Training in Teams
Training is instruction aimed at procedural knowledge and proficiency, at knowing how to execute the procedures necessary to do a job. It can be distinguished from declarative knowledge, which is knowledge of facts or static information. Training programs have three general goals: successful training, transfer to the work situation (generalization), and long-term use on the job of what was learned (maintenance). Training can be structured for trainees to learn individually (either in competition with peers or on their own) or in teams. Training can focus on individuals who are selected from their job situations (or from a general population of potential job applicants), assigned to training teams and given training, and then returned to their job situations (where the trainees work alone or as part of a team, which may be nested in a network of teams). Or training can focus on cohort, intact teams that are given training as a team at a training site and then returned to the job site. Team training can be defined as training in which teams are used to increase individual procedural knowledge and proficiency in doing a job (taskwork), individual procedural knowledge and proficiency in functioning as part of a team (teamwork), and overall team performance. Team training can be differentiated from team building, which takes place at the job site and focuses on analyzing a work team's procedures and activities to improve the team's productivity (see Chapter 6). Team training can also be differentiated from cooperative learning, which usually takes place in schools and is focused primarily on declarative knowledge, rather than procedural knowledge and proficiency (see Chapter 5).
When team training takes place at a training site, it has four elements (see Figure 7-1): inputs (individual trainees or cohort teams, resources,
task characteristics, preparatory activities); training processes (training for taskwork, teamwork, and team performance); mediating variables (positive interdependence, promotive interaction, individual accountability, and team processing); and outcomes (individual and team proficiency on the job; improved relationships; individual psychological health; changes in a team's norms, roles, mental models, and communication patterns; and team activities after training). This chapter discusses each of these elements in turn, which are presented in terms of a framework that serves to organize the available research. The discussion results in a set of guidelines for research and practice, but the guidelines must be tempered by awareness of some limitations of the research.
The research on team training covers two interrelated literaturesa professional literature and a scientific literature. The professional literature tends to emphasize applied studies that have demonstrated that team training works in specific settings. This literature consists primarily of case studies that present descriptions of the effects of using specific training programs in specific situations. In contrast, the scientific literature consists of carefully controlled research studies conducted to validate or disconfirm theory. Most of them have been conducted in either laboratory or field experimental settings. They often lack external validitythat is, they are not conducted in workplace settingswhich reduces their credibility among practitioners. The professional and scientific studies, however, complement each other and strengthen each other's findings.
There are a number of problems with the professional literature on team training, some of which spill over into the scientific literature. First, the definitions of teams and team training are not unambiguous. It is difficult to differentiate between what is a team, a small group, a working group, an organization, or even a family. Without a clear conceptualization of what is and is not a team, it is difficult to build a coherent program of research. A related second problem is that it is difficult to summarize what is known about the use of teams for training purposes because of the many different kinds of teams. Some researchers have even studied "teams" made up of coacting individuals who never see or interact with each other. Often the units that have been studied are not described very clearly.
Third, much of the literature on team training lacks a theoretical base. This leaves research on team training without a direction. In 1982 McGrath and Kravitz concluded that the area of team performance was dominated by "atheoretical" (even antitheoretical) viewpoints, and things have not improved since then. A related fourth problem is that much of the professional research on team training was designed to show that a particular type of training is effective, but not why, when, and for whom it is so: that is, the variables mediating the effectiveness of team training programs have been relatively ignored. Fifth, the nature and purpose of the team training
being studied is rarely completely described. Researchers have tended to consider all training the same, without regard to the purpose of the training or the type of learning involved. To compare the results of different studies, reviewers need clear descriptions of the training methods used and the nature and purpose of the training. Finally, many of the studies lack methodological rigor, focusing far more on external validity (such as being conducted in a real-world setting) rather than internal validity (such as experimental control).
This chapter emphasizes the scientific literature. Some of the professional literature is cited to illustrate points, but the review covers primarily the knowledge generated in social and organizational psychology about some consequences of training individuals in teams and the effects of team structures on performance. The review of evidence in this chapter does not lead to firm conclusions. Rather, we draw from the literature general guidelines that support team training and specific ideas that merit further investigation.
For team training at a site, the inputs include the trainees, the resources required to conduct the training, the characteristics of the task, and preparatory activities of the team.
Trainees can come as intact, cohort teams from a work site, individuals from different work sites, or as individuals who wish to qualify for a certain job. There are advantages to training cohort teams rather than individuals. According to the evidence reviewed by Katz and Kahn (1978), a traditional error in training programs is to train individuals while ignoring the systemic properties of the organization in which they work. Taking members out of the organization, giving them special training, and then returning them to the organization is a poor strategy for changing their performance because it ignores the power of role expectations, organizational norms, and other organizational variables in determining organizational behavior. When isolated individuals are trained and then returned to the job setting, the organizational pressures on them tend to be towards going "back to practice as usual," which causes an abandonment and decay in what is learned (Katz and Kahn, 1978). Training intact teams seems to mitigate such pressures by providing a mutual redefinition of role responsibilities and job procedures and social support for implementing and maintaining the procedures learned in the training program.
Resources and Task Characteristics
The resources required to conduct training include personnel (instructors and support personnel), materials, and equipment. Besides classrooms, desks, overhead projectors, and so forth, equipment includes computers, electronic networking systems, and simulators. Although there has been research on the effectiveness of training for different types of tasks, perhaps the most important distinction is whether the task is simple or complex. Different types of training procedures are usually needed for simple and complex tasks.
Prior to formal sessions, trainees can be prepared for a team training experience in a number of ways. First, trainees can be helped to set goals for what they will learn and how hard they will work during the training. Cohen (1990) found that trainees who set goals prior to training began with higher levels of motivation to learn. Second, an accountability system can be provided to motivate trainees to take the training seriously and to use what they learn on the job. Trainers can inform members that how hard they work and how much they learn during training will be directly assessed after the training is over. Baldwin and Magjurka (1991) found that trainees who entered training expecting some form of follow-up activity or assessment afterward reported stronger intentions to transfer what they learned to the job. The fact that an organization would require them to prepare a posttraining report or undergo an assessment meant that they were being held accountable for their learning and apparently conveyed the message that the training was important.
Third, trainers can lessen constraints to using what is learned on the job. Trainees can be assured that what they learn in training will be used on the job and shown how constraints for doing so will be reduced. Mathieu et al. (1990) found that trainees who reported many situational constraints in their jobs (e.g., lack of time, equipment, and resources) entered training with lower motivation to learn. The trainees had little incentive to learn new skills when they worked in an environment in which the skills could not be applied.
Fourth, trainers can create positive expectations toward the training. In a study of five companies, Cohen (1990) found that trainees with more supportive supervisors entered training with stronger beliefs that training would be useful. Supervisors can show their support for training by discussing it with an employee, establishing training goals, providing time to prepare, and generally encouraging the employee.
Fifth, the more members value the training, the more likely they are to volunteer to participate in it. Volunteering to participate in the training, compared with mandatory attendance, tends to be related to higher motiva-
tion to learn, greater learning, and more positive trainee reactions (Cohen, 1990; Hicks and Klimoski, 1987; Mathieu et al., 1990). An exception may be when the history of training has been very positive. Baldwin and Magjurka (1991) found that engineers who had several previous positive training experiences, and who perceived the training to be mandatory, reported greater intentions to apply what they learned back on the job than engineers who did not have positive experiences, and who viewed their attendance as voluntary. Mandatory attendance may, however, be demoralizing when training is not valued.1
Training programs try to ensure that trainees gain proficiency in both taskwork procedures and teamwork procedures and that the performance of the team as a whole is enhanced. In most work situations, taskwork and teamwork are so interrelated that it is difficult to separate them. No matter how great individual taskwork skills may be, they will be ineffective if they are not coordinated with the supporting efforts of teammates. In work (performance) situations, both individual technical skills and skills in interacting and coordinating efforts with other team members are needed to succeed and, therefore, both have to be learned and practiced. In reviewing the research, however, we consider separately each of the parts of the training process.
