Group Exercises

Forsyth: Chapters 11 – 12

Donelson Forsyth

Forsyth, D. (2018).  Group Dynamics (7th ed.). Cengage Learning US.  https://mbsdirect.vitalsource.com/books/9798214344799

Chapter 11. Teams

When the work to be done is difficult, complicated, and important—such as building a bridge, flying a spacecraft to Mars, or performing cardiac surgery—people turn to teams. A teams’ potential depends on the people who the team joins together, but also the dynamic processes that transform those individuals into a task-performing system. Teams create reliable relations among their members, who find that by working collectively they can reach goals that would elude them if they worked alone.

What are teams and what are their various forms?

How does the team’s composition influence its effectiveness?

What processes mediate the input–output relationship?

How effective are teams, and how can they be improved?

Chapter Outline

Working Together in Teams

What Is a Team?

When to Work in Teams

Varieties of Teams

A Systems Model of Teams

Input: Building the Team

The Team Player

Knowledge, Skill, and Ability (KSA)

Diversity

Men, Women, and Teams

Process: Working in Teams

Interlocking Interdependence

Coordinated Interaction

Compelling Purpose

Adaptive Structure

Cohesive Alliance

Output: Team Performance

Evaluating Teams

Suggestions for Using Teams

Chapter Review

Resources

Groups That Work: The Mountain Medical Cardiac Surgery Team

The cardiac surgery team was ready to undertake the most technically challenging of all surgeries: repair of the heart. Only last week they had been using the traditional, open-heart procedure that requires splitting the patient’s chest at the breastbone, stopping the heart and transferring its duties to a heart–lung bypass machine, clamping off arteries and valves, isolating and repairing the damaged portions of the heart, and then closing the eight-inch long wound in the chest. But they would not be using those methods today. Instead, the team would be carrying out a minimally invasive surgical procedure. The surgeon would make a small incision between the patient’s ribs and snake high-tech instruments to the heart, guided by feedback from a network of technicians, computers, cameras, and ultrasound scanners.

These new procedures would make entirely new demands of the surgical team. Traditional surgical teammates work closely with one another, but they are not continually interdependent. The anesthesiologist sedates the patient and monitors his or her breathing. The perfusionist is the technician who operates the heart–lung machine. The surgeon makes the incision, splits the chest, repairs the heart, and then closes the incision. The scrub nurse or technician prepares the sterile field, suctions blood from the site, and passes instruments to the surgeon as needed. The new procedure is not so modularized. The surgeon can no longer see the heart, but must rely on the computer-enhanced images provided by the perfusionist and the anesthesiologist. Because the surgeon cannot apply clamps directly to the heart to stop the flow of blood, that work is done by the anesthesiologist who threads a catheter into the aorta through the femoral vein. The scrub nurse monitors and maintains pressures and vital signs and attaches, when needed, forceps, scissors, scalpels, and other surgical tools to the surgeon’s operating mechanicals.

The new procedures require an unprecedented degree of teamwork, but the Mountain Medical team was ready for the challenge. They had practiced for months to learn the new method, and their diligence showed in their level of coordination and communication in the operating room. The operation took somewhat longer than they had expected it would, but there were no surprises: Their first patient recovered fully but also more quickly because of their use of the minimally invasive, and team-intensive, technique (Healey, Undre, & Vincent, 2006; Pisano, Bohmer, & Edmondson, 2001).

The Mountain Medical surgical team is not the first team we have encountered in our analysis of groups and their dynamics. Peak Search and Rescue, the U.S. Olympic hockey team, the Old Christian rugby team stranded in the Andes, and the crew of Flight 1549 were all teams. These groups created reliable social alliances among the members, and these interpersonal bonds patterned their communications, their influence, and their interdependencies. These groups became more and more unified over time, as they embedded themselves in members’ daily lives and in their identities. These teams included leaders and followers, interpersonal stars and isolates, and hard workers and loafers, and they excelled at the many and varied tasks they attempted.

Those who understand groups are well on their way to understanding teams, but not all groups are teams. Teams require more from the members in the way of collaboration and coordination. Teams are often spawned when one or more individuals confront an obstacle, a problem, or a task they wish to overcome, solve, or complete, but they recognize that the solution is beyond the reach of a single person. Such situations require that members combine their personal energies and resources in such a way that the group, and not just the individuals in the group, reaches its goals. This chapter examines these unique aspects of teams—their nature, design, processes, and effectiveness.

11-1. Working Together in Teams

In the past, most teams were either pulling plows or playing games. Groups assembled for work that required many hands and much muscle, but less physically demanding labor was given over to skilled craftsmen and artisans. Over time, however, the complexity of the tasks that humans undertook grew, and so did their need to work in teams in order to achieve their ends. A very talented person could potentially perform coronary surgery, design a new telecommunication device, create an online database of all knowledge, or pilot a spacecraft to the moon, but such tasks are now done by people working in teams.

11-1a. What Is a Team?

Nowadays the word team is used to describe a wide assortment of groups. In business settings, work units are sometimes referred to as production teams or management teams. At a university, professors and students form research teams to conduct research cooperatively. In the military, small squads of soldiers train as special operations teams. In schools, teaching teams may be responsible for the education of hundreds of students. In online multiplayer games, people join carefully composed teams to attempt challenges that require the skills of many types of characters. Can all these groups be considered teams?

Despite this diversity in terms of focus, composition, and design, teams are fundamentally groups—two or more individuals who are connected by and within social relationships—with the qualities one can expect in any group: boundaries, interaction among members, interdependence, structure, goals, cohesion, and so on (see Chapter 1). But the word team is usually reserved for a particular type of group—one whose members are working together in the pursuit of a shared goal. While no one definition of team will satisfy everyone who studies them, most would agree with team expert Richard Hackman (2002) when he explained the members of teams “work together to produce something—a product, a service, or decision for which members are collectively accountable and whose acceptability is potentially assessable” (p. 42).

Teams usually have other important characteristics, particularly if they are effective teams. For example, the members often have “complementary skills” (Katzenbach & Smith, 2003, p. 45), “different roles and responsibilities” (Kozlowski and Ilgen, 2006, p. 79), and the “authority to manage their own work and internal processes” (Thompson, 2014, p. 4). Many teams, too, have clear boundaries, time pressures, and standards for performance. They are usually part of a larger social organization. Just as groups vary in their degree of “groupness,” some teams—those with higher levels of interdependence and integration of members’ efforts—seem more team-like than others. But the sine qua non of a team is task-oriented interdependence. This specialized form of interdependence even has a name: teamwork. Not all teamwork is good teamwork, for it can range along a continuum from skilled and effective to ungainly and ineffectual, but at minimum teamwork requires two or more individuals trying to get something done by working together instead of separately.

11-1b. When to Work in Teams

Not all tasks require the skills, attentions, and resources of a group of people working in close collaboration. Teams, with their greater resources, goal-focus, and vast potential are becoming the default choice in a variety of performance settings, but some caution is needed before rushing to form a team to solve a problem. Studies of group performance and decision making (see Chapters 10 and 12, respectively) suggest that groups are not all gain without loss. As tasks become more challenging, complex, and consequential, the more likely a well-organized team will succeed where an individual may fail (Salas, Cooke, & Rosen, 2008).

Level of Difficulty

In some circumstances, people are faced with tasks that are well beyond the skills and resources of a single individual. No one person, no matter how talented, can compile a dictionary of all the words in the English language, construct a nuclear power plant, or overthrow a political dictator. Other tasks are difficult ones because they require enormous amounts of time, effort, or strength. One talented individual could build a car or dig a 100-yard-long trench, but a crew of workers will accomplish these tasks far more quickly and with better results. Projects that take months or years to complete are best attempted by multiple individuals, so that the work continues even when specific individuals leave the group.

When Were Teams Invented?

Teams may be de rigueur in most organizations today, but for centuries humans did not work in teams—the word was reserved for harnessed animals. The first documented use of the word team to describe groups of humans working collectively did not occur until the 1600s, when Ben Jonson wrote in Bartholomew Fayre, “Twere like falling into a whole Shire of butter: they had need be a teeme of Dutchmen, should draw him out” (OED Online, 1989). The word team apparently derives from the old English and Norse word “for a bridle and thence to a set of draught animals harnessed together” (Annett & Stanton, 2001, p. 1045).

It was not until the second half of the twentieth century that teams began their ascension to prominence (Sundstrom et al., 2000). In the early 1960s, concerns about the inflexible, autocratic nature of most large organizations prompted a search for alternatives (Likert, 1967; McGregor, 1960). Heeding the call for worker autonomy and participation in decision making, a number of companies began experimenting with employees working in small groups. General Motors, for example, used teams rather than an assembly line in one of its truck factories; General Foods set up autonomous work teams at its Topeka, Kansas plant; the Banner Company, a large manufacturer, set up work groups with varying levels of authority and organizational overlap; and Volvo and Saab both began using teams in their production plants.

From these initial beginnings, organizations began relying on teams for production, management, distribution, and general decision making. Half of the workers in the United States now belong to at least one team at work. Teams are used by a majority of all larger organizations in the United States, and in countries like Sweden and Japan, the use of teams approaches 100% (Devine et al., 1999). Nonprofit organizations, such as health care organizations and public service corporations, are particularly heavy adopters of team approaches to work (81%), followed by blue-collar industries such as construction, manufacturing, and retail sales (50%) and white-collar industries such as banking, real estate, and insurance (34%). In many organizations, employees serve not on one team, but on many; recent estimates suggest between 65% and 95% of knowledge workers are members of more than one team (Maynard et al., 2012). The modern organization is no longer a network of individuals, but rather a network of interconnected teams (DeChurch & Mathieu, 2009).

Complexity and Interdependence

A single person cannot perform Beethoven’s Fifth Symphony or compete against the New York Yankees. Individuals may be able to carry out specific assignments with great skill, but tasks that require the “integrated action of group members” (Shaw, 1981, p. 364) create dependencies among the members. Individuals who agree with such statements as “I have a one-person job; I rarely have to check or work with others” have little need to work in groups on tasks, whereas those who depend on their colleagues to complete their work are more likely to tout the benefits of working collectively rather than individually (Haines & Taggar, 2006). In general, as interdependency increases, group members feel more responsible for successfully completing their work (Hoegl & Gemuenden, 2001).

Importance

Changing a flat tire or reformatting a document pale in importance when compared to rescuing the wildlife harmed by an oil spill or landing a jet airliner whose engines have failed. Important tasks are those that have significant effects on many, rather than a few people, and these effects are long-lasting rather than temporary. Important tasks are also those that concern health, safety, and survival, as well as those that have significant financial consequences. When circumstances are dire and the consequences of a mistake would be catastrophic, the wisdom of a team is preferred to intervention of the lone expert (Cannon-Bowers & Bowers, 2011). Groups may also help individuals deal with the stress of decision making and the consequences of their actions should they fail. Even individuals who question the wisdom of a group’s judgment may turn to groups in an effort to shield themselves from liability should the outcome be a negative one.

The Romance of Teams

Teams, with their greater resources, goal-focus, and promise of increased productivity, are becoming the default choice in a variety of performance settings, but some caution is needed before rushing to form a team. Teams are sometimes used because they are popular, rather than effective or appropriate. Just as the “romance of leadership” describes people’s tendency to put too much faith in their leaders as saviors who will rescue them when they face difficult circumstances, the romance of teams is a “faith in the effectiveness of team-based work that is not supported by, or is even inconsistent with, relevant empirical evidence” (Allen & Hecht, 2004, p. 440). As the industrial/organizational psychologist Edwin Locke and his colleagues (2001, p. 501) put it: “the emphasis on groups and teams has gone far beyond any rational assessment of their practical usefulness. We are in the age of groupomania.”

A team approach may be the best choice in a given situation, but teams perform remarkably badly when given a task to do that is so simple, routine, or individualistic that collaboration is both unnecessary and irritating. Teams should also be avoided when constraints in the situation are such that the basic requirements for a highly effective team (e.g., high rates of interaction, interdependence, and consensual goals) cannot be met.

11-1c. Varieties of Teams

Teams, like groups in general, vary considerably in form and function. Some teams require members to be experts and specialists, but others require less differentiation in terms of members’ skills and experience levels. Some teams meet in face-to-face contexts, whereas others rely on communication technologies to conduct their work. Some teams work for brief periods of time, before disbanding when the project is complete, but others continue their work for many years. Some distinctions, though, can be made between teams based on the type of task their members are attempting (see Figure 11.1). For some teams, their work involves solving problems, planning, and making decisions. Other teams, in contrast, are more action-oriented, for they make products and perform services. (For alternative approaches to classifying teams see Cohen & Bailey, 1997; Devine, 2002; Hollenbeck, Beersma, & Schouten, 2012; Wildman et al., 2012.)

Work Teams

When people think about teams, work teams are probably the groups that first come to mind; these are the millions of groups that provide or produce specialized, and often highly desired, goods and services through their members’ coordinated actions. Mining crews, workers glazing the windows of a skyscraper, and the rock band at a concert are all work teams who, through coordinated actions, create a product. Other work teams, such as the Mountain Medical surgical team and Peak Search and Rescue team, provide services to others.

Management Teams

Many groups carry out both operational and strategic functions within an organization. These management teams identify and solve problems, make decisions about day-to-day operations and production, and set the goals for the organization’s future. They coordinate the actions of others and deal with unexpected issues and difficulties. These groups usually have the authority to make decisions that influence other individuals and groups in the organization.

Project Teams

When groups and organizations encounter problems, issues, or challenges that are out of the ordinary in some way, they often respond by forming a project team. These teams, because they are given a specific task to accomplish, are usually time limited. They do not, in most cases, move on to the next assignment once they reach their goal—instead, they disband. For example, commissions make judgments on specific issues, in situations that require both objectivity and sensitivity to the needs of many parties. Design teams grapple with ill-defined problems—ones that require creativity and innovation rather than those that are routine and have a demonstrably correct solution. An expeditionary team will only exist for the duration of the exploration that it is undertaking.

Advisory Teams

Such groups as review panels, steering committees, investigatory teams, and some judiciary boards are advisory teams, for they study issues and make recommendations to other individuals and groups in the organization. These teams often produce a set of conclusions or an official report that analyzes a specific, and sometimes contentious, issue, and then offer recommendations regarding how things should be done differently. These teams are sometimes called parallel teams because they work in parallel with existing groups and organizational structures (Cohen & Bailey, 1997).

Self-Managing Teams

Most groups require the services of a leader, and teams are no different. But whereas some work best by recognizing one person within the group as the leader, others adopt a distributed leadership structure whereby members share the responsibilities among themselves on a rotating basis. Some teams, too, work under the guidance of an external leader, who may or may not monitor the team’s procedures closely, regulate its activities, and determine its overall purposes (Stewart, 2006).

Hackman (2002) developed his authority matrix model to account for these variations in team autonomy. This model, summarized in Table 11.1, describes (in the left column) four critically important team and executive responsibilities: execution of the task itself, managing the work process, designing the team within the organizational context, and determining the team’s overall mission and objectives. The model also identifies, along the top row of the chart, four types of teams that differ in their degree of responsibility and autonomy. In a manager-led team, members provide a service or generate a product, but that is their sole responsibility: An external leader or manager monitors the work, designs the team, and sets the team’s direction. Members of self-managing teams have more autonomy, for they are charged with both executing the task and managing the team’s work. These teams can gauge the quality of their work product, but only an authority external to the team can adjust its procedures and structures. Self-designing teams enjoy more discretion in terms of control over their team’s structure, for they have the authority to change the team itself. The team’s leader sets the direction, but the members have full responsibility for doing what needs to be done to get the work accomplished. Finally, members of self-governing teams have responsibility for all four of the major functions listed in Table 11.1. They decide what is to be done, structure the team and its context, manage their own performance, and actually carry out the work. The Mountain Medical cardiac surgery team was a self-governing team. The surgeon who founded the team was the one who lobbied the hospital to try the new procedure, and he worked closely with the staff to design the team.

Cross-Functional Teams

Many organizations create connections among different units, groups, and teams within their organizations by forming cross-functional teams. These teams bring together representatives with different backgrounds and responsibilities to reduce the insularity of each segment of the organization and increase the pool of available information needed to make an informed decision. The representatives to such groups function as boundary spanners by building relationships that stretch beyond their own subgroups and divisions.

Cross-functional teams are not the most stable, equable, or effective of groups, however. They provide organizations with the means to break down communication barriers between isolated units and increase collaboration, but in many cases the members of these teams do not have sufficient authority to make decisions for their units. Also, when these teams must identify ways to solve problems, identify ways to reduce costs, or suggest new initiatives, the members’ commitment to the groups they represent prevents them from cooperating fully with their other team members. Since their loyalties lie with the group they represent and not with the team, cross-functional teams are too often unable to reach their goals. As one participant in such a team explained, “Very few of these meetings actually lead to creative problem solving…The result is that the group doesn’t take collective responsibility, and that can be very demotivating” (Denison, Hart, & Kahn, 1996, p. 1012).

What Is Holacracy?

MTSs—multiteam systems—are not some new thing, for rare is the organization that does not rely on the coordinated actions of interdependent teams, particularly in production and the delivery of services. Recently, however, a number of companies have begun to use teams to reorganize their work more fully, shifting from traditional hierarchically organized layers of authority with free-standing divisions, departments, and branches to systems of interconnected groups and subgroups. Their teams go by various names, such as “circles,” “cabals,” “pods,” or just “teams,” but they share one basic quality: they are teams of collaborating teams working together to achieve shared goals (Bernstein et al., 2016).

Zappos, for example, is a billion-dollar online shoe and apparel retailer that is experimenting with a multiteam organizational system called holacracy (Robertson, 2015). Zappos shifted from a traditional, hierarchical structure with a relatively small number of departments responsible for such tasks as customer service, product development, merchandizing, and finance to approximately 500 self-managed and interdependent teams. In holacracy, these teams are called circles, and each one is defined not by the individuals who belong to that circle, but by the roles within that circle. Each role has a specific set of “accountabilities”: activities and responsibilities that the person who occupies that role continually adjusts over time. Each circle includes one role that is responsible for monitoring the connections among the roles—the “lead link”—but the individual who takes on that role has no more authority than any others in the circle. Most people at Zappos occupy multiple roles in the organization and belong to many circles, which overlap and are nested: circles contain subcircles, and supercircles contain any number of smaller circles. Holacracy also sets guidelines for regulating the system, including governance standards, prescriptions for tactical meetings, specifications for defining a role’s responsibilities, and an organizational constitution.

This emphasis on interlocking, multiteam systems balances two desirable organizational necessities that too often compete with each other: the need for reliable execution of responsibilities and the need to adapt to changing circumstances. Too much order and standardization means that organizations can’t respond quickly to new opportunities or deal with unexpected problems, but too little structure can result in very little work getting done. These new organizational strategies, by fully utilizing teams at all levels of the organization, promise to undo that tension between standardization and flexibility. Time will tell if this innovative, team-centered approach to organizational design can deliver on this promise.

Multiteam Systems

Given the prevalence of teams in organizational settings, many teams interact regularly with other teams. When these teams cooperate with one another in the pursuit of common goals, then these teams become “teams of teams,” or multiteam systems (MTSs). Some tasks are not just too demanding for individuals, but too demanding for a single team—and so require the coordinated engagement of multiple teams to complete them. The cardiac surgical team, for example, did not work alone in providing care for patients at Mountain Medical. Dozens of teams staffed the hospital—nursing teams, the recovery room teams, the emergency room crews, the patient management teams, and so on. These teams pursued their own team’s goals, but also goals that were common across all the hospital’s teams (Marks et al., 2005).

As might be expected, the dynamics of MTSs are even more complicated than the dynamics of any one team. The teams in some MTSs, with sufficient experience, learn to work well together—they coordinate their efforts, communicate necessary information across the team boundaries, and respond appropriately to the organizational interventions of team leaders. However, features and processes that promote effectiveness for isolated teams, such as high levels of cohesion, trust, and goal striving, may be dysfunctional when teams must reach across their team’s boundaries to work with the members of other teams. Groups that have developed high levels of coordinated interdependence when members work together may find that they must change their structures and procedures to align their actions with the members of other groups. Social identity processes, which work to increase members’ commitment to their own teams, may also prompt them to view the members of other teams negatively (Shuffler, Jiménez-Rodríguez, & Kramer, 2015).

11-1d. A Systems Model of Teams

When Ludwig von Bertalanffy, the originator of open systems theory, defined a system to be a “complex of mutually interacting components,” he could have just as well have been describing a team. Teams emerge from and then sustain patterns of coordinated interdependencies among individual members. Teams, because of their emphasis on the achievement of desired goals, are more likely than most groups to plan a strategy to enact over a given time period, seek feedback about the effectiveness of the plan and implementation, and make adjustments to procedures and operations on the basis of that analysis. Teams strive for regularity and order, but they also adapt spontaneously to deal with new circumstances or internal changes. Their internal relations are always drifting toward entropy—disorganization—but they counter this tendency through monitoring outcomes and efficiencies (Arrow et al., 2000; Kozlowski et al., 1999).

The Input–Process–Output Model

Rather than assuming that variables in the system are linked to one another in simple, one-to-one relationships, systems theory recognizes factors that set the stage for teamwork (inputs), that facilitate or inhibit the nature of the teamwork (processes), and a variety of consequences that result from the team’s activities (outputs). This assumption is the basis of the well-known input–process–output (IPO) model first introduced in Chapter 2 and shown in Figure 11.2.

Inputs include any antecedent factors that may influence, directly or indirectly, the team members and the team itself. These antecedents include individual-level factors (e.g., who is on the team, and what are their strengths and weaknesses), team-level factors (e.g., how large is the team, and what resources does it control), and contextual- and environmental-level factors (e.g., what is the culture of the larger organization, and how does this team work with other units within the organization).

Processes are operations and activities that mediate the relationship between the input factors and the team’s outcomes. These processes sustain the members’ interdependence, insuring that members’ actions are coordinated and focused on the group’s goals. Teams tend to be well-structured groups, and so members’ interactions are often guided by role requirements, normative standards, and networks of relationships that standardize interactions and productivity. Many team processes also pertain to interpersonal aspects of the team’s system, such as dealing with conflict and increasing the team’s level of cohesion.

