The Delphi Technique

Roger Lee
Arkansas State University
EDFN 8773
Doctoral Seminar, Research and
Evaluation
Delphi defined…..
 Delphi defined…..
 An “organized method” for collecting views and
information pertaining to a specific area;
 A method that allows dialogue between
geographically separated experts while serving an
effective means for learning;
 Gathering a group of experts to forecast events
and assess complex issues;
 Collective human intelligence;
 A process of exploring… assessing… and
evaluating.
 Delphi… an historical perspective
 Developed in 1953 by Olaf Helmer &
Norman Dalkey to address military
Delphi Studies
forecasting.
 1960’s – A forecasting instrument in 1200

many American industries & human 1000

services. 800
600
 1970’s – Critics questioned the
400
scientific respectability of Delphi
200
saying it lacked a scientific research
paradigm. 0
1960's 1970's 1980's 1990's
 Early 1980’s – In education alone,
over 441 Delphi dissertations.
 Late 1980’s – an additional 361
dissertations reported.
 Delphi..when is it appropriate?
 In situations where there is  What kind of group
no clear-cut resolution of a communication process
given policy issue; is desirable in order to
explore the problem at
 When time & cost constraints hand?
make frequent face-to-face  Who are the people with
meetings difficult to arrange. expertise on the
 When the heterogeneity of problem & where are
the participants must be they located?
preserved and anonymity  What are the alternative
assured. techniques available?
 Delphi… why use it?
 To construct a 4-year baccalaureate degree program;
 Automotive Technologist Curriculum;
 Design new curriculum in social work, nursing, and medical
education;
 Metal fabrication;
 Literacy campaigns;
 Land use policies;
 Future roles & organizational arrangements of hospitals and other
health services;
 Recreation & Park Administration
 To predict the impact of socio-economic developments on future
school systems, training, and research needs.
Delphi – Basic Findings
 A majority of the applications are in the field
of education;
 Education goals and objectives;
 Curriculum & campus planning;
 Development of educational criteria and policies.
 Output in quantity not matched by quality;
 Research investigating Delphi is low in
relation to its application.
Delphi… potential dangers
 Individuals may misinterpret the Delphi
exercise to be a policy “decision” tool
rather than a policy “analysis” tool;
 Once a Delphi exercise has started
there is no way to guarantee or control
for a specific pattern of response;
 The Delphi has been used as a political
tool rather than an analytical tool.
 Delphi… +’s
 Opportunities for large number of people to
participate;
 Focus is on “ideas” rather than “individuals”;
 Anonymity for participants which make
contributions of ideas a safe activity;
 Opportunities for participants to reconsider their
opinions;
 Allows for identification of priorities.
 Delphi … -’s
 Large amount of time to conduct several rounds;
 The complexity of data analysis;
 The difficulty of maintaining participant enthusiasm throughout
process;
 The problem of keeping statements value free and clearly
defined;
 Self reporting data is subject to respondent’s biases and
memories;
 The bandwagon effect of a majority opinion;
 The power of persuasion or prestigious individuals to shape
group opinion;
 The vulnerability of group dynamics to manipulation;
 The unwillingness of individuals to abandon publicly stated
positions.
Who participates…?
Designing the “expert” team
 A national survey of directors of social
work….
 69 respondents generated 220 names

 a nominee had to receive 5 or more votes

 14 experts were identified

How Consensus is Reached...
 Experts submit a
statement regarding the
issue at hand...
 Participants vote on
statements…
 a statement is retained
if 70% of the panelists
rated the issue “3” or
higher on a scale of 1-5.
How Consensus is Reached...
 Statistical consensus can be
derived by using the quartile
deviation.
 The quartile deviation is
calculated by determining the
quartile range, subtracting the
lowest number, and dividing by 2.
 A quartile deviation of 1.00 or less
indicates consensus by the group.
 The Delphi Method
Delphi Stages Delphi Processes
Round 1: Participants Participants present their
are chosen. Initial data views on the policy.
is gathered.

Round 2: A list of Ideas are synthesized

possible alternatives is and a smaller number of
compiled and distributed possible policy recommenda-
to participants. tions are compiled.

Round 3: An amended These “policy” ideas are fine

list of alternatives is tuned by the participants.
distributed.

Study One collective policy

Results is established.
References...
 Adler, M. & Ziglio, E. (Eds.) (1996). Gazing into the Oracle:
The Delphi Method and its Application to Social Policy and
Public Health. Jessica Kingsley Publishers.
 Bramante, A.C. (1988). Establishing a Basis for the
Development of an Undergraduate Curriculum in Recreation
and Leisure Studies in Brazil: A Delphi Approach. A thesis
submitted for partial fulfillment for Ph.D. degree.
Pennsylvania State University.
 Raskin, M.S. (1994). The Delphi Study in Field Instruction
Revisited: Expert Consensus on Issues and Research
Priorities. Journal of Social Work Education. V30 n1 pp 75-
89.
References - continued...
 Shih, C. & Gamon, J.A. (1997). Using the Delphi
Technique to Assess Educational Needs Related to
Extension’s 4-H Beef Program. Journal of Agricultural
Education. v38 n1 pp 14-20.
 Webber, C.F. (1995). School Board Members’ Perceptions
of the Future: Patterns and Implications. (EA026769)
 Winzenried, A. (1997). Delphi Studies: The Value of
Expert Opinion Bridging the Gap -- Data to Knowledge.
Paper presented at the Annual Conference of the
International Association of School Librarianship.
Vancouver, BC