This section considers in detail nine components of taskwork that are relevant to the likelihood of successful training, transfer, and maintenance in team training:
1. conceptual understanding;
2. applying conceptual understanding;
3. procedural learning;
5. social support;
6. relevant attitudes;
7. use of technology;
8. positive professional identity among team members; and
9. behavioral models.
Transfer to the actual job situation and long-term maintenance of the learned procedures and skills depend on trainees' conceptually understand-
ing what they are learning for several reasons. First, jobs are becoming more complex, and work requirements at all levels are becoming cognitively more demanding. Second, learning of a complex task can be facilitated by helping a learner develop an accurate and efficient mental conceptualization or mental model of the material that must be understood before the task is undertaken. Conceptual models are useful when they allow a trainee to infer the exact procedures for operating a device or completing a procedure or when the model is necessary to generalize the procedures to situations that were not trained explicitly (see Chapter 3; also see Gentner and Gentner, 1983; Kieras, 1988; Kieras and Bovair, 1984). Conceptual models are most helpful for dealing with complex problems when they organize material hierarchically rather than linearly (Eylon and Reif, 1984). Learning a conceptual model may not be useful when procedures are easily learned by rote mastery, when a particular device is so simple that the trainee does not need to make inferences about how the device works, when the conceptual model is too difficult or complicated for a trainee to acquire, or when the model does not support inferences that the trainee needs to make (Gentner and Gentner, 1983; Kieras, 1988; Kieras and Bovair, 1984).
Third, in job settings, complex work is done primarily by teams. In order for team members to work together effectively, trainees must develop conceptual understandings that are shared by all team members. Cannon-Bowers et al. (1990) and Orasanu and Salas (1993) suggest that shared or overlapping mental models among team members (created by such practices as cross-training) should enhance the use of implicit coordination strategies by enabling team members to anticipate behavior and information needs more accurately. Fourth, when trainees are provided with general principles governing successful or competent performance in both the training and real-world settings, transfer of training is generally facilitated (see Chapter 3).
In order to maximize cognitive understanding, training programs may emphasize competition among trainees, individualistic work by trainees, or cooperative efforts by learning teams. Miller and Hamblin (1963) concluded from a review of research that the superiority of cooperative teams over competitive and individualistic learning increases as the learning task becomes more conceptual and difficult. This conclusion was also supported by more comprehensive reviews (D.W. Johnson and R.T. Johnson, 1974, 1989; D.W. Johnson et al., 1981). In addition, in their meta-analysis, Johnson and Johnson (1989) found that both quality of individual reasoning strategies and level of individual cognitive processes were higher when persons learned in cooperative teams rather than competitively or individually (effect sizes of 0.79 and 0.97, respectively).2Taken together, the evidence indicates that when complex conceptual learning is required, individual performance is enhanced more by learning in teams than by learning individually (competitively or individually).
Applying Conceptual Understanding
Besides engaging in abstract, conceptual thinking about what is being learned, trainees need to develop the ability to apply that capacity for abstract thought to complex real-world job problems that involve the use of scientific and technical knowledge, are nonstandard and full of ambiguities, and have more than one right answer. In a recent meta-analysis, Qin (1992) identified 63 studies, conducted between 1929 and 1989, that examined the relative success of cooperative and competitive efforts on individual problem solving. She classified the studies into several kinds of problem solving: linguistic, solved through written and oral language; nonlinguistic, solved through symbols, math, motor activities, actions; well defined, clearly defined with well-defined operations and solutions; and ill defined, lacking clear definitions, operations, solutions. She found that members of cooperative teams outperformed individuals who worked competitively on all four types of problem solving (effect sizes of 0.37, 0.72, 0.52, and 0.60, respectively). Adults and secondary students working cooperatively performed better on linguistic, nonlinguistic, and well-defined problem solving than did children.
The ultimate aim of training is procedural learning, that is, for trainees to be competent in performing a job. Only a few studies have compared team training with individual instruction on tasks that appear to predominantly rely on procedural learning. Two studies conducted in a naval job training program have found that, compared with traditional naval training involving competitive and individual activities, training in teams resulted in greater conceptual mastery of the procedures taught, greater independent functional ability to perform job functions, a zero failure rate, and lower attrition from the program (Holubec et al., in press; Vasquez et al., 1993). Performance on a computer-assisted problem-solving task involving map reading and navigational skills was found to be higher in cooperative than in competitive or individual instruction (R.T. Johnson et al., 1985, 1986). A related study found similar results for the learning of map reading skills (Yager et al., 1985). Martino and Johnson (1979) compared cooperative and individual instruction on learning how to swim; students taught in cooperative pairs learned how to perform more swimming skills than did students taught individually. Similarly, subjects taught the golf skill of putting performed better when they were taught cooperatively than when they were taught competitively or individually (R.T. Johnson et al., 1983).
A basic principle of training design is that, whatever is being taught, trainees must actively ''produce" that during training. Production facilitates
both learning and retention (Perry and Downs, 1985). Performing the actual procedures and skills being taught during training complements the learning of more general concepts to improve posttraining performance (see Chapter 3).
Feedback can come from many sources, including oneself, technological devices, and other people. Interpersonal feedback from teammates may have a number of benefits. First, it personalizes the learning situation. Personal feedback has been found to increase performance to a greater extent than impersonal feedback (Acheson, 1964; Fuller et al., 1969; Morse et al., 1970; Steiner, 1967; Tuckman et al., 1969). Second, to the extent that interpersonal feedback tends to be subjectively vivid, the information provided may be attended to more highly (Borgida and Nisbett, 1977; Hamill et al., 1980; Nisbett et al., 1976; Taylor and Thompson, 1982). Statistical data summaries and impersonal information sources are less vivid than faceto-face interactions. Third, peers can provide immediate and sustained remediation after a trainee gives an incorrect answer, supportively probing, providing cues, repeating the question, rephrasing the question, or allowing more time for the trainee to answer, all of which is important for achievement (Rowe, 1974; Webb, 1982). Fourth, teams provide a cooperative context for feedback. Trainees process the feedback they receive from collaborators differently than the feedback they receive from competitors (or trainers). Teams are especially helpful in continually monitoring members' learning and providing immediate interpersonal feedback and remedial help and assistance.
Social support is the exchange of resources intended to enhance mutual well-being and the existence and availability of people on whom one can rely for assistance, encouragement, acceptance, and caring (D.W. Johnson and R.T. Johnson, 1989). A social support system consists of others who collaboratively share trainees' tasks and goals, who provide trainees with emotional concern (such as caring, reassurance, trust), instrumental aid (such as materials, tools, skills), information and advice, and feedback about the degree to which certain behavioral standards are met, all of which help trainees mobilize their psychological resources in order to complete the training program successfully. Social support directly and indirectly promotes achievement and productivity, psychological well-being, physical health and management of stress.
Social support is related to achievement, successful problem solving, persistence on challenging tasks under frustrating conditions, lack of cognitive
interference during problem solving, satisfaction, high morale, attendance or lack of absenteeism, retention, academic and career aspirations, more appropriate seeking of assistance, and greater compliance with regimens and behavioral patterns that increase health and productivity (Bowlby, 1969, 1973, 1980; Cohen and Willis, 1985; Sarason et al., 1983). And social support directly reduces stress by providing the caring, information, resources, and feedback individuals need to cope with the stress inherent in participating in challenging training (see LaRocco et al., 1980; Seers et al., 1983). Indirect reduction of stress occurs when social support decreases the number or severity of stressful events during training (see, e.g., Kessler et al., 1985).