Outputs are the consequences of the team’s activities. The team’s emphasis on outcome means that the tangible results of the team effort draw the most attention—did the team win or lose, is the team’s product high in quality or inadequate, did the team successfully complete the operation, or did it kill the patient—but other outcomes are also important, including changes in the team’s cohesiveness or modification of the team’s structure and processes that improve its overall efficiency.

Beyond the IPO Model

The IPO model, despite years of steady service to researchers studying teams, is not without its limitations. First, the model, with its categorization of factors as inputs, processes, or outputs, understates the complex interdependencies among the variables that influence team performance. Second, some of the so-called “processes” within the process category are not actually processes at all, but rather characteristics of the team that emerge over time as members interact with one another. These emergent states certainly influence the team’s outcomes, but it would be more accurate to call them mediators of the relationship between inputs and outputs rather than processes. Third, given that the IPO model is a systems theory, it is essential to always consider feedback processes that occur over time. The model is often interpreted as a sequential one, with inputs leading to processes/mediators and these leading to outcomes; but the reverse causal sequences are also a part of the complete model. In consequence, some suggest that the IPO model should be reconfigured into an input–mediator–output–input (I-M-O-I) model to indicate the diversity of elements in the process stage and the fact that the outputs feedback to become inputs (Ilgen et al., 2005; Marks, Mathieu, & Zaccaro, 2001).

These limitations notwithstanding, the IPO model provides a heuristic framework for this chapter’s examination of teams. The next section considers inputs, with a focus on who is recruited to the team and how their personal qualities shape the team’s interactions. The chapter then turns to issues of process before considering ways to evaluate the effectiveness of teams.

11-2. Input: Building the Team

In 1996, hospitals around the United States began considering adopting noninvasive surgical methods for cardiac surgeries. Technological developments ensured that the procedure was a safe one, but each hospital needed to determine how to change from the traditional method to the newer procedure (Pisano et al., 2001).

Nearly all hospitals settled on a team approach: They would create teams of physicians, nurses, and technicians who would study the method and implement it locally once they had mastered its demands. One hospital, given here the fictitious name of Chelsea Hospital, put the chief of cardiac surgery in charge of building the team. He was an extremely skilled surgeon, but he did not view the new surgery as much of a challenge. He was also very busy and did not get involved in selecting the members of his team. The composition of the Chelsea team was determined by seniority and who was available to attend the three-day offsite training session.

Mountain Medical did things a little differently from Chelsea. A young surgeon, who was new to the hospital, volunteered to get the team started. He talked with the staff in all the departments, and he picked people for the team “based on their experience working together” rather than their seniority (Edmondson, Bohmer, & Pisano, 2001, p. 128). He was part of the team during the training sessions and held meetings with physicians in other departments to share information about the procedure and to identify the best patients for referrals. The members of the team met regularly, prior to the procedure, to walk through the basic steps and to share information about what each of them would be doing and how their actions fit with what the other members of the team were doing.

When organizational behavior researchers Gary Pisano, Richard Bohmer, and Amy Edmondson (2001) studied 16 hospitals that used the new method, they discovered that things worked out differently for Chelsea Hospital and Mountain Medical. The Chelsea team did not lose patients, but the operations took longer than they should have, even after they gained experience with the procedure. Mountain Medical, in contrast, performed the first few operations slowly, but then became one of the fastest and most effective surgical teams in the group of 16 studied—despite being led by one of the least experienced surgeons.

11-2a. The Team Player

Mountain Medical, like most teams, owes much of its success to its composition: the individuals who were selected to make up the team. All teams are composites formed by the joining together of multiple, relatively independent individuals. Each member of the group brings to the team a set of unique personal experiences, interests, skills, abilities, and motivations, that merge together with the personal qualities of all the other individual members to form the team as a whole (Mathieu et al., 2014).

Personality Traits

Mountain Medical deliberately sought out “team players” for their surgical team. Such people are often identified on the basis of their personalities, for some people, by temperament, make better teammates than others. For example, the personality traits identified in the five-factor model of personality—extraversion, agreeableness, conscientiousness, emotional stability (or low neuroticism), and openness—all reliably predict how people respond in group settings. Teams, as task-focused groups, seek and reward the efforts of those who are conscientious; people who are dependable, dispatch their duties, are achievement oriented, and confident are very much appreciated when work demands are unpredictable and success depends on each person completing his or her portion of the group’s task. Extraversion, too, is consistent with a number of desirable qualities in a teammate (dominance, affiliation, social perceptiveness, and expressivity), as is agreeableness. Even emotional stability and openness are likely associated with success working with others, since they are indicators of adjustment, confidence (self-esteem), and flexibility. These general tendencies are, however, partly moderated by the nature of the team and the type of work it does. In a work or project team, for example, conscientiousness may be critically important, particularly if the group is self-managing. In action teams that require higher levels of cohesion, in contrast, agreeableness and emotional stability may be more critical (Driskell et al., 2006; Judge & LePine, 2007).

These team members’ personality traits not only influence the effectiveness of individual members but also the overall performance of teams themselves. Meta-analytic reviews of the relationship between personality and team performance report that groups with members who were, on average, more agreeable and conscientious outperformed teams whose members were less agreeable and conscientious—but only for teams working in professional contexts rather than educational or laboratory settings (Bell, 2007; Peeters et al., 2006; see Figure 11.3 ).

Team Composition

Each member brings to the team a set of unique personal experiences, interests, skills, abilities, and motivations, and those personal qualities will influence how they act as team members. But a group-level analysis considers teams to be composites formed when the personal qualities of all the individual members merge to form the team as a whole. Certain combinations of people, given their personal motivations, are more effective than others (Mathieu et al., 2014). Teams composed of all highly dominant individuals are less stable and less productive than groups that include a mix of people who are dominant and less dominant (Groysberg, Polzer, & Elfenbein, 2011). An individual who is conscientious may fit in well with a team when other members are highly motivated to perform well, but that individual may not fit the requirements of other, less task-focused teams (Bell et al., 2015).

In one illustrative study, researchers examined the complex interplay of personality, group composition, and performance by taking into account the demands of the tasks the teams were performing and the personality traits of each member. As noted in Chapter 10, some tasks require more in the way of coordination than others. Additive tasks, for example, only require summing together each team member’s work to yield a total, group product. Conjunctive tasks, in contrast, require each member contribute to the group’s product, so the group’s performance is determined by the group’s least productive member. When researchers tracked group performance, they discovered increases in team-level extraversion went hand in hand with increases in performance on the additive task, but not the conjunctive task. Performance on the conjunctive task, in contrast was better explained by emotional stability: groups with members who were more emotionally stable did well at conjunctive tasks, but not additive ones (Kramer, Bhave, & Johnson, 2014).

What’s Your Type: A or B?

Studies of personality and health have identified two basic orientations to work: time pressure and productivity. Some of us are very busy people; these Type A personalities tend to be temperamental, competitive, and time-oriented, but they are also high in their achievement orientation. Type B individuals, in contrast, are more relaxed and slow going. To determine what would happen when these two types mixed in teams, researchers experimentally manipulated team memberships to create all Type A teams, all Type B teams, and some teams with a mixture of both types. After they worked together for a time, the members of these teams were asked to indicate their level of satisfaction with their team and its members. In general, people were more satisfied when their teammates were similar in terms of personality. Teams composed of all Type As or all Type Bs were rated as more satisfying by their members than were teams when Type As and Bs were mixed together. Teams of only Type As did, however, get a lot more done (Keinan & Koren, 2002).

Team Orientation

Even though some people value working in an organized, task-focused group, others are quick to announce “I don’t like working in teams” at any opportunity. Individuals who have negative beliefs about groups, for example, generally perform more poorly when they are forced to work in them (Karau & Elsaid, 2009). Individualists, in contrast to collectivists, tend to respond less positively when part of a team (see Chapter 3). Teams with many such individuals in their ranks perform more poorly, as a group, than teams composed of fewer anti-team types (Bell, 2007).

This negativity toward teams is sustained, in part, by two interrelated components: a preference for working alone rather than with others and an unwillingness to accept input from other people. Those who expressed the most negative attitudes toward working in groups agreed strongly with such statements as “I would rather take action on my own than to wait around for others’ input” and “I prefer to complete a task from beginning to end with no assistance from others.” But they also tended to exhibit a low opinion of other people as a source of useful information. They agreed with such statements as “When I have a different opinion than another group member, I usually try to stick with my own opinion” and “When others disagree, it is important to hold one’s own ground and not give in.” Individuals with these beliefs tended to perform poorly on a series of team-based tasks, particularly those that called for high levels of interdependence among the team members (Driskell, Salas, & Hughes, 2010).

11-2b. Knowledge, Skill, and Ability (KSA)

Some teams fail because they simply do not include people with the qualities and characteristics needed for success at the task. A team struggling to generate solutions to math puzzles may not have any mathematicians at the table. A soccer team made up of slow-moving defensive fullbacks but no offensive goal scorers will likely lose. A team’s performance depends, in part, on its members’ knowledge, skills, and abilities (KSAs). Those KSAs are generally of two types: task-relevant proficiencies and interpersonal skills.

Task-Specific Proficiencies

A team of mediocre individuals can, with enough practice, good leadership, and determination, reach lofty goals, but teams cannot work miracles; mediocre members make for mediocre teams (Ellis et al., 2003). Careful design and leadership cannot take a group beyond the limits set by the skills and capabilities of the individual members.

Studies of sport teams indicate that “the best individuals make the best team” (Gill, 1984, p. 325). In many sports, the players’ offensive and defensive performances can be tracked so that their skill levels can be identified accurately. These qualities can then be used to calculate the statistical aggregation of the talent level of the team, which can be compared to the team’s outcomes. Such analyses indicate that the correlation between the aggregation of individual members’ ability and team performance is very strong: .91 in football, .94 in baseball, and .60 in basketball (Jones, 1974; Widmeyer, 1990). The relationship is somewhat reduced in basketball because this sport requires more coordination among members and the teams are smaller in size. Hence, the team members’ ability to play together may have a larger impact on the outcome of a basketball game, whereas the sheer level of ability of players has a greater impact on a football or baseball game’s outcome.

The close connection between members’ proficiency and performance is not limited only to sports teams. Teams that succeed in creating new products and solutions to long-standing problems are generally staffed by individuals of high intelligence, motivation, and energy. For example, investigators who studied high-performance teams in such organizations as Disney Studios, Lockheed, and the Palo Alto Research Center (PARC) traced much of the success of these teams back to their composition. Many of the members of these groups were individually highly motivated: “fueled by an invigorating, completely unrealistic view of what they can accomplish” (Bennis & Biederman, 1997, p. 15). But their most essential characteristics were their talent and expertise.

A team of experts does not always make an expert team, however (Almandoz & Tilcsik, 2016). In one study, researchers had four people work on a task that simulated the work of teams in the intelligence services, detecting possible terrorists from email messages, images taken from surveillance cameras, archives and databases, and so on. Each member of the group reviewed substantial quantities of information individually before meeting collectively to make a team decision. The researchers screened members of the groups, identifying individuals with the particular qualities that would enhance their performance in this kind of situation (good verbal memories and facial recognition). A group with experts in these two areas should have helped their teams perform more effectively—and they did, but only if the teams had spent time organizing their team’s procedures. Groups with two experts in them performed particularly poorly—even worse than groups with no experts—if the teams immediately began their work without discussing how they would collaborate to make a decision (Woolley et al., 2008).

Collective Intelligence

Most high-performance teams specialize: they are more effective when working at one type of task rather than another. The Mountain Medical team, for example, was skilled in the operating room, but their proficiency at this task does not guarantee success when they attempt some other type of task. Similarly, groups in military contexts may be skilled when dealing with hostile forces and protecting the group members from harm, but this same group may prove inadequate when asked to act as a peacekeeping police force. The group may be quite capable when dealing with situations rife with conflict—contests, battles, and competition—but not when the situation requires cooperation and creative problem-solving (Weiss, 2011).

But some groups may be jacks-of-all-trades—skilled at many types of tasks. One team of researchers examined groups’ generalized proficiency—which they called collective intelligence, or c-factor—by assembling groups of two to five members and having them work for up to five hours on tasks from all the quadrants of the McGrath (1984) task model examined in Chapter 1. The groups spent time brainstorming possible uses for a brick, solving problems taken from intelligence tests, discussing an issue that required moral sensitivity, planning a shopping trip, typing a shared Google document, playing a game of checkers against a computerized opponent, generating as many words as possible that started with “s” and ended in “n”, and so on. The researchers then added together the scores on all these different tests to generate each group’s “collective intelligence” (only the morality test was unrelated to all the other tasks). Like individuals, some groups had higher levels of “intelligence” than others, for they were able to master most of the diverse problems they faced. Groups with more women performed better than those with few or none, as did groups with higher average scores on social sensitivity. The best performing groups were also those where members contributed to the tasks more equally (Woolley, Aggarwal, & Malone, 2015; Woolley et al., 2010). More recent work, however, affirms only the strength of the relationship between the intelligence of each member and team’s collective intelligence: teams that have lots of smart members tend to be smarter at the collective level (Bates & Gupta, 2017).

Interpersonal Skills

On the social side, teams function best when the members have sufficient social skills to get along well with other people. Social skills are those basic cognitive and behavioral competencies that allow people to interact with other people in an effective, respectful, and supportive way. They include skill in understanding other people, communicating with them effectively, and responding appropriately during social exchanges. Basic social skills include conversational skill, emotional sensitivity, maintaining self-control, making appropriate self-disclosures, giving praise and encouragement as needed, and expressing agreement (Riggio & Kwong, 2009).

In addition to these basic skills—ones that help people work with others in a range of situations—team members must also understand the unique interpersonal dynamics of teams themselves—they require team knowledge (Stevens & Campion, 1994). Although different teams require different skills of their members, many performance settings reward individuals who are skilled in conflict resolution, can collaborate with others to solve problems, and are good communicators. Conflict resolution KSAs, for example, include the ability to distinguish between harmful and constructive conflicts and an emphasis on integrative dispute resolution skills rather than a confrontational orientation. Collaborative problem-solving KSAs involve skill in using group approaches to decision making. Communication KSAs require a range of finely tuned listening and messaging skills, including the capacity to engage in small talk: “to engage in ritual greetings and small talk, and a recognition of their importance” (Stevens & Campion, 1994, p. 505).

How, for example, would you respond if you found yourself in the following situation: You and your coworkers do not agree about “who should do a very disagreeable, but routine task” (Stevens & Campion, 1999, p. 225)? Should you:

have your supervisor decide, because this would avoid any personal bias?

arrange a rotating schedule so everyone shares the chore?

let the workers who show up earliest choose on a first-come, first-served basis?

randomly assign a person to do the task and not change it?

Or what if you wanted to improve the quality and flow of conversations among the members of the teams. Should you:

use comments that build upon and connect to what others have already said?

set up a specific order for everyone to speak and then follow it?

let team members with more to say determine the direction and topic of conversation?

do all of the above?

According to the Teamwork-KSA Test (Stevens & Campion, 1999), the best choice, in the situation in which you were arguing with others about who must do the unpleasant chore, is option B. In contrast, the best choice in terms of KSAs for interpersonal skill for the second question is option A. An individual who scores well on the Teamwork-KSA Test is more likely to cooperate “with others in the team,” “help other team members accomplish their work,” and talk “to other team members before taking actions that might affect them” (Morgeson, Reider, & Campion, 2005, p. 611).

11-2c. Diversity

The Mountain Medical team was, in some ways, a relatively homogeneous team. Members were similar in terms of ethnicity, skill level, age, motivation, background, and experience with the new procedure. They were, however, heterogeneous with regard to sex, status in the hospital, and training. Would their similarities and dissimilarities influence their team’s processes and performance?

The diversity of a team is determined by the extent to which members are different from one another. A sample of the many ways that people do, in fact, differ from each other is shown in Table 11.2 that identifies six general clusters of differences: social categories, knowledge and skills, values and beliefs, personality, status, and social connections (Mannix & Neale, 2005). Some of these differences pertain to demographic qualities of people, such as race and sex. Others refer to acquired, functional differences between the members, such as variations in experience, knowledge, and skills.

Table 11.2 Categories and Types of Diversity

Categories Types of Diversity

Social-category differences Race, ethnicity, gender, age, religion, sexual orientation, and physical abilities

Differences in knowledge or skills Education, functional knowledge, information, expertise, training, experience, and abilities

Differences in values or beliefs Cultural background, ideological beliefs, and political orientation

Personality differences Cognitive style, affective disposition, and motivational factors

Organizational or community-status differences Tenure or length of service, and title

Differences in social and network ties Work-related ties, friendship ties, community ties, and ingroup membership

SOURCE: Mannix, E. and Neale, M. A. Psychological Science in the Public Interest, 6, 31–55, copyright © 2005 Association for Psychological Science. Reprinted by Permission of SAGE Publications.

Diversity and Team Performance

From a strictly informational perspective, diverse teams should win out against less diverse ones. Diversity brings variety to the team and with that variety should come a broader range of expertise, knowledge, insight, and ideas. If a team is composed of highly similar individuals, they bring the same information and insights to the team so that they are less able to identify new strategies and solutions. A diverse team, in contrast, should maximize performance, particularly in situations where success is not determined by the capacity to apply traditional solutions. For example, the team of researchers who conducted the studies of Mountain Medical was a diverse group, at least in terms of sex and disciplinary training. One had a background in organizational behavior and engineering. Another was an economist and the third a physician. When the researchers first started their investigation, each one expected his or her discipline’s theories and models would explain why some of the cardiac units were more successful in learning the new procedures than others. But their final conclusions were based on the combined insights drawn from the intertwining of all three disciplinary outlooks (Edmondson, 2011).

But diversity has a possible downside. Diversity can also separate members of the team from one another. As social categorization theory suggests (Chapter 3), individuals are quick to categorize other people based on their membership in social groups. Although the members of a team should think of each other as “we” or “us,” when members belong to a variety of social categories, some members of the team may be viewed as “they” and “them.” Diversity may therefore create faultlines within the team, and when the team experiences tension, it may break apart along these divisions (Lau & Murnighan, 1998). As Chapter 5 noted, because people are attracted to those who are similar to them, homogeneous teams tend to be cohesive teams and so members may be more willing to perform the supportive, cooperative actions that are so essential for team success (for reviews, see Jackson & Joshi, 2011; Meyer, 2017; Van Knippenberg & Schippers, 2007).

Studies of Team Diversity

Diversity is a mixed blessing for teams, contributing to gains in performance but at the same time adding the potential for process loss. Diversity, when based on information and expertise, tends to improve team outcomes, particularly on difficult tasks. When members vary in ability, then by definition the team will include at least one individual with high ability. Some homogeneous teams will be uniformly unskilled, so these teams will perform particularly badly at their task. As studies of social compensation discussed in Chapter 10 suggest, heterogeneous teams may also become more productive because the low-performing members are motivated by the high standards set by the others in the team, and the others in the team may also be a source of help and assistance as the low performers work to increase their performance.

But other types of diversity, such as variations in ethnicity, race, age, and sex, influence performance less reliably. Teams of researchers are more productive when they join with researchers from other scientific fields, but not always: in some cases interdisciplinary research teams never agree on the goals they are seeking or the means they will take to reach those goals. Diverse top management teams navigate challenges more successfully than teams that are very homogenous, but these teams also tend to experience heighten levels of conflict and turnover. Studies of various types of teams making decisions and solving problems, such as project teams, policy advisors, and even students working to solve problems in research labs, are more innovative when they include members with different backgrounds and orientations—but not when working on very challenging problems. When researchers weighed the relative costs and gains of team diversity in a meta-analytic review of 108 studies of over 10,000 teams, they discovered that process gains associated with diverse groups were offset, to a degree, by process loss (Stahl et al., 2010). Conflict was greater in diverse groups, and social integration was reduced. However, diverse groups were more creative than less diverse ones, and—somewhat unexpectedly—people enjoyed membership more in diverse groups compared to less diverse ones.

Designing for Diversity

These conflicting findings attest to the mixed benefits and limitations offered by diversity in teams. Diverse teams may be better at coping with changing work conditions because their wider range of talents and traits enhances their flexibility. Diverse teams, however, may lack cohesion because members may perceive one another as dissimilar. Heterogeneity may increase conflict within the team (Mannix & Neale, 2005; van Knippenberg & Schippers, 2007; Williams & O’Reilly, 1998).

Steps can, however, be taken to minimize the negative side effects of diversity and maximize diversity’s gains. First, diverse teams will need time to work through the initial period in which differences between people based on their surface-level qualities—race, sex, and age—lower the team’s overall level of cohesiveness. Intervention may also be required when, after time, members have discovered that these surface-level differences are unimportant, but that their deep-level differences in values and principles are causing unexpected turbulence in the team (Harrison et al., 2002). Second, because teams exist in an organizational context, the nature of that organization’s culture will influence how teammates respond to diversity. If the organization’s culture encourages collectivistic values and minimizes distinctions based on tenure and status, then diverse teammates tend to behave more cooperatively than they would in more traditional organizations (Chatman & Spataro, 2005). Third, to minimize conflict between team members from different social categories, steps should be taken to minimize any tendency to draw distinctions between people based on their category memberships (Cunningham, 2005; Homan et al., 2010). Team leaders should remind members of the importance of involving all members of the team in the process and make certain that individuals in the minority do not become isolated from the rest of the team (see Chapter 7‘s analysis of minority influence).

11-2d. Men, Women, and Teams

Same-sex teams are becoming increasingly anachronistic. Whereas women were once barred from many types of teams in business and organizational settings, changes in the social climate—and in employment law—have increased sex-based diversity in the workforce.

The Myth of Male Bonding

These changes are not welcomed as progress in all quarters of society or recognized as adaptive by all theories of collective action. Some evolutionary anthropologists, for example, argue that the presence of women in previously all-male teams may disrupt the functioning of such teams in substantial ways. This perspective suggests that it was males, and not females, who affiliated in same-sex groups for adaptive reasons, so that over time male bonding became a stronger psychological force than female bonding. In consequence, heterogeneously gendered teams may be less productive than same-sex teams, since all-male teams would be more cohesive than mixed-sex teams. Bonding theorists also suggest “the difficulty females experience in male work groups is not that males dislike females but rather that the force of their enthusiasm for females can disrupt the work and endanger the integrity of groups of men” (Tiger & Fox, 1998, p. 145).

How Productive Are Online “Crowds”?

At this moment, teams are working on thousands of tasks, but these teammates will never meet each other: They are scattered around the world, linked by an online network of distributed production called crowdsourcing. This term combines both the idea of a crowd—a large and often dispersed group of people—with the idea of outsourcing—assigning a task to individuals outside of the group or organization. Crowdsourcing, as used by businesses, harnesses the great diversity of Internet users around the world to review, evaluate, create, develop, and even market new products, procedures, and applications. Wikipedia, for example, is a crowdsourced resource, for this online encyclopedia is edited by volunteers who continually upgrade and monitor the entries. The open source software program Linux also makes use of crowdsourcing, as programmers work on its code and applications voluntarily and without direct compensation. Uniting all these forms of crowdsourcing is the reliance on a great variety of people—diverse in terms of expertize, training, and skill—to perform group-level tasks.

Crowdsourced products, even though they are generated by large numbers of relatively anonymous and independent individuals, often rival the work of dedicated (and often highly paid) experts and professionals. One team of researchers, seeking to classify a large set of documents for a research project, turned to the workers of the Mechanical Turk for assistance. This service links requesters—those who are seeking people to perform human intelligence tasks (HITs)—with people who are interested in performing these tasks (turkers) in an online network. Turkers are paid for their time, if the work is adequate, although the amount paid is usually quite small—in this study, for example, two cents per HIT. The investigators provided the turkers with a statement of the criteria for rating the documents and a document to review and rate on a scale from 0 (not relevant) to 3 (excellent). When the ratings were returned from the hundreds of turkers who did the work, the researchers just averaged the ratings and compared these ratings to those provided by experts. In all but four cases, the experts and turkers agreed, but when the researchers reviewed those four, they concluded that the turkers were right on three of them (Alonso & Mizzaro, 2009). The crowd was wiser than experts.

The data do not support either the idea that males bond more cohesively in all-male groups than females bond in all-female groups or that, in consequence, male teams outperform female teams. Social psychologist Wendy Wood (1987), after reviewing 52 studies of sex differences in group performance, identified two factors that influence the effectiveness of all-male and all-female teams—task content and interaction style. First, in the studies that favored men, the content of the task was more consistent with the typical skills, interests, and abilities of men than of women. Groups of men were better at tasks that required math or physical strength, whereas women excelled on verbal tasks. Second, Wood suggested that sex differences in performance are influenced by the different interaction styles that men and women often adopt in groups. Men more frequently enact a task-oriented interaction style, whereas women tend to enact an interpersonal-oriented interaction style. Thus, men outperform women (to a small extent) when success is predicated on a high rate of task activity, and women outperform men when success depends on a high level of social activity (Mendelberg & Karpowitz, 2016).

Heterogeneously Gendered Teams

But what of mixed-gender teams—teams that include both men and women? Studies of men and women working together in teams suggest that such teams, because of their diversity, have greater information resources than same-sex teams and so excel at tasks that require a broad range of expertise, experience, and information. However, sexism, sexual harassment, and stereotyping continue to dog such teams (Bell et al., 2015). As with other forms of diversity, sex-based diversity can create subgroups within the team and increase levels of conflict. Diverse teams must also deal with problems of proportion, particularly when very few men are entering into groups that were traditionally staffed by women and vice versa. Teams that achieve diversity by adding only one or two members of a social category, such as a team with one woman and many men, tend to encounter more problems than homogeneous teams. When work groups include a single token or “solo” woman, for example, coworkers are more likely to categorize each other in terms of their sex. Solo members are also scrutinized more than other group members, and this unwanted attention may be emotionally depleting and may contribute to stereotype threat (Johnson & Richeson, 2009). Token members are more often targets of sexism and prejudice and must, in many cases, work harder and express higher levels of commitment to the group to overcome other members’ biases (see Chapter 8 ).

In some cases, teams with token members will outperform homogeneous teams, even when the teams attempt tasks that are traditionally reserved for homogeneous teams. For example, one team of researchers watched groups working on a wilderness survival exercise—an activity that favors people who have knowledge of the outdoors. Groups of men generally outperformed women, but groups of men that included one woman performed best of all. The researchers speculated that the addition of a woman to the otherwise all-male groups may have tempered the men’s tendency to compete with one another and, thus, helped them to function as a team (Rogelberg & Rumery, 1996). Other research, however, confirms one of the speculations offered by the proponents of male-bonding theory: Men exhibit impaired cognitive functioning when working in the presence of the opposite sex—women do not (Karremans et al., 2009). In fact, only men show declines in cognitive ability (they make more errors when taking the Stroop test) when they are led to believe that they are being observed by a member of the opposite sex (Nauts et al., 2012).

Hackman and his colleagues have explored the complex relationships of gender diversity, the proportion of men and women, and the organizational context in their studies of a particular type of team: the concert orchestra (Allmendinger, Hackman, & Lehman, 1996; Hackman, 2003). As noted in Chapter 2, these orchestras were in the midst of a transition from all-male groups to groups that included both men and women. Some orchestras were only beginning this transition, for they included very few women (2% was the lowest), whereas others were more heterogeneous (up to 59% women). When they measured members’ work motivation and overall satisfaction with their orchestras, they discovered that orchestras with a larger proportion of female members were viewed more negatively. This tendency was more pronounced among the men in the group and also in countries with traditional conceptions of the role of men and women in society. Hackman wrote:

Life in a homogeneously male orchestra surely is not much affected by the presence of one or two women, especially if they play a gendered instrument such as a harp. Larger numbers of women, however, can become a worrisome presence on high-status turf that previously had been an exclusively male province, engendering intergroup conflicts that stress all players and disrupt the social dynamics of the orchestra. (2003, p. 908)

11-3. Process: Working in Teams

The members of the Mountain Medical surgical group had the experience, skills, ability, and drive needed to function as a highly effective team, but they needed to combine these raw materials expeditiously to maximize their team’s performance. A group of people who join together to get something done may qualify as a team, but will they be an effective team? Although researchers have identified dozens of processes that work to transform team inputs into performance outputs, here we consider five that are noted with regularity across most models of highly effective teams: reliable interdependence, coordinated thought and action, a compelling purpose, adaptive structures, and cohesion (Hackman & Katz, 2010; Kozlowski & Ilgen, 2006; Marks et al., 2001; Salas et al., 2015).

11-3a. Interlocking Interdependence

Chelsea Hospital and Mountain Medical both faced the same problem, and they both decided to solve the problem by forming a team. But they designed their teams differently. Both teams included a scrub nurse, a perfusionist, an anesthesiologist, and a cardiac surgeon, and each trained carefully that they were skilled at the tasks they needed to perform. But the leaders of the two teams had different views about how they should work together. Chelsea’s head surgeon believed that the team’s members should focus on their own subtasks: as he explained, “Once I get the team set up, I never look up [from the operating field].” The young surgeon who headed the team at Mountain Medical, in contrast, stressed the team members’ connections to one another: “you really do have to change what you’re doing [during an operation] based on a suggestion from someone else on the team. This is a complete restructuring of the [operating room] and how it works” (Edmondson et al., 2001, p. 128). He asked that during the operation everyone communicate with everyone else and not focus on only their own duties.

All group members are interdependent to a degree, but members of teams are so tightly coupled that no member can determine his or her own outcome. As Figure 1.3 in Chapter 1 suggested, groups create many types of interdependencies among members. For example, a team on an assembly line may pass work from one person to the next in a fixed order, so the level of interdependence is equal and sequential. In other teams, however, all members interact jointly to complete their tasks, so they are fully rather than partially interdependent, with members reliably and substantially influencing one another’s outcomes over a long period of time and in predictable ways (Saavedra, Earley, & van Dyne, 1993). In a traditional operating room, for example, surgeons have been known to control the entire operation, so the outcome was largely determined by the surgeon’s skills rather than the effectiveness of the “support staff.” In the Mountain Medical team, in contrast, the contributions of each member of the team were vital to the successful outcome. The interdependencies in a team tend to be equal and reciprocal rather than asymmetric and unequal (see Figure 1.3 ).

Shared Mental Models

Physicians, before undertaking an operation, have a clear understanding of the complexities of what they are about to do. Each one has a mental model that serves as a cognitive representation of all they know about the case, the procedure, and the challenges. But when physicians are no longer working alone, but as part of a team with others, their mental model must be shared with those others. Because of differences in prior experiences, knowledge, expectations, and so on, each team member may have a differing view of the case, the procedures, and even their specific duties as a part of the team. Some of these differences may lead to misunderstandings and inefficiencies as the team does its work, so the emergence of social sharedness (Tindale, Talbot, & Martinez, 2013), a shared mental model (Cannon-Bowers & Salas, 2001), or a collective problem orientation (Bonner, Soderberg, & Romney, 2016)—will facilitate the group’s functioning.

Some semblance of this shared mental model is present nearly from the team’s inception, but as the team practices, differences among the members in terms of their understanding of their situation and their team diminish (Tindale, Stawiski, & Jacobs, 2008). In one study of this process, members of groups completed two geography quizzes about U.S. cities, with such questions as “What city is known as the Crescent City?” and “Through what city does the Trinity River run?” Unbeknown to the group, one of their members was a confederate who had been prepped with the answers, and he answered seven of the eight questions correctly on the first test. The group used some of his answers (60.3%) on the first test, but when they were given feedback and a chance to do a second quiz, they used his answers almost exclusively (84.7%). They had learned to rely on his expertise (Littlepage, Robison, & Reddington, 1997; see, too, Littlepage et al., 2008; Littlepage & Silbiger, 1992).

Transactive Memory

Teams also need time to develop transactive memory systems (Wegner, 1987). In the complex world of the operating room during heart surgery, there is too much information about the equipment, the proper settings, the instruments, the heart–lung machine, and so on, for a single individual to retain it all with any degree of accuracy. The surgical team therefore distributes the information to specific members of the team, depending on their role and responsibilities. Then, when the information is required, the team consults with the member known to be the “expert” on that particular matter who then supplies the necessary information to the best of his or her ability (see, too, Chapter 12‘s discussion of collective memory).

Social psychologist Richard Moreland and his colleagues (Moreland, Argote, & Krishnan, 1996) examined the development of transactive memory systems by training volunteers to build radios from hobby kits. Each kit included a circuit board and dozens of components that had to be put in the correct locations and connected before the radio would function. All the participants received the same training in the first session, but some of them worked alone practicing building the radio whereas others practiced in three-person teams. One week later, the participants returned and assembled a radio, this time with an offer of a cash prize if they performed well. All the subjects worked in teams, but only some of them were assigned to the same team they had worked with originally. These individuals outperformed the subjects who were trained individually, apparently because they were able to form a collaborative, transactive memory for the procedures in the first session. Moreland and his colleagues discovered that teams that performed the best showed signs of

(a)

memory differentiation—some of the team members were better at remembering certain parts of the assembly procedures than others;

(b)

task coordination—the team-trained teams worked with less confusion; and

(c)

task credibility—the teams with stronger transactive memories trusted one another’s claims about the assembly process.

11-3b. Coordinated Interaction

Before Mountain Medical carried out its first surgery the members of the team had already worked, for weeks, as a team. They met regularly to discuss the procedure, and all had trained together for three days offsite in a simulated operation procedure. They had discussed the sequence of steps that would begin with an anesthetized patient and end with a repaired heart, so that when it was time to work together, they functioned as a team.

Teams do things as single, coordinated units. The members of all groups engage in a mixture of task and relational interactions, but in teams the interaction rate is higher and the flow more continuous. In addition, teams are work-focused groups, so a greater proportion of their interactions pertains to group tasks: monitoring of progress, improving coordination, structuring the work process, assisting one another, strategizing, and so on (Aubé & Rousseau, 2005). Members of teams are attentive to each other’s interpersonal needs—they continuously maintain, build, and even question the quality of their social connections—but they spend the bulk of their time on their work. When researchers used the Bales (1999) Interaction Process Analysis to record the communication among team members, they discovered relational statements were used only rarely. In the groups Bales studied, nearly 40% of the commentary was classified as relational, whereas teams made relational comments (e.g., shows solidarity, shows tension) only 11% of the time (Gorse & Emmitt, 2009).

Organizational experts Michelle Marks, John Mathieu, and Stephen Zaccaro (2001), in their analysis of coordination in teams, identify three processes that high-functioning teams display as they work: transitioning, acting, and managing interpersonal relations among members. During the initial phase of their work, teams plan what they will do in later stages, set their goals, and plan strategy. The group then transitions to the actual action stage when it carries out its assigned tasks through coordinated activity. Once this action phase is completed, the team reenters the transition phase and begins preparing for subsequent tasks. Across all phases, the members are also managing the interpersonal aspects of the team in order to minimize conflict and maximize coordination. Thus, as Figure 11.4 indicates, Marks and her associates break teamwork down into three fundamental components: action processes, transition processes, and interpersonal processes (see, too, LePine et al., 2008).

Figure 11.4

Details

SOURCE: “A Temporally Based Framework and Taxonomy of Team Processes,” by Michelle A. Marks, John E. Mathieu, and Stephen J. Zaccaro (2001). Academy Of Management Review, 26, 356–376. Reprinted by permission of Academy of Management via Copyright Clearanc Center.

The teamwork process model.

Action Processes

When teams are at work, their task-related actions are so perceptually vivid that the action processes that make up the teamwork portion of their activities often go undetected. When, for example, the Mountain Medical team began to repair the patient’s heart, an observer watching the team would see a physician incising and suturing, a nurse monitoring the patient’s vital signs, and an anesthesiologist sedating the patient. But Marks, Mathieu, and Zaccaro’s teamwork process model suggests that four other, teamwork-related actions are also taking place during the action period. First, the group is monitoring progress toward its goals as members implicitly check their own actions as well as those performed by others. Second, systems monitoring involves keeping track of the resources the team needs, whether they be physical resources, time, or even energy. Third, team monitoring and backup behavior, considered by some to be a key difference between teams and task groups, occurs when one member of the team delivers assistance to another member, simply because that team member needs help. Finally, coordination of action involves a change in the behaviors of the team members so that each one’s actions mesh with other’s actions, resulting in synchrony.

Transition Processes

Often, teams attempt tasks that are so complex that they cannot be completed, at least with any degree of success, without advance planning. The first type of transition process, mission analysis, focuses on the current situation: the tasks and subtasks that must be completed, the resources available to the team, and any environmental conditions that may influence the team’s work. Teams also engage in goal specification and strategy formulation between action episodes, since experience working together will provide the members with a clearer idea of the team’s potential and limitations. Strategy formulation is particularly essential if the team is unable to reach the goals it has set for itself, for, by reviewing the causes of failure, team members may find ways to improve their efficiency and outcomes (Cannon & Edmondson, 2005).

Interpersonal Processes

Consistent with studies of work groups in general, during both the transition and action periods, teammates must spend some of their time tending to the relational side of their team. To reach a high level of effectiveness, teams require a degree of unity; yet the pressures often encountered by groups as they strive to reach their goals can produce tension within the group. Members of effective teams tend to reduce the threat of such conflict to the group’s cohesion through conflict management. Other types of interpersonal work required of the group members include motivation and confidence building and affect management.

11-3c. Compelling Purpose

The sine qua non of teams is their pursuit of goals—and collective ones at that. Teams have been defined in many different ways, but nearly all definitions suggest teams “work toward shared and valued goals” (Salas et al., 2009, p. 39); they seek a “common purpose” (Hackman, 2011, p. 51); and team members are “committed to a common purpose, performance goals, and approach for which they hold themselves accountable” (Katzenbach & Smith, 2001, p. 7). All these definitions stress the consensual nature of a team’s goals: members share an understanding of the group’s recognized purpose and—in effective teams, at any rate—willingly contribute their time and energy in the group’s pursuit of its goals. Shared goals increase coordination within the group and reduce the tendency for members to work at cross-purposes to one another.

Team expert Richard Hackman (2002, 2011) stresses the importance of goals in his real teams model. Real teams, he suggests, embrace shared goals that guide the work of the group and heighten members’ motivation. According to Hackman, a team’s purposes should be clear, challenging, and consequential, but not overly specified, impossibly difficult, or so daunting that team members are motivated by a fear of failure. Teams stress outcomes to such an extent that their very existence is threatened should they fail to achieve their agreed-upon goals. Team members’ high level of interdependence, combined with the team’s pursuit of consensual goals, means that the members of a team cannot succeed unless their group succeeds. The members of a team may strive to outperform each other or achieve personally important goals, but each member’s outcomes are tied to the team’s outcomes, such that if the team is successful, so are the individual members. But, should the team fail, the members do as well.

11-3d. Adaptive Structures

Teams—effective ones, at least—are usually well organized, for their roles, norms, and intermember relations are defined rather than nebulous. The members of a baseball team, for example, play different positions, and the actions of each player are determined, not only by the skills and abilities of the person who occupies that position, but also by the standards that define what a person in that role should and should not do: the pitcher will pitch, the outfielders will catch fly balls, the infielders will cover the bases and field grounders, and so on. Similarly, in the Mountain Medical surgical team, each member played a specific role in the operation, and the outcome of the surgery depended on each team member meeting the demands of his or her role. These structures are adaptive, for they serve to improve the degree of coordination of the members’ actions.

Other structural elements, including performance norms, status, communication networks, and so on, also tend to be better defined in teams than in more informal types of groups. The actions and interactions of the team members are governed by implicit and explicit social norms that prescribe the appropriate way to respond in the situation as well as proscribed actions members should avoid if at all possible. These norms are what transform an aggregate of individuals into hardworking, engaged team members who are committed to the group’s goals. Run-of-the-mill teams become high-performance teams when team norms encourage mutual respect and support, open communication, commitment to effectiveness, and high levels of engagement in the work (Katzenbach & Smith, 2003).

Teams, although they tend to develop formalized procedures and a fixed division of labor, are not miniature bureaucracies. Bureaucracies are highly organized systems, but they tend to become inefficient as structures that are no longer needed are retained, and changes that would improve the system are not implemented. Teams, in contrast, continuously review and revise their structures to improve their functioning. Researchers confirmed the relationship between adaptive team structures and performance by first calibrating the extent to which members of one hundred teams working for a Fortune 100 technology firm agreed when describing their teams’ degree of role specialization, responsibilities, and work scheduling. They then asked outside observers to rate each team’s capacity for self-improvement. Supporting the adaptive function of structure, they found that teams with clearer structures were more, rather than less, likely to “continually look for more efficient ways to accomplish our assigned tasks” and “learn from one another as we do our individual jobs” (Bunderson & Boumgarden, 2010, p. 616). Ratings of the team’s structure were also highly correlated with level of information sharing among members, absence of conflict in the team, and members’ sense of psychological safety.

11-3e. Cohesive Alliance

Teams are cohesive groups: The relations linking members to the group are strong, rather than weak and the group tends to remain intact over time and in difficult circumstances. Teams are united in their pursuit of a common goal, so a team’s unity usually springs from its task cohesion, but teams may exhibit one or more of the other forms of cohesion considered in Chapter 5. Teammates are often socially and emotionally close to each other (social cohesion), they strongly identify with their team (collective cohesion), and they are affectively bonded (emotional cohesion). The team’s dense network of interdependencies, combined with its stability in membership and clear boundaries, may also heighten its structural cohesion. External pressures may magnify this unity, for teams usually work under some kind of pressure, such as a heavy workload, limited time, or competition with other teams.

Cohesion’s Benefits

An increase in a team’s level of cohesion results in both affective (emotional) and instrumental gains for the individual members and the group as a whole. Working in a cohesive team is, at core, more satisfying for members—the experience is more enjoyable, the relationships more positive, and the satisfactions gained from success are more pronounced. If, as many have argued, people have a strong need to belong to groups, membership in a cohesive, successful team will likely satisfy that need. But cohesion also yields practical advantages for teams. Cohesive groups retain their members, and so one of the core requirements for group efficiency—a predictable roster of membership—is more likely in cohesive teams than less cohesive teams. Conflict, although inevitable, is usually managed more effectively by the members of cohesive teams, and so results in fewer long-term negative consequences. As noted in Chapter 5, when a group lacks cohesion, the members are less likely to act in a coordinated, efficient way: members’ actions will not mesh with those of other members. If the group’s tasks do not require synchronization and coordination, a lack of cohesion may not undermine performance, but as coordination demands increase, so does the need for cohesion (Severt & Estrada, 2015).

Is Your Team Psychologically Safe?

Google, the fabulously successful technology company, has always relied on teams to maintain its effectiveness. So, when the company wanted to know more about why some teams were more effective than others, it initiated its own project—code-named Aristotle—and began collecting all sorts of data from hundreds of Google teams. Initially, they expected to find the answer in the composition of the groups—the talents of each individual members—but the data told a different story. The keys to success in Google groups were the norms that took hold in each team. But what mattered most was not so much the content of the norms, but their clarity: when all agreed on the norms of the group, these norms sustained the group’s workflow, creativity, and commitment, and the teams tended to prosper (Duhigg, 2016).

Clear norms also contributed to what team researcher Amy Edmondson (1999) calls a sense of psychological safety: “people’s perceptions of the consequences of taking interpersonal risks in a particular context such as a workplace” (Edmondson & Lei, 2014, p. 23) In some teams, members felt insecure and uncertain; they were so constrained that they did not speak up and express concerns and dissatisfactions. In other groups, in contrast, the members felt safe, that their voices were heard, and that others were concerned for their well-being.

Instructions. Focus on a specific team or group to which you belong (a club, class, organization, work group, or even your family). Put a check by those statements with which you agree.

Members of this team are able to bring up problems and tough issues.

It is safe to take risks on this team.

No one on this team would deliberately act in a way that would undermine my efforts.

If you make a mistake on this team, it is often held against you.

It is difficult to ask other members of this team for help.

Scoring. Do you feel psychologically safe in this group or team? If you agreed with the first three items, then the level of psychological safety in your group is high. But if you agreed with the final two items, then you described your group as low in safety. (These items are only a subset of Edmondson’s Team Psychological Safety measure; please see Edmondson, 1999, for more details.)

Because of these positive benefits, a cohesive team will usually outperform a less cohesive team. However, and as discussed in detail in Chapter 5, the cohesion–performance relationship is a complex one. Meta-analytic studies suggest that cohesion improves teamwork among members, but that performance quality influences cohesion more than cohesion influences performance (Mathieu et al., 2015). The work group may not be successful because it is cohesive, but instead it may be cohesive because it has succeeded in the past. Cohesive teams can also be spectacularly unproductive if the group’s norms stress low productivity rather than high productivity.

Cohesion and Trust

Cohesion also benefits teams because it contributes to the development of interpersonal trust: the mutual assurance that other members of the group will do what they are supposed to do and do so without too much supervision, pestering, or application of pressure (Kramer, 1999). As sociologist Gary Alan Fine concludes, “trust, which originates in confidence in information provided by groups and individuals and builds on personal commitment to the group, is translated into a ‘pure’ relationship that, when generalized to the collectivity, produces organizational loyalty. Trust anchors cohesion” (2003, p. 189).

Do the Members of Online Teams Trust Each Other?

Not all teams meet in face-to-face settings, but instead they collaborate on shared tasks using information technologies. Variously termed e-teams, virtual teams (VTs), and distributed teams, e-teams interact via computer-based communication technologies, and so the nature of these teams has changed as technology has changed. Whereas these teams once used e-mail and telephone-based conference calls as their primary means of communicating, most now augment these tools with videoconferencing, decision support software, file store systems, and even virtual-world conferencing.

When teams first turned to these technologies, they often encountered difficulties in planning and strategizing relative to offline groups, and, in some cases, performance outcomes were disappointing. Over time, however, improvements in technological tools have increased the richness of e-team interactions, and members have become more skilled in using them. This improvement is due, in part, to a generational shift in team composition: The newest members of teams grew up using technology in all aspects of their lives (Maynard et al., 2012).

How do members of e-teams learn to trust each other as they collaborate on projects from a distance? Sirkka Jarvenpaa (2016), a researcher in business and management, examined this question by arranging for students in college classes at dozens of universities located around the world to work together virtually (Jarvenpaa, Knoll, & Leidner, 1998; Jarvenpaa, Shaw, & Staples, 2004). The groups were given eight weeks to complete a series of tasks, including a written paper and a working website that would count toward their course grade. These groups varied considerably in their overall levels of productivity, but also in their levels of trust. The high-trust groups expressed a high degree of confidence in each other and in their group, in part because members believed that their partners in the work had the skills they needed to be successful. Then, as the group worked on the project, members of high-trust groups also attributed integrity (e.g., “I am never doubtful about whether the other team members will do what they promised”) and benevolence (e.g., “the other team members will do everything within their capacity to help the team perform”) to the group and its members.

Team B, for example, was a high-trusting group. Members communicated with one another at much higher rates to clarify the group’s task and procedures, but to also exchange personal information. They “communicated their excitement and optimism in their first messages (‘I am very excited about working on the project with all of you … ‘; ‘I am really looking forward to work with you … the assignments do look very interesting’)” (Jarvenpaa et al., 1998, p. 46). They did not quickly divide up the work into parts and assign each part to a member, but instead worked continuously together on their tasks. They communicated with one another continuously.

Team F, in contrast, was low in trust, for members showed little initiative or concern for others’ work. Communication rates were low throughout the process, and much of the work was done (grudgingly) by just two of the members:

Although the team members expressed their interest and commitment in early messages, no member was willing to take charge. Each time something was needed, a member would ask who was going to do the activity rather than volunteering. This began with the first exercise in which a member said that someone needed to coordinate the activity and asked for a volunteer. No one volunteered. (Jarvenpaa et al., 1998, p. 51)

Team F never completed the assignment, whereas members of Team B stayed in touch with each other even after the project ended.

Trust develops gradually as members interact with one another, for members require time to gather the information they need to estimate the strength of the relationships before they risk testing those relationships. Many factors influence trust in groups, but one particularly influential theory of trust—the organizational trust model—ties trust in others to perceptions of ability, benevolence, and integrity (Mayer, Davis, & Schoorman, 1995; Schoorman, Mayer, & Davis, 2007). At first, trust is largely determined by appraisals of the skills and abilities of other group members; individuals who are thought to be incompetent or untrained cannot be trusted to complete their tasks. Integrity, too, is also salient to group members early in the life of the group, as members determine if others accept the group’s norms and standards as their own. Members will also, in time, add to ability and integrity a third quality—benevolence—as they contribute unstintingly to the group (Aubert & Kelsey, 2003; Yakovleva, Reilly, & Werko, 2010). In teams with high levels of trust, members share information, expend more effort, and provide others with more support, and as a result: trust leads to improved team performance (Fiore, Carter, & Asencio, 2015).

11-4. Output: Team Performance

Organizational experts recommend using teams to achieve excellence. No matter what system the experts propose—job enrichment, balanced scorecard management, business process reengineering, activity-based management, or an updated version of management by objectives—most will tout the benefits of using teams to get work done. But do teams offer the best means to maximize human potential? This section examines the final segment of the input–process–output model of teams: What do teams generate by way of direct and indirect outcomes? The analysis raises the question of evaluation—how effective are teams?—and also considers ways to improve teams.

11-4a. Evaluating Teams

Viewed from an evolutionary perspective, teams are highly successful social organisms. From relatively humble beginnings in athletics, farming, and agriculture, teams have spread out to populate much of the world. Teams are gaining popularity as preferred approaches to management, and “how to” books on team methods continue to make the bestseller lists. Teams have also taken the place of some traditional groups as people’s source of social connection, for more people report belonging to teams than they do to hobby, community, and social groups. Teams now have only one group to overtake in terms of popularity: religious groups. But do teams live up to their promise as systems for increasing productivity and members’ well-being?

The Success of Teams

Anecdotal evidence and research findings converge on a verdict that favors teams, but with reservations. Case study approaches are generally, but not uniformly, positive (Applebaum & Blatt, 1994). Texas Instruments, for example, increased productivity when it organized its employees into small groups whenever possible, took steps to build up team cohesiveness, and went to great lengths to establish clear goals based on realistic levels of aspiration (Bass & Ryterband, 1979). When a manufacturer in the United States shifted to teams, supportive supervision, participant leadership, organizational overlap among groups, and intensified group interaction, employee satisfaction increased and turnover decreased (Seashore & Bowers, 1970). Case studies have, however, uncovered examples of spectacularly ineffective teams. For example, Hackman (1990), after examining the effectiveness of 33 teams, had to revise the proposed title of the book he had planned: Groups That Work was given the subtitle (and Those That Don’t) because he found considerable variation in performance quality across the teams he studied.

Field studies of the use of groups and team development generally support the wisdom of relying on teams (Sundstrom et al., 2000). The Harley-Davidson Motor Company, for example, dramatically transformed their production methods by shifting from a traditional command-and-control culture to one based on self-managing work teams, and the positive results of this conversion appear to depend in large part on the high level of cohesiveness maintained by these groups (Chansler, Swamidass, & Cammann, 2003). When researchers, through meta-analysis, examined the link between organizational change and performance, they found that companies that made multiple changes usually improved their performance and that group-level interventions were more closely linked to productivity than individual-level interventions (Macy & Izumi, 1993). A recent survey of people’s satisfaction with their team memberships, however, suggests that members themselves are not so happy with their teams. Only 13% of the 23,000 managers, workers, and executives in one survey agreed that their “teams work smoothly across functions” (Covey, 2004, p. 371).

Beyond Productivity

Teams are task-focused groups, and so the major criterion for determining their success is their performance: Do they reach the goals they, and others, set for them? By this standard, Mountain Medical was a success. The team learned to perform the new surgery quickly and safely, and this efficiency meant a better recovery for the patients and substantial savings for the hospital. The team needed less time in the operating room, and its efficiency was so high that it could do more operations than other teams. At a price of approximately $36,000 per case, the team proved to be both medically sound and economically profitable.

A team’s productivity, however, is only one of the outputs that should be considered when determining its effectiveness. Mountain Medical may have become a crack surgical team, but what if the demands of the task were so great that members, feeling great pressure, decided to leave the group? What if the team was productive, but over time members grew to dislike working with each other? What if Mountain Medical became stagnant—repeating the motions required for the operation with each case, but losing the capacity to adapt and change that had made them a high-performance team in the first place?

Hackman (2002) suggests three key factors that should be considered when evaluating the success of a team. Task performance is the first and foremost criterion. Teams are created for the purpose of generating results, and a successful group is one that meets or exceeds agreed-upon “standards of quantity, quality, and timeliness” (Hackman, 2002, p. 23). But Hackman adds to this criterion two other, more indirect, outcomes: adaptive growth of the team as a whole and individual development of the members. Many teams can perform their basic work effectively, but over time they fail to profit from their experiences of working together. A truly successful team is one that grows stronger over time so that it can undertake even more challenging tasks in the future. Hackman (2002, p. 28) also feels that a high-performing team should contribute, in positive ways, “to the learning and personal well-being of individual team members”:

If the group prevents members from doing what they want and need to do, if it compromises their personal learning, or if members’ main reactions to having been in the group are frustration and disillusionment, then the costs of generating the group product were too high. (Hackman, 2002, p. 29)

Team Learning

Because these cognitive foundations of teamwork develop as the teammates experience working together, teams require group rather than individual practice. Although in years past, organizations often sent their personnel offsite to individually receive training in team skills at institutes and workshops, team members need to be trained together—as a unit—rather than separately. Only by confronting the learning situation as a group can the team engage in team learning, a “process in which a group takes action, obtains and reflects upon feedback, and makes changes to adapt or improve” (Sessa & London, 2008, p. 5).

The success of the Mountain Medical Center’s cardiac surgery team illustrates the importance of learning as a team. The 16 hospitals that Pisano and colleagues (2001) studied all used the same equipment, and the operating room staffs were all trained by the equipment’s manufacturer. These highly trained surgical teams performed their work well, and nearly all of the patients fully recovered after their surgery. Some, however, recovered more rapidly and with fewer complications than others, and this gain was indicated by the speed of the operation. None of the teams operated too quickly, but some were relatively slow. With each patient, the teams improved—minimizing the amount of time that the patient was on the heart–lung machine is an indicator of recovery time—but some teams learned more quickly than others. Surprisingly, the educational backgrounds and surgical experience of the teams did not predict learning rates nor did the overall support for the new procedure by the hospital’s administrative staff. The status of the head surgeon on the team was also unrelated to learning rate, as was the amount of time the teams spent in formal debriefing sessions after each case.

What did predict learning rates? The way the teams were designed and trained. In the slow-to-learn teams, the surgeons assigned to the team happened to be the ones who were available to attend the training session. They showed little interest in who was on their surgical team—in fact, the members of the team varied from case to case, violating a basic rule of good team design (Hackman, 2002). These teams did not fully realize how intense the new surgical methods would be in terms of coordination demands, and the surgeons did not explicitly discuss the need for greater attention to teamwork.

At places like Mountain Medical, in contrast, the team surgeon was usually an advocate for the procedure, and he or she was actively involved in selecting all the other members of the team. These individuals worked together during the training sessions as a team, and they remained together longer during the first cases using the new methods. The surgeons in these teams also stressed the importance of working together as a team rather than stressing the acquisition of new individual skills: “They made it clear that this reinvention of working relationships would require the contribution of every team member” (Edmondson et al., 2001, p. 130). These fast learners also continued to increase their efficiency, as they developed an open pattern of communication where all felt free to make suggestions for improving the work. As noted earlier in the chapter, this team began slowly, taking longer to finish the procedure than most teams. By the fifth case, however, this team was performing at the same speed as most other teams, and they continued to improve their rate with each new case until they were able to conduct the operation faster than most.

11-4b. Suggestions for Using Teams

Teams are perhaps the most popular of all groups. No longer does a lone mechanic change your car’s engine oil; most automotive shops claim a team of technicians will take care of your car’s needs. Powerful moguls once ran all the large companies, but now executive leadership teams are in charge. Even scientists, long portrayed as loners working in solitude to unlock the secrets to the universe, are doing their work in groups: Science is now “team science” (Fiore, 2008).

But even the most optimistic appraisal of the available data on team effectiveness would suggest that there is room for improvement in the use of teams in performance settings. Teams are a group with extraordinary promise, but to fulfill that promise they must be implemented correctly, and members must be given assistance to use them to their full advantage (Cordery, 2004; Kozlowski & Ilgen, 2006).

Fidelity of Team Innovations

The popularity of team approaches has brought with it a significant drawback—in the rush to claim that they are using team methods, individuals sometimes call work groups “teams” even though they lack the defining features of real teams. More than 80% of the executives, managers, and team members surveyed in one study reported that their teams lacked clear goals; that their members did not engage in creative discussion; that team members did not hold each other accountable for their assigned tasks; and that members of their team rarely initiated actions to solve problems (Covey, 2004). These are basic, essential qualities of teams, and, if they are lacking, these work groups likely are not actually teams.

These responses may indicate that the very concept of a team—individuals joining together in unified groups to pursue shared goals—is unworkable, but it may also be that team-based methods have not been properly implemented. Members of true teams cannot complete their work without interacting with each other. That interaction may involve exchanging information, sharing resources, or even lifting, carrying, or moving something together rather than individually, but the work requires each team member to contribute in some way (Aubé & Rousseau, 2005; Kauffeld & Lehmann-Willenbrock, 2012). Members of successful teams are also committed to group-level goals, and the rewards they receive should be based on attaining those goals rather than individual ones. Teams are also relatively well structured and cohesive. If a team fails because it lacked these key ingredients, then the blame most likely rests with those who built the team rather than the team itself.

Training in Teamwork

Too many organizations create teams but then do little to help team members develop the skills they need to work in those teams. Only 29% of the organizations in one survey gave their teams any kind of training in teamwork or interpersonal relations, and only 26% based compensation (salary, bonuses) on team performance (Devine et al., 1999). Given the complexity of interpersonal and cognitive demands that teams require, members will likely need assistance in learning how to work effectively in them.

Fortunately, team training has robust effects on team effectiveness (Kozlowski & Ilgen, 2006). When team expert Eduardo Salas and his colleagues examined the effectiveness of several types of training interventions using meta-analysis, they concluded that

(a)

most methods work, but

(b)

the best ones focus on improving member coordination rather than communication strategies (Salas, Nichols, & Driskell, 2007).

Cross-training, which involves rotating members throughout the various positions within the group, was particularly helpful, in that it provided members with a clear understanding of the demands associated with each role and the interconnections among members’ responsibilities. Crew resource management (CRM), developed for training flight crews in teamwork procedures, has also been successfully applied to teams working in many other settings with considerable success. Salas, summarizing the available data, concludes that training team members yields demonstrably positive results, and he encourages organizations to make use of scientific principles and findings to improve their training of team members (Salas et al., 2012; Shuffler, DiazGrandados, & Salas, 2011).

Team Building

When well-meaning leaders and managers wish to help their teams work more effectively, they often turn to team building exercises, social events, and off-site training experiences. In the name of team building, organizations often place their teams in challenging environments so that the members will learn teamwork skills but also develop a sense of unity as a result of surviving the ordeal. Team building adventures, such as backpacking together in the wilderness, spending the day on a ropes course, or playing a paintball game against a rival team, continue to be popular methods for increasing team unity.

These activities often function as group-level rewards for participating in teams, but they are no substitute for research-based team building interventions. Unlike team training—which is skill-focused and usually involves practice and feedback—team building is less structured and targets general relational skills. Team building, however, when properly implemented, does tend to improve team functioning (Klein et al., 2009). Salas and his colleagues identified four basic approaches to team building that target more specific problems that teams often face: goal setting, interpersonal relations (e.g., trust, communication, and teamwork), role clarification, and problem-solving procedures (Salas et al., 1999; Salas, Priest, & DeRouin, 2005). A meta-analytic review of studies of these types of team building methods suggests that all are relatively effective, but that goal-setting and role-clarification interventions led to more significant improvements (Klein et al., 2009).

Situational Support

A final condition for implementing teams is the degree of organizational support available to the teams (Kennedy et al., 2009). Organizations may, in the rush to implement teams, create them but then fail to provide them with the support they need to flourish. Features of the organizational context, such as support for technologically based group support systems, development of group-level reward systems to supplement or complement individual rewards, degree of collectivism in the organizational culture, and the availability of external coaches who can assist the team to navigate trouble spots, will increase the probability that team-based approaches will be successful (Mathieu et al., 2008). Other organizational features, such as traditional leadership styles, hierarchical patterns of organization, and individually based compensation systems, will increase the likelihood that team approaches will not prosper.

The case of quality circles (QCs) provides a lesson in the importance of providing support for group-level innovations. QCs were popular in the 1980s. These small, self-regulated decision-making groups usually included five to ten employees who performed similar jobs within the organization. The groups were often led by a supervisor who had been trained for the role, but participation in the circle was often voluntary, and no monetary incentives were offered to those involved. These groups were thought to be excellent ways to increase workers’ participation in the management of the organization and to increase productivity, efficiency, quality, and job satisfaction. Yet, by the 1990s, most of these groups were gone—the failure rate was between 60% and 70% (Tang & Butler, 1997). What happened?

QCs were not teams, and they had their own unique limitations—participants volunteered and were not compensated, and, in many cases, conflicts developed between participants and nonparticipants. Worse, however, was the lack of support provided the QCs. They were originally viewed as an easy means of increasing involvement and satisfaction, but the suggestions of QCs were rarely heeded by management. They were essentially powerless, and members soon realized that they were an ineffective means of achieving valued outcomes. A few transformed from QCs into true self-managing teams, but most were just abandoned (Lawler & Mohrman, 1985).

The lesson of QCs should not be ignored. As many as 90% of Fortune 500 companies implemented such methods in their plants, factories, and meeting rooms at the peak of their popularity, but the method did not take. Without institutional support or proper design, QCs rapidly disappeared. It would be unfortunate if teams went the way of quality circles, due to failures to implement them correctly, failures to train individuals to work effectively in them, and failures to support them.

Resources

Chapter Case: Mountain Medical’s Cardiac Surgery Team

“Organizational Differences in Rates of Learning: Evidence from the Adoption of Minimally Invasive Cardiac Surgery” by Gary P. Pisano, Richard M. J. Bohmer, and Amy C. Edmondson (2001) examined how the surgery teams at 16 different medical centers adjusted to a new surgical procedure that required a higher degree of teamwork (see, too, Edmondson et al., 2001).

The Nature of Teams

Making the Team: A Guide for Managers by Leigh L. Thompson (2014) provides a detailed analysis of all the core topics in the analysis of teams, with chapters pertaining to internal dynamics (e.g., communication and conflict) and external dynamics (e.g., social networks and multiteam contexts).

“Team Development and Functioning” by Janis A. Cannon-Bowers and Clint Bowers (2011) is a comprehensive review of the current state of research into team processes and dynamics that examines historical approaches, taxonomies, team selection processes, team performance, and emerging issues.

“Unraveling the Effects of Cultural Diversity in Teams: A Meta-Analysis of Research on Multicultural Work Groups” by Gunter K Stahl, Martha L Maznevski, Andreas Voigt, and Karsten Jonsen (2010) synthesizes the findings from 108 studies of over 10,000 teams to draw conclusions about the relationship between team diversity and performance.

Working in Teams

Team Cohesion: Advances in Psychological Theory, Methods, and Practice, edited by Eduardo Salas, William B. Vessey, and Armando X. Estrada (2015), provides extensive theoretical and empirical details about one of the core qualities of most successful teams: cohesion.

The Wiley Blackwell Handbook of the Psychology of Team Working and Collaborative Processes, edited by Eduardo Salas, Ramon Rico, and Jonathan Passmore (2017), provides extensive coverage of team work processes, with 25 chapters dealing with antecedents to team effectiveness, team work processes, and performance outcomes.

Looking Back, Moving Forward: A Review of Group and Team-based Research, edited by Margaret A. Neale and Elizabeth A. Mannix (2012), includes chapters examining fundamentally important topics in the study of groups and teams, including power, leadership, composition, and conflict.

Improving Teams

Smarter Faster Better: The Secrets of Productivity in Life and Business by Charles Duhigg (2016) is an engaging integration of core concepts in the scientific study of team processes with cases of high-functioning teams in military, business, and entertainment settings.

Holacracy: The New Management System for a Rapidly Changing World, by Brian J. Robertson (2015) describes one way teams can be used to structure collective effort in business and industry.

Chapter 12. Decision Making

People turn to groups when they must solve problems and make decisions: They are better informed, they can review and appraise ideas, information, and alternatives through discussion, and they use consensus-based standards when making their final choices. But groups are not perfect and neither are their decisions. People sometimes make more extreme decisions when in groups than they would if they were alone, and in rare cases they seek concurrence instead of the best decision possible, and so fall victim to groupthink.

How do groups make decisions?

What problems undermine the effectiveness of decision making in groups?

Why do groups make riskier decisions than individuals?

What is groupthink, and how can it be prevented?

Chapter Outline

The Decision-Making Process

Orientation

Discussion

The Difficulty of Discussion

Making the Decision

Implementation

Decisional Biases

Judgmental Biases

The Shared Information Bias

Group Polarization

Victims of Groupthink

Symptoms of Groupthink

Defective Decision Making

Causes of Groupthink

The Emergence of Groupthink

Alternative Models

Preventing Groupthink

Chapter Review

Resources

The Bay of Pigs Planners: Victims of Groupthink

The decision makers are meeting in the Cabinet Room of the White House, just down the hall from the Oval Office. The group has gathered at the request of President John F. Kennedy to discuss a covert para-military operation code-named Zapata. The Central Intelligence Agency (CIA) had developed operation Zapata after Fidel Castro came to power in Cuba. During this period in U.S. history the CIA used various methods to curtail the spread of communism in Latin and South American countries, and it hoped to apply these covert operational strategies to overthrow Castro and his communist government.

Operation Zapata proposed training, transporting, and supporting a regiment-sized group of guerrilla fighters who would invade Cuba at the Bahía de Cochinos (Bay of Pigs). The men would then launch raids and encourage civilian revolt in the country’s capital city, Havana. The CIA needed Kennedy’s approval, but Kennedy did not want to make this decision by himself. So, he assembled his advisors, and together they discussed the strengths and weaknesses of the proposal. The Bay of Pigs advisory group, as diagrammed in Figure 12.1, included White House senior advisors and staff members, cabinet members, the CIA and their subject matter experts, and military advisors from the Department of Defense—all highly skilled individuals well trained in making critically important policy and military decisions. This group, after a thorough review, advised the president to give the CIA the go-ahead.

