VI . A . Introduction HAROLD

Cross-impact analysis is only one of the interesting paths being pursued in the exploration of the frontiers of Delphi. In this chapter we consider other significant prospects uncovered in recent research. The problem of "reduction" occurs again and again in Delphi as well as other analysis efforts. How similar are the goals, programs, and objectives? Do these items overlap? Dalkey in his article illustrates the use of "cluster" analysis techniques to narrow the number of unique variables making up the "Quality of Life" as viewed by a Delphi panel. It is a very real problem in many Delphis to be able to reduce the scope of the items the respondent group must consider. One must carefully weigh the effort to be expected of a respondent group and establish a consistency with what actual time the panelists may have to devote to the effort. Techniques such as the one discussed by Dalkey can be of benefit to many applications. Another technique to reduce the necessary range of considerations is that of Multidimensional Scaling. Although this technique is widely used in marketing research, it has yet to see application in a Delphi study. However, since we feel it has tremendous potential for application in Delphi design, we have asked Carroll and Wish to supply a review article on this subject. In essence, this is a method of systematically exposing underlying or hidden considerations (i.e., dimensions) about a set of multiple but related questions. In another sense this technique constitutes the development of a more sophisticated form of cluster analysis. The third article in this chapter cannot `strictly be subsumed under the label Delphi. However, it introduces' a very novel' questionnaire concept of exploration of the future which may well have significant impact on the Delphi communication process. In one way it represents a rather pragmatic demonstration of some of the concepts put forth in Scheele's article: Adelson and Aroni use a set of images in `the form of printed pictures rather than words. The latter are obviously inadequate in communicating holistic insights about complex systems. We fail dismally when we attempt to describe Picasso's "Guernica," Van Gogh's "Starry Night," Ravel's Daphnis and Chloe, or Olivier's Othello in words. Similarly we cannot depict a Soleri city of the future, Skinner's Walden II, Huxley's Brave New World, or Bell's post-industrial society satisfactorily by narratives. Holism is lost in a linear communication system. Pictures are not the ultimate answer; we may need life-size "live-in" models (a la Disneyland) to experience the true meaning of a future alternative. But pictorial images are a step in this direction and there is no justification for ignoring them in the development of Delphi as a communication concept. In the final article John Sutherland constructs a formal process for scenario building which relies strongly on Delphi. Both normative (futures-creative) and