A predictable human tendency is to seek social support in times of stress, fear, sorrow, or high excitement. One of the earliest researchers to examine this tendency experimentally was Schachter (1959). He frightened subjects with the threat of shock, then examined the extent to which they chose to be with others. He argued that fear promoted affiliation for two reasonsto gain information about the stressful situation and to directly reduce the fear. In difficult training programs, therefore, where trainees experience considerable anxiety and pressure to achieve, teams may be effective by providing members with support that reduces anxiety (Schachter, 1959) and increases feelings of self-control and self-esteem (Cutrona and Troutman, 1986). A metaanalysis of over 106 experimental studies comparing the relative effects of teams versus individual efforts on social support provides evidence that team experiences promoted greater social support than did competitive or individual experiences (D.W. Johnson and R.T. Johnson, 1989). Teams can provide a social support system for their members as they work together to maximize their own and each other's productivity.
Such attitudes as commitment to quality work, commitment to continuous improvement, and self-efficacy can be inculcated or strengthened during a training program. It is through interpersonal influences that attitudes are typically acquired and behavioral patterns are changed. Teams thus have several advantages over working alone when a training program is designed to inculcate attitudes that may promote the acquisition of training.
The more a team becomes a reference group for members, the more it will influence members' attitudes and values (D.W. Johnson and F. Johnson, 1994; Watson and Johnson, 1972). Attitudes and commitments to engage in specified behavior patterns that are made public and discussed in small groups are more likely to be adopted than are those that are private, especially when peers hold one accountable to fulfill one's commitments (D.W. Johnson, 1981; Lewin, 1943; Radke and Caso, 1948; Radke and Klisurich, 1947). During discussion, a norm of commitment seems to be created. This norm prescribes
that individuals carry out those actions that they have promised or committed themselves to perform. It is strengthened when group members are unanimously committed to the course of action. Orbell et al. (1988), for example, observed that high rates of promising to cooperate during group discussion are followed by high rates of actual cooperation, when all members agreed to follow through on the action. When as few as one member disagreed, the correlation between promises and actions was negligible.
People are most likely to accept new attitudes and behaviors when they come into contact with visible and credible models who demonstrate the recommended attitudes and behavior patterns and directly discuss their importance (Aronson and O'Leary, 1982-1983; Goldman, 1940; Nisbett et al., 1976; Rogers and Shoemaker, 1977). Discussing information with peers in ways that promote active cognitive processing often produces attitude change and induces commitment to behave in prescribed ways. Active cognitive involvement in the persuasion situation is crucial for attitude change (Cook and Frey, 1978; Petty, 1977). In addition, people are particularly prone to increase their own commitment to a cause that they have attempted to persuade others to adopt (Nel et al., 1969).
In addition to the general attitude effects, learning as part of a team tends to promote more positive attitudes toward the task being worked on and the experience of doing so than does learning by oneself competitively or individually (D.W. Johnson and R.T. Johnson, 1989). Most individuals, furthermore, prefer cooperative over competitive and individual training experiences (DeVries and Edwards, 1973; D.W. Johnson and R.T. Johnson, 1976; D.W. Johnson et al., 1976; R.T. Johnson, 1976; R.T. Johnson et al., 1973, 1974; Wheeler and Ryan, 1973).
Use of Computers
The use of computers in training requires trainees concurrently to learn how to use the technology and to master the information, skills, procedures, and processes being presented within the technology. There is reason to believe that teams better manage the dual demands of the learning situation and make better use of computers than do individuals. Computer-assisted team learning, compared with individual efforts at the computer, promoted higher quantity and quality of daily achievement, greater mastery of factual information, greater ability to apply a learner's factual knowledge in test questions requiring application of facts, greater ability to use factual information to answer problem-solving questions, and greater success in problem solving (D.W. Johnson et al., 1986a, 1989, 1990; R.T. Johnson et al., 1985, 1986). In addition, working as part of a team at the computer promoted greater motivation to persist on problem-solving tasks and greater success in operating the computer program.
Technology such as electronic mail, bulletin boards, and conferences can be used to create teams made up of individuals who are widely separated geographically. In an electronically networked team, interaction no longer is face to face, and trainees may depend on one another differently than they do in face-to-face teams. "Meetings" only require that team members be at their terminals, and communication between meetings can be asynchronous and extremely fast in comparison with telephone conversations and interoffice mail. Participation may be more equalized and less affected by prestige and status (McGuire et al., 1987; Siegel et al., 1986). Electronic communication, however, relies almost entirely on plain text for conveying messages, text that is often ephemeral, appearing on and disappearing from a screen without any necessary tangible artifacts. It becomes easy for a sender to be out of touch with his or her audience. And it is easy for the sender to be less constrained by conventional norms and rules for behavior in composing messages. Communicators can feel a greater sense of anonymity, detect less individuality in others, feel less empathy, feel less guilt, be less concerned over how they compare with others, and be less influenced by social conventions (Short et al., 1976; Kiesler et al., 1984). Such influences can lead both to more honesty and more "flaming" (name calling and epithets).
While electronic communication has many positive features, face-toface communication has a richness that electronic communication may never match. There is evidence that up to 93 percent of people's intent is conveyed by facial expression and tone of voice, with the most important channel being facial expression (Druckman et al., 1982; Meherabian, 1971). Harold Geneen, the former head of ITT, believed that his response to requests was different face to face than through teletype. "In New York, I might read a request and say no. But in Europe, I could see that an answer to the same question might be yes . . . it became our policy to deal with problems on the spot, face to face" (cited in Trevino et al., 1987). For this and other reasons (such as lack of effective groupware), training programs may be most effective when they use face-to-face rather than electronic teams. Team members may also benefit more from the social support function and vividness of feedback provided directly by other members.
Positive Professional Identity
An important aspect of training programs is to create, modify, and extend a trainee's professional identity and esteem. Attending Marine bootcamp, for example, is supposed to result in a trainee's adopting a new identity as a "Marine." Being trained to be an air-traffic controller is supposed to result in a trainee adopting a new identity as an "air-traffic-controller." Most training programs are aimed at inculcating an identification with the job the person is
being trained to do. Identity has at least four properties (Miller and Harrington, 1990): relational, who one is in relation to others; situational; interactive, in interaction with other people; and affective. A person's social identity is substantially determined by the groups to which he or she belongs (Miller and Harrington, 1990). The more positive a person perceives the group to be, the more positive the implications of membership in that group for the person's self-identity. Team training typically builds a stronger, healthier, and more positive professional identity than does individual training (D.W. Johnson and R.T. Johnson, 1989). The new identity resulting from the training program helps trainees establish their membership in subsequent work teams.
Within learning teams there are visible and credible models who demonstrate the procedures being learned. Although the research on behavioral modeling is not consistent (Druckman and Bjork, 1991), observing skilled team members engage in a procedure or skill can facilitate learning.
In many work situations, problems and even the work is so complex that multiple experts and multiple sources of information are needed. To deal with that complexity, people doing many jobs have to work in teams. Teamwork competencies thus become essential aspects of those jobs. The taskwork cannot be done unless the individuals function as team members. This function can be aided by the development of teamwork skills.
Like all skills, teamwork skills need to be learned to an automatic level. If teamwork skills are not emphasized during training and if individuals do not learn the interpersonal and small group skills they need to function as team members, the taskwork will suffer.
