The "Delphi Technique" as described in the link is indeed insidious and its use is fraudulent.
However, you may from time to time read that the Delphi Technique was developed at the RAND Corporation. In fact, I recently read that, with the further statement that it was developed as a mind control technique.
WRONG!
Two of my friends at RAND did indeed develop something they called Delphi. However, it was not a method for forcing consensus. It was instead a means for efficiently extracting information and opinion from a group, without the frequent committee problems of dominant personality, social pressure from a majority, and settling on a least-common-denominator simply to get agreement.
RAND Delphi was conducted as follows.
Round one: people were asked to submit written statements relevant to the issue the group was to take up. These statements were summarized.
Round two: the consolidated list of statements was submitted to the group as a whole, with the identity of the originator of each statement eliminated (anonymity). Each group member responds, again anonymously (degree of agreement on a numerical rating scale; numerical response to the issue, etc.). These numerical responses are described statistically (usually median and upper and lower quartile).
Round three: The statements, and the statistical description, go to each group member, again anonymously. Each member again responds with a numerical estimate. Those whose estimates fall outside the quartiles must provide a reason why they believe 3/4 of the group is wrong.
Round four: each group member again receives the statements, and the statistics from round three. Each member gain provides a numerical response. Again, the "outliers" must provide reasons. In addition, anyone can provide an argument for or against the numerical estimates of the rest of the group.
The final result is the statements, the median estimates, and the quartiles.
Note this is not at all like what the group dynamics people call Delphi. The anonymity, the requirement to make numerical responses (even if on a rating scale), and the statistics that take into account the entire range of group opinion, eliminate forced consensus.
I have frequently used RAND Delphi in my profession as a technological forecaster. It is an excellent means for obtaining the opinions of a group of experts, giving both the central tendency of the group, and the degree of disagreement within the group. I usually conduct these studies by mail, as the technique lends itself to that quite readily.
When someone tries to tell you that the group dynamics session you're in was invented at RAND, they're telling you a lie. It's that simple.
What year did your two friends develop that Delphi Technique and why did they choose that particular name for it?
Just curious.
Jen