Delphi
The Delphi Technique is an established, accepted method for achieving a broad convergence of opinion from a diverse group of subject specialists who are posed questions on a certain topic area. Developed by the Rand Corporation in the 1950s, the method involves a structured discussion on a specific area with the purpose of guiding the discussion towards a broad consensus. The method seeks to move beyond the limitations of “what is” to begin a conversation of “what could/should be” (Miller, 2006).
The method was also determined by the area of inquiry. The Delphi method was chosen because expert judgment rather than statistical method was appropriate (Jolson & Rossow, 1971). Rowe and Wright (2001) reviewed studies contrasting Delphi against using individual experts, against traditional group consultations and against statistical analysis of survy consultation and concluded that the Delphi results, when properly administered, were 5 times more accurate than traditional groups and 12 times better than statistical surveys. The core features of the Delphi approach are:
- iteration – to try to build a consensus among the experts
- controlled feedback – where only information and opinions relevant the question is fed back to the panellists
- some form of statistical aggregation of their responses – using frequencies rather than probabilities; and
- anonymity – to control for the halo effect due to the differing areas of expertise.