There are several explanations for potential interference between taskwork and teamwork. Our proposal is that, if teamwork skills are not well learned, team members have an internal conflict between attending to the task and learning how to coordinate efforts with others (Baron et al., 1978). If a team member tries to attend to more than he or she can manage, the resulting conflict leads to drive (arousal) and stress, which in turn produces social impairment on difficult tasks.
Second, trying to learn teamwork skills while doing taskwork may cause attention overload. Attention overload occurs when a team member tries to attend to more things (such as taskwork and teamwork both) than he or she has the capacity to process. When people are bombarded with attentional demands, their focus of attention actually shrinks (see, e.g., Geen, 1976, 1980), causing deterioration of taskwork efforts.
Third, when one or more team members are not socially skilled and do not participate effectively in the team, they may distract the other team members from effective taskwork.
Fourth, trying to learn teamwork skills while doing taskwork may make team members self-conscious and self-aware. Self-awareness generally results in team members' "trying harder" to perform well (Duval and Wicklund, 1972; Carver and Scheier, 1981). One possible result of self-awareness is withdrawal from the task. By paying close attention to how well they are performing, self-conscious team members may experience a decrease in their desire to excel, as in the frequently observed phenomenon of stage fright.
Fifth, trying to learn teamwork skills while doing taskwork may interfere with team members' presenting a positive image of themselves. Self-presentation involves projecting a positive self-image ("looking good") to onlookers (e.g., Bond, 1982). Social impairment sometimes occurs on difficult tasks because initial failures produce embarrassment, which then disrupts performance.
Since both taskwork and teamwork proficiencies are required in many job situations, it is important that training programs emphasize both. The failure to teach teamwork skills in these situations can result in subsequent interference between taskwork and teamwork job demands.
Because training in teams has been found to be have certain advantages, compared with individual training, does not mean that it will be more effective under all conditions. There are a number of variables that tend to mediate the effectiveness of team training. They include the degree of positive interdependence, face-to-face promotive interaction, individual accountability, and team processing.
One apparent constraint on the observation of higher achievement for team training than individual training is that the team should have high positive interdependence. Positive interdependence is defined to exist when team members perceive that they are linked with teammates so that they cannot succeed unless their teammates do (and vice versa) or that they must coordinate their efforts with the efforts of their groupmates to complete a task (Deutsch, 1962; D.W. Johnson and R.T. Johnson, 1989). When positive interdependence is clearly understood, it highlights that each member's efforts is required and indispensable for group success (i.e., there can be no "free riders") and each member has a unique contribution to make to the joint effort because of his or her resources, or role, and task responsibilities.
There are a number of ways that positive interdependence can be structured within a team. One view is that a team needs to have positive goal interdependence: team members should realize that they have a mutual set of goals that all are striving to accomplish and that success depends on all members reaching the goal; otherwise, it is not a team (D.W. Johnson and R.T. Johnson, 1989).3Other methods of structuring positive interdependence that supplement and support the positive goal interdependence include: positive reward interdependence, when each team member group is given the same reward for successfully attaining the team's goals; positive role interdependence, when team members are assigned complementary and interconnected roles; positive task interdependence, when a division of labor is created so that the actions of one team member have to be completed if another team member is to complete the next actions; positive resource interdependence, when each member has only a portion of the information, resources, or materials necessary for the task to be completed; and identity interdependence, which exists when a team establishes a mutual identity through a name, flag, motto, or song (D.W. Johnson et al., 1991).
In a series of studies to determine the relative effects of the types of interdependence on performance after a group has studied as a team, it was found that team membership is not enough to promote higher individual achievement; it also took positive goal interdependence (Hwong et al., in press). Knowing that one's performance affects the success of teammates seems to create "responsibility forces" that increase one's efforts to achieve. Interaction among individuals is not enough either. In a series of studies that investigated whether the relationship between teamwork and individual achievement was due to the opportunity to interact with peers or to positive goal interdependence, researchers consistently found that individuals achieved more under positive goal interdependence than when they worked individually but had the opportunity to interact with others (Lew et al., 1986a, 1986b; Mesch et al., 1986, 1988). Hagman and Hayes (1986) found that teams of two, three, and four members working together within an individual structure achieved at a lower level than did individuals who studied as cooperative teams. They concluded that without positive interdependence, there was no advantage to having individuals interact with one another while they work.
The results of the Mesch and Lew studies also indicate that, while positive goal interdependence is sufficient to produce higher individual achievement after studying as a team rather than individually, the combination of goal and reward interdependence is even more effective. The impact of the two types of outcome interdependence seems to be additive. Frank (1984) demonstrated that working both to achieve a reward and to avoid the loss of a reward produced higher individual achievement after studying as a team than after studying individually. D. W. Johnson and R. T. Johnson (1994)
found goal interdependence promoted higher achievement than did resource interdependence. A study by D.W. Johnson et al. (1990) demonstrated that goal interdependence alone increased achievement, and the combination of goal and resource interdependence increased achievement even more. The use of resource interdependence alone seemed to decrease individual achievement and lower productivity after studying as a team rather than studying individually.
Finally, there is evidence that positive interdependence does more than motivate individuals to try harder. Positive interdependence seems to facilitate the development of new insights and discoveries through promotive interaction (Gabbert et al., 1986; D.W. Johnson and R.T. Johnson, 1981; D.W. Johnson et al., 1980; Skon et al., 1981).
Face-to-Face Promotive Interaction
In order for team training to promote higher achievement than individual training, team members must promote each other's success. Promotive interaction can be defined as individuals encouraging and facilitating each other's efforts to achieve, complete tasks, and produce in order to reach the team's goals (D.W. Johnson and R.T. Johnson, 1989). Positive interdependence can result in promotive interaction. Although positive interdependence in and of itself may have some effect on outcomes, the face-to-face promotive interaction among individuals fostered by positive interdependence appears to be the most powerful influence on efforts to achieve, caring and committed work relationships, and psychological adjustment and social competence.
Promotive interaction is characterized by team members' behavior in at least eight ways: (1) providing each other with efficient and effective help and assistance; (2) exchanging needed resources, such as information and materials, so that information can be processed efficiently and effectively; (3) providing each other with feedback in order to improve their subsequent performance on assigned tasks and responsibilities; (4) challenging each other's conclusions and reasoning, which promotes higher quality decision making and greater insight into the problems being considered; (5) advocating the exertion of effort to achieve mutual goals; (6) influencing each other's efforts to achieve the team's goals; (7) acting in trusting and trustworthy ways; and (8) being motivated to strive for mutual benefit.
For promotive interaction to occur, however, the development of appropriate social skills is required. A team's function is impaired if members do not have and use the interpersonal and small group skills needed to coordinate their efforts, namely, communication, leadership, decision-making, and conflict-management skills (Cartwright and Zander, 1968). A number of studies have demonstrated that the more these skills are used, the more productive a
team is (Lew et al., 1986a, 1986b; Mesch et al., 1986, 1988) and the more positive the relationships among group members (Putnam et al., 1989). Cooperation, furthermore, promotes more frequent, effective, and accurate communication than either competitive or individual situations (D.W. Johnson, 1974). In cooperative situations, communication is relatively open, effective, and accurate; in competitive situations, communication tends to be closed, ineffective, and inaccurate (Bonoma et al., 1974; Crombag, 1966; Deutsch, 1962; French, 1951; Grossack, 1953; Krauss and Deutsch, 1966).
