that as much as possible.

No doubt about that. In my opinion we have to distinguish between unavoidable uncertainty and ‘fabricated’ subjectivity.

There are always uncertain and subjective aspects in projects analysis, such as weather, people opinion, inflation, interest rates, demand, geological features, etc. Therefore, what we do is to refine that uncertain information as much as we can using for instance Monte Carlo, statistics, surveys, fuzzy logic, etc. This is unavoidable, and we have to live with it.

However, introducing at purpose subjective factors such as personal comparisons and weights is a different matter, and this is at what I am aiming to. I am not saying that we must not input comparisons and weights, what I am saying is that we have to choose the moment to do it, and that moment is when a result, based on the reliable and unreliable information we have, is obtained. That is we have a base, imperfect of course, but is the best we have.

 After the result is known, purely by mathematical means, the DM and experts may have the opportunity to analyze the sensitivity of the solution, and correct it as needed and according to preferences and even adding weights. Otherwise, if weights are established from the very beginning and the model run, when the result is reached, the DM and experts are applying subjective values on subjective values.

The powerful of some of MCDM such as AHP is that you can incorporate both quantitative and qualitative scores in the model and check the inconsistency.

Most models allow for both types of data, but it does not change the fact that some values are uncertain.

 Regarding inconsistency, I would not like to be in the skin of a DM, in front  of stakeholders for one thousand million Euros hydro project, and telling them that because inconsistencies there could be an error of 100 million Euros between selected options.

It is totally depend on the objective of the study and the DM who will design the problem and the availability of the data.

Sorry, what is totally dependent of the objective? I agree that the DM designs the problem, regarding the alternatives he has, which give him information about which criteria to choose, and the MCDM to apply

From your question, in my opinion it is better to go with  run the model with the initial reliable data and apply judgment at the end.

Fine, you agree with me then

 Moreover, I recommend to do a scenarios and a sensitivity analysis to consider experts inputs and their impact on the final decision to present holistic approach in your model.

Agreed on in both counts.  However remember that normally you cannot design scenarios although it is possible in other cases. That is, a real an uncertain scenario may be demand decreasing, prices increasing, weather suddenly changing, the government imposing a new tax, and all of them simultaneously and in different degrees.

This is for me the challenge we have. Create a model that can give information on the potential effects of various external factors acting simultaneously

Thank you for your valuable comments

Nolberto

Thank you for your comments Prof. Munier.

1. conference paper can be found at 

https://www.researchgate.net/publication/309666424_An_analysis_of_the_process_in_deriving_further_benefits_of_an_AHP_model

2. I meant "kept me up late last night" in a figurative manner - No need for any guilt feelings ...

actually, the sub-conscious impact of the question, and the random walk of my activities in the day led me to Ragner Frisch's work (which I was not previously aware)

3. Thank you for your note on "seed value". agree with you that  "MCDM is conducted nowadays is a little too flexible, if not naïve".  Agreed. When I work with senior project managers, unfortunately expediency takes over, there is another factor affecting decisions - "who is to make the decision? (Kahneman and co-workers)" is usually violated i.e., the competent person with the right data is not in an influencing position of the decision! 

There are number of other behavioral aspects that also cloud decisions (in infrastructure construction sector in India) - that I link to Harvey's Abilene paradox and Seligman's Learned Pessimism.  Both issues have counter-measures - Harvey suggests one himself and Esther Duflo suggests another for the Pessimism. My effort is only at "case studies" level right now.  

I try individual versus group (Delphi etc.); multiple group assessments as a way of highlighting that risks in decisions are subjective.  My attempt in training to senior decision makers is to get them out vested interests (moral hazard) or comfort zones and out of overly judgmental (without data) stances.

There are practical limits to how much of the subjectivity can be removed - given "requisite holism" and "bounded rationality" will also kick in at some point.

4. The way I understand Frisch's sufficiency condition for optimization - is when data becomes available , say after decision, it can be used to strength the decision. 

5. In your response to Hasan, you state - "Create a model that can give information on the potential effects of various external factors acting simultaneously".  This is not possible with a static model, it needs a dynamic model which is "learning" as well.  At other places, you are clamoring for a decision support system as well.