The Bay of Pigs invasion took place on April 17, 1961. The assault that was so carefully planned was a disaster. For the plan to succeed, the attacking force needed to secure and hold the beachhead at Playa Girón, but Castro’s forces’ counterattack overwhelmed them. Several key elements of the plan, including air support and supplying the ground forces with munitions, were either aborted or poorly executed. The entire attacking force was killed or captured within days, and the U.S. government had to send food and supplies to Cuba to ransom them back. Group expert Irving Janis described the decision as one of the “worst fiascoes ever perpetrated by a responsible government” (1972, p. 14), and President Kennedy lamented, “How could I have been so stupid?” (quoted in Wyden, 1979, p. 8).

The Bay of Pigs advisory committee, like many other groups, faced a problem needing a solution. Through discussion, the members pooled their expertise and knowledge. They sought out information from available sources, and they thoroughly weighed alternatives and considered the ramifications of their actions. When their alternatives were narrowed down to two—to invade or not to invade—they made the decision as a group. But the committee was typical in another way. Like so many other groups, it made the wrong decision.

We owe much to groups. Groups put humans on the moon, built the Empire State Building, performed the first symphony, and invented the personal computer. But groups also killed innocent civilians at My Lai, marketed thalidomide, doomed the space shuttles Challenger and Columbia, and decided that the best way to deal with the communist regime in Cuba was to invade it. This chapter examines both the pros and the cons of making decisions in groups before examining one potentially catastrophic group process—groupthink—in detail.

12-1. The Decision-Making Process

In office buildings, executives hold conferences to solve problems of management and production; at the dinner table, families talk over moving to a new neighborhood; in courthouses, juries weigh evidence to determine guilt and innocence; on the battlefield, a combat squad identifies a target and plans an attack. In these and thousands of other similar settings, interdependent individuals make decisions in groups.

Why turn to a group when a decision must be made? Even though groups are far from perfect, their choices, judgments, estimates, and solutions are usually superior to those tendered by lone individuals (Hinsz, 2015). The work of scientists who collaborate is superior to the work of scientists who work alone (Uzzi et al., 2013). Groups’ perceptions of other people are more accurate than individuals’ impressions (Ruscher & Hammer, 2006). Small groups working together on Google searching for information will find more relevant information more quickly than a single searcher can (Lazonder, 2005). Teams of physicians making a diagnosis are more accurate than single physicians (Glick & Staley, 2007). Students who take a test in groups get better grades and learn more than individual students (Vogler & Robinson, 2016). Burglars who work in groups are less likely to be caught than are thieves who work alone (Warr, 2002). Groups solve difficult logic problems (e.g., the Wason selection task) faster than individuals, and when the members encounter similar problems later as individuals they outperform those who did not have a group learning experience (Maciejovsky et al., 2012). Even very powerful leaders—presidents of the United States, for example—rarely make decisions without consulting others. Instead, they rely on groups, for the weighty problems that they must handle on a daily basis would overwhelm a lone individual. In most situations, the wisdom of the many is greater than even the genius of the one.

What is the secret to groups’ superiority in making decisions? A process model of group decision making suggests that groups engage in a sequence of activities, operations, and practices as they move from uncertainty to decisional conviction and that each step in the series serves some purpose. When group members convene to make a decision or solve a problem, they don’t just vote and then adjourn. Instead, they define their purpose, share information through discussion and deliberation, consider alternatives and solutions, and identify implications. No two groups reach their decisions in precisely the same way, and no two theorists agree on the definitive list of processes that determine those decisions, but the four phases identified in the ODDI process model shown in Figure 12.2—Orientation → Discussion → Decision → Implementation—are often in evidence when groups make decisions and solve problems (Simon, 1997). The group defines the problem, sets goals, and develops a strategy in the orientation phase. Next, during the discussion phase, the group gathers information about the situation and, if a decision must be made, identifies and considers options. In the decision phase, the group chooses its solution by reaching consensus, voting, or using some other social decision process. In the implementation phase, groups put the decision into action and assess the consequences of their choice. These processes need not occur in this order or in every case, but they often do. Groups that follow these four stages are more likely to make better decisions than those who do not (Wittenbaum, Hollingshead, & Botero, 2004).

12-1a. Orientation

Decisions begin with a problem that needs a solution. A group of students in a class are required to complete a project that includes a written paper and a presentation. The president of the United States is briefed by the CIA on the invasion of Cuba. The police unit must take into custody a criminal who is residing in a heavily defended residence. Such situations trigger a decision-making process that often begins with a period of orientation, as the group reviews its objectives and organizes the procedures it will use in its work. As that great expert on human behavior, baseball player and manager Yogi Berra (2002, p. 53), once warned, “If you don’t know where you’re going, you might not get there.”

Goal Clarity and Goal-Path Clarity

Some groups may know exactly what they want to achieve and how they will go about doing it, but most must first clarify both the goals they seek and the path they will take to reach those goals (Shaw, 1981). Goal clarification requires not only setting specific, attainable goals, but also the review of the group’s overall mission, the problems it is dealing with and the decisions it must make, the results it intends to deliver, and the criteria it will use to evaluate the quality of its performance and results. Goal-path clarification, in contrast, requires spelling out just how the group will do its work, including identifying tasks and subtasks, organizing members’ roles and responsibilities, specifying how the members will work together, determining how the group will make its decisions, and setting milestones and deadlines.

Both groups and the individuals who are members of those groups are more successful—they make better decisions, solve problems more efficiently, and waste less time—when they have clarified their goals and their paths to those goals (Weingart, 1992; Weldon, Jehn, & Pradhan, 1991). The importance of this orientation process is so great that in some cases it is the only thing that differentiates successful groups from unsuccessful ones (Hirokawa, 1980). In a study of six conferences in which panels of experts evaluated new medical technologies, participants were more satisfied when the decisional procedures had been discussed in advance (Vinokur et al., 1985). Similarly, in a project that experimentally manipulated the use of process planning, groups were more productive when they were encouraged to discuss their performance strategies before working on a task requiring intermember coordination (Hackman, Brousseau, & Weiss, 1976). Effective teams, as noted in Chapter 11, take time to develop a shared mental model as the members review their tasks, roles, goals, and procedures (Bonner, Soderberg, & Romney, 2016).

Researchers examined the benefits of this orientation phase by studying groups that were responsible for making critically important decisions for the companies where they worked—such as launching an e-commerce website or developing a technologically sophisticated training program. They discovered that some of these groups performed marvelously, meeting their goals, and impressing high-ranking leaders in the company. Others did not. But what set the high and low performers apart was how they mobilized their resources during their initial meetings. Groups that spent considerable time clarifying their goals and reviewing their procedures in a series of sessions were more effective than those who did not. These investigators also found, however, that in the best teams the leader did not orchestrate the planning process, but let the group members themselves explore their purposes and procedures (Ericksen & Dyer, 2004).

Problems with Planning

Given the clear benefits of spending time setting goals and making plans, it is unfortunate that few groups show much interest in planning their procedures. When a group member raises the issue of planning, very rarely do any of the other group members respond positively (Hackman et al., 1976). When groups are given a problem to solve or a decision to make, their first tendency is to get started on the task itself rather than consider process-related issues. Even when enjoined to plan, groups believe that planning is less important than doing (Shure et al., 1962). Even Kennedy’s group moved through the orientation stage too hastily. Kennedy had just become the president of the United States, and his advisors had not worked together before, so the members should have spent several meetings talking about the problem and the strategy they would take in solving it. Instead, they immediately began to discuss logistics and operations (Stern, 1997). This anti-planning bias stems, in part, from the tendency of groups to apply whatever method they used in the past to current and future projects (Hackman & Morris, 1975).

Yet, groups that recognize that their time is limited plan out their work better than groups that do not (Sanna et al., 2005). In one survey of 48 self-managing teams, those who spent time during their initial stages with temporal planning developed strong norms about time, and these norms helped these groups perform better than groups that did not put enough time into time planning (Janicik & Bartel, 2003). Groups that spend time setting deadlines and reviewing problems they might encounter as they pursue their goals are also more likely to avoid the planning fallacy: the tendency to underestimate just how long a task will take to finish (Kahneman & Tversky, 1973). This fallacy occurs because most people assume that the future will be pleasant rather than bleak, that issues that come up along the way will be handled quickly and without great expenditure and effort, and that one’s choices will be right rather than wrong. People are basically optimists, so when they envision the future, they tend to construct mental scenarios that err on the positive side and underestimate the possibility of negative, time-draining problems and missteps (Dunning, 2007).

Both individuals and groups are prone to this bias, but groups are even less accurate than individuals. When college students were asked to make predictions about the time needed to complete a written case analysis that would count as 25% of their final grade, their estimates were less accurate after they talked the project over as a group. Even though the projects had been divided into subtasks, each with its own distinct deadline, the group members grossly underestimated the time they would need to complete each one. During their discussion of deadlines, the groups focused primarily on factors promoting successful task completion and overlooked possible problems (Buehler, Messervey, & Griffin, 2005).

Are Group Discussions a Waste of Time?

The humorist C. Northcote Parkinson’s (1957) time management laws are particularly apt when applied to groups. Parkinson’s first law, which he modestly named Parkinson’s Law, states that a task will expand to fill the time available for its completion. Hence, if a group gathers at 1 PM for a one-hour meeting, the group will likely adjourn at 2 PM, no matter how simple or routine the issues they examine.

Parkinson’s second law, the law of triviality, states that the time a group spends on discussing any issue will be in inverse proportion to the consequentiality of the issue (Parkinson, 1957, paraphrased from p. 24). Parkinson described a hypothetical finance committee dealing with Item 9 on a long agenda: a $10-million allocation to build a nuclear reactor. Discussion is terse, lasting about 2½ minutes, and the committee unanimously approves the item. However, when the group turns to Item 10, the allocation of $2350 to build a bicycle shed to be used by the office staff, everyone on the committee has something to say. As Parkinson (1957, p. 301) explained:

A sum of $2350 is well within everybody’s comprehension. Everybody can visualize a bicycle shed. Discussion goes on, therefore, for forty-five minutes, with the possible result of saving some $300. Members at length sit back with a feeling of achievement.

Research does not entirely support Parkinson’s bleak assessment. Groups are sensitive to the challenges that various items on their agendas pose, and they tend to allocate more time to more difficult tasks—although their attention and interest does tend to wane the longer the meeting lasts (Littlepage & Poole, 1993).

12-1b. Discussion

On February 17, 1961, President Kennedy asked his advisors a simple question: Should the United States support a paramilitary invasion of Cuba? To provide the president with an answer, the group examined all kinds of issues through discussion. They reviewed the issue in a series of meetings by sharing information, setting priorities, expressing misgivings, and studying alternatives until they felt confident enough to make a recommendation.

If information is the lifeblood of decision making, then the discussion phase must be the heart of that process (Kowert, 2002). An information processing model of decision making assumes that people strive, in most cases, to make good decisions by acquiring the information that is relevant to the issue and processing that information thoroughly, so that its implications are clearly understood. A collective information processing model also assumes that people seek out and process relevant information but that they do this cognitive work during the group discussion (Hinsz, Tindale, & Vollrath, 1997). When people discuss the problem as a group, they improve their memory for information, exchange information with each other, process the information more thoroughly, and identify errors and mistakes (see Figure 12.2).

Collective Memory Processes

Two heads are better than one because groups have superior memories for information relative to individuals. Arthur Schlesinger, for example, knew a great deal about international relations, but he could not compete with the combined informational resources of all the Bay of Pigs planners. Their memories, when combined, contained a vast assortment of information about Cuba, Castro, weaponry, and even the terrain of the beach where the troops would land (see Harris, Paterson, & Kemp, 2008, for a review).

A group’s collective memory is the shared reservoir of information held in the memories of two or more members of a group. Groups remember more than individuals, because groups draw on more memories that contain different types of information. The CIA operatives who met with the Bay of Pigs planners knew all about the weapons, tactics, and the morale of Castro’s troops, but Rusk was an expert on the relationship between Cuba and the Soviet Union. When they joined together, they could pool their individual expertise to form the group’s decisions. (Unfortunately, no one in the group knew that the Bay of Pigs was Castro’s favorite fishing spot, so he was thoroughly familiar with every path, road, and hill in the area.) Similarly, when students are permitted to take examinations as a group, they usually outperform individuals, for the student who is stumped by the question “Name four common phases of group decision making” is saved by a group member who remembers the mnemonic acronym ODDI: Orientation, Discussion, Decision, and Implementation (Bonner & Baumann, 2012). Groups can also get more information than individuals can. In many cases, decision-making groups are staffed by individuals who have widely differing experiences, backgrounds, and associations, so each one can contribute unique information to the discussion (Henningsen & Henningsen, 2007).

Groups’ memories are also superior to individuals’ memories because only groups can store and retrieve information in a transactive memory system (Wegner, 1987). If the human mind is an information processing network, then a group is a “network of networks” (Van Overwalle & Heylighen, 2006, p. 606). A group has not just one system for storing memories, but many such systems, so the group can improve both the capacity and durability of its memories by dividing data among the members. Members working in the same group often specialize, to a degree, in different areas. These individuals not only have more information on a given topic, but they are also the ones who should be more responsible for storing any new information that is relevant to their area of expertise. In the committee, for example, the CIA was recognized as the source of all information about the invasion force, so other group members spent little effort deliberately storing information on that topic. When anyone needed to check a fact pertaining to the commandos, they turned to the CIA and their memory stores (Hollingshead, 2001a). But the Bay of Pigs planners had not spent enough time together to develop a strong transactive memory system. As discussed in Chapter 11, transactive memory is enriched by the experience of working as a team and by trust among members. The Bay of Pigs planners had neither.

Groups can also improve their access to information stored in members’ memories through cross-cuing. This process occurs when one member says something that jogs other members’ memories. For example, President Kennedy may not remember where the force will land, but perhaps he will say, “I think it’s a bay.” This cue may trigger someone else’s memories, so that the name “Bay of Pigs” is retrieved by the group, even though none of the members could generate this name initially (Meudell, Hitch, & Kirby, 1992). Unfortunately, if a group member offers up a misleading cue—instead of saying, “I think it’s a bay,” a committee member said, “I think it’s near a lagoon”—then such cueing can inhibit memory retrieval rather than facilitate it (Andersson, Hitch, & Meudell, 2006).

Information Exchange

Groups do not merely draw on a larger pool of information than individuals. They can also exchange information among the members of the group, thereby further strengthening their access to information as well as their recall of that information. For example, the discussion groups in one study were asked to make simple estimates, such as “What is the population of the state of Utah?” These groups exchanged, on average, 27 pieces of information before drawing a conclusion (Bonner & Baumann, 2012). The general discussion groups that sociologist Robert F. Bales (1955) studied exchanged, on average, 960 pieces of information in each of their sessions. More than 50% of all comments made by members were suggestions, expressions of opinion, and attempts at orientation (see Figure 12.3). Group members also shared information about the problem, expressed agreement or disagreement, and asked for more information and clarification. The proportion of comments in each category will vary depending on the nature of the group discussion and its level of intensity, but in most groups communication peaks during the discussion phase.

Processing Information

Groups not only recall and exchange information more effectively than individuals, but they also process that information more thoroughly through discussion. Members ask questions, and others offer answers. Alternative options are discussed, and the strengths and weaknesses of each option are considered. Group members analyze each other’s ideas and offer corrections when they note errors. Members dialogue with one another, sharing viewpoints and seeking a shared meaning. Ideas are debated, with some group members seeking to convince others that their position is better. The group members also monitor their work and intervene as necessary to bring the group back on task. Most group discussions also include an interpersonal element that complements the focus on the work to be done. Members of decision-making groups not only share and evaluate information, but they also encourage each other, express commitment to the group, and help each other (Jehn & Shah, 1997; Weingart & Weldon, 1991).

Just as the orientation period is essential to effective decision making, so the time spent in active discussion increases the quality of the group’s decision (Katz & Tushman, 1979). When researchers monitored group members’ communications while working on a problem that could be solved only by properly sequencing individuals’ responses, they found that the group’s use of essential information through discussion proved to be the best predictor of success (Lanzetta & Roby, 1960). Groups working on collective induction problems—tasks that require a cycle of hypothesis generation and testing—performed best when members discussed the problems actively and focused their analysis on evidence rather than on hypotheses (Laughlin & Hollingshead, 1995). Flight crews that confront sudden emergencies often overcome the problem if they share information with one another; but those crews that do not take advantage of group discussion often make errors in judgment that are not corrected by the group (Paris, Salas, & Cannon-Bowers, 1999). Physicians trying to diagnose an illness were more successful when they spent more time “talking to the room”—expressing ideas about the problem openly, so the entire group can hear them (Tschan et al., 2009). Studies of online groups have found that the online format substantially hampers the group’s ability to make an informed decision if the rate of information exchange is too low and too slow (Baltes et al., 2002). When researchers watched groups make decisions, they found that information sharing (talking a great deal, free expression of ideas, thoughts, and feelings) and critical evaluation of ideas (critically evaluating each other’s ideas or works, differences of opinion, disagreement among group members, and disagreements on who should do what or how something should be done) were correlated with judgmental accuracy (Jehn & Shah, 1997).

Error Detection and Correction

As groups discuss information, they appraise the validity of ideas being shared, seeking increased accuracy, and identifying any errors of fact or implications. This error-checking process was identified as a crucial determinant of successful decision making by psychologist Marjorie Shaw (1932) in one of the earliest experimental studies of group performance. She examined the sagacity of groups by putting 21 individuals and 5 four-person groups to work on several intellective tasks, including the famous (at least to people who study groups) missionary–cannibal dilemma:

Three missionaries and three cannibals are on one side of the river and want to cross to the other side by means of a boat that can only hold two persons at a time. All the missionaries can row, but two cannibals cannot. For obvious reasons, the missionaries must never be outnumbered by the cannibals, under any circumstances or at any time, except where no missionaries are present at all. How many crossings will be necessary to transport the six people across the river?

When the groups and individuals finished the first set of problems, Shaw reorganized them, so that those who worked alone initially solved several new problems in groups and those who initially worked in groups solved several new problems individually.

Shaw’s groups performed well for two reasons. First, compared to individuals, her groups generated more correct solutions. Second, however, the groups were better at checking for errors in calculations and identifying faulty inferences about the problems. If a group member recommended a solution that was inaccurate, groups were more likely to reject that solution. Groups, when they did make mistakes, also erred later in the decision process than did individuals, in part because groups were more proficient at noticing and correcting errors. Groups, however, took longer to complete the task than did individuals. (The answer to the missionary–cannibal problem, by the way, is 13 crossings! Note, too, that this study was conducted by Marjorie E. Shaw—no relation to Marvin E. Shaw who also studied groups and whose classic 1981 text, Group Dynamics, is this book’s intellectual progenitor.)

12-1c. The Difficulty of Discussion

Group discussion is an excellent method for solving problems, making decisions, resolving disputes, and so on, but discussion requires skill, motivation, and practice, and not all groups can muster all three of these key ingredients. Even the Bay of Pigs planners, despite their great expertise, failed to use the group’s resources to examine fully the military, political, social, and ethical limitations of the CIA plan. People assume their skills are sufficient for dealing with problems through discussion, but their confidence is often misplaced.

Groups Are Forgetful

Groups can make use of cognitive mechanisms such as cross-cueing and transactive memory to improve their recall of information, but groups are not mnemonic marvels (Rajaram & Pereira-Pasarin, 2010). When researchers compared the memories of collaborative groups, nominal groups (groups of noninteracting individuals), and individuals, collaborative groups outperformed both the average single individual and the best single individual. Collaborative groups did not, however, perform as well as nominal groups, and the groups displayed many of the characteristics typically seen in individual memory. Individuals, for example, generally have better memory for information that they process more deeply and better memory for pictures than for words. Groups displayed these same tendencies when their memories were tested (Weldon & Bellinger, 1997). Groups also reported words that were not on the original list, and their memories were also less well structured (Hyman, Cardwell, & Roy, 2013).

The loss of motivation that occurs in groups—social loafing (see Chapter 10)—partly explains this decline in processing efficiency (Weldon, Blair, & Huebsch, 2000). But the complexity of the group setting also disrupts members’ ability to organize information in memory so that it can be retrieved later. In consequence, collaborating groups perform particularly poorly when trying to remember unorganized information, but they perform the same as noninteracting (nominal) groups when trying to remember organized information. When group members spend time sharing memories of previous events—reminiscing—each member’s memory of that information is improved, but otherwise: groups are more forgetful than individuals (Marion & Thorley, 2016). These inadequacies in collective memory may be so substantial that groups cannot remember their decisions unless they keep a written record of them (minutes). Although few group members relish the role of recorder, without minutes, details of the group’s actions may be forgotten.

Group Members Misunderstand

Group members make decisions by exchanging information, but they often make mistakes both when expressing their message and when listening to what others say. On the sender side, many group members lack the skills needed to express themselves clearly. They fail to make certain that their verbal and nonverbal messages are easily decipherable and so unintentionally mislead, confuse, or even insult other members. One study of college students reported that 33% could not give accurate directions, 49% could not summarize the points made by a person who disagreed with them, and 35% could neither state their point of view clearly nor defend it (Rubin, 1985). On the receiver side, inaccuracies also arise from the information processing limitations and faulty listening habits of human beings. Listeners tend to level (simplify and shorten), sharpen (embellish distinctions made by the speaker), and assimilate (interpret messages so that they match personal expectations and beliefs) information offered by others during a discussion (Campbell, 1958b; Collins & Guetzkow, 1964).

These limitations are apparent to anyone who attends meetings with any frequency. When researchers asked 569 full-time employees who worked at jobs ranging from clerical positions to upper-level management to describe “in their own words what happens during a meeting that limits its effectiveness,” they received nearly 2500 answers. Poor meeting planning was a common complaint (24%; see below), but the majority complained that too many members either

(a)

did not have the skills needed to work well with others in a group setting or

(b)

they did not bother to make use of their skills once in the group.

Poor meeting planning: the meeting was poorly organized; for example, it included the wrong people, the time needed for the work was inadequate, and the agenda used to structure the meeting was unhelpful or inaccurate.

Lack of skill in communication: not only did members frequently fail to listen to what others had to say, but they rarely communicated their ideas and opinions in engaging and informative ways.

