The numbers game...

Delphi Survey is a method developed, IIRC by the Rand Corporation. It is most often used as a tool for predictions and planning, usually by surveying a panel of experts in a field, but can also be used in most any situation.The basic premise is that if you ask a large body of people a question where they DO NOT know the answer, the responses will tend to clump around the correct answer. The idea is to remove the influence of a strong supporter that might happen in a meeting or gathering so that you get the individual responses. For example you might ask a million people when the Norman Conquest happened. Some few will know the date, but many will not. Yet when all the answers are analyzed you'll find there is a clump around 1066.The reason I bring this up is that there may well be some validity to the numbers.

---

Aslan is not a Tame Lion

IIRC, there is a bit more to the method. There is a second questioning where everyone is given the first pass answers and invited to change theirs.

Yes. It can be iterative in nature. The key point is to give only the results and not the sources. If you expose the sources and one is a person of Authority, it skews the results towards the opinion of the Authority.One of the interesting places where Delphi was used in a novel was in Shockwave Rider. In the novel it had been adopted and adapted as a functional merger of Wall Street and Lotto, where you could vote by placing a wager on specific questions, "Who will be the next President of Paraguay?" or "Time travel will become reality in 2026?".

---

Aslan is not a Tame Lion

Just done a spot of reading on the Delphi method, and to me it seems, upon a quick scan, that it is more for correlating predictions of specific occurrences garnered from a large body (about 50 or so) of responders with experience in that field, but done in such a way that the usual pecking order is side stepped (though how you avoid people recognising the ideas and individual styles of their contemporaries is another matter)The key thing does seem to be a stressing of knowledge of the responders and iteration of questionnaires to shake-out a consensus of opinion on what might happen, not what is.Sure, the responders don’t know the answer, because they are being asked to speculate on future events based upon their own knowledge of the subject. Where as you can’t speculate on the date of the Norman invasion, that is an established ‘fact’ you either you know it or you don’t. Any clumping in the data in regard to the Norman Conquest survey is probably due more to people guessing against their partially remembered knowledge of history lessons. Also the feedback in the second, third, fourth questionnaires, as the date of the Norman Conquest isn’t a matter of speculation, should nudge the memories those who are un-clear at to their knowledge and convince those who had no clue as to the correct date, there should be very little spread as the correlation should almost be absolute. [if they had no knowledge of this subject, why are they responders in the first place seeing that the knowledge of the responders is paramount to the Delphi Method?]So I don’t see how Delphi adds validity to the idea that if more people believe a concept to be true that somehow makes the concept to be true.You could use Delphi to quiz experts on religion to garner some kind of predictions on whether secularism will grow yet further or whether there will be a strong theist back lash, but not whether the concepts of theism are true based on responders or clumped response.Though I did say I only glanced at info on this so I could be arguing from limited evidence.

The reason that I mentioned the Norman Invasion is that it is a known. Before placing any reliance on projections of something in the future it's good to have some confirmation that the method actually works. Tests can only be run against a known result, sorta a standard. It's like testing radiometric dating against something of known age.The amazing thing about the Delphi Method is that when a statistically large sample is used and questions asked where the respondents cannot know the answer, there really does seem to be some correlation between the responses and fact. The problem when we consider applying it to religion is that frankly, it's unlikely we will ever be able to arrive at a conclusion that can be transfered to others still living. If though, the method really works, then there is the possibility that similar value might be placed on such questions even though we will never know the answer for sure.

---

Aslan is not a Tame Lion

So you are saying if you randomly select a number between 1-1,000,000 and place a card marked with that number in sealed box then run a series of Delphi surveys (with say one million respondents) to determine a prediction of the number on the card, when none of the respondents could possibly know what the number on the card is, you are saying that the survey is likely to demonstrate a statistically significant correlation between the results and the number on the card?Either such a large respondent group will produce, by the nature of the task, a good number of clumps which will in turn be enhanced by the iteration, and as there are likely to be a few well defined clumps one is likely to seemingly correlate to the hidden answer. That or a significant number of those responders actually know what the number in the box is, and if that is the case this does seem to be condoning ESP. And If it’s ESP then isn’t there a large prize for proving such?