6. I do an assessment of subjectivity by quantizing David Hillson's risk attitude model.  To me this is a bound on the person's "subjectivity".  Same can be done with Kahneman-Tversky's risk appetite model.

7. Are there any estimates on the bounds of subjectivity that you are seeing even when no new information comes up?  This is a simple problem I think.  In the conference paper and its aftermath - I attempted a robustness estimate of AHP decisions (did not finish this work). This line of thinking may help.

Conference Paper An Analysis of the Process in Deriving Further Benefits of a...

Dear Chakradahar

Thank you very much for sending me your paper.

2. I meant "kept me up late last night" in a figurative manner - No need for any guilt feelings ... actually, the sub-conscious impact of the question, and the random walk of my activities in the day led me to Ragner Frisch's work (which I was not previously aware)

However because your comments it looks that you were well awake and knowledgeable………

3. Thank you for your note on "seed value". agree with you that  "MCDM is conducted nowadays is a little too flexible, if not naïve".  Agreed. When I work with senior project managers, unfortunately expediency takes over, there is another factor affecting decisions - "who is to make the decision? (Kahneman and co-workers)" is usually violated i.e., the competent person with the right data is not in an influencing position of the decision! 

Unfortunately this is quite true and I have examples about that. However, remember that unless that a large percentage of stakeholders agree on something, supported or not by what the DM tells them, that decision discharges responsibility from the DM, because he is an advisor, he express what he thinks is the right choice and if he did a good job and has solid arguments to support his decision, there is nothing he can do if the majority of stakeholders do not agree with him. There are many reasons, political, economical, of risk, of vested interests, etc. that perhaps the DM is not aware of but that have a large weight in stakeholders’ final decision.

Suppose for instance - and this happened to me - that a City Major commissions a MCDM study to decide what undertakings will take place in his four years mandate. Assume that the DM submits his recommendation as follows: B>A>D>C, which responds to an analysis that included limited resources. The Major then says ‘Why project K is not included?’ The DM answers that it has not been selected by the model and his analysis shows that it does not have enough merits to compete with others.

The Mayor then says: ‘Well, I understand what you are saying but project K has to be included as the best, because I promised to execute it in my pre-election campaign’

As you can see I know first hand this problem, but, I was not asking about how to make transparent and documented management decisions. It is an aspect of management that is not included in my original question. I was merely asking for the MCDM process by itself.

There are number of other behavioral aspects that also cloud decisions (in infrastructure construction sector in India) - that I link to Harvey's Abilene paradox and Seligman's Learned Pessimism.  Both issues have counter-measures - Harvey suggests one himself and Esther Duflo suggests another for the Pessimism. My effort is only at "case studies" level right now.

I believe that you will agree with me that this actual example enters in the category of Harvey’s Abilene paradox, because some stakeholders that in private support the DM decision don’t dare to express their opinions. The reason is quite obvious.

I try individual versus group (Delphi etc.); multiple group assessments as a way of highlighting that risks in decisions are subjective.

I wouldn’t say that risks are subjective; they are possible and perhaps probable.. Remember than Risk = Probability of occurrence x Impact, and both can be assessed in most cases

 My attempt in training to senior decision makers is to get them out vested interests (moral hazard) or comfort zones and out of overly judgmental (without data) stances.

Naturally, I agree with you

There are practical limits to how much of the subjectivity can be removed - given "requisite holism" and "bounded rationality" will also kick in at some point.

Well, Herbert Simon proposed bounded rationality as an alternative to mathematical modeling in decision making. Attractive as it might be remember that there is not a measure for rationality and thus what is A for somebody is D for another. This is the reason by which I support the top-down process in MCDM because once you know the ‘right’ answer given by a mathematical model it can enlarge the rationality bounds of the DM

4. The way I understand Frisch's sufficiency condition for optimization - is when data becomes available, say after decision, it can be used to strength the decision. 

I share your opinion

5. In your response to Hasan, you state - "Create a model that can give information on the potential effects of various external factors acting simultaneously".  This is not possible with a static model, it needs a dynamic model which is "learning" as well.  

Regarding the bold statement I believe that when you can express your results for several scenarios, it is no longer static. If you are able to draw a curve that allows you to see how the utility of your project vary by different variations of key parameters along time, I think that you can feed information into the model as often as necessary and see the results.

