Argumentative Delphi Surveys: Lessons for Sociological Research

Abstract This contribution explains new variants of Argumentative Delphi surveys that can also be used in sociological research, some examples and the learnings from and limits of argumentative surveys with feedback. Argumentative Delphi surveys are not new. As Christian Dayé explains in his book, the early expert surveys and especially the Delphi surveys used explanations and arguments for exchanging knowledge - but always without direct interation (Dayé, C. (2020). Experts, Social Scientistss and Techniques of Prognosis in cold war of America. Socio-Historical Studies of the Social and Human Sciences, palgrave McMilan, Switzerland:41, see also Cuhls, K. (1998). Technikvorausschau in Japan. Ein Rückblick auf 30 Jahre Delphi-Expertenbefragungen. Physica. [Technology Foresight in Japan]). The very first approaches of Delphi surveys did not only make use of expert knowledge in judging issues under uncertainty or were trying to make accurate predictions with statistical analysis, but there were also tests in groups of students. In some of them, the groups did not only choose and tick boxes, but gave reasons or comments for their judgments. Modern Argumentative Delphi surveys do ask for comments AND use a variety of open questions for adding information to the statistical findings. This way of performing a Delphi survey gets more and more ground and can be analysed in a fast way by new means of text mining and Delphi software tools. But they have their limits - especially as they are very demanding for the participants and the analysts. If many people participate, many arguments are given, and they can quickly go beyond the limits of the participants‘ understanding and their time availability. Some lessons learned from recent Delphi projects are reported. This is closing the cycle to Dayé’s description of expert knowledge inclusion in policy-making - a way of integration of expert opinion without direct interaction.