On Group Decision-Making, or E Pluribus Unum

From 2003 through 2005 I lived in Cedar Rapids, IA and frequently drove to our offices in Madison, WI or Des Moines, IA. Since these were regular drives through parts of the country that rapidly become repetitive I looked for ways to occupy myself. At the time downloadable audio books were a relatively new phenomenon, and I became a member of Audible.com where I found unabridged copies of books to listen to while driving. One of the books I remember clearly from that period was James Surowiecki’s The Wisdom of Crowds, a fairly non-academic book about how combining information from large groups of people can sometimes lead to better insights or decisions. I thought about this book while researching group decision-making methods for this discussion board post, as Surowiecki suggests that one approach to group decision-making involves simply aggregating the decisions of individuals. To avoid suggesting that crazed mobs represent group wisdom, Surowiecki identified several criteria which differentiate wise crowds from crowds acting without intelligence (“The Wisdom of Crowds,” 2023). In this post I will summarize the Delphi method for decision-making and will contrast it with the technique for order performance by similarity to ideal solution, or TOPSIS, method.

Similar to Surowiecki’s view of crowd wisdom, the Delphi method is based on the belief that many experts working together in a rigorous process are likely to deliver valuable insights towards solving a complex problem. There are a few key characteristics of this method, including the use of structured group communication, a feedback cycle to disseminate individual contributions among the broader group, an opportunity for the group to judge each perspective, and the opportunity for on-going revision and improvement in the contributions (Turoff & Linstone, 2002). This method uses an iterative approach that is well-suited to predicting future trends, but which can be applied to any complex problem which is not solvable through analytical techniques. In the typical approach to facilitating a group decision using this method, several steps are used which iterate until the group reaches consensus or some predetermined criteria is met. The first step involves each expert receiving a questionnaire on a specific policy or strategy topic related to their expertise. Responses are collected, often anonymously, and then in the second step the think tank researchers analyze and summarize the group’s responses before sharing them with the experts. This allows the experts to consider how the perspectives of others might inform their own thinking, leading to additional iterations of the process of collecting input and summarizing the evolving perspectives, which continues either until a consensus is reached or some stopping criteria is met.

The TOPSIS approach to decision-making attempts to quantify the distance between each considered solution and the visualized ideal solution using mathematical calculations against measured or scaled solution criteria (Shih et al., 2007). While this sounds complex in practice it involves each participant scaling the degree to which a specific proposed solution satisfies the criteria previously agreed by the group to be representative of the ideal solution. For example, in evaluating candidates for an open job the participants would identify the attributes of the ideal candidate such as experience within a specific field, number of published papers, contributions to important discoveries, and so on. Then when each participant reviews candidates they scale the degree to which each candidate matches those idealized attributes, after which the decision-making model identifies the best candidate based on the shortest distance between the ideal and the group rating of attributes.

The Delphi method uses an open ended and iterative approach to solving complex problems which are not easily quantifiable, through the repeated peer review of expert perspectives which are aggregated and combined as appropriate. By contrast TOPSIS seems to be a more closed and analytical approach best-suited for problems which have attributes that can be scaled and ranked by experts. The two approaches should be leveraged for different types of problems, with TOPSIS likely a faster process with explicit decision criteria and Delphi an lengthier process that produces a more comprehensive solution.

 

References

Shih, H.-S., Shyur, H.-J., & Lee, E. S. (2007). An extension of TOPSIS for group decision making. Mathematical and Computer Modelling, 45(7–8), 801–813.

The Wisdom of Crowds. (2023). In Wikipedia. https://en.wikipedia.org/w/index.php?title=The_Wisdom_of_Crowds&oldid=1167044587

Turoff, M., & Linstone, H. A. (2002). The Delphi method-techniques and applications.