Suppose that our project deals with copper export. The international price of metallic copper varies continuously but within certain limits and normally in small amounts in plus and minus. Therefore, if you can compute the utility of your project you can compute it as often as you want, since you only need to replace values and run de model, and you can check that it is safe as long as prices oscillate within the allowed range of variation

At other places, you are clamoring for a decision support system as well.

And you are right, I think that MCDM is half of the picture. If the DM cannot extract useful information from the result, that is, if there is not a decision support system that allows him to justify what the model suggests,  then it is not very useful, because decisions are taken by the DM, not by a model, whatever powerful it might be.

6. I do an assessment of subjectivity by quantizing David Hillson's risk attitude model.  To me this is a bound on the person's "subjectivity".  Same can be done with Kahneman-Tversky's risk appetite model.

I agree with David Hillson philosophy especially when he says that it is necessary to have ‘an appropriate and mature risk culture’, concept which by the way I applied in a book of mine ‘Risk Management for Engineering Projects’, some years ago

Regarding the other researchers you mention unfortunately I am not aware of their work

7. Are there any estimates on the bounds of subjectivity that you are seeing even when no new information comes up?  

I believe so if you appropriately investigate the information that a model provides and that your research reveals

For instance a hydro project is related with uncertain water flow rates. From here you can widen these uncertain bounds by determining the origin of water and once you know it by studying the influence of atmospheric factors such as rainfalls and snowfalls. Then you can examine statistically which and in what extent each atmospheric phenomenon participates. You can also investigate how the operational regime of other hydro projects upstream may affect your project, as  well as determining what industries and undertakings take water from the river, etc.

This is a simple problem I think.  In the conference paper and its aftermath - I attempted a robustness estimate of AHP decisions (did not finish this work). This line of thinking may help.

In your work of which I had a glance you promote working with more that one DM, computing criteria weights for each one and then using the same weights to find alternatives ranking. I find the method interesting but also points out to one of the MCDM models deficiencies since two different DM find different values. This of them to select is in itself a new decision problem!

I am sure you understand after this very fruitful interchange of comments between you and me, my reservations about the AHP method. For me it is useful to solve simple problems where decision are purely subjective as is the classical example of purchasing a car, but I would not use or recommend it to treat serious, complicated problems, let alone complex, not only because the extensive subjectivity and the heavy load work involved, but also considering than in AHP criteria and alternatives are taken isolated, as independent entities, and that is not real (for instance a criterion price is not independent of criterion quality or contort).

In my opinion the method is flawed, and I am not the first to say so. Naturally if you care to challenge my statement I will be glad to oblige.

Saaty corrected this by developing ANP which links everything with everything and then results are more reliable, albeit the subjectivity in preferences subsists.

Nolberto

1.  Kahneman got the Nobel Prize in 2002. Landmark papers freely available are-

• Judgment under Uncertainty: Heuristics and Biases, Amos Tversky; Daniel Kahneman, Science, New Series, Vol. 185, No. 4157. (Sep. 27, 1974), pp. 1124-1131.
• Prospect Theory: An Analysis of Decision under Risk by Daniel Kahneman and Amos Tversky , Econometrica, 47(2), pp. 263-291, March 1979

2.  The only statement, I would like to quarrel with you is on the matter of probability assessments.  In construction projects - these are magnificently subjective.

3. we have not addressed induced subjectivity from estimation of an individual's bias or even emotions.

4.  will be shortly procuring your books ... the table of contents are great.

I am writing to you privately and see if we can work out a research program. 

Chakradhar Iyyunni · 6.61 · Larsen and Toubro

The only statement, I would like to quarrel with you is on the matter of probability assessments.  In construction projects - these are magnificently subjective.

Construction projects is a subject on which I have some experience on both the planning and feasibility and in the execution stages. I worked in three large hydro projects, as well as two projects in mining and chemical and in a project for the automotive industry, sometimes as Planning and Control Engineer and others as Project Manager. These projects involved hundred of millions of dollars and were located in Argentina, Ecuador, Mexico, Peru, United States and Canada. My book on Project Management reflects in a great extent that experience. Therefore, if you think that we could do something about the construction industry it will fall within my expertise.  

