Sunday, January 12, 2014

Predictions

I recently joined a psychological study, now in its third year, called the Good Judgment Project. The project asks groups of people of diverse backgrounds to make predictions about the the occurrence of various events, mostly in the realm of foreign affairs. This study has already had remarkable success in forecasting the likelihood of various world political and economic events, in contrast to the often dismal record of even highly renowned individual experts and pundits. Similarly, various trading exchanges that allow the public to purchase (with real or imaginary money) interests in various possible outcomes, have also had notable success in accurately predicting election results and such. These kinds of experiments demonstrate something about the wisdom of crowds.

But the Good Judgment Project might demonstrate more than the value of harnessing the wisdom of crowds. In the training materials I had to read to participate in the project, there is a great deal of emphasis on using a collaborative approach to working on problems as a team. The project is designed to foster competition between teams, and even competition among individuals on a team. But team members are also encouraged to share information freely with the team, to rely on the strengths of various members in different topics, to explain the basis for their judgments and comment on the judgments of others, to reveal to others the extent and limitations of each members' knowledge of various topics, and to listen respectfully and take into consideration the diverse opinions within the group. Each team also has a facilitator to assist the team with developing the most productive strategies for answering the questions being posed to all the participants in the study. So far the results of the competition suggest that teams that communicate effectively do better in the competition. Teams that have more internal disagreements may also perform better. Thus, conflict does not impede success. It's how you manage that conflict that matters.

The materials teach that constructive conflict promotes accuracy, while destructive conflict just makes people angry and defensive. Teams that did poorly in the first rounds of the competition, and that created unhappiness among team members, tended to resort to blaming individuals in the group for problems. They also ignored problems in the group, and also allowed negative emotions to fester.

The results so far would seem to suggest that a process in which each competing point of view tries to prove those who disagree wrong, while digging in to their own positions, is going to be less successful in accurately predicting outcomes than a process in which people with divergent points take account of conflicting forecasts, respectfully point out the flaws in one another's analyses, and work together to solve a problem.

Readers of this blog can probably see where I am going with this analogy. Typically, in a litigated or other highly-contested dispute, both sides are highly invested in proving their own positions right and unwilling to credit the other side's point of view. To the extent parties in destructive conflict do listen to the other side, it is mainly with a view toward buttressing their own position so as to counter the other sides' arguments, or constructing ammunition to throw back at them. As a result, neither side is in a very good position to make an accurate prediction about the likely or best resolution of their conflicting positions. Both are apt to over-confidently predict their own sides' success. And although a neutral, whether a judge or mediator, can be brought in to try to make a more objective evaluation after listening to the exaggerations being tossed out by both sides, that neutral still may not be working with the best information available. Research seems to indicate that a more effective process would encourage both sides to work collaboratively to explore the strengths and weaknesses of their initial assumptions. That kind of process will probably make more accurate predictions, and also seems more likely to arrive at solutions that are more fair and more satisfying to both sides.




No comments: