Tag Archives: Lu Hong

Ethnic Diversity Reduces “Groupthink,” Economic “Bubbles”

Despite progress in raising awareness about implicit bias and stereotypes, most people are less likely to trust recommendations and evaluations from people of different ethnic groups.

Sheen Levine

Sheen Levine

However, this bias may reduce the “herd mentality” that characterized recent price “bubbles” in U.S. housing and global financial markets, reported Columbia’s Sheen S. Levine, Evan P. Apfelbaum of MIT, Goethe University’s Mark Bernard, Texas A&M’s Valerie L. Bartelt, Edward J. Zajac of Northwestern, and University of Warwick’s David Stark.
They concluded that, “Diversity facilitates friction that enhances deliberation and upends conformity.”

Economic “bubbles” occur when the majority of traders, probably influenced by a type of “groupthink,” set inaccurate prices, leading to a mismatch between market prices and true asset values.

Irving Janis

Irving Janis

Groupthink can occur when three conditions interact, according to Yale’s Irving Janis:

  • Group Cohesiveness
    • Deindividuation,” when group cohesiveness becomes more important than individual dissenting views,
  • Group Structure
  • Context
    • Stressful external threats,
    • Recent failures,
    • Decision-making difficulties,
    • Moral dilemmas.
Scott E. Page

Scott E. Page

A mathematical model, developed by University of Michigan’s Scott E. Page and Lu Hong of Loyola University, demonstrated that a wider range of viewpoints leads to less groupthink and more balanced decisions.

Diverse groups ran into fewer “dead ends” in developing solutions than homogenous groups full of smart individuals, who tended to think similarly.

David A. Thomas

David A. Thomas

Likewise, additional experimental evidence by Georgetown’s David A Thomas and Robin J. Ely of Harvard confirmed that identity-diverse groups can outperform homogeneous groups.
Group errors depended on group member ability and member diversity, expressed in the formula:

Collective Accuracy = Average Accuracy + Diversity.

To test the impact of group diversity on market “bubbles,” Levine’s group constructed experimental markets in Singapore and Texas, USA, in which participants traded stocks to earn money.

Evan Apfelbaum

Evan Apfelbaum

More than 175 volunteers with backgrounds in business or finance were randomly-assigned to groups of six ethnically-homogeneous or ethnically- diverse participants.

Traders knew the ethnic composition of their groups, but they couldn’t communicate with each other.
In addition, their “trades” of dividend-paying stock during 10 rounds were anonymous.

Homogeneous groups set inflated selling prices, yet traders in those groups still bought the stock, resulting in increasing stock prices.

Mark Bernard

Mark Bernard

In contrast, traders in diverse groups refused inflated selling prices, so the stock price fell to approximately the price in an “ideal” market with “rational” traders.

When traders and other decision-makers come from similar ethnic, social, and attitudinal backgrounds, they tend to place undue confidence in others’ opinions and decisions, and tend not to subject them to rigorous analysis and scrutiny.

Valerie Bartelt

Valerie Bartelt

As a result, they may be more likely to accept prices and deals that deviate from actual underlying values.
Levine’s group concluded that “homogeneity…imbues people with false confidence in the judgment of coethnics, discouraging them from scrutinizing behavior.”

  • How do you mitigate “groupthink” in organizational decision-making?

RELATED POSTS:

Twitter  @kathrynwelds
Google+
Facebook Notes

©Kathryn Welds

When Do Women Talk More than Men?

-*Are women really more talkative than men?
-*Do women in business meetings not claim as much “talk time” as male colleagues?

Kay Deaux

Kay Deaux

More than 25 years ago, NYU’s Kay Deaux and Brenda Major of University of California Santa Barbara proposed that context and expectations of the individual and others determine when females talk more than males.

Brenda Major

Brenda Major

More recently, participants with digital “sociometers” recorded identities of people nearby and talk volume during a work collaboration project, and during lunchtime social conversations in a study by Harvard’s Jukka-Pekka Onnela and Sebastian Schnorf, with David Lazer of Northeastern and MIT colleagues Benjamin N. Waber and Sandy Pentland.

Jukka-Pekka Onnela

Jukka-Pekka Onnela

During the work project women talked significantly more than men, except when groups included seven or more people.
Larger group size suppressed women’s verbal contributions to the project.
In addition, women sat closer to other women in these groups.

Sebastian Schnorf

Sebastian Schnorf

In contrast, during social conversations, women talked the same amount as men, and even more than men when the group was large.
As a result, group size seems to affect women’s verbal participation in groups depending on the task focus vs. social focus.

Matthias Mehl

Matthias Mehl

This finding supports earlier reports of equal verbal participation by women and men by University of Arizona’s Matthias R. Mehl, collaborating with Simine Vazire of Washington University in St. Louis and University of Connecticut’s Nairán Ramírez-Esparza.
Together with Richard B. Slatcher of Wayne State and University of Texas’s James W. Pennebaker, they analyzed voice recordings from more than 390 participants, and concluded that women and men both spoke about 16,000 words per day.

David Lazer

David Lazer

In addition, large group social settings seemed to enhance women’s verbal participation, in contrast to the opposite effect in collaborative work projects, found Onnela’s team.
The strongest difference in gender participation related to relationship strength and group size.

Scott E. Page

Scott E. Page

Contributions from all members of diverse work groups are required to produce the largest number and most innovative solutions, according to Loyola University’s Lu Hong and Scott E. Page.
They found that diverse work groups produce superior solutions compared with homogenous groups, even if groups were composed of uniformly top performers.

In fact, a group’s “general collective intelligence factor” is most closely associated with:

Anita Wooley Williams

Anita Wooley Williams

This “collective intelligence factor” is not related to the average or maximum individual intelligence of group members, found Carnegie Mellon’s Anita Williams Woolley, Christopher F. Chabris of Union College, with MIT colleagues Sandy Pentland, Nada Hashmi, and Thomas W. Malone.

Diverse groups, including women, can most effectively produce innovative solutions when all participants contribute divergent views.
Women who  consciously increase verbal participation establish visibility and professional credibility, while contributing to improved group performance.

-*How do you determine your degree of verbal contribution in work groups?

Follow-share-like www.kathrynwelds.com and @kathrynwelds


RELATED POSTS:

Twitter  @kathrynwelds
Google+
Facebook Notes

©Kathryn Welds