Research-In-Progress: Deanna Kennedy’s Super-Groups

By Jasleena Grewal

Imagine you’re undertaking a daunting group project, and your professor assigns your colleagues: Albert Einstein, Oprah Winfrey, and Bill Gates. You and Mr. Gates brainstorm a list of innovative project topics, while Oprah organizes the presentation aspect. All the while, Einstein is fervently and quietly fact checking away, not to get in the way of the charismatic Winfrey-Gates pair.  With an all-star team and a super-powered group dynamic, you start to feel confident about your project’s success.

This week’s Research-In-Progress seminar profiles Deanna Kennedy’s research, entitled “Having the Right People: A Study of Team Composition and Performance.” Her results suggest that such “super-group” fantasies may soon materialize in a classroom or office near you, excluding the celebrities.

Kennedy is an Assistant Professor of Business at the University of Washington, Bothell. She received her PhD in Management Science from the Isenberg School of Management, University of Massachusetts Amherst. Upon moving to the Seattle area, Kennedy paired up with Michael Johnson, a researcher at UW-Seattle, to interview groups of professionals who had participated in collaborative projects. When the interviewees were asked what contributed to project success, Kennedy noticed that the common response was that “[they] just had the right people.” The response led Kennedy to wonder what specific characteristics of group members contributed to a successful project dynamic and outcome. She asked herself, “What’s the secret sauce to making the best project teams?”

With a background in Operations Management, Kennedy realized that employers assign group members based on project cost and members’ availability, but there is a lack of consideration of task performance based on human characteristics. To enhance group performance, Kennedy’s goal was to design a tool which would support team-staffing decisions. The tool would take prior measured characteristics of each team member and weigh them against each other, eventually matching up compatible members. The meat of Kennedy’s research outlines the mathematical machinery of such a tool.

To conduct her research, Kennedy assigned groups of undergraduates a task, wherein performance was evaluated on completion time and quality of work. She conducted the study over the course of four years.

First, Kennedy and her colleagues established that performance depends on diversity of characteristics, as well as the characteristics of the strongest and weakest members of the team. In order to statistically measure performance, each factor was scored on quantifiable terms. The diversity of characteristics was quantified as variance. The characteristics of the strongest and weakest members were quantified as maximum and minimum values, respectively. Characteristics were measured within two pooled categories: knowledge, skills, and abilities; and personality and demographics. Within the knowledge, skills, and abilities category, group members’ college tenure, class grade, and area of expertise were scored and combined. Likewise, within the personality and demographics category, group members’ “Big Five Personality Traits*,” age, and gender were scored and combined. Each score reflected the overall minimum values, maximum values, and variance of each group.

By mathematically organizing group characteristics, Kennedy was able to analyze group performance based on how low scores versus high scores of variance affected performance, which characteristics of the strongest members were advantageous to performance and which characteristics of the weakest members were disadvantageous to performance. Using statistical analysis with a T-Test, Kennedy concluded that groups with high performance had lower variance in grade, age, tenure, gender, neuroticism and extraversion; and they had higher variance in openness, agreeableness and conscientiousness. This result implies that people with similar demographics, but complementary personality traits have high group performance. Furthermore, groups with higher maximum values in neuroticism, extraversion, male presence, and older age were associated with high performance.

To further develop a group assignment tool, Kennedy seeks to explore the implications of mixing and matching groups with different variance, minimum value, and maximum value scores, rather than aggregating groups based on one blanketing criterion. The features of such a tool may be novel, as virtually every professional field relies on team composition to varying degrees. Kennedy quips that military officials assign task groups based on similar characteristics between individuals, and marketing schemes often rely on strong group dynamics.

In an academic scope, the prospect of a computerized team-assigning algorithm would enhance the ability of students to create quality collaborative work, a skill that is emphasized for UW-Bothell students involved in IAS programs. Students’ dread of group projects will be alleviated, and team efficacy will be catalyzed.  The “super-group” concept might be a utopian fantasy of the past, and a reality of the future.

The Research-In-Progress seminars feature the progressive research of UW Bothell faculty. They will be held every Tuesday in the library (LB1-205) at 4pm. T

he next seminar will be held on Nov. 8, featuring collaborative research by Shauna Carlisle, Karen Ericson, Karen Gourd, and Keith Nitta on "Social and Civic Development Through Engaged Scholarship: Building Capacity and Assessing Outcomes."

Students who are interested in research opportunities are encouraged to attend. Beverages and snacks will be provided.

*Big Five Personality traits measure openness, agreeableness, conscientiousness, extraversion, and neuroticism. Source: http://psychology.about.com/od/personalitydevelopment/a/bigfive.htm