Why is group decision-making superior to individual decision-making?
As the proverb says, 'Three heads are better than one,' it is generally believed that when thinking about a particular topic, a better solution is achieved when multiple people think about it together rather than alone. New research has focused on the 'decision-making process of multiple people' and revealed some of the reasons why group decision-making is superior.
Collective problem decomposition improves the wisdom of deliberative crowds | Nature Communications
The Real Reason Group Decisions Beat Solo Ones (Hint: It's Not Averaging) : ScienceAlert
https://www.sciencealert.com/groups-really-do-make-better-decisions-but-not-for-the-reason-you-think
Cases where collective decision-making is superior to individual decision-making have been observed in a wide range of situations, from geopolitical judgments and financial market forecasts to medical diagnoses. However, it is not always the case that collective decision-making is superior in every situation.
Some studies have concluded that the accuracy of estimations on a problem decreases when people influence each other rather than thinking independently. On the other hand, some studies have reported that averaging estimates based on the consensus of small groups consistently yields better results than averaging independent estimates from much larger groups. For this reason, researchers have spent many years trying to understand why collective decision-making sometimes works and sometimes doesn't.
One important question is, 'What happens in the process of a successful discussion?' For example, one method of discussion involves each member of a group bringing their own viewpoint on a problem and then taking the average. Another method involves the group thinking logically about the problem and arriving at a completely new answer that they would not have thought of individually.
In this study, a research team at the University of Torquato di Terra in Argentina hypothesized that 'instead of averaging the initial estimates brought by all members of the group, performing
To test this hypothesis, the research team conducted three experiments involving a total of approximately 900 participants to investigate how small groups reach consensus. In the first experiment, small chat rooms were set up with four people each, and participants were asked to answer numerical estimation problems based on common sense. The participants were not given any instructions on how to conduct the discussion.
Analysis of the discussions and the group's answers revealed that groups that broke down the problem into smaller parts and conducted logical group discussions arrived at more accurate answers than groups that initially only offered individual estimates. Furthermore, in the groups that arrived at more accurate answers, the discussion was not dominated by a single outstanding individual, but rather all members participated in the discussion almost equally.
The second experiment aimed to determine whether it was possible to teach a thinking method of 'breaking down a problem into smaller parts and thinking about them collectively.' The research team divided the participants into two groups: one instructed them to 'share their initial estimates and calculate the average,' and the other instructed them to 'break down the problem into even smaller parts, make estimates for each part, and combine them to arrive at a solution.' The results showed that the group instructed to think logically as a group consistently produced more accurate answers.
In the third experiment, all participants were instructed to use Fermi estimation, but half worked on the problem individually, while the other half worked on it in a group. Analysis of the results showed that while there was indeed improvement at the individual level, more accurate results were obtained when working in a group. These results suggest that sharing and verifying estimates with each other in a group setting provides some kind of benefit that cannot be obtained individually.
Even more interestingly, the research team's analysis of the language used by the groups during their discussions revealed that the most successful group tended to discuss concepts that were more closely related to the problem they were trying to solve.
For example, in a problem asking to estimate the number of steps on a monument, better teams are more likely to use terms like 'stairs,' 'building,' and 'number of floors.' This shows that breaking down the problem into smaller parts makes it easier to think through.
It should be noted that this study only dealt with numerical estimation tasks, so further research is needed to determine whether the same strategy can be applied to different types of group decision-making. However, the research results suggest that dividing problems is a more efficient method for group decision-making.
In their paper, the research team stated, 'This study demonstrates that collective reasoning, particularly reasoning by approximation, underlies the improvement of collective accuracy in deliberations. Providing tools to detect and facilitate such processes will contribute to a fundamental understanding of human collective decision-making and will be useful for practical applications aimed at improving crowdsourcing strategies.'
Related Posts:
in Science, Posted by log1h_ik