Egocentric behavior: Some members dominated the meeting or used the meeting to flaunt idiosyncrasies, intimidate others, grandstand, and complain.

Low engagement: Too many members did not take part in an active way in the sessions; they did not speak, volunteer to join with others working on projects, and sometimes just stared silently at others.

Sidetracked: The meeting did not stay on the topic, with many tangents and irrelevancies.

Interruptions: The meeting’s flow was disrupted as members interrupted each other, took phone calls, or engaged in side conversations.

Inadequate leadership: the leader was not organized, and so did not facilitate or control the meeting’s process.

Negative attitudes and emotions: Members did not communicate in a civil, respectful way, they expressed anger, grumbled, and so on.

Follow-up: Nothing ever happened as a result of the meeting; no implementation of decisions.

The participants in this research suggested that the groups failed more frequently than they succeeded. As one respondent explained, “No one, not even the leader, ever listens to what anybody else has to say” (Di Salvo, Nikkel, & Monroe, 1989, p. 557).

Are Meetings a Waste of Time?

When Kennedy needed to make a decision about the CIA’s plan, he called a meeting, and in doing so he was not doing anything particularly extraordinary. In many organizations, people spend much of their day scheduling meetings, preparing for meetings, traveling to meetings, sitting in meetings, recovering from meetings, and planning the next meeting (Allen, Lehmann-Willenbrock, & Rogelberg, 2015b). One source estimates the number of meetings held in the United States in a year’s time to be about three billion (Nunamaker et al., 1997).

Meetings may be essential tools for making decisions and organizing productivity, but those who attend often consider them to be boring, uninteresting, and inefficient. They can, in some cases, be filled with conflict so that they are not just boring but also threatening. When meetings are not used for their basic purpose—to make decisions and solve problems—but instead are just information sharing sessions, some participants may consider them to be a waste of time. This reaction is particularly likely if participants do not work closely with others, and so the meetings interrupt their solitary pursuits.

Industrial/organizational psychologist Steve Rogelberg and his colleagues explored the downside of meetings by asking workers in England, Australia, and the United States about their involvement in meetings and to rate them on a scale from 1 (extremely ineffective) to 5 (extremely effective). The researchers also measured such variables as job satisfaction, stress (e.g., tension, anxiety, worry, gloom, depression, and misery), and the degree of interdependence required by their job. People who thought their meetings were effective felt more enthusiastic about their work, more satisfied, and more productive. But if they rated their meetings as ineffective, they were more depressed, more anxious, and more likely to be thinking about quitting—particularly if they did not feel that they needed to work closely with other people to accomplish their work-related tasks. Many saw meetings as interruptions—not as ways to get more work done, but as obstacles to productivity (Luong & Rogelberg, 2005; Rogelberg et al., 2006). These findings suggest that most people would appreciate two things: a reduction in the number of meetings and an improvement in the quality of those that take place.

12-1d. Making the Decision

By early April, the Bay of Pigs committee was ready to make its decision. The members had spent days examining the CIA’s plan, and even though many questions remained unanswered, the group could delay no longer. Word of the plan had leaked to the press, and the group was worried that Castro might begin to shore up his defenses. They needed to make up their minds.

A social decision scheme is a group’s method for combining individual member’s inputs in a single group decision. Some groups have clearly defined ways of making a decision—their bylaws may state, for example, that they will follow a particular rule of order (such as Robert’s Rules). In many cases, though, the social decision scheme is an implicit one that is taken for granted by groups. Not until someone says, “Let’s take a vote” does the group realize that a decision must be made about how to make decisions. Some common social decision schemes are averaging, voting, reaching consensus, or delegating (Hastie & Kameda, 2005; Kameda et al., 2011). Each decision scheme has strengths as well as weaknesses.

Averaging: Statisticized Decisions

In some cases, groups make decisions by combining each individual’s preferences using some type of computational procedure. If the group must select the best option among five alternatives, for example, each member could make his or her decision individually (either before or after a group discussion), and these private recommendations could then be averaged to yield a group decision. As with compensatory tasks discussed in Chapter 10, such decisions do not necessarily require any interaction among members, and they are often surprisingly accurate. They weight equally all the group members’ opinions, and errors tend to cancel each other out. Unfortunately, there are equally important disadvantages associated with the timing of the ranking process. If the group just averages without discussion, then the benefits of discussion will be lost. If, however, the group averages after a discussion, biases and inaccuracies introduced during the discussion may skew the group’s conclusions. Members may also feel that the process is too arbitrary and so feel little responsibility for implementing the decision.

Voting: Plurality Decisions

Most groups, at least in Western cultures, use some type of voting procedure to make decisions. Members express their individual preferences publicly or, to reduce social pressure, by secret ballot. In most cases, the group selects the alternative favored by the majority of the members (the very common majority-rules scheme), but in some cases, a more substantial plurality (such as a two-thirds majority scheme) is needed before a decision becomes final. Some groups also use ranking methods with more points awarded to alternatives that are ranked higher than others (the Borda count method). In even rarer cases, the group’s decision rules may give single individuals the authority to rule against any impending decision (veto scheme).

When researchers compared these decision rules, plurality was both easy and effective (Hastie & Kameda, 2005). Even groups that do not vote officially implicitly adopt a plurality decision scheme. When people are asked to estimate what the final decision a group will make given the distribution of various opinions within the group, their predictions coincide with the majority-rules decisional scheme. People assume that if more group members think option A is better than option B, the group will be picking option A—even if it is the wrong one (Ladbury & Hinsz, 2009). But plurality, despite its overall effectiveness, has limitations. When the vote is close, some members of the group may feel alienated and defeated. In consequence, they become dissatisfied with their membership and are less likely to lend support to the decision (Castore & Murnighan, 1978). Voting can also lead to internal politics, as members get together before meetings to apply pressure, form coalitions, and trade favors to ensure the passage of proposals that they favor. Also, if the vote is taken publicly, individuals may

(a)

conform to others’ opinions rather than expressing their personal views or

(b)

refuse to change their public opinion so as to maintain the image of independence and consistency (Davis et al., 1988)

.

Reaching Consensus: Unanimous Decisions

The Bay of Pigs advisory committee took several polls of the members, but, in the final review, the group’s position was unanimous—all agreed that the covert invasion of Cuba should proceed as planned. Some groups, such as juries, officially adopt a consensus rule, but, in many cases, groups that adopt a plurality decision scheme reach consensus at the final vote. This outcome can occur when all in the group actually agree, but it also happens when the minority who disagree vote for the favored solution for various reasons, including not wanting to be on the side that loses.

Consensus decision schemes are often galvanizing and can lead to high levels of commitment to the decision and to the group; people usually express more satisfaction with this procedure than any other decision-making method (Schweiger, Sandberg, & Ragan, 1986). However, consensus building takes a good deal of time, and if rushed, the strategy can misfire. When most of the members of a decision-making group favor an invasion, for example, the three individuals who think that it is a terrible idea may hold back information that they believe would cause dissent within the group (Kameda et al., 2002). Groups often prefer to reach consensus on sensitive questions, such as issues of morality, but they favor a majority-rules voting scheme on problem-solving tasks (Kaplan & Miller, 1987).

Consensus, no matter if achieved by voting or through prolonged discussion ending in unanimity, has long-term positive effects on performance. When researchers combined the results of dozens of previous studies that examined the level of consensus among decision makers and the performance of their groups and organizations, they discovered that higher levels of agreement went hand in hand with superior performance. This relationship was strongest, however, when the group was discussing strategic priorities, rather than more specific issues, such as the means to implement those strategies (Kellermanns et al., 2011).

Delegating: Sharing Decisions

The group as a whole does not make the decision when the decision is made by the group’s leader, or is delegated to one of the members, a subgroup within the group, or someone outside of the group. Under an authority scheme, the leader, president, or other individual makes the final decision with or without input from the group members. When an oligarchy operates in a group, a coalition speaks for the entire group. Other forms of delegation include asking an expert to answer (the best-informed member) or forming a subcommittee made up of a few members to study the issue and reach a conclusion. Delegation saves the group time and is appropriate for less important issues. Delegation is also used, in some cases, as a means to avoid responsibility and blame if the decision goes awry (Steffel, Williams, & Perrmann-Graham, 2016).

When Should the Group, and Not the Leader, Make the Decision?

Industrial/organizational psychologist Victor Vroom, recognizing the pros and cons of using groups to make decisions, developed the normative model of decision making to determine when the group’s input on an issue is needed (Vroom, 2003; Vroom & Jago, 2007). Although decisional methods can fall anywhere along the continuum from leader-centered to group-centered, Vroom’s (2003) most recent model identifies these five basic methods:

Leader decides: The leader solves the problem or makes the decision and announces it to the group. The leader may rely on information available to him or her at that time, but may also obtain information from group members. The members only provide information to the leader, and the leader may not tell the group members why the information is needed.

Leader consults members: The leader shares the problem with the group members individually, getting their ideas and suggestions one-on-one without meeting as a full group. The leader then makes the decision, which may not reflect the group members’ influence.

Leader consults the group: The leader discusses the problem with the members as a group, collectively obtaining their input. Then the leader makes the decision, which may not reflect the group members’ influence.

Facilitated group discussion: The leader coordinates the group’s analysis of the problem, helping the group reach consensus on the issue. The leader is active in the process, but does not try to influence the group to adopt a particular solution. The leader accepts the will of the group and implements any decision that is supported by the entire group.

Group decides: If the group already functions independently of the leader, then the group can make its decision without the leader’s direct involvement. The leader can provide support, direction, clarification, and resources as the group deliberates, but does not influence the group’s decision directly.

Vroom does not advocate for any one decision-making method. Rather, he recommends examining the situation before selecting an approach that is most suited to the given context. One of the most important of all factors to consider is the significance of the decision itself; if the problem is not very important, then it can be solved using a method that involves the least amount of time and the fewest individuals. But when the problem becomes increasingly important, other situational factors must also be considered: Does the leader have substantial knowledge about the issue? Does the group know even more about the problem? Will the group be committed to the solution and its implementation if it does not get involved in the decision-making process, and does that even matter? Is conflict so high in the group that members may not be able to work together on the problem? In general, when problems are simple ones, the leader is well informed and the consequences for a poor decision are relatively minor, then in the interest of time the leader should decide. A group-focused approach, in contrast, is best whenever a high-quality solution is needed, along with support from the group to implement it. However, choosing between an individual and a group approach is so complex that Vroom and his colleagues have developed a computer program that guides the choice between deciding, consulting, facilitating, and delegating (Vroom, 2003).

12-1e. Implementation

When the die is cast and the decision made, two significant pieces of work remain to be done. First, the decision must be implemented. If a union decides to strike, it must put its strike plan into effect. If a city planning commission decides that a new highway bypass is needed, it must take the steps necessary to begin construction. If an advisory committee approves an invasion, its members must mobilize the necessary military forces. Second, the quality of the decision must be evaluated. Was the strike necessary? Did we put the highway where it was needed the most? Was invading Cuba really such a good idea?

Social Justice

Implementation is, in some cases, more or less successful depending on perceptions of the fairness of the decision. Fairness judgments are determined by two forms of social justice: distributive and procedural. Distributive justice concerns how rights, resources, and costs are granted to, shared with, and imposed on (distributed across) a group’s members. When the group decides that your project will only be given a grade of C, that the Cubans who live near the Bay of Pigs will be subjected to aerial bombardments, or all of the group members will be getting raises, one wonders about distributive justice. Procedural justice, in contrast, is concerned with the methods used to make decisions about the allocation of resources. Procedural justice asks, “Did we make the decision in a fair way?” (Tyler, 2013).

Successful implementation depends on both forms of justice. People generally consider decisions that benefit them personally to be superior to those that run counter to their interests; they define them as fair and worthy of their support. How the decision was made also matters. When fair procedures are followed, group members feel more satisfied with the decision and will be more likely to perform the tasks that are required to implement the decision. For example, many of the members of the president’s advisory group were against the Bay of Pigs plan, but they believed that the group had examined the issue in a fair, impartial way, and so, when the decision was made, they supported the group's choice. People are more likely to regard a decision as a fair one if the decisional procedures are implemented “

(a)

consistently,

(b)

without self-interest,

(c)

on the basis of accurate information,

(d)

with opportunities to correct the decision,

(e)

with the interests of all concerned parties represented, and

(f)

following moral and ethical standards” (Brockner & Wiesenfeld, 1996, p. 189).

The group that uses procedurally just methods for making decisions will be more successful during the implementation stage (Colquitt et al., 2013).

Participation and Voice

Many factors influence perceptions of procedural fairness, but when people believe that they had a voice in the matter—that they could have expressed any concerns they had and others would have listened and responded—then they tend to be far more engaged in the implementation of the final decision. This voice effect was examined in an early study by psychologists Lester Coch and John R. P. French (1948). The management of a clothing mill asked Coch and French to identify a way to improve employees’ commitment to new production methods. Coch and French suspected that employees would respond more positively if they were involved in planning changes, so they devised three different training programs. Employees in the no-participation program were just given an explanation for the innovations. Those in the participation-through-representation program attended group meetings where the need for change was discussed openly and an informal decision was reached. A subgroup was then chosen to become the “special” operators who would serve as the first training group. Employees in the third program—total participation—followed much the same procedures as those in the second program, but here all the employees, not a select group, took part in the training system.

Confirming the voice effect, hostility, turnover, and inefficiency was highest in the no-participation group; 17% quit rather than learn the new procedures, and those who remained never reached the goals set by management. Those in the two participation conditions, in contrast, learned their new tasks quickly, and their productivity soon surpassed prechange levels and management goals. Morale was high, only one hostile action was recorded, and none of the employees quit in the 40 days following the change. Furthermore, when the members of the no-participation control condition were run through a program of increased voice and involvement several months later, they, too, reached appropriate production levels (cf. Bartlem & Locke, 1981). Autonomous work groups and self-directed teams, which were discussed in Chapter 11, are the modern-day counterparts to Coch and French’s total-participation groups (Cascio, 1995).

Evaluating the Plan

The Bay of Pigs invasion plan was implemented on April 17, 1961. It was a high-risk plan to begin with, but its implementation was fraught with its own ambiguities and inadequacies. The invasion failed not only because it was based on incorrect information about the region and the forces that were involved in the conflict, but also because the planners did not agree on the basic purpose of the mission. As the attack unfolded, changes were made to deal with circumstances, and these choices did not have the desired effects. The mission failed.

Kennedy and his staff did not execute the planning, discussion, and decision stages skillfully, but they did do well during the implementation stage. When it became clear that the decision was a faulty one, the group took steps to evaluate the cause of the fiasco. Kennedy created a task force headed by an outside expert rather than a member of the advisory group and provided considerable support to the fact-finding group. Kennedy then used that report to change the way his advisors made decisions.

Dysfunctional Postdecision Tendencies

Groups should expend considerable time and energy reviewing their efforts and outcomes. They should gather and weigh information about their performance and review the strategies they used to make their decision and implement their solution. But groups are often so confident that their solution is the correct one that the postdecision review process is unsystematic, attenuated, biased, or skipped altogether (Wilson, 2011).

Groups also prefer to deny responsibility should their decisions result in failure (Leary & Forsyth, 1987). The officers at the Enron Corporation, for example, when explaining their company’s abuses, blamed their accountants for using the wrong auditing methods. When the Challenger space shuttle exploded, NASA scientists blamed the engineers who designed the O-rings, but the engineers blamed NASA for launching the shuttle in cold weather. In the aftermath of the failed invasion of Cuba, President Kennedy publicly took the blame for the failure, but he also ended the careers of several group members who he believed were actually at fault.

12-2. Decisional Biases

Making a decision in a group offers a number of advantages over making a decision alone. Groups, with their greater informational resources and capacity to process that information, are better able to identify solutions and to detect errors in reasoning. Members may also find a group’s decision more satisfying than that of a single individual, particularly if the group uses a consensus-building decision process. But even though groups have the potential to outperform individuals, they do not always reach that potential. Groups do not always correct for the biases that plague individual decision makers—in fact, in some cases, they amplify them.

12-2a. Judgmental Biases

Psychologists have identified a number of cognitive and motivational biases that systematically distort our judgments and decisions. For example, we often use the information we have available to us inappropriately, putting too much emphasis on interesting information and ignoring statistical information. We sometimes form conclusions very quickly and then do not sufficiently revise those conclusions once we acquire additional information. When we cannot easily imagine an outcome, we assume that such an outcome is less likely to occur than one that springs easily to mind. We overestimate our judgmental accuracy because we remember all the times our decisions were confirmed and forget the times when our predictions were disconfirmed (Kahneman, 2011).

Decisional Sin

Groups, unfortunately, are not immune from these judgmental biases. When social psychologist Norbert Kerr and his colleagues reviewed the research literature looking for studies of these mental glitches in decision making, they identified the three general categories of potential bias summarized in Table 12.1:

Sins of commission: misusing information in some way, including continuing to based judgments on false or irrelevant information.

Sins of omission: failing to seek out information, overlooking useful information, or not checking for errors and mistakes.

Sins of imprecision: relying inappropriately on mental rules of thumb, or heuristics, that oversimplify the decision or introduce errors into the decision process.

Table 12.1 Types of Information Processing Errors Made by Individuals and by Groups When Making Decisions

Type of Error Examples

Sins of commission

Belief perseverance: reliance on information that has already been reviewed and found to be inaccurate.

Sunk cost bias: reluctance to abandon a course of action once an investment has been made in that action.

Extra-evidentiary bias: the use of information that one has been told explicitly to ignore.

Hindsight bias: the tendency to overestimate the accuracy of one’s prior knowledge of an outcome.

Sins of omission

Base rate bias: failure to pay attention to information about general tendencies.

Fundamental attribution error: stressing dispositional causes when making attributions about the cause of people’s behaviors.

Confirmation bias: failure to seek out information that may disconfirm a conclusion.

Sins of imprecision

Availability heuristic: basing decisions on information that is readily available.

Conjunction bias: failing to recognize that the probability of two events occurring together will always be less than the probability of just one of the events occurring.

Representativeness heuristic: excessive reliance on salient but misleading aspects of a problem.

SOURCE: Adapted from Kerr, N. L., MacCoun, R. J., & Kramer, G. P. (1996a). Bias in judgment: Comparing individuals and groups. Psychological Review, 103, 687–719.

After reviewing studies that compared individuals’ and groups’ resistance to these types of biases, Kerr and his colleagues cautiously ruled against groups: Groups amplify rather than suppress these biases. For example, they use information that has already been discredited or they have been told to ignore; they overlook statistical information about general tendencies; they overemphasize personality as a cause of behaviors that are due, in part, to the pressures of the situation; and they base decisions on information that is readily available rather than actually diagnostic. More so even than individuals, groups know decisional sin (Kerr, MacCoun, & Kramer, 1996a, 1996b).

Restoring Rationality

Groups can err in the ways noted in Table 12.1, but they can take steps to minimize their susceptibility to these biases. Discussion gives group members the opportunity to test their initial inclinations, but this review tends to be a biased one: People usually seek out information that confirms their preferences, and they avoid disconfirming evidence. Groups, however, can minimize this confirmation bias if they deliberately ban any public statements of initial preferences (Dawes, 1988).

Groups can also escape the confirmation bias by fully exploiting the diversity of members’ experiences and expertise. Researchers studied this possibility by creating three kinds of groups:

(1)

unanimous groups, composed of individuals who shared the same initial preference;

(2)

groups with one member who took a minority position on the issue; and

(3)

groups with two minority members.

Participants in these three conditions were given the opportunity to select and review background readings which either supported or opposed their initial preferences. A fourth set of participants made these choices as individuals. Everyone—both groups and individuals—sought out data that confirmed their initial preferences—a very robust confirmation bias. But the diverse groups were less biased than the groups composed of individuals who all agreed with each other. Groups with two dissenting members, in fact, even bested individual decision makers, reversing the tendency for groups to be less rational than individuals (Schulz-Hardt et al., 2000; see also Kerschreiter et al., 2008; Mojzisch et al., 2008).

Groups can also maximize the rationality of their members in one simple way: pay them when they make good decisions. Whereas social psychologists frequently find that the groups they study are error-prone, experimental economists find that their groups are more rational than individuals, particularly when “incentivized”: given a reward (usually monetary) for making good decisions (Charness & Sutter, 2012; Kugler, Kausel, & Kocher, 2012; Rockenbach, Sadrieh, & Mathauschek, 2007). Consider, for example, the famous Linda problem and the well-documented conjunction error (Tversky & Kahneman, 1983, p. 297).

Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

(a)

Linda is a bank teller.

(b)

Linda is a bank teller and is active in the feminist movement.

Many people get this question wrong. Distracted by the lack of representativeness between Linda’s attributes and their stereotypes about bank tellers, they pick (b). But the laws of probability tell us that the likelihood of a conjunction—two things occurring together—is always less probable than the likelihood of just one of the things occurring. People who are bank tellers and feminists are just a small subset of people who are bank tellers, so certainly the probability of (a) is greater than (b).

When investigators tested both individuals and groups with the Linda problem, they found that two-person and three-person groups outperformed individuals—groups of two made fewer errors than individuals, and 3-person groups even bested the 2-person groups. Both individuals and groups became even more rational, though, when accuracy earned them a reward—a $4 bonus if they answered correctly. This incentive essentially eliminated the conjunction error when individuals collaborated in small groups in selecting their answers (Charness, Karni, & Levin, 2010; see Figure 12.4).

12-2b. The Shared Information Bias

The Bay of Pigs planners spent much time talking about the incompetence of Castro’s forces and how U.S. citizens would react to the invasion. They did not spend as much time talking about the weapons that the troops would carry, the political climate in Cuba, or the type of communication system used by Cuban military forces. Only the CIA representatives knew that the morale of the invasion force was low, but they never mentioned that information during the discussion. Kennedy, as a result of other briefings, was privy to many facts relevant to the initiative, but he kept this information to himself (Kramer, 2008, 2016).

