Collective intelligence in Delphi procedures - an experiment (acronym: KID)

The basic idea behind Delphi techniques is that a group of experts makes more accurate judgments about an epistemic issue than individual experts. This is why Delphi techniques are typically used for foresight or consensus building. The idea behind this comes from the theoretical approach of collective intelligence according to Surowiecki (2004). To date, there is insufficient evidence that the concept of collective intelligence can be meaningfully applied to Delphi techniques, even though some experimental studies suggest that it can. With our research project we want to investigate this assumption experimentally.

Our research questions are:

  1. How close is the aggregated group judgment after three rounds of questioning to a true value that the respondents do not know?
  2. What influence does the respondents‘ inherent expertise have on the accuracy of the aggregated group judgment?
  3. What influence does the feedback have on the accuracy of the aggregated group judgment? 

The project is funded by the University of Education Schwäbisch Gmünd. Ph.D. students Jessica Dieudonné (M.Sc.) and Julia Schifano (M.Sc.) serve as research assistants on this project, under the leadership of Prof. Dr. Marlen Niederberger.

Run time: 2025 to 2026