Promotive interaction may have a number of effects. First, there are cognitive insights and understandings that can come from explaining one's conclusions and views to others (e.g., Spurlin et al., 1984; Murray, 1983; Yager et al., 1985). Second, it is within face-to-face interaction that the opportunity for a wide variety of social influences and patterns emerge (see, e.g., Coleman, 1961; Hulten and DeVries, 1976). Third, the verbal and nonverbal responses of group members can provide important feedback concerning each other's performance (e.g., Lockhead, 1983). Fourth, it can provide an opportunity for peers to pressure unmotivated team members to do their share of the work (e.g., Deutsch, 1949; Crombag, 1966). Finally, the interaction involved in completing the work allows team members to get to know each other as persons, which in turn forms the basis for caring and committed work relationships.
Individual accountability exists when the contributions that each team member makes to the team effort are assessed and the results are given back to the team and the individual. It minimizes the likelihood of social loafing, which is one of the major obstacles to group effectiveness. When groups work on tasks for which it is difficult to identify members' contributions, when there is an increased likelihood of redundant efforts, when there is a lack of group cohesiveness, and when there is lessened responsibility for the final outcome, some members will contribute less than otherwise to goal achievement (Harkins and Petty, 1982; Ingham et al., 1974; Kerr and Bruun, 1981; Latane et al., 1979; Moede, 1927; Petty et al., 1977; Sanna, 1992; Williams, 1981; Williams et al., 1981). Social loafing can be avoided by making each team member individually accountable for doing his own share of the work.
Individual accountability can be enhanced in several ways. First, social loafing can be avoided by clearly structuring positive interdependence so feelings of personal responsibility for the group's final outcome are high. Personal responsibility adds the concept of "ought" to members' motivation-one ought to do one's part, pull one's weight, contribute, and satisfy peer norms. Second, since individual accountability increases as the size of
the team decreases (Messick and Brewer, 1983), teams can be kept small. Third, when it is made clear how much effort each member is contributing (i.e., there is individual assessment of each member's contributions), individual accountability is high. The more that group members are provided with information about the level of productivity and helpfulness of each member, the more individually accountable each member will be and the better members will be able to support and assist each other. Finally, when team members are assigned roles to ensure a lack of redundancy, individual accountability is higha member cannot depend on another member to accomplish his or her assignment. However, this advantage is offset by the safety implications of a lack of redundancy. The challenge is to balance the motivating effects of lack of redundancy with the obvious protection that redundancy provides against individual error or unexpected occurrences.
As defined here, process is an identifiable sequence of events taking place over time. Team processing occurs when members discuss the sequence of events instrumental in achieving their goals and maintaining effective working relationships among members (D.W. Johnson and F. Johnson, 1994). Effective teamwork is influenced by whether or not teams reflect on and discuss how well they are functioning. Team members process when they describe what member actions were helpful and unhelpful and make decisions about what member actions to continue or change. Such processing enables teams to focus on group maintenance, facilitates the learning of social skills, ensures that members receive feedback on their participation, and reminds members to practice social skills consistently.
Although it is well accepted in the group dynamics literature that to be productive, teams have to ''process" how well they are working and take action to resolve any difficulties members have in collaborating productively, there is actually very little evidence on the relationship between team processing and team productivity. Most of the research examines effects on individual outcomes. Yager et al. (1985) found that team members scored higher on individual measures of learning and retention when team processing occurred than did individuals in teams that did not process or individuals who studied alone. The combination of team processing and feedback from a supervisor produced greater individual and group problem-solving success than did supervisor feedback alone or neither (D.W. Johnson et al., 1990). Feedback to individual team members about their performance is more effective than overall team feedback in increasing productivity (Archer-Kath et al., in press).
Team processing may promote individual self-monitoring, which can promote a sense of self-efficacy (i.e., the expectation of successfully obtaining
valued outcomes through personal effort) rather than helplessness. Sarason and Potter (1983) examined the impact of individual self-monitoring of thoughts on self-efficacy and successful performance and found that having individuals focus their attention on self-efficacious thoughts is related to greater task persistence and less cognitive interference. They concluded that the more that people are aware of what they are experiencing, the more aware they will be of their own role in determining their success. The greater the sense of self-efficacy and joint-efficacy promoted by group processing, the more productive and effective group members and the group as a whole become.
Monitoring one's own and one's collaborators' actions begins with deciding on which behaviors to focus one's attention. Individuals can focus either on positive and effective behaviors or negative and ineffective behaviors. Effective processing focuses primarily on positive rather than negative behaviors.4Sarason and Potter (1983) found that when individuals monitored their stressful experiences they were more likely to perceive a program as having been more stressful than did those who did not; but when individuals monitored their positive experiences, they were more likely to perceive the group experience as involving less psychological demands, were more attracted to the group and had greater motivation to remain members, and felt less strained during the experience and more prepared for future group experiences. When individuals are anxious about being successful and are then told they have failed, their performance tends to decrease significantly, but when individuals anxious about being successful are told they have succeeded, their performance tends to increase significantly (Turk and Sarason, 1983).
Although team training is used extensively, there is little direct evaluation of its effectiveness on team outcomes. Lassiter et al. (1990) found that teams in a skill-oriented training program demonstrated better communication skills than those in a lecture-based training program or a no-training control group. Several commercial airlines have established team training programs for air crews, but the results on training effectiveness have been mixed (Cannon-Bowers et al., 1991; Helmreich and Wilhelm, 1989; Helmreich et al., 1990). One reason that more research has not been conducted is the lack of methods for analyzing team tasks, behaviors, and skills (Modrick, 1986; Morgan and Salas, 1988; O'Neil et al., 1992). Without valid and reliable dependent (outcome) measures, programmatic research on team training is difficult. Team training, furthermore, has tended to ignore the development of teamwork skills and behaviors that are demanded by the interaction requirements of the team task. What is indicative of the effectiveness
of team training is the research comparing the relative effects of team and individual training on subsequent knowledge and proficiency. This section first considers that work and then reviews other areas of research that have compared outcomes for individuals in individual, competitive, and team or group training.
Individual Proficiency and Team Productivity
A recent meta-analysis has been conducted on the effectiveness of team versus individual training for adults (18 years and older) (D.W. Johnson and R.T. Johnson, 1994). The studies included were divided into those using individual achievement and proficiency as the dependent (outcome) measure and those using team productivity as the dependent measure. Although a variety of training tasks were used, all studies separated the training session from the evaluation of proficiency or productivity.5 The individual training programs were classified as either structuring interpersonal competition among trainees or having trainees work on their own. The results are reported as effects obtained on outcomes across many studies. High effect sizes indicate strong relationships between type of training programs and proficiency: for example, subjects trained in teams out-performed those trained individually in most or many, but not necessarily all, studies.
More than 120 studies have compared the relative efficacy of team and individual training on individual knowledge and proficiency (D.W. Johnson and R.T. Johnson, 1994). Although the first study was conducted in 1924, 70 percent of the studies have been conducted since 1970. A meta-analysis of these studies suggests that overall, team learning promoted higher individual knowledge and proficiency than did individual learning structured competitively or individually (effect sizes of 0.54 and 0.51, respectively). When only the methodologically high quality studies were included, team learning still promoted greater individual achievement than did competitive or individual efforts (effect sizes of 0.61 and 0.35, respectively). These results indicate that there is greater team-to-individual transfer than individual-to-individual transfer. The meta-analysis results, however, do not define the conditions under which team-to-individual transfer will be greatest.