Concerning your comment, it appears that we have not a quarrel but two different points of view, because especially in construction projects there is no too much room for subjectivity, but unfortunately some sort of ‘planned subjectivity’ may occur. For instance, data for cost and duration of a task is obtained as you know from updated databases, however, some times the cost estimator, adds some additional little percentage ‘just in case’, It is a subjective action but we cannot call this subjectivity, but incorrect procedure.

This ‘illegal’ action will probably be reflected during construction when it is realized that consistently, costs and durations have been unduly increased. The evidence is normally reflected in the control stage of the project when you monthly compute the cost performance index and the schedule performance index.

If you refer to external factors that can affect a project, you normally have statistics that can give information about weather, labor productivity for repetitive tasks such as rock excavation or for pouring concrete slabs (for instance from the ‘Learning curve’), suppliers delays, etc. In my opinion there are many uncertainties or facts that you can do nothing about such as interest rates, government import measures, bank loans, geology of the terrain, etc. that cause that sometimes the Project Manager has to guess what can happen, using supporting data.

For me, in general, in construction it is not a problem of subjective appraisal but a matter of a rational planning and working together with diverse contractors and suppliers and the ability to foresee potential problems. As an example, if you need that a European manufacturer fabricates for you a very costly and complex equipment (for instance a Tunnel Boring Machine, or large press,  or turbines), I wouldn’t be satisfied with what he promises; I, as a Project Manager, would visit his factory, evaluate the infrastructure he has, consult with some former clients about compliance, reliability, performance of similar equipment, etc. The DM and the PM have to work together to detect where potential problems may appear related with prices, delivery time, transportation, equipment assembly, etc. In my opinion it is necessary to work by preventing what can go wrong in time, performance, transportation, etc.

A cost of equipment for instance is not subjective. Either you purchase the equipment from the supplier on a CIF price at the job site, or you defer the purchase and make an economic analysis of the convenience to defer the future cost of the equipment that most certainly will be higher, and it is unknown.

3. we have not addressed induced subjectivity from estimation of an individual's bias or even emotions.

True, we didn’t, and I think that my example with the estimator addresses this issue.

There could be bias of course, as the example of the Mayor of a City that insists that a project he promised during his campaign must be included as the best, even if wasn't selected However, I wouldn’t call this bias but dishonesty.

4.  will be shortly procuring your books ... the table of contents are great.

Thank you very much for your evaluation

Nolberto

I would like to do some work with you for bringing your best practices to the Indian Context. Our center works mostly with construction project personnel (from 4years to 40!).  A colleague of mine (Sunil Kumar) is a construction manager with exposure to 6 hydro dam sites in India - 3 of which he was site in-charge. 

There are three other thing , I am concerned about - it is my worry for my project personnel:

1. the band-width available to a DM or PM (more so for an untrained one) for being clear-headed and focused.

2. I am really troubled with the mis-use of the Pareto principle - commonly stated - "the vital few and (trivial) many" - The vital few kills your energy and trivial many kills your project. 

3. More and more, I am able to corroborate Flyvbjerg's observation that projects social (human) behavior  is a reflection of the social structure in the region or country.  Sunil think and I think that the observation holds not only for projects but also for organizations - making it surprisingly rare to find strong project sponsors.

Thank god that exceptions do exist to my above predicament even in the Indian context - One Mr. E. Sreedharan who was the overall in-charge for the Delhi Metro Project.  He got a mandate that the Chief Minister of Delhi will not interfere; he also moved the arbitration/ jurisdiction location moved outside Delhi - so, frivolous law-suits are reduced etc.  There is a small booklet put out Delhi Metro Rail Corporation, authored by Anuj Dayal which catalogs much of this. I will try to send you this booklet.

Dear Chankradahr

I would like to do some work with you for bringing your best practices to the Indian Context. Our center works mostly with construction project personnel (from 4years to 40!).  A colleague of mine (Sunil Kumar) is a construction manager with exposure to 6 hydro dam sites in India - 3 of which he was site in-charge.

I think it is an interesting proposal.

 I can perhaps help in preparing a list and corresponding explanation and if possible, with actual examples. However, I can bring only experience from the American (North America, Central and South America) context, but I do not know how they can be embedded into the Indian context. For instance, I remember that a Canadian company was engaged in building a harbour in Eastern India. They devised a fully automated conveyor system that could be managed by five persons. However, Indian Port Authorities rejected it on the grounds that the need the harbour, but also they need to employ people, and therefore, instead of an automated conveyor they prefer to use people. How can you fight that?

