[RUME] UM News - Scott Page and Lu Hong paper published in PNAS (fwd)

[mbrown at umich.edu]

---------- Forwarded Message ----------
Date: Tuesday, November 16, 2004 10:43 AM -0500
From: Howard Oishi <hoishi at umich.edu>
To: Complex Systems Mailing List <Complex.Systems.Mailing.List at umich.edu>
Subject: UM News - Scott Page and Lu Hong paper published in PNAS

Diverse group is the best solution for problem-solving tasks

ANN ARBOR, Mich.---A diverse group of problem solvers is more likely to
outperform a team of the best and brightest problem solvers, a new
University of Michigan study shows.

Individuals chosen from a diverse, randomly selected pool will offer
different perspectives that could result in better solutions.
Conversely, a group comprised of the best problem solvers is likely to
take similar approaches, said Scott Page, a U-M political science,
economics and complex systems professor.

"If the best problem solvers tend to think about a problem similarly,
then it stands to reason that as a group, they may not be very
effective," he said.

Page conducted the research with Lu Hong, a visiting professor in U-M's
Stephen M. Ross School of Business and a faculty member at Loyola
University in Chicago. The Proceedings of the National Academy of
Sciences (PNAS) published the paper this month.

In the study, diversity wasn't necessarily meant to indicate identity
diversity---differences in race, gender, age, or life experiences---but
differences in how problem solvers encode problems and attempt to solve
them. A person's value to solving problems depends on his or her ability
to improve the collective decision, the researchers said.

"A person's expected contribution is contextual, depending on the
perspectives and heuristics of others who work on the problem," said
Page, who is also a senior research scientist at the Institute for
Social Research at Michigan. Heuristics are the variations in how people
encode and search for solutions to problems.

Page and Hong tested their theory using computational and mathematical
models, with each determining the best performance occurred when the
problem-solving group was diverse.

Page noted that the collection of problem solvers could sit in a room
together making a joint decision. These individuals might operate in a
hierarchy, where each person works on a problem and passes his/her
solution on to the next person, he said.

The researchers said the study's results have implications for
organizational forms and management styles, especially for
problem-solving firms and organizations.

"In an environment where competition depends on continuous innovation
and introduction of new products, firms ... that take advantage of the
power of functional diversity should perform well," Hong said.

Page teaches an undergraduate course, "Theories of Diversity," that
focuses on the many implications of diversity. His course touches on
topics ranging from the stability of political systems and ecosystems to
the collective wisdom of crowds.

For more information on Page and the project, visit
http://www.cscs.umich.edu/diversity
Additional information on PNAS can be found at http://intl.pnas.org/


-- 
          Howard Oishi, Research Secretary        Center for the Study of
Complex Systems
               University of Michigan
	    4485 Randall Lab 48109-1120
        Ph. 734/763-3301    Fax 734/763-9267
cscs.umich.edu





---------- End Forwarded Message ----------