Hagman and Hayes (1986) conducted two studies in which they demonstrated that the superiority of team-to-individual transfer over individual-to-individual transfer increased as subjects worked toward a team (rather than an individual) goal and as the size of the team was reduced. Teams in which members interacted with each other and discussed the material being
learned and received a team reward had the greatest amount of team-to-individual transfer. In a study involving children as subjects, learning in a team resulted in greater individual transfer than did learning as an individual for complex higher-level tasks, but not for simple lower-level tasks (Gabbert et al., 1986). The studies comparing team and individual training used a wide variety of tasks. The tasks were classified into those that required verbal skills to complete (such as reading, writing, and oral presentation), mathematical skills to complete, or procedural skills to present (such as sports like swimming, golf, and tennis). When the results were analyzed for type of task, team training showed better transfer than individual training, structured either competitively or individually, on verbal tasks (effect sizes of 0.36 and 0.66, respectively), on mathematical tasks (effect sizes of 0.45 and 1.32, respectively), and on procedural tasks (effect sizes of 0.95 and 1.06, respectively). These results may be taken to indicate that training in teams promotes higher individual proficiency and knowledge than does individual training on all three types of tasks. The adult studies did not include simple rote or decoding tasks, but the research on children and adolescents indicates that individual training (structured competitively) may be just as effective as team training when very simple tasks are used (D.W. Johnson and R.T. Johnson, 1989).
Since 1928, more than 57 studies have been conducted on the relative effectiveness of team and individual training on team productivity (D.W. Johnson and R.T. Johnson, 1994). Twelve percent of the studies were conducted before 1930, 15 percent from 1930-1960, 43 percent in the 1960s, and 30 percent in the 1970s. A meta-analysis of the results indicated that team training led to higher productivity than did individual training, structured either competitively or individually (effect sizes of 0.63 and 0.94, respectively). When only the methodologically high quality studies were included in the analysis, team learning was still more effective than competitive or individual efforts (effect sizes of 0.96 and 0.66, respectively). Teams tend to make better decisions and solve problems better than do individuals. When the results were analyzed for type of task, team training resulted in better performance than individual training, structured competitively and individually on verbal tasks (effect sizes of 0.73 and 1.47, respectively), on mathematical tasks (effect sizes of 0.26 and 0.86, respectively), and on procedural tasks (effect sizes of 1.37 and 0.95, respectively). These results indicate that team training promotes higher team performance than does individual training on all three types of tasks.6Other evidence suggests, however, that on brainstorming tasks individuals may do just as well as teams (see Dunnette et al., 1963).
Positive Relationships and Social Support
Training is often aimed at affecting relationships among trainees, as well as learning. Both long-term networking of trainees and increasing the positiveness of relationships among diverse trainees may be important goals of training programs. Since the 1940s research conducted on individual proficiency has also examined the quality of relationships among trainees (D.W. Johnson and R.T. Johnson, 1994). A meta-analysis of this research indicated that team training tended to promote greater interpersonal attraction among trainees than did individual efforts structured competitively (effect size, 0.68) or individualist (effect size, 0.55). Similar results were found in studies focusing on group productivity (effect sizes, 0.64 and 0.39, respectively).
Building positive relationships among trainees may be especially important when the trainees are heterogeneous in terms of ethnicity, handicapping conditions, and gender. More than 61 studies have been conducted comparing the relative effects of team and individual training (structured either competitively or individually) on interpersonal attraction between majority and minority individuals (D.W. Johnson and R.T. Johnson, 1989). Team training promoted significantly better relationships between white and minority individuals than did competition (effect size, 0.52) or individual efforts (effect size, 0.44). Individual efforts promoted more positive cross-ethnic relationships than did competition (effect size, 0.65), but since only three studies have been conducted, this finding is only suggestive. When only the high-quality studies were included in the analysis, the advantage of team over competitive and individual efforts was even more apparent (effect sizes of 0.68 and 0.53, respectively). In 41 studies comparing the relative effects of team and individual training on interpersonal attraction between handicapped and nonhandicapped individuals (D.W. Johnson and R.T. Johnson, 1989), team training resulted in greater interpersonal attraction between handicapped and nonhandicapped individuals than did competitive training (effect size, 0.70) or individual training (effect size, 0.64). Individual experiences tended to promote somewhat more positive cross-handicapped relationships than did competitive training (effect size, 0.16) in the five studies conducted. Only a few studies have compared the relative effects of team and individual training on interpersonal attraction between males and females. The results indicated that team training promoted more positive cross-gender relationships than did individual training (Warring et al., 1985), and cross-gender friendships may be more difficult to promote than cross-ethnic friendships between trainees of the same sex (Cooper et al., 1980).
In addition to interpersonal attraction or liking, relationships among trainees may be characterized by social support. Social support involves
the exchange of resources intended to enhance mutual well-being and the existence and availability of people on whom one can rely for emotional, instrumental, informational, and appraisal aid. In a meta-analysis of studies focusing on individual performance, team training promoted greater social support than did individual efforts structured competitively (effect size, 0.60) or individually (effect sizes, 0.61 and 0.51, respectively) (D.W. Johnson and R.T. Johnson, 1994). Studies focusing on team performance found that cooperation promoted greater social support than did competitive efforts (effect size, 0.13) or individual efforts (effect size, 0.38).
Social support is not a trivial factor. It is associated with several positive outcomes: achievement and productivity; physical healthindividuals involved in close relationships live longer, get sick less often, and recover from illness faster than do isolated individuals; psychological healthby preventing neuroticism and psychopathology, reducing distress, and providing resources such as confidants; and constructive management of stressby providing the caring, resources, information, and feedback needed to cope with stress and by buffering the impact of stress on the individual. These relationships are bidirectional, with support and health influencing each other in a recursive fashion. Social support and stress are related in that the more social support individuals have, the less the stress they experience, and the better able they are to manage the stresses they do have (Cohen and Willis, 1985; Kessler and McLeod, 1985).
In research on team training, two types of social support have been examined. Social support can be aimed at promoting a person's task success or it can be aimed at the trainee as a person. Caring about how much a person achieves and wanting to be a person's friend were perceived to go hand in hand; there was little difference between the levels of task and personal support perceived from peers and superiors (D.W. Johnson and R.T. Johnson, 1989). There seems to be a "halo effect" in team situationssuperiors as well as peers are perceived to be more supportive of the trainee as a person as well as of the trainee's task success.
Psychological Health, Social Competence, and Self-Esteem
Training programs are often aimed at inculcating a professional identity, self-esteem, and the ability to cope with adversity under highly stressful conditions. Training for managing emergencies within nuclear power plants or pilot training for commercial airlines, for example, are aimed at making sure the learned procedures will be correctly used when one's own life and the lives of others are at stake. Thus, training is aimed at promoting trainees' psychological health and ability to adjust psychologically to changing conditions and situations. And since training programs often assume that the knowledge and procedures learned will be used in teams in
work situations, trainees are expected to pick up the social competencies needed to coordinate and integrate their efforts with those of others.
Several studies have directly measured the relationship between social interdependence and psychological health (Crandall, 1982; Haynes, 1986; N. James and Johnson, 1983; S. James and Johnson, 1988; D.W. Johnson et al., 1984, 1986b; D.W. Johnson and Norem-Heibeisen, 1977). The samples studied included university students, older adults, suburban high school seniors, juvenile and adult prisoners, step-couples, and Olympic hockey players. The results indicated that cooperative attitudes are highly correlated with a wide variety of indices of psychological health, competitiveness was in some cases positively and in some cases negatively related to psychological health, and individual attitudes were negatively related to a wide variety of indices of psychological health.