However, if we describe best practices for personnel preparing and managing the project, perhaps I can contribute.

There are three other thing , I am concerned about - it is my worry for my project personnel:

1. the band-width available to a DM or PM (more so for an untrained one) for being clear-headed and focused.

You lost me with your reference to band-width since I don’t see the relationship between this expression and your personnel being clear-headed. Do you mean that in general your personnel do not have clear ideas or objectives and see them out of focus?

2. I am really troubled with the mis-use of the Pareto principle - commonly stated - "the vital few and (trivial) many" - The vital few kills your energy and trivial many kills your project.

I use it sometimes to make a prioritization of items that must be closely checked, especially related with costs from suppliers for equipment and services rendered. I consider that as in the same way as the Critical Path identifies critical items, Pareto principle is able to identify critical cost items.

Sorry I don’t understand the last sentence, the one that is italicized

3. More and more, I am able to corroborate Flyvbjerg's observation that projects social (human) behavior  is a reflection of the social structure in the region or country.  Sunil think and I think that the observation holds not only for projects but also for organizations - making it surprisingly rare to find strong project sponsors.

I am not familiar with the text you mention from Flyvbjerg. I read from his author a paper entitled (Bent Flyvbjerg and Cass R. Sunstein). ‘The Principle of the Malevolent Hiding Hand or, the Planning Fallacy Writ Large’

I was discussing its text in an interchange of communications between the Karlsruhe Institute of Technology, in Germany and myself, where I lectured. I am here replicating my answer regarding this article by the head of the Risk Department

“The paper is really interesting especially because the two opposite concepts such as ‘Benevolent’ and ‘ Malevolent ‘ the hiding hands, from Hirschman and dismissed by Flyvbjerg, with who I coincide. If you have problems in a job may be sometimes you can use your aptitude as problem solver and remediate it, but not in construction work. You cannot replace with some contraption the absence of necessary inputs or inefficiency of personal.

However, I coincide in many aspects with Flyvbjerg in the sense that many failures that cause time and costs overrun are a consequence of short seeing when the project is prepared, because planners do not consider risks along the project and because they take as certain things which are not.

This is precisely the core of the paper that you kindly requested me to write, where I believe that the whole project must be prepared based on the concept that risk exists everywhere because normally things do not develop in the way one expect. Uncertainty is always present and then it has to be taken into account, since concepts such deterministic times and cost are a fallacy”

Apparently you and I agree on that part of your message that I italicized.

Thank god that exceptions do exist to my above predicament even in the Indian context - One Mr. E. Sreedharan who was the overall in-charge for the Delhi Metro Project.  He got a mandate that the Chief Minister of Delhi will not interfere; he also moved the arbitration/ jurisdiction location moved outside Delhi - so, frivolous law-suits are reduced etc.  There is a small booklet put out Delhi Metro Rail Corporation, authored by Anuj Dayal which catalogs much of this. I will try to send you this booklet.

Yes agree, you can find honest people once in a while, but in all honesty I don`t quite understand what you are trying to say in this long paragraph

At any appropriate level, according to the task! Example: Conference Paper FOREST ECOLOGICAL MANAGEMENT AND EXHAUSTIBLE TIMBER PRODUCTI...

Data Value (utility) based control and decisions

Dear Yuri

Yes, you are correct, however the key word here is 'appropriate'

Decision making is not a matter of preferences but a conscious analysis based on actual facts in case of quantitative criteria and on as much as reliable information from statistics, polls and surveys in qualitative criteria.

However, this is not universal. In trivial problems such as selecting an apartment there is no doubt that preferences play a paramount role, for the simple fact that nobody knows better his/her needs than the person who is going to live there.

But a very different scenario is when the projects involve millions of dollars, affect the environment, modify existent structures or affect millions of people.

In these cases nobody can assess the needs, importance, benefits and drawbacks for the millions of person that are going to be benefitted or hurt because the project. Here, it does not make any sense to have a person (DM) deciding or those millions

Now, when results are know using an unbiased procedures, is where the DM plays his most important role, by examining results, based on realities and analyzing and informing the stakeholders about the advantages, disadvantages and especially risks of the best selection