The Common Knowledge Effect

The good news is that groups can pool their individual resources to make a decision that takes into account far more information than any one individual can consider. The bad news is that groups often exhibit the shared information bias: they spend too much of their discussion time examining details that two or more group members know in common rather than information known by only one of them (Stasser, 1992; Wittenbaum et al., 2004). If all the groups’ members know that the majority of U.S. citizens oppose communism, they talk that over at length. But if only the CIA representative knows that the invading troops are poorly trained or only Kennedy knows that Cuban citizens support Castro, these important—but unshared—pieces of information might never be discussed. This tendency is also known as the common knowledge effect (Gigone, 2010; Gigone & Hastie, 1997).

The Hidden Profile Problem

The harmful consequences of this shared information bias are substantial when the group must consider carefully the unshared information to make a good decision. If a group is working on a problem when the shared information suggests that alternative A is correct, but the unshared information favors alternative B, then the group will only discover this so-called hidden profile if it discusses the unshared information. Social psychologists Garold Stasser and William Titus (1985) studied this problem by giving the members of four-person groups 16 pieces of information about three candidates for student body president. Candidate A was the best choice for the post, for he possessed eight positive qualities, four neutral qualities, and four negative qualities. The other two candidates had four positive qualities, eight neutral qualities, and four negative qualities. When the group members were given all the available information about the candidates, 83% of the groups favored candidate A—an improvement over the 67% rate reported by the participants before they joined their group. But groups did not fare so well when Stasser and Titus manipulated the distribution of the positive and negative information among the members to create a hidden profile. Candidate A still had eight positive qualities, but Stasser and Titus made certain that each group member received information about only two of these qualities. Person 1, for example, knew that candidate A had positive qualities P1 and P2; Person 2 knew that he had positive qualities P3 and P4; Person 3 knew that he had positive qualities P5 and P6; and Person 4 knew that he had positive qualities P7 and P8. But they all knew that candidate A had negative qualities N1, N2, N3, and N4.

Had the group members pooled their information carefully, they would have discovered that candidate A had positive qualities P1 to P8 and only four negative qualities. But they oversampled the shared negative qualities and chose the less qualified candidate 76% of the time (Stasser & Titus, 1985, 1987). Subsequent studies have repeatedly confirmed groups’ inability to dig deeper than the information known to all: Groups working with hidden profiles are eight times less likely to find the solution than are groups having full information (Lu, Yuan, & McLeod, 2012).

What Causes the Shared Information Bias?

The shared information bias reflects the dual purposes of discussion. As a form of informational influence, discussions help individuals marshal the evidence and information they need to make good decisions. But as a form of normative influence, discussions give members the chance to influence each other’s opinions on the issue. Discussing unshared information may be enlightening, but discussing shared information helps the group reach consensus. Hence, the bias is strongest when group members are seeking closure, they want to convince the group to back their initial preferences, or the problems are complex and have no clear right or wrong answer. The bias is attenuated if groups are working under time pressure and have relatively few alternatives to consider (for reviews, see Nijstad, 2009; Reimer, Reimer, & Czienskowski, 2010).

The shared information bias also reflects the psychological and interpersonal needs of the group members. If group members enter into the group discussion with a clear preference, they will argue in favor of their preference and resist changing their minds. If the shared information all points in one direction—as it did in Stasser and Titus’s (1985) study of hidden profiles—then all the group members begin the discussion with a negative opinion of candidate A. The group’s final choice reflects these initial preferences. Discussion should have caused them to change their opinions, as additional (unshared) information came to light, but it did not (Nijstad, 2009).

At the interpersonal level, discussions aren’t only about making good decisions. Members are striving to reach the best decision possible, but they have other motivations as well: They are trying to establish reputations for themselves, secure tighter bonds of attraction with others, and possibly compete with and succeed against other group members (Mojzisch et al., 2014; Wittenbaum et al., 2004). Therefore, they are selective regarding when they disclose information and to whom they disclose it, often emphasizing shared information to express their agreement with others in the group. Ironically, people consider shared information to be highly diagnostic, so they mistakenly believe that people who discuss shared information are more knowledgeable, competent, and credible than are group members who contribute unshared information to the discussion (Wittenbaum, Hubbell, & Zuckerman, 1999). Members, to make a good impression with the group, dwell on what everyone knows rather than on the points that only they understand. Group members who anticipate a group discussion implicitly focus on information that they know others also possess, instead of concentrating on their unique informational resources (Wittenbaum, Stasser, & Merry, 1996).

Can the Shared Information Bias Be Avoided?

Even though groups prefer to spend their time discussing shared information, experienced members avoid this tendency, and they often intervene to focus the group’s attention on unshared data (Wittenbaum, 1998). When researchers studied medical teams making decisions, they noted that the more senior group members repeated more shared information, but they also repeated more unshared information than the other group members. Moreover, as the discussion progressed, they were more likely to repeat unshared information that was mentioned during the session—evidence of their attempt to bring unshared information out through the discussion (Larson et al., 1996). Groups can also avoid the shared information bias if they spend more time actively discussing their decisions. Because group members tend to discuss shared information first, groups are more likely to review unshared information in longer meetings (Bowman & Wittenbaum, 2012; Winquist & Larson, 1998). Other methods of avoiding the bias include increasing the diversity of opinions within the group (Smith, 2008), using an advocacy approach rather than a general discussion (Greitemeyer et al., 2006), emphasizing the importance of dissent (Klocke, 2007), and introducing the discussion as a new topic (new business) rather than a return to a previously discussed item (Reimer, Reimer, & Hinsz, 2010).

Technology also offers a solution to the bias. Group decision support systems (GDSS) offer members a way to catalog, more comprehensively, the group’s total stock of information and then share that information collectively. Depending on the GDSS, the group would have access to an array of decision-making tools, such as databases, search engines for locating information, communication tools for sending messages to specific individuals and to the entire group, shared writing and drawing areas where members can collaborate on projects, and computational tools that will poll members automatically and help them to estimate costs, risks, probabilities, and so on (Hollingshead, 2001b). The value of a GDSS has not, however, been confirmed consistently by researchers. Some investigators have found that even a simple GDSS, that only promotes communication, reduces the biasing effects by a half (Lam & Schaubroeck, 2000). Other studies, however, suggest that the common knowledge effect is as common in online groups as it is in face-to-face groups (Lu, Yuan, & McLeod, 2012).

12-2c. Group Polarization

When President Kennedy took the invasion plan to his advisory group, was he acting on the intuitively appealing notion that groups have a moderating impact on individuals? Did he assume that a group, if faced with a choice between a risky alternative, such as “Invade Cuba,” and a more moderate alternative, such as “Use diplomatic means to influence Cuba,” would prefer the moderate route? Unfortunately for Kennedy, for his advisers, and for the members of the attack force, groups’ decisions are often more extreme than individuals’ decisions. Groups do not always urge restraint; instead, they often polarize.

Risky Shifts in Groups

At about the time that Kennedy’s committee was grappling with the problems inherent in the invasion plan, researchers were initiating the first experimental studies of groups making risky decisions. Many of these early studies used the Choice Dilemmas Questionnaire to measure the level of risk people were willing to accept (Stoner, 1961, 1968). This self-report measure asks respondents to read a series of scenarios that may or may not yield financial, interpersonal, or educational benefits (see Figure 12.5). They then indicate what the odds of success would have to be before they would recommend the course of action. The researchers expected that groups would be more cautious than individuals, but they instead discovered the risky-shift effect: Groups’ choices tended to be slightly riskier. This shift also occurred when individual judgments were collected after the group discussion and when the individual postdiscussion measures were delayed two to six weeks. (The delayed posttests were collected from male participants only.) Participants in a control condition shifted very little (Wallach, Kogan, & Bem, 1962).

Cautious Shifts in Groups

Researchers soon replicated the risky-shift effect in countries around the world, but during this research period some studies hinted at the possibility of the opposite process—a cautious shift. For example, when the early risky-shift researchers examined the amount of postdiscussion change revealed on each item of the Choice Dilemmas Questionnaire, they frequently found that group members consistently advocated a less risky course of action than did individuals on one particular item (Wallach et al., 1962). Intrigued by this anomalous finding, subsequent researchers wrote additional choice dilemmas, and they, too, occasionally found evidence of a cautious shift. Then, in 1969, researchers reported some groups moving toward risk after a discussion, but others becoming more cautious, suggesting both types of shifts were possible (Doise, 1969).

Researchers also discovered that group discussions can cause shifts in members’ attitudes, beliefs, values, judgments, and perceptions (Myers, 1982). In France, for example, where people generally like their government but dislike Americans, group discussions improved their attitude toward their government but exacerbated their negative opinions of Americans (Moscovici & Zavalloni, 1969). Similarly, strongly prejudiced people who discussed racial issues with other prejudiced individuals became even more prejudiced. However, when mildly prejudiced persons discussed racial issues with other mildly prejudiced individuals, they became less prejudiced (Myers & Bishop, 1970).

Polarization in Groups

Researchers eventually realized that risky shifts after group discussions were a part of a more general process. When people discuss issues in groups, they sometimes draw a more extreme conclusion than would be suggested by the average of their individual judgments. The direction of this shift depends on their average initial preferences. When supporters gather to discuss a candidate’s strengths and weaknesses, by the meeting’s end their opinions will likely become even more favorable toward the candidate. A gathering of students who are moderately negative about a professor’s teaching methods will become openly hostile after a discussion. Social psychologists David Myers and Helmut Lamm called this process group polarization because the “average postgroup response will tend to be more extreme in the same direction as the average of the pregroup responses” (Myers & Lamm, 1976, p. 603; see also Lamm & Myers, 1978). Group discussion polarizes, in the sense that it tends to push people away from the center toward the extremes.

Imagine two groups of four individuals whose opinions vary in terms of preference for risk. As Figure 12.6 indicates, when the average choice of the group members before discussion is closer to the risky pole of the continuum than to the cautious pole (as would be the case in a group composed of persons A, B, C, and D), a risky shift will occur. If, in contrast, the group is composed of persons C, D, E, and F, a cautious shift will take place, because the pregroup mean of 6.5 falls closer to the cautious pole. This example is, of course, something of an oversimplification, because the shift depends on the distance from the psychological rather than the mathematical midpoint of the scale. As Myers and Lamm (1976) noted, on choice dilemmas, an initial pregroup mean of 6 or smaller is usually sufficient to produce a risky shift, whereas a mean of 7 or greater is necessary to produce a cautious shift. If the pregroup mean falls between 6 and 7, a shift is unlikely.

Causes of Polarization

How do groups intensify individuals’ reactions? Early explanations suggested that groups feel less responsible for their decisions and are overly influenced by risk-prone leaders, but in time, investigators recognized that polarization results from social influence processes that operate routinely in groups, including social comparison, persuasion, and social identity processes (Friedkin, 1999).

Social comparison: When we make choices together, we use others as reference points to evaluate our own preferences and positions (Myers, 1978). If, for example, we discover the majority of the group likes risk, then to make a good impression we might claim that we really like risk (Weigold & Schlenker, 1991): “To be virtuous … is to be different from the mean—in the right direction and to the right degree” (Brown, 1974, p. 469).

Persuasive arguments: We change our opinions because we are persuaded by others’ arguments and ideas, and groups usually generate more arguments that support the position endorsed by the majority of the group or the position that is most consistent with dominant social values. As a result, the group persuades itself, as more arguments favoring the dominant viewpoint are brought up during the discussion (Burnstein & Vinokur, 1973, 1977; Vinokur & Burnstein, 1974, 1978). As we all restate our own and other members’ arguments during the discussion, we will gravitate toward which ever position is best supported by the pool of available arguments (Brauer, Judd, & Gliner, 1995).

Social identity: Curiously, at least for persuasive arguments theory, we sometimes shift our opinions when we discover others’ positions but not their arguments (Blascovich, Ginsburg, & Howe, 1975, 1976). Why? Social identity theory suggests that we are not persuaded by the content of other’s arguments, but by consensus of opinion. If, through discussion, we come to believe that the prototypical group member holds a relatively extreme attitude on the issue, if we identify with the group we will shift in that direction (Haslam, 2004). This shift causes the diversity of opinions in the group to decrease, as we all converge on what we hold to be the opinion of the prototypical group member. This conception of the prototype may also shift toward more extreme positions to differentiate the ingroup from other groups (Hogg, Turner, & David, 1990). We also respond more positively to the arguments offered by ingroup members than outgroup members, and so we end up persuading each other to take more extreme positions (Mackie & Queller, 2000).

The Consequences of Polarization

Would people who believe that environmental pollution is a serious problem be more likely, after their discussion, to insist on severe measures to prevent pollution? Would bringing together two sides in a community conflict and allowing them to meet separately for an hour before a joint meeting create even more tensions between the two groups? Would a group of government experts who slightly favored an invasion plan enthusiastically endorse the plan after they discussed it? Do groups amplify group members’ shared tendencies? Studies of polarization say yes (Sunstein, 2002).

Polarization may, in some cases, yield positive effects. Groups, when viewed from an evolutionary perspective, were designed to monitor risk, hence, they are sensitive to threats and urge caution when alarmed (Kameda & Tamura, 2007). A group’s collective efficacy may rise as individually optimistic members join together and discuss their chances for success. The members of a support group may become far more hopeful of their chances for recovery when they gather together with others who are moderately optimistic. Innovations and new ideas may be adopted by large numbers of people as polarization amplifies enthusiasm for the new products, methods, or outlooks. Polarization may also encourage the strengthening of positions within the group that might go unexpressed or even be suppressed. Thus, polarization, though sometimes a source of error and bias, can in some cases have a beneficial impact on the group and its members.

12-3. Victims of Groupthink

Social psychologist Irving Janis was intrigued by President Kennedy’s advisory group. The group, like so many others, failed to make the best decision it could. Its failure, though, was so spectacular that Janis wondered if something more than such common group difficulties as faulty communication and judgmental biases were to blame.

Janis pursued this insight by searching for other groups that made similar errors in judgment. And he found many that qualified: Senior naval officers who ignored repeated warnings of Japan’s aggressive intentions regarding Pearl Harbor and took few steps to defend it; President Truman’s policy-making staff who recommended that U.S. troops cross the 38th parallel during the Korean War, prompting China to ally with North Korea against the United States; President Nixon’s staff who decided to cover up involvement in the break-in at Watergate. After studying these groups and their gross errors of judgment, he concluded that they suffered from groupthink—”a mode of thinking that people engage in when they are deeply involved in a cohesive ingroup, when the members’ strivings for unanimity override their motivation to realistically appraise alternative courses of actions” (Janis, 1982, p. 9). During groupthink, members try so hard to agree with one another that they make mistakes and commit errors that could easily be avoided.

Janis identified both the causes of groupthink, as well as the symptoms that signal that a group may be experiencing this malady. As Figure 12.7 indicates, Janis identified three key sets of antecedent conditions that set the stage for groupthink, including cohesion, structural faults of the group or organization, and provocative situational contexts. These conditions cause members to seek out agreement with others (concurrence-seeking tendency), which in turn leads to two classes of observable consequences: symptoms of groupthink and symptoms of defective decision making. In this section, we will work backward through the model shown in Figure 12.7: We will start with the symptoms of groupthink and then consider the antecedent conditions.

12-3a. Symptoms of Groupthink

Like a physician who searches for symptoms that signal the onset of the illness, Janis identified a number of recurring patterns that occur in groupthink situations. He organized these symptoms into three categories: overestimation of the group, closed-mindedness , and pressures toward uniformity (Janis, 1972, 1982, 1989; Longley & Pruitt, 1980; Wheeler & Janis, 1980).

Overestimation of the Group

Groups that have fallen into the trap of groupthink are actually planning fiascoes and making all the wrong choices. Yet the members usually assume that everything is working perfectly. They even express enthusiasm in their public statements about their wrong-headed decisions (Tetlock, 1979). Janis traced this unwarranted optimism to illusions of invulnerability and illusions of morality.

Illusion of invulnerability: The Bay of Pigs planners overestimated their group’s decisional savvy. Members felt that they were performing well, even though they were not. Feelings of assurance and confidence engulfed the group. Such feelings of confidence and power may help athletic teams or combat units reach their objectives, but the feeling that all obstacles can be easily overcome through power and good luck can cut short clear, analytic thinking in decision-making groups (Silver & Bufanio, 1996).

Illusions of morality: The planners believed in the inherent morality of their group and its decisions. The plan to invade Cuba could unsympathetically be described as an unprovoked sneak attack by a major world power on a defenseless neighbor. Although groups are capable of reaching admirable levels of moral thought, this capability is unrealized during groupthink (Muehlheusser, Roider, & Wallmeier, 2015).

Closed-Mindedness

Groups that are overtaken by groupthink are not open-minded groups searching for new ideas and perspectives. Rather, they are closed-minded—rigidly shut off from alternatives, merely seeking to bolster their initial decision through rationalization and stereotypes.

Collective rationalization: Once the group began to lean in the direction of endorsing the plan, members began to discount information and opinions that called into question their preliminary decision: they failed to “reconsider their assumptions” since there was no need. They just dismissed objectives out of hand through shared (but unwarranted) justifications (Janis, 1982, p. 174).

Stereotyping: The members of the planning group shared an inaccurate and negative opinion of Castro and his political ideology. Castro was depicted as a weak leader, an evil communist, and a man too stupid to realize that his country was about to be attacked. His ability to maintain an air force was discredited, as was his control over his troops and the citizenry. The group participants’ underestimation of their enemy was so pronounced that they sent a force of 1,400 men to fight a force of 200,000 and expected an easy success.

Pressures toward Uniformity

The struggle for consensus is an essential and unavoidable aspect of life in groups, but in groupthink situations, interpersonal pressures make agreeing too easy and disagreeing too difficult. Tolerance for any sort of dissent is nil, and groups may use harsh measures to bring those who disagree into line. In the president’s committee, criticism was taboo, and members who broke this norm were pressured to conform. Janis highlighted four indicators of this pressure: self-censorship, the illusion of unanimity, direct pressure on dissenters, and self-appointed mindguards.

Self-censorship: In the planning group, many of the members of the group privately felt uncertain about the plan, but they kept their doubts to themselves. Some even sent private memorandums to the president before or after a meeting; but when the group convened, the doubting Thomases sat in silence. As Schlesinger (1965, p. 225) later wrote, “In the months after the Bay of Pigs I bitterly reproached myself for having kept so silent during those crucial discussions.”

Illusion of unanimity: The members seemed to agree that the basic plan presented by the CIA was the only solution to the problem. In later discussions, they appeared to just be “going through the motions” of debate. A “curious atmosphere of assumed consensus” (Schlesinger, 1965, p. 250) characterized the discussion, as each person wrongly concluded that everyone else liked the plan. As Janis (1972) explained, the planners apparently felt that it would be “better to share a pleasant, balmy group atmosphere than be battered in a storm” (p. 39).

Direct social pressure: The group’s easygoing, supportive atmosphere did not extend to those who disagreed with the group. As the group’s support for the plan grew stronger, individuals who disagreed were pressured to keep their doubts to themselves. Kennedy, the group’s leader, displayed strong support for the CIA representatives, and would permit them to counterargue any objections that other group members raised.

Mindguards: Some members of the group shielded the group from information that would shake the members’ confidence in themselves or their leader. These so-called mindguard diverted controversial information away from the group by losing it, forgetting to mention it, or deeming it irrelevant and thus unworthy of the group’s attention. Alternatively, the mindguard would take dissenting members aside and pressure them to keep silent. Kennedy, for example, withheld memorandums condemning the plan from both Schlesinger and Fulbright. Rusk suppressed information that his own staff had given him. One extreme example of this mindguarding occurred when Rusk, unable to attend a meeting, sent Undersecretary of State Chester Bowles. Although Bowles was said to be horrified by the plan under discussion, President Kennedy never gave him the opportunity to speak during the meeting. Bowles followed bureaucratic channels to voice his critical misgivings, but his superior, Rusk, did not transmit those concerns to the committee, and he told Bowles that the plan had been revised. Bowles was fired several weeks after the Bay of Pigs defeat—partly because a scapegoat was needed, but also because President Kennedy disliked him intensely (Kramer, 2008).

When Is Agreement Difficult to Manage?

The day was hot and dusty, as was often the case in July in the small town of Coleman, Texas. Jerry Harvey, his wife, and his wife’s parents were fanning themselves on the back porch, playing dominoes and drinking lemonade. Suddenly, Jerry’s father-in-law suggested, “Let’s get in the car and go to Abilene and have dinner at the cafeteria” (Harvey, 1988, p. 13). Abilene was 53 miles away, it was 104 degrees in the shade, and the only available means of transportation was an unairconditioned 1958 Buick. So they drove all the way to Abilene, had a miserable time, and only when they were back on the porch did they realize that none of them had wanted to go in the first place.

After blaming each other for the bad decision, we all sat back in silence. Here we were, four reasonably sensible people who—of our own volition—had just taken a 106-mile trip across a godforsaken desert in furnace-like heat and dust storm to eat unpalatable food in a hole-in-the-wall cafeteria in Abilene, when none of us had really wanted to go. To be concise, we’d just done the opposite of what we wanted to do. (Harvey, 1988, p. 14)

Groups sometimes make decisions that veer far from the plans, desires, and preferences of their individual members. Organizational expert Jerry Harvey’s Abilene paradox aptly illustrates this tendency, highlighting two factors that can cause members to mismanage their group’s agreement. First, the Abilene group suffered from a severe case of pluralistic ignorance (see Chapter 6). The group members mistakenly believed that their private opinion about the Abilene outing was discrepant from the other group members’ opinions. Therefore, each group member, wishing to be seen as a cooperative member of the family, publicly conformed to what they thought was the group’s norm, each one erroneously assuming that he or she was the only one with misgivings. Jerry went to Abilene because that is what everyone else wanted to do—or so he thought. Unfortunately, everyone else was thinking the same thing, so the group mismanaged its consensus.

Second, the group committed to its decision quickly and did not reconsider its choice when negative consequences—the heat, the cost, and the discomfort—mounted. This process is sometimes termed entrapment—a special form of escalation that occurs when the group expends “more of its time, energy, money, or other resources than seems justifiable by external standards” (Pruitt & Kim, 2004, p. 165). Entrapment occurs when groups become so invested in a course of action that they refuse to reverse their decisions (Brockner, 1995; Brockner & Rubin, 1985). Such situations often lure in groups by raising concerns over investments the group has already made in the choice (Arkes & Blumer, 1985). If a group discovers that the costs for a project are escalating, the members will rarely consider canceling the project altogether. Instead, they will continue to fund the project, because the initial investment, or sunk cost, must be honored. Unfortunately, the money and time invested in a plan of action is already spent and should no longer be considered in weighing the ultimate value of the project. Sunk cost, however, can cause groups to continue to expend resources on projects that will ultimately fail. Analyses of truly massive, much-criticized projects that cost millions of dollars—such as the Millennium Dome in London, EuroDisney, and the Denver International Airport—can be traced to entrapment (Nutt, 2002).

