Sunday, February 13, 2011

What is the Delphi Technique?

     The Delphi technique is “a judgemental forecasting procedure for obtaining, exchanging, and developing informed opinion about future events” (Dunn, 2008, p. 180).  More specifically, Linstone and Turoff (2002) describe it as “a method for structuring a groups’ communication process so that the process is effective in allowing a group of individuals as a whole, to deal with a complex problem” (as cited in Vernon, 2009, p. 69).

     The Delphi technique was developed by researchers of the Rand Corporation in 1948 and was "named after Apollo's shrine at Delphi, where Greek oracles sought to foresee the future" (Dunn, 2008, p. 180).  It was originally utilized to forecast problems with military strategies but has since been applied to the areas of: education, healthcare, technology, media, research, technology, etc (Dunn, 2008).

     The Delphi technique is one example of a group of research procedures known as the formal consensus development methods, which are considered when there is limited evidence or when evidence is contradictory in a specific area (Vernon, 2009). Consensus is considered to be an opinion, which is accepted by more than a basic majority, or by those with power (Vernon, 2009). The purpose of these methods is to identify the extent of agreement over a certain issue (consensus measurement) and in doing so, to overcome the disadvantages normally found with group decision making (Vernon, 2009).


The Delphi technique was designed to avoid the following group communication failures:
·         domination of one or more members of the group;
·         pressures to conform to the group’s opinion;
·          personality or interpersonal conflicts; and
·         the difficulty of two opposing individuals of power (Dunn, 2008).

Characteristics of the Delphi Technique

Five basic characteristics of the conventional Delphi technique:
1.      Anonymity: Respondents can remain completely unknown to one another when they are responding to a series of surveys.
2.      Iteration: Delphi’s generally occur over more than one round of surveys, and following each round the responses which have been received are communicated back to the participants.
3.      Controlled feedback: The information received is collated and communicated back to the participants in the form of a collective group response.
4.      Statistical group response: The collated information is statistically summarized for each item under consideration and then is presented back to all of the participants. This allows each participant to compare their personal viewpoint with that of the collective group.
5.      Expert consensus: The idea is to have a panel of experts (who have a level of expertise in the area in question), with the aim of obtaining consensus among the experts. 
      (Dunn, 2008; Vernon, 2009)


     The conventional Delphi technique dominated the field until the late 1960s, at which time the policy Delphi was constructed in response to the limitations of the conventional Delphi (Dunn, 2008).  The conventional Delphi was designed to deal with technical topics and to seek consensus from a panel of homogeneous experts, whereas the policy Delphi seeks to obtain an extensive collection of opposing views on one policy issue in which there are no defined experts, simply a collection of knowledgeable individuals (Dunn, 2008).
   
The policy Delphi is based on two of the above conventional Delphi characteristics (iteration and controlled feedback) however it is also based on the following:
1.    Selective anonymity: Participants remain anonymous only during the initial rounds of the questionnaires, as they are brought together at the end to discuss and debate their opinions.
2.    Informed multiple advocacy: Participant selection is based on interest and knowledge of the area in question, rather than an expertise.
3.    Polarized statistical response: When summarising individual responses, measures are utilized that intentionally ignite disagreement and conflict.
4.    Structured conflict: Conflict is introduced in order to encourage a broader discussion and debate.
5.    Computer conferencing: By having continuous access to computer terminals there is ability for continuous anonymous interactions among participants who are geographically separated.  
    (Dunn, 2008)