Studies focusing on individual proficiency that compared teams and individual efforts on self-esteem found that team training promoted higher self-esteem than did individual training structured competitively or individually (effect sizes, 0.47 and 0.29, respectively). Studies focusing on group productivity found that team training promoted higher self-esteem than did competitive efforts (effect size, 0.86) or individual effort (effect size, 0.68). The level of self-esteem is affected not only by being part of a team effort, but also by the process by which individuals make judgments about their self-worth. In four studies of 821 white, middle-class, high school seniors in a midwestern suburban community, Norem-Hebeisen and Johnson (1981) found that cooperative experiences promoted basic self-acceptance, freedom from conditional acceptance, and seeing oneself positively compared to peers. Competitive experiences were related to conditional self-acceptance, and individual attitudes were related to basic self-rejection, including anxiety about relating to other people. Cooperative, team-based experiences seem to result in internalizing perceptions that one is known, accepted, and liked as one is; internalizing mutual success; and developing multidimensional views of oneself and others that allow for positive self-perceptions (D.W. Johnson and R.T. Johnson, 1989).
Changes in Team Structure and Procedures
When intact teams are taken from a job situation, given training as a team, and then returned to the work setting, the training often focuses on changing the way the team functions and operates, as well as increasing the proficiency of its individual members. On the basis of the broad literature on group dynamics, it is assumed that such team training will be more effective in changing a team's norms, roles, communication patterns, shared mental models, and team processing procedures than will individual training (Baron et al., 1992; Cartwright and Zander, 1968).
One of the most important procedures for teams to learn is how to make effective decisions. Decision making typically involves considering possible alternatives and choosing one (D.W. Johnson and F. Johnson, 1994). Doing so often requires that team members be advocates for points of view and information that the team is not considering. Airplane cockpit crews are a good example. A factor in more than two-thirds of air accidents is a failure of crew members to use their information expeditiously to cope with safety hazards (Cooper et al, 1979). Among the factors involved in crashes are the reluctance of first officers to disagree with pilots and the focus of crew members on a minor problem so that they forget to pilot the plane (Blake et al., 1989; Foushee, 1984). Crews that communicated extensively, acknowledged each other's communication attempts, made commands, disagreed, and felt less angry and embarrassed have been found to make fewer errors and crashes in simulated flights (Cooper et al., 1979; Foushee et al., 1986; Foushee and Manos, 1981; Ruffell-Smith, 1979; Tjosvold, 1990c).
By definition, all decision-making situations involve some conflict as to which of several alternatives should be chosen (D.W. Johnson and F. Johnson, 1994; Tjosvold, 1991). During training, trainees need to be taught the procedures and skills required to advocate their points of view when decisions have to be made. The procedure for doing so is controversy. Controversy exists when one person's ideas, information, conclusions, theories, and opinions are incompatible with those of another, and the two seek to reach an agreement (D.W. Johnson and R.T. Johnson, 1979, 1989). Controversy can be a powerful decision-making tool. The alternative solutions to the problem being faced can be identified, with each major alternative assigned to an advocacy group of team members who have the responsibility of preparing, presenting, and advocating the "best case" for each of the alternatives. The team then synthesizes the best reasoning from all perspectives into the group's decision. Research indicates that controversy tends to increase quality of decision making and problem solving, creativity of problem solving, more efficient exchange of expertise, more involvement in the task, greater interpersonal attraction among participants, and higher self-esteem (D.W. Johnson and R.T. Johnson, 1989). For example, electronic discussion groups with a planted group member who criticized others produced more new ideas and achieved more than did groups whose planted member was highly supportive (Connolly et al., 1990).7
Team Activities After Training
Teams can influence the transfer and maintenance of what is taught in a training program by highlighting implementation goals, encouraging immediate use, holding trainees accountable for using what they have learned, providing social support for using what was learned, providing a climate for
transfer, and giving feedback and social rewards for transfer. In a series of studies, Tjosvold (1990a) and Tjosvold and McNeely (1988) found that the more a team is focused on cooperative goals of implementing new practices, the greater the innovation and restructuring of work that takes place, even when team members have quite diverse perspectives.
Teams can ensure that opportunities to use what is learned are available and the new procedures and skills are immediately practiced. In a study of managers at the Internal Revenue Service, Pentland (1989) found that attempts to practice trained computer skills immediately on returning to the job had a major impact on long-term retention of the skills. In a study of Air Force technical trainees, Ford et al. (1991) noted that there were significant differences in opportunity to apply the training and wide variations in the length of time before trainees first performed the tasks for which they had been trained. Supervisor and peer support were found to be related to the extent to which airmen had opportunities to perform the trained tasks.
Teams create an expectation that each trainee will be held accountable for using the training. Baldwin and Magjurka (1991) found that trainees who expected to be held accountable for what they learned left training with stronger intentions to transfer. Marx and Karren (1990) found that trainees were more likely to apply time management skills when follow-up occurred 3 weeks after a time management course. Teams can provide social support for using what trainees have learned by meeting regularly to ensure that members are implementing what was learned regularly, appropriately, and with fidelity, and receiving the help they need to solve implementation problems (Baldwin and Ford, 1988).
Teams can provide a positive transfer climate for trainees. Transfer climate can be defined as reminding trainees to use their training through goal cues, social cues, and task and structural cues. Rouillier and Goldstein (1991) conducted a study of assistant managers who completed a week-long training program and were then randomly assigned to 1 of 102 organization units. In units with a more positive transfer climate, trainees demonstrated significantly more trained behaviors, even when the outcomes were controlled for learning and for unit performance. Teams are ideal for providing trainees with the goal, social, and task cues conducive to transfer and maintenance.
Teams can provide feedback about and reinforcement for transfer of what was learned from the training to the job situation. After training, paraprofessionals at a facility for the handicapped were assigned to pairs and instructed to provide their partners with feedback and reinforcement (Fleming and Sulzer-Azaroff, 1990). The researchers concluded that the peer support increased the maintenance of the procedures and skills learned during training. Stable and enduring performance of newly learned skills in application settings, furthermore, is very much dependent on real-life reinforcement contingencies. Supplemental programs, such as follow-up teams, can provide the rewards trainees need to maintain their new behaviors.
Transfer of what is learned during training to the job is affected by how the work organization structures the posttraining situation. If, for example, the organization has clear goals, pay incentives, and job aids for using what was learned, transfer is encouraged (Tjosvold, 1990a, 1990b, 1990d; Tjosvold and McNeely, 1988). But if using the new skills is ridiculed by peers who did not take the training and if job responsibilities have not been modified to require the use of the new competencies, then transfer is discouraged. And if necessary equipment is lacking, the use of the new competencies on the job may be actually prohibited. Although it seems logical that such posttraining influences affect transfer, there is very little empirical evidence on the issue. Baldwin and Ford (1988) found only seven studies that examined the influence of the work environment on the transfer of training. In these seven studies, social support seemed to be the most important influence on transfer.
As noted at the beginning of this chapter, training is instruction aimed at procedural knowledge and proficiency. Teams are used in training programs to increase members' procedural knowledge and proficiency of taskwork, members' procedural knowledge and proficiency of teamwork, and team productivity. These training goals can be achieved through a process construed in terms of the framework developed in this chapter (see Figure 7-1). Team training programs can take place at a training site or a job site. The training process is a combination of inputs, training processes, mediating variables, and outcomes. The inputs include trainees, resources, task characteristics, and pretraining activities of teams. The training processes are the procedures used to teach both taskwork and teamwork and to increase team performance. The mediating variables are positive interdependence, face-to-face promotive interaction, individual accountability, and team processing. The outcomes are individual proficiency and team productivity, positive relationships and social support, psychological health and self-esteem as well as social competencies, team changes, and posttraining team activities.
Guidelines for Research and Practice
On the basis of the research reviewed in this chapter, a number of suggestions can be made about using teams in training. The suggestions are also guidelines for trainers on how best to structure training in teams. And they can be regarded as hypotheses to be explored in further research.