12-3b. Defective Decision Making

Janis did not consider the group to be one overtaken by groupthink only because it made a bad decision, but because it displayed symptoms of groupthink and symptoms of defective decision making. The committee, for example, discussed two extreme alternatives—either endorse the Bay of Pigs invasion or abandon Cuba to communism—while ignoring all other potential alternatives. The group actively avoided any information that pointed to limitations in its plans, while seeking out facts and opinions that buttressed its initial preferences. The group members did not just make a few small errors. They committed dozens of blunders. The invasion was a fiasco, but it was the faulty decisional strategies of the group that indicated that the group suffered from groupthink.

12-3c. Causes of Groupthink

To Janis, groupthink is like a disease that infects healthy groups, rendering them inefficient and unproductive. The symptoms of this disease signal the group’s decline, but they are not the root causes of groupthink. These processes contribute to poor judgments, but Janis (1989) distinguished between symptoms of groupthink and its causes: cohesiveness, structural faults of the group or organization, and provocative situational factors (see Figure 12.7).

Cohesiveness

The members of the president’s committee felt fortunate to belong to a group that boasted such high morale and esprit de corps. Problems could be handled without too much internal bickering, personality clashes were rare, the atmosphere of each meeting was congenial, and replacements were never needed, because no one ever left the group. However, these benefits of cohesiveness did not offset one fatal consequence of a close-knit group—group pressures so strong that critical thinking degenerates into groupthink.

Of the many factors that contribute to the rise of groupthink, Janis emphasized cohesiveness above all others. He agreed that groups that lack cohesion can also make terrible decisions—”especially if the members are engaging in internal warfare”—but they cannot experience groupthink (Janis, 1982, p. 176). In a cohesive group, members refrain from speaking out against decisions, avoid arguing with others, and strive to maintain friendly, cordial relations at all costs. If cohesiveness reaches such a level that internal disagreements disappear, then the group is ripe for groupthink.

Measures of cohesiveness were, of course, never collected for the president’s committee. But many signs point to the group’s unity. The committee members were all men, and they were in many cases close personal friends. These men, when describing the group in their memoirs, lauded the group, suggesting that their attitudes toward the group were exceptionally positive. The members also identified with the group and its goals; all proudly proclaimed their membership in such an elite body. Robert Kennedy’s remarks, peppered with frequent use of the words we and us, betrayed the magnitude of this identification:

It seemed that with John Kennedy leading us and with all the talent he had assembled, nothing could stop us. We believed that if we faced up to the nation’s problems and applied bold, new ideas with common sense and hard work, we would overcome whatever challenged us. (quoted in Guthman, 1971, p. 88; italics added)

Other evidence, however, suggests that the advisory group was not as unified as Janis believed. The membership of the group was not stable, so different people were present at different meetings, and therefore it is likely that no strong sense of identity actually developed. Also, like many groups composed of influential, successful individuals, personalities and differences in style and strategy caused tension within the group. To a large extent, group members were not motivated by group-centered motives, but by their own political ambitions (Kramer, 2008, 2016).

Structural Faults of the Group or Organization

Cohesion is a necessary condition for groupthink, but the syndrome is more likely to emerge when the group is organized in ways that inhibit the flow of information and promote carelessness in the application of decision-making procedures. Insulation of the group from other groups, for example, can promote the development of unique, potentially inaccurate perspectives on issues and their solution. The Bay of Pigs planners worked in secret, so very few outsiders ever came into the group to participate in the discussion. The committee was insulated from criticism. Many experts on military questions and Cuban affairs were available and, if contacted, could have warned the group about the limitations of the plan, but the committee closed itself off from these valuable resources.

President Kennedy’s leadership style also shaped the way the Bay of Pigs planners worked and may have contributed to groupthink. By tradition, the committee meetings, like cabinet meetings, were very formal affairs that followed a rigid protocol. The president could completely control the group discussion by setting the agenda, permitting only certain questions to be asked, and asking for input only from particular conferees (Stasson, Kameda, & Davis, 1997). He often stated his opinion at the outset of each meeting; his procedures for requiring a voice vote by individuals without prior group discussion paralleled quite closely the methods used by Asch (1952) to heighten conformity pressures in discussion groups. Ironically, Kennedy did not give his advisors opportunities to advise him (Kowert, 2002).

Provocative Situational Context

A number of provocative situational factors may push the group in the direction of error rather than accuracy. As humans tend to be reluctant decision makers in the best of circumstances, they can unravel when they must make important, high-stakes decisions. Such decisions trigger greater tension and anxiety, so group members cope with this provocative decisional stress in less than logical ways. Through collective discussion, the group members may rationalize their choice by exaggerating the positive consequences, minimizing the possibility of negative outcomes, concentrating on minor details, and overlooking larger issues. Because the insecurity of each individual can be minimized if the group quickly chooses a plan of action with little argument or dissension, the group may rush to reach closure by making a decision as quickly as possible (Callaway, Marriott, & Esser, 1985). Janis also suggested that any factors that work to lower members’ self-esteem, such as a history of mistakes or prior lapses of morality, may further increase the possibility of groupthink.

12-3d. The Emergence of Groupthink

Because of the complexity of the groupthink model, few tests of the entire model have been conducted. Researchers have, however, attempted to replicate Janis’s findings through archival case studies of other historical and political groups. They have also examined specific aspects of the theory—such as the impact of cohesion and stress on decision-making groups—to determine if its key assumptions hold up under empirical scrutiny. These studies, which are reviewed briefly here, sometimes support, sometimes challenge, and sometimes clarify Janis’s theory.

Archival Case Studies

Janis, using an archival method, compared groups that made very poor decisions to groups that made excellent choices to determine if error-prone groups exhibited more of the symptoms of groupthink. In later work, he enlarged his pool of cases to a total of 19 decision-making groups and had external raters who worked from the same historical texts rate the groups’ symptoms. As predicted, the higher the number of groupthink symptoms, the more unfavorable the outcome of the group’s deliberations (

𝑟

=

.64

; Herek, Janis, & Huth, 1987, 1989; Welch, 1989).

Other archival studies, too, have supported the key elements of Janis’s groupthink model (Esser, 1998; Turner & Pratkanis, 1998b). One study of the content of leaders’ public speeches when in groupthink and vigilant decision-making situations, for example, revealed signs of reduced complexity and ingroup favoritism in leaders of groupthink groups (Tetlock, 1979). Another indicated that the structural faults identified by Janis were related to groupthink, but cohesiveness and provocative situational context factors were not (Peterson et al., 1998; Tetlock et al., 1992). When political scientists coded governmental groups that made both good decisions (e.g., Clinton’s handling of the proliferation of nuclear capabilities in South Asia) and bad decisions (e.g., Reagan’s funding of the Contras) on the variables specified by Janis in his theory, they found considerable correspondence between variables Janis identified and symptoms displayed by poor-performing groups—although, their analysis suggested leadership was even more influential than what Janis suggested (Schafer & Crichlow, 2010).

Cohesion and Groupthink

Janis maintained that groupthink was a characteristic of cohesive groups only. If a group lacked cohesion, it might make poor decisions, but those decisions would be due to processes other than groupthink. His basic prediction was that cohesion, combined with one or more of the other potential causes of groupthink (e.g., structural faults and provocative situational context), would trigger groupthink. He admitted that cohesive groups are not necessarily doomed to be victims of groupthink, but “a high degree of group cohesiveness is conducive to a high frequency of symptoms of groupthink, which, in turn, are conducive to a high frequency of defects in decision-making” (Janis, 1972, p. 199).

A meta-analytic review of the results of seven studies involving more than 1,300 participants provided some support for this prediction (Mullen et al., 1994). High cohesiveness impaired decision making, provided that one or more of the other triggering conditions for groupthink were present in the situation. If the other causes of groupthink were absent, then cohesiveness increased the quality of a group’s decision-making processes. The cohesion–groupthink relationship may also depend on the source of the group’s cohesion. In a laboratory study of women college students, groups that derived their cohesiveness from their members’ commitment to the task displayed significantly fewer symptoms of groupthink, whereas groups that were interpersonally cohesive displayed more symptoms of groupthink (Bernthal & Insko, 1993). However, a case study of two groups that made catastrophic errors and displayed many of the symptoms of groupthink described by Janis displayed little social cohesion, but instead were united by task cohesion (Burnette, Pollack, & Forsyth, 2011).

Structural Faults and Groupthink

Janis identified several structural features of groups that can contribute to groupthink, but researchers have concentrated most of their attention on the group leader (Chen et al., 1996; Flowers, 1977). In one project, group members discussed evidence pertaining to a civil trial. Researchers told some of the groups’ assigned leaders to adopt a closed style of leadership: They were to announce their opinions on the case prior to the discussion. Open-style leaders were told to withhold their own opinions until later in the discussion. Groups with a leader who adopted a closed style were more biased in their judgments, particularly when many of the group members had a high need for certainty (Hodson & Sorrentino, 1997). Groups with leaders with a strong need for power also performed less effectively, irrespective of the group’s level of cohesion (Fodor & Smith, 1982). Other evidence, however, suggests that leaders who are highly directive improve their group’s decisions, provided that they limit their control to the group’s decisional processes rather than the group’s decisional outcomes (Peterson, 1997).

Provocative Situational Context

Studies of groups under stress suggest that they are more likely to make errors, lose their focus on their primary goals, and make use of procedures that members know have not been effective in the past. Social psychologist Janis Kelly and her colleagues, for example, have documented the negative impact of time pressures on both group performance and process (Kelly & Karau, 1999; Kelly & Loving, 2004). They find that when groups work under time pressure, members focus much of their attention on the task, but doing so leaves them at risk of overlooking important contextual information. They also tend to concentrate on getting the task completed quickly and so become more concerned with efficiency and quick results rather than accuracy and quality. In consequence, time pressures cause groups to “produce a less creative, less adequate, and less carefully reasoned decision. However, when a decision is routine or straightforward, these strategies can lead to adequate or even good decision making” (Kelly & Loving, 2004, p. 186).

12-3e. Alternative Models

Groupthink is not an obscure idea known only to those who study groups. A mere three years after the publication of Janis’s 1972 analysis, the term groupthink appeared in Webster’s New Collegiate Dictionary (Turner & Pratkanis, 1998b). The theory offers insight into very puzzling groups—those that make wrong-headed decisions—and has been applied to political decision makers, cults, businesses, and communities. In 2004, for example, the U.S. Senate Select Committee on Intelligence concluded that the intelligence community of the U.S. government had displayed a number of the symptoms of groupthink when it erroneously concluded that the country of Iraq was assembling weapons of mass destruction (U.S. Senate Select Committee on Intelligence, 2004). The theory serves as a reminder that if we are to understand political events that change the lives of people the world over, we must understand groups.

Researchers, however, continue to debate the validity of the model itself (Baron, 2005). Some suggest that it should be drastically revised. Others feel that the jury is still out and encourage more research. Others have proposed alternative models.

Group-Centrism Theory

Social psychologist Arie Kruglanski and his colleagues (2006), like Janis, have identified a syndrome that characterizes groups and often causes them to make faulty decisions. They term this syndrome group-centrism, because it springs primarily from the group members’ striving to maintain and support their group’s unity. Group-centric groups tend to rush to make judgments on the basis of insufficient information, particularly if they face situations that interfere with their capacity to process information—time pressures, severe ambiguity, noise, or fatigue. They are more likely to reject a member who disagrees with the group, and they express a strong desire for agreement with other members. Stereotyped thought and tendencies to favor the ingroup over the outgroup increase and willingness to compromise in order to reach integrative solutions during bargaining decreases. The group also strives for cognitive closure—”a desire for a definite answer to a question, any firm answer, rather than uncertainty, confusion, or ambiguity” (Kruglanski et al., 2002, p. 649)—and so adopts a more centralized structure with autocratic leaders. These groups’ discussions are dominated by high-status group members who have a much greater impact on the group’s communications and decisions than the rank-and-file members. These consequences of group-centrism are consistent with the symptoms of groupthink identified by Janis (De Dreu, 2003).

Are You Quick to Reach a Decision?

Some group members have more tolerance for long-winded, drawn out discussions of relatively mundane topics: others don’t. Social psychologist Arie Kruglanski and his colleagues (1993) developed the Need for Closure Scale to measure people’s desire to wrap things up quickly and move on.

Instructions. Put a check by each item that you feel accurately describes you in most group situations.

Preference for Order

I find that a well-ordered life with regular hours suits my temperament.

I hate to change my plans at the last minute.

I think that having clear rules and order at work is essential for success.

Preference for Predictability

I don’t like to be with people who are capable of unexpected actions.

I don’t like to go into a situation without knowing what I can expect from it.

I prefer to socialize with familiar friends because I know what to expect from them.

Decisiveness

I usually make important decisions quickly and confidently.

When faced with a problem I usually see the one best solution very quickly.

I do not struggle with most decisions.

Discomfort with Ambiguity

I don’t like situations that are uncertain.

In most social conflicts, I can easily see which side is right and which is wrong.

It’s annoying to listen to someone who cannot seem to make up his or her mind.

Closed-Mindedness

I feel irritated when one person disagrees with what everyone else in a group believes.

I dislike questions which could be answered in many different ways.

I do not usually consult many different opinions before forming my own view.

Interpretation. Only a sample of the 42 items from the Need for Closure Scale are included here, and the original scale uses a different response scale. If, however, you checked any two of the three items in any set, you may be predisposed to prefer order, predictability, and decisiveness, but you are uncomfortable with ambiguity. The final set of items measures the tendency to close your mind to possible alternatives (Kruglanski, Webster, & Klem, 1993).

Social Identity and the Ubiquity Model

Social psychologist Robert Baron (2005), after reviewing much of the existing research on Janis’s theory, agrees with Janis that members of groups often strive for consensus and that, in doing so, they tend to limit dissent, denigrate the outgroup, and misjudge their own group’s competence. Baron’s ubiquity model of groupthink, however, suggests that these qualities are ubiquitous features of groups, rather than rare ones. They only lead to problems, Baron suggests, when three conditions are met. First, it is not group unity per se that increases groupthink symptoms, but rather a threat to a shared social identity that may result should the group fail (Haslam et al., 2006; Turner & Pratkanis, 1998a). Second, the group must be one that has developed a set of norms that constrains members’ opinions with regard to the topic under discussion. Third, groupthink is more likely if group members lack self-confidence. In such cases, they are likely to rely on others’ judgments, with the result that the group does not adequately consider its alternatives (Sniezek, 1992).

12-3f. Preventing Groupthink

Kennedy did not take his Bay of Pigs failure lightly. In the months following the defeat, he explored the causes of his group’s poor decision making. He fired those he felt had misled him, put in place improved procedures for handling information, and learned how to decipher messages from his military staff. These changes prepared him for the next great issue to face his administration—the Cuban missile crisis of 1962. When Kennedy learned that the Soviet Union was constructing a missile base in Cuba, he called in his advisors again. But this time, the group made the right decision. Many of the same people meeting in the same room and guided by the same leader worked equally hard under similar pressures. Both crises occurred in the same area of the world, involved the same foreign powers, and could have led to equally serious consequences. Why did the missile crisis advisors succeed when the Bay of Pigs committee had failed?

Limiting Premature Seeking of Concurrence

If conformity was the norm in the Bay of Pigs group, dissent was championed by the group during the missile crisis. Kennedy deliberately suspended the rules of discussion that guided such meetings; agendas were avoided, and new ideas were welcomed. Although pressures to conform surfaced from time to time during the discussion, the members felt so comfortable in their role as skeptical, critical thinkers that they were able to resist the temptation to go along with the consensus. In fact, the group never did reach 100% agreement on the decision to turn back the Soviet ships.

The atmosphere of open inquiry can be credited to changes designed and implemented by Kennedy. He dropped his closed style of leadership to become an open leader as he

(1)

carefully refused to state his personal beliefs at the beginning of the session, waiting instead until others had let their views be known;

(2)

required a full, unbiased discussion of the pros and cons of each possible course of action;

(3)

convinced his subordinates that he would welcome healthy criticism and condemn “yea-saying”;

(4)

arranged for the group to meet without him on several occasions; and

(5)

encouraged specific members of the group to take the role of dissenter, or devil’s advocate, during the group discussions.

Kennedy also reduced pressures to conform by sometimes breaking the advisors into separate groups to reduce their size and spur increased debate (Wheeler & Janis, 1980). Kennedy also gave the group an official name and identity. In a memorandum dated October 22, 1962, Kennedy established, named, and populated the “Executive Committee of the National Security Council” and mandated that this group would meet regularly for the purpose of conducting governmental operations.

Correcting Misperceptions and Biases

The new and improved advisory committee did not take success for granted. The group recognized the challenge they faced and did not try to ease their discomfort by overestimating American superiority, belittling the Russians, and denying the magnitude of the dangers. According to the official version of this incident, no trace of the illusion of superiority that had permeated the planning sessions of the Bay of Pigs invasion was in evidence during the executive committee meetings. Each solution was assumed to be flawed and, even when the blockade had been painstakingly arranged, the members developed contingency plans in case it failed.

As members admitted their personal inadequacies and ignorance, they willingly consulted experts who were not members of the group. No group member’s statements were taken as fact until independently verified, and the ideas of younger, low-level staff members were solicited at each discussion. Participants also discussed the group’s activities with their own staffs and entered each meeting armed with the misgivings and criticisms of these unbiased outsiders.

The committee discussed the ethics of the situation and the proposed solutions. For example, although some members felt that the Russians had left themselves open to any violent response the Americans deemed appropriate, the majority argued that a final course of action had to be consistent with “America’s humanitarian heritage and ideals” (Janis, 1972, p. 157). Illusions of morality and invulnerability were supposedly minimized along with biased perceptions of the outgroup (see, for an alternative interpretation of this incident, Alterman, 2004).

Using Effective Decision-Making Techniques

The executive committee is not an example of an effective decision-making body simply because its solution to the missile crisis worked. Rather, just as the decision-making methods used by the Bay of Pigs committee ensured its failure, the executive committee’s use of effective decision-making techniques increased its chances of success (‘t Hart, 1998). Members analyzed a wide range of alternative courses of action, deliberately considered and then reconsidered the potential effects of their actions, consulted experts, and made detailed contingency plans in case the blockade failed to stop the Russians. Many initially favored military intervention, but the majority of the group’s members insisted that other alternatives be explored. This demand led to an expanded search for alternatives, and soon the following list emerged:

Do nothing.

Exert pressure on the Soviet Union through the United Nations.

Arrange a summit meeting between the two nations’ leaders.

Secretly negotiate with Castro.

Initiate a low-level naval action involving a blockade of Cuban ports.

Bombard the sites with small pellets, rendering the missiles inoperable.

Launch an air strike against the sites with advance warning to reduce loss of life.

Launch an air strike without advance warning.

Carry out a series of air attacks against all Cuban military installations.

Invade Cuba.

Once this list was complete, the men focused on each course of action before moving on to the next option. They considered the pros and cons, fleshed out unanticipated drawbacks, and estimated the likelihood of success. During this process, outside experts were consulted to give the members a better handle on the problem, and contingency plans were briefly explored. Even those alternatives that had initially been rejected were resurrected and discussed, and the group invested considerable effort in trying to find any overlooked detail. When a consensus on the blockade plan finally developed, the group went back over this alternative, reconsidered its problematic aspects, and meticulously reviewed the steps required to implement it. Messages were sent to the Russians, military strategies were worked out to prevent any slipups that would escalate the conflict, and a graded series of actions was developed to be undertaken should the blockade fail. Allies were contacted and told of the U.S. intentions, the legal basis of the intervention was established by arranging for a hemisphere blockade sanctioned by the Organization of American States; African countries with airports that could have been used by Russia to circumvent the naval blockade were warned not to cooperate. To quote Robert Kennedy, “Nothing, whether a weighty matter or a small detail, was overlooked” (1969, p. 60).

Resources

Chapter Case: The Bay of Pigs Planners

Bay of Pigs: The Untold Story by Peter Wyden (1979) offers a wealth of detail about the group that planned the invasion and draws on personal interviews with many of the original group members.

“Ships in the Night: The CIA, the White House and the Bay of Pigs” by Piero Gleijeses (1995) relies on interviews with group members to develop an explanation for the advisory committee’s error in judgment that emphasizes mistakes in communication and misunderstanding.

Making Decisions in Groups

“A Look at Groups from a Functional Perspective” by Andrea A. Hollingshead, Gwen M. Wittenbaum, Paul B. Paulus, Randy Y. Hirokawa, Deborah G. Ancona, Randall S. Peterson, Karen A. Jehn, and Kay Yoon (2005) uses a functional perspective to examine group composition, structure, and task performance.

“Decision Making in Groups and Organizations” by R. Scott Tindale and Katharina Kluwe (2015) systematically reviews the basic processes that groups use to make decisions, including aggregation, decision markets, advocacy, and full discussion to consensus.

The Cambridge Handbook of Meeting Science, edited by Joseph A. Allen, Nale Lehmann-Willenbrock, and Steven G. Rogelberg (2015a), provides a solid foundation for expanding the study of meetings, with sections devoted to premeeting activities, meeting dynamics, and tools and models for making meetings more successful.

“Team Decision Making” by Tom W. Reader (2017) is a concise analysis of decision making in groups, but with a focus on highly organized, purposeful groups (teams) in applied settings.

Groupthink

“Presidential Leadership and Group Folly: Reappraising the Role of Groupthink in the Bay of Pigs Decisions” by Roderick M. Kramer (2008) uses declassified documents and historical records not available to Janis before concluding that many aspects of the Bay of Pigs decision are inconsistent with the groupthink theory.

The Polythink Syndrome: U.S. Foreign Policy Decisions on 9/11, Afghanistan, Iraq, Syria, and ISIS by Alex Mintz and Carly Wayne (2016) offers a different explanation for the many mistakes made by policy makers: a plurality of opinions that leads to disunity and fragmented policies.