Delphi Process

The Delphi method implies an iterative consultation of experts. Each participant supplements a questionnaire and receives then a feedback on all his/her answers. In the light of this information, it fills out once again the questionnaire; this time, by explaining each one of its opinions which strongly diverge from those of the other participants. These explanations will prove very useful for the others. Moreover, it can change opinion, according to its own evaluation of new furnished information by the other participants. This process is repeated as much once than it is necessary it, the idea being that the whole of the group can examine dissenting opinions based on privileged or rare information.

Consequently, in the majority of the Delphi processes, the degree of consensus increases with each turn.

Traditionally organized by mail, other alternatives of Delphi methods can take place on line or face to face. In original the Delphi process, the main features of this method were 1) the structuring of flow of information, 2) the feedback given to the participants and 3) the anonymity of the participants. In a Delphi face to face, anonymity disappears. Another alternative: the Delphi Policy, whose main objective consists in exposing all the options and opinions concerning a topic and the main arguments for and against to the regard of these opinions.

So the Delphi process is used to forecast the success of an idea using the opinion of experts. However, how many times experts predicted the success of ideas and they are wrong, what might be the result? Many times ideas are pursued for long period of time and commitment can be made on a unsuccessful one.

Sackman (1974) criticized the Delphi method for being unscientific and Armstrong (1978) criticized it for its inaccuracy (as cited in Illinois Institute of Technology, n.d.). The general complaints against the Delphi method were summarized by Makridakis and Wheelright (1978) in terms of: (a) a low level of reliability in the collective opinion which leads to a dependency on experts selected; (b) ambiguity in the questionnaires can taint the results; and (c) it is difficult to assess the degree of expertise incorporated into the forecast (ibid.). Martino (1978) lists six major concerns about the Delphi method (ibid.):
1. Discounting the future: Future (and past) happenings are not as important as
the current ones, therefore one may have a tendency to discount the future
events.
2. The simplification urge: Experts tend to judge the future of events in
isolation from other developments.
3. Illusory expertise: Some of the experts may be poor forecasters.
4. Sloppy execution: There are many ways to do a poor job.
5. Format bias: It should be recognized that the format of the questionnaire
might be unsuitable to some potential societal participants.
6. Manipulation of Delphi: The responses can be altered by the monitors in the
hope of moving the next round responses in a desired direction.

Ref: http://amynelson.efoliomn.com/uploads/rm502epaper.pdf
http://is.njit.edu/pubs/delphibook/delphibook.pdf