Team training may work better if attention is paid to how the inputs to training are assembled. Sending cohort, intact teams to be trained who will
work together subsequent to the training will tend to increase the positive effects of the training. Team training will be more successful if pretraining teams are organized and focused on helping trainees set goals for what they will learn, alerting members that what they learn will later be assessed on the job, reducing the constraints for using the procedures and knowledge learned, creating positive expectations toward the training, and ensuring that supervisors support the training.
Team training can promote the simultaneous learning of both taskwork and teamwork skills: every session on taskwork becomes also a session on teamwork. In order to complete the tasks designed to teach participants taskwork procedures and proficiencies, trainees must clarify and implement the assignments (develop clear team goals), ensure that there is clear two-way communication among all team members, hold all members accountable for participating and providing their skills, and matching the decision-making procedures to the needs of the situation (including constructive conflict among group members). Trainees in individual training settings do not provide opportunities to develop these skills.
Team training also provides opportunities to better master the taskwork procedures being taught and to increase the likelihood of transferring the procedures to the job situation and maintaining their use. Team training provides a setting in which trainees can practice and master many aspects of taskwork: construct and extend conceptual understanding of what is being learned through explanations and discussion; use the shared conceptual models learned in flexible ways to solve problems jointly; jointly perform the procedures learned; receive other team members' feedback as to how well the procedures are performed; receive social support and encouragement to extend one's competencies; be held accountable by peers to practice until the procedures and skills are well learned; acquire the attitudes (such as continuous improvement) needed to refine the procedures learned; jointly master the use of the relevant technological devices; experience the stimulus variability inherent in working as part of a team; establish a positive and shared identity with other team members and with future colleagues, and observe the most outstanding team members as behavioral models to be emulated. These opportunities are not guaranteed, and they do not automatically occur in every team, but the likelihood of their occurrence exists when teams are used for training. The likelihood of their occurrence is very low when teams are not used.
Team training creates opportunities for a variety of positive outcomes. Individual proficiency tends to be increased, as well as productivity. As trainees work together to complete assignments, positive and supportive relationships tend to develop, even among trainees from different ethnic and gender groups. Working together and developing positive relationships also contributes to increased psychological health, self-esteem, and social competencies.
Completing a training program together can change a team's norms, roles, communication patterns, and decision-making procedures. Individual training programs do not have the potential for resulting in such multiple outcomes.
If teams are to be effectively used in training programs, the essential mediators should be structured in a disciplined and diligent way. Instructors must carefully structure positive interdependence to ensure that all trainees are committed to each other's success and each other as persons. Trainees must meet in teams and promote each other's learning face to face. Each trainee must be held accountable by his or her peers to maximize the effort to learn. In addition, each team must be held accountable for meeting the criteria set for successful performance. The teamwork skills required to coordinate efforts to complete joint assignments must be directly taught and learned. Finally, team members must gather data on their progress and plan how to improve the process they are using to learn. Without the disciplined application of these mediators of team effectiveness, the potential of training teams may not be realized.
Teams should be used in posttraining work situations to ensure that trainees engage in actions conducive to transfer and long-term maintenance of what was learned. Posttraining teams can be organized to ensure that trainees provide each other with opportunities for using what they learned, hold each other accountable for doing so, support each other's efforts to engage in a continuous improvement process to increase the quality of their application, and generally create a positive transfer climate.
Future research on the use of teams in training programs should be neither atheoretical nor nonpractical. Research is needed that directly tests the theoretical propositions about what makes team training programs effective. Validated theories can then be used to redesign training programs.
Obstacles to Team Training
The use of teams in training is not a panacea that solves all problems experienced by training programs. There are many ways that the use of teams in training programs can go wrong. There are tasks, for example, on which individuals may outperform a group (such as driving a car).
There are conditions under which the use of teams in training programs may be counterproductive. The most obvious is the use of poorly structured teams. Placing several trainees in the same room and calling them a team does not make them one. There is no advantage to telling trainees to work together if there is no positive interdependence, promotive interaction, individual and team accountability, use of social skills, and group processing. If teams are going to be used effectively in training programs, the mediating variables must be structured in a disciplined way. A second obstacle is lack of instructor training. Using teams effectively is a complex process that requires a clear understanding of the basic mediating variables and their
application for a specific training purpose. Instructors who have spent years training individuals cannot be expected to be automatically knowledgeable and skilled in using teams for training purposes.
If training promotes a very strong team identity, several unfortunate outcomes can result. Members can overestimate the value of their group and its position and underestimate the value of other teams and their positions. For example, Blake and Mouton (1962), demonstrated that team members knew and understood their own position in negotiations far better than they knew and understood the opposing position, even after studying both for some time. When the ingroup-outgroup bias is extreme, teams can become closed brotherhoods or microsocieties that lose touch with external realities and the organization's goals (see Chapter 6). A fourth obstacle to effective team training is the possibility that team goals may be valued more than an organization's overall goals. If team goals are isolated from and not made interdependent with the goals of the rest of the organization, team members may be more committed to the short-term objectives of the team than to the organization's goals. Fifth, when training becomes a discrete event separated from the ongoing work situation and organizational life, the training can become irrelevant or even counterproductive. Of course, this phenomenon is not unique to team training.
Finally, as noted above, a major obstacle to team training can be the lack of trainees' teamwork skills. The inability of trainees to work as part of a team can interfere with their taskwork, through such mechanisms as attention overload, distraction, self-consciousness, and embarrassment.
There is a face validity to the use of teams in training that may be deceptive. Since most tasks impose greater mental and physical demands than one individual can perform in isolation, work teams may become the rule, not the exception. When work teams are used, complex mixtures of taskwork and teamwork are usually required. This does not mean, however, that teams should be used for every assignment in training programs. To practitioners, this cautionary note may seem strange, as the majority of training is currently structured so that trainees work alone, either independent from or in competition with their peers. Although the use of teams in training programs should be increased dramatically, this does not mean that individual work should be stopped or that all training activities should take place in teams. Individual practice, for example, is usually necessary. Instructors need to be trained to integrate their use of team and individual procedures. Since poorly structured teams may be counter-productive, furthermore, instructors have to know how to be judicious consumers who can tell the difference between teams in which the mediating processes have been carefully structured and groups in which they have not. Thus, training program staff have the difficult responsibility of increasing the use of well-structured teams while stopping the use of poorly structured ones.
1 It should be noted that these facilitatory practices during the preparatory period need not be limited to training in teams. Trainers of individuals could also profit from goal setting, and trainees can benefit from reduced constraints or application, positive expectation, and motivation.
2 Unlike the effect sizes reported in Chapter 6, those reported here are not correlation coefficients. Each is calculated as the difference between the experimental and control group averages divided by the pooled sample standard deviation. This ratio can be greater than 1.00 (see Hedges and Olkin, 1985).
3 Most of the contemporaneous research on positive interdependence has been done by D.W. Johnson and his colleagues at Minnesota. The research has extended the original work on this topic by Deutsch (1949). Another view on the role of interdependence in training, referred to as process interdependence, is discussed in Chapter 5.
4 We note, however, that making and correcting errors during training has been shown to have positive effects on long-term performance (Druckman and Bjork, 1991:Chapter 3).
5 Typically, however, these studies do not bring subjects to a criterion of performance before the training is completed.
6 The teams in these studies were organized and trained to the task. The advantage for teams may largely dissipate for ad hoc freely interacting teams. Indeed, early research by Taylor (1954), Marquart (1955), and Lorge and Solomon (1955) showed that group problem-solving efforts were not routinely superior to individual efforts when an appropriate baseline for comparison was used. (This issue is discussed in Druckman and Bjork, 1991:Ch.12).
7 Another way to analyze team decision making is to examine rules, either implicit or explicit, by which individual judgments are combined to yield a team decision (e.g., Smoke and Zajonc, 1962).