Getting back to the OP . . .

quote:

---

In the movie the voting is to choose the one representative of earth to ride in this machine that earth is building from a set of plans ‘faxed’ down to us from some unknown aliens.

---

In this context, I think it is appropriate for one of the "95%" to be a REPRESENTATIVE of human culture. Religion or a belief in a deity is a defining characteristic of many peoples lives. I think it is legitimate to have a deist as a representative, or at least someone who is sympathetic to the mindset. Jodie Foster's character was not sympathetic to the religious mindset so the comment was not inappropriate.When arguing about "truth", an appeal to authority is totally inappropriate.

I agree I would agree with this also. To me the operative word is representative. One serving as an example of a larger group.And to avoid perpetuating a falsehood or in the least, an exaggeration, here is a graphic dated 2002 that indicates the 95% number is actually closer to 86% as previously posted by ohnai. The data is a division of the worlds population and is footnoted as an "approximate estimate".

No, not really. That is a straight lottery with no influence from society on the outcome. Delphi survey method is concerned with things where there is input from society on the outcome. For example, "In what year will the US become a Theocracy?" might be a question that could be answered through a Delphi survey.

---

Aslan is not a Tame Lion

A really good a question suited for Delphi analysis, no question there. But is Delphi analysis really suited for confirming the truth of a concept held by the majority?This message has been edited by ohnhai, 09 March 2005 22:51 AM

It can't confirm anything. Map & Territory. But it might give an indication. Confirmation comes from the item itself. But it might be useful in some ways. Remember, it is generally used to predict future events. It is also for questions where none of the participants KNOW the answer for sure. So it could be used on concepts and might even give a reasonable prediction. The example I used might well be one that would be both valid and of significant importance.

---

Aslan is not a Tame Lion

Ok so we have decided to send a religious person to represent us to aliens, as they represent the majority of the population. So In this regard, which religious belief, do you choose to represent everyone else? Remember there is only one seat.As they are the biggest group do you send a Christian? If so which Christian faith do you choose to be representative of, not only Christianity as a whole but, all religious views on the planet, and on top of that to represent the entire human race including the Atheists?If you take Christianity and cut it into its representative slices (of which there are apparently over 9,000 Christian denominations) does Islam then become the biggest single religious group and thus becomes worthy of representing the entirety of human kind? If you were a Catholic would you be willing to let a JW or LDS represent you and all other faiths. Or would you be happier with a Muslim, Jew, Buddhist or Scientologist? Would you trust any of them to fairly represent you? Would you trust an Atheist more or less? What about an Agnostic? What about Jedi?

Why would you need to subdivide the characteristic of religous belief? Why must there be a choice? Suffice to say that a significant majority of the population of this blue marble has a belief in a deity or deities, and the selected representative has a likewise belief.

Because there is one seat. So yes anyone you choose will represent the theists out there, but which one will represent the whole world the best?If you make the distinction between Theist and Atheist claiming the theist is more proportionally representative then surely you must carry that thinking on into the religions themselves. Sending a Jew would not be any good as they only represent a small percentage of the world’s population, Though as they don’t believe that Jesus is the son of god they do represent the majority in that respect.What about YEC’s ? they don’t represent the majority of the human population as their beliefs fly in the face of science and it’s evidence and as science is such an important part of human existence you would have to choose some one who follows not only a religion whose beliefs tie most strongly with the majority but also who accept the findings of mainstream science.If you make the distinction between Theist and Atheist then you have to apply that rational through all aspects of selection.If you are going to play the numbers game then play it fairly.This message has been edited by ohnhai, 10 March 2005 08:54 AM

I'd choose a Buddhist, as their theism seems least affected by voluntary ignorance. That is, while IMO there is a significant quantity of obscurantism in Buddhism, its probable that a Buddhist would be more able to brong back a reasonably factual, uninterpreted account fo their experiences.