How we conducted the Delphi study

The method
The Delphi method is an iterative process through which expert opinions are refined through several rounds of feedbacks. The goal is to show consensus and dissent on a given question. An important aspect is that expert inputs are anonymous. This anonymity is to focus on the message rather than on the messenger to prevent opinion leaders from overtaking the discussion.
For further information consult e.g. pages 41 to 50 in McDonald D, Bammer G, Deane P 2009. Research Integration Using Dialogue Methods. Canberra: ANU E-Press. open access

Number of rounds
This Delphi study consisted of three rounds:
  • In the first round we asked the experts: "Which publications would you recommend as first readings to a (young) scientist that wants to become familiar with the key concepts and thinking in the field of inter- and transdisciplinarity?". The experts were invited to mention up to 5 publications and to explain in 2 to 3 sentences why they recommend the publications. 10 experts answered the first round that lasted from June 4 to July 31 2013.
  • For the second round we assigned the mentioned publications to four categories: A) collections of introductory texts; B) introductory monographs; C) introductory articles and D) introductory texts in the arts. Within each category the publications were listed in alphabetical order and were presented with the comments from the participating experts with no further editing. We then asked the experts to assess all listed publications within each category separately (0: not relevant; 1: somewhat relevant; 2: very relevant; -: I am not able to assess this publication). Experts were also asked to give a short rationale for each of their ratings. 10 experts participated in the second round that lasted from August 15 to October 31 2013.
  • For the third round we averaged the ratings and ranked the publications based on these values within each category. We dropped section D) (introductory texts in the arts) because there were not enough ratings. We coded and grouped the given rationales by qualitative contend analysis. For each publication we collected the pro and contra arguments and counted how often an argument was used. We then asked the experts to reassess the publications taking into account their personal ratings from the second round and the rationales of the other experts. 8 experts participated in the third round that lasted from January 24 to March 15 2014. As two experts from the second round did not participate in the third round their ratings of the second round were included in the final ranking.
The Delphi study was possible thanks to the input of the following 14 experts:
Gabriele Bammer, The Australian National University, Australia
Frédéric Darbellay, Institut Universitaire Kurt Bösch, Switzerland
Florian Dombois, Zurich University of the Arts, Switzerland
Gerd Folkers, Collegium Helveticum, Switzerland
Robert Frodeman, University of North Texas, USA
Gertrude Hirsch-Hadorn, ETH Zürich, Switzerland
Bernard Hubert, French National Institute for Agricultural Research, France
Roderick Lawrence, University of Geneva, Switzerland
Catherine Lyall, University of Edinburgh, Scotland
Merritt Polk ,University of Gothenburg, Sweden
Patricia Rosenfield, Rockefeller Archive Center, USA
Dan Stokols, University of California, USA
Rick Szostak, University of Alberta, Canada
Bianca Vienni, Universidad de la República, Urugway

The td-net team initiated and moderated the processes by designing the rounds, compiling, summarizing and structuring expert inputs and ratings after each feedback round; and the td-net was responsible for editing the final outcome of the Delphi process.

House of Academies
Laupenstrasse 7
P.O. Box
3001 Bern

Tel +41 (0)31 306 93 60
Twitter: @tdnetCH

Contact directly
our collaborators