This chapter presents the key findings and conclusions of our work, organized into four categories, as is the book: learning and remembering, learning and performing in teams, mental and emotional states, and new directions. Together, the various lines of research covered in this volume reveal both the complexity of the problems involved in enhancing human performance and some approaches that can improve performance of both individuals and teams. For some topics, conclusions follow from a large body of research findings and clearly indicate approaches to enhancing performance: for example, the effect of context on training and hypnosis. For other topics, however, the evidence accumulated to date leads primarily to suggestions for further research: for example, cooperative learning and team development. In some cases, a small body of research has accumulated; for example, thought suppression. In other cases, the research has not made direct connections with issues of performance; for example, the chapter on socially induced affect.
LEARNING AND REMEMBERING
Chapters 3 and 4 deal with issues related to individual learning and remembering. The work reviewed in Chapter 3 has implications for questions about how best to structure learning programs to optimize transfer to other settings and about the value of what is known as situated learning. The work reviewed in Chapter 4 has implications for judgments of performance made during training and on the job. The performance setting of interest is usually a workplace, although implications are also derived for classrooms and sports settings.
In many domains, fundamental skills are important to acquire before special skills that are relevant to particular settings are learned. The evidence obtained to date does not support the hypothesis that performance is enhanced to the extent that skills are learned in the setting in which they will be performed. Although concrete experience is very important, the teaching of abstract principles plays a role in acquiring skills over a broad domain of tasks. It is clear that learning need not be situated in the performance context to be effective.
A training approach that provides learners with varied contexts and general procedures allows them to adapt to new situations not encountered during training. In certain settings, training techniques based on the idea of part-to-whole transfer would be useful to develop. Although longer training sessions may improve performance, they can also lead to inflexible performance if the training fails to anticipate the variability that exists in the performance settings.
In its second report the committee concluded that immediate and constant feedback may fail to optimize performance. Delayed and intermittent feedback may produce superior performance because it allows learners to detect and correct errors (Druckman and Bjork, 1991:Ch.3). The evidence reviewed since that report supports this conclusion; it also shows that delayed and intermittent feedback diminishes reliance on extrinsic feedback.
Simulators are frequently used in training. At issue is the extent to which the simulated environment should resemble the performance setting: this is referred to as task fidelity. The research shows that the level of fidelity needed in simulators depends on the results of a task analysis of the specific skill to be taught. Such an analysis can identify the critical elements in the skill that should be identical in the training and performance environments. Advances in understanding the way that skills transfer from training to performance environments depend on a careful analysis of the social characteristics of these settings.
Illusions of Comprehension and Control
Subjective experience is a compelling basis for judgment. It is also an error-prone source for judgment. The paradox suggested by these statements is understood when one realizes that people do not regard their own experience as being error prone. Feelings of knowing and actual comprehension of a topic are often disparate. Reasons for this disparity are people's failure to recall material or solutions presented in the past, confusing the way material is presented with understanding of the material, confusing general knowledge about a domain with the specific knowledge represented
in a text, and a confusion of memories for imagined events with events that really happened. Also misleading is people's assumptions about other people's subjective experience: people often fail to realize that any event can have multiple interpretations. Indeed, most sources of knowledge are subject to debate and "resolved" through a better understanding of the reasons for different interpretations.
The committee has found that teachers, trainers, and students can reach a judgment that they are more proficient at a task than they are. Subjective experience can be misleading. Regular challenges are needed to provide learners with experiences that reveal the actual extent of their understanding of the task or material they are learning. In fact, benefits may derive from making the conditions of learning more demanding than the anticipated real-world (performance) conditions.
LEARNING AND PERFORMING IN TEAMS
Chapters 5-7 cover factors that influence performance in team settings. Recognizing that the operating unit in most organizations is the team, the committee considers what factors contribute to improved team performance.
Cooperative learning consists of peers working together to enhance their individual acquisition of knowledge and skills. Although some techniques are tailored to particular content areas and institutional settings, most are general-purpose "packages" with the following characteristics: precooperation activities (specifying the technique, creating goals and incentives, group assignment and, often, training on the cooperative script and roles to be enacted); the cooperation episode, which may include such aids as worksheets, visual displays, and computer assistance; postcooperation activities in which group members examine how they have functioned; and outcome assessments of knowledge and skill acquisition. Little empirical attention has been given to the effect of pretraining in the precooperation stage, to the use of task aids in the cooperative phase, to the effects of self-analysis on subsequent activities, or to transfer to other tasks during the outcome assessment. (The lack of attention to transfer is also characteristic of evaluations of simulated role exercises, discussed in Chapter 6.)
Nowhere in this volume is the interplay between social and individual processes more evident than in work on cooperative learning. The framework developed by the committee combines insights from a social-behavioral perspective that emphasizes task and incentive structure with those derived from a cognitive-developmental perspective that emphasizes cooperative activities. The cooperative experience is an attempt to foster indi-
vidual learning and social interactions, both of which lead to enhanced performance.
The evidence accumulated to date supports the effectiveness of the approach, at least under some conditions. But there are limitations. One important limitation is that most of the research has been done with children. Adults may not benefit from cooperative approaches in the same manner as do children, and there is a need to take into account the complexity of most adult learning tasks and to augment the cooperative experience with appropriate aids. Unlike many children, adults may be skeptical of attempts to manipulate the learning environment; they need to understand precisely the purpose of the intrusion and to perceive the benefits likely to result from the experience.
Research on cooperative learning with adults is limited, and additional research is needed to validate several tentative conclusions. If implemented appropriately, adult cooperative learning apparently can be more effective than individual learning across a variety of topics and tasks. Participation in well-constructed cooperative learning exercises can result in the acquisition of skills that can be used in a variety of settings. During the early stages of cooperation, participants benefit from such aids as information displays, worksheets, computers, and tutoring. For participants who lack experience with cooperative learning, it is useful to expose them to scripts that encourage productive cognitive and social activities. However, for adults, group rewards are unnecessary and may even hinder script enactment.
Further research is needed to understand why cooperative learning effects occur and the conditions that determine when they will occur. Effects of a number of potentially critical variables have not been explored to date. Examples include effects of pretraining on group skills, the availability of various communication aids, the effects of group processing, and assessment of transfer to other group tasks. Particular attention needs to be paid to the cognitive activities that are salient during the cooperative interactions, many of which are largely unobservable during individual training.
The Performance and Development of Teams
Critical to an understanding of the factors that influence team performance is the idea of ''throughputs" or mediating processes. The structural aspects of tasks and team members affect performance because they affect such mediating processes as interaction, coordination, and cohesiveness. Although various team-building interventions have been shown to influence these processes, team performance is also affected by the organizational context within which it occurs. Recent findings make clear the key role of the external environment in decisions made by teams.
A broader framework for thinking about teams emphasizes the way
they relate to each other across boundaries as a "network" of interacting entities that can contribute in positive or negative ways to organizational effectiveness: for example, team representatives can amass important resources through effective negotiation, and they can also develop "blind spots" and perceptual distortions due to the pressures on them to identify strongly with team products and positions. Yet even expanded perspectives on team performance do not capture the learning process that must be understood by trainers. More useful approaches to team learning come from other directions, including work on developmental phases, metacognitions, and shared mental models.
Particularly compelling is the notion of transition points in a team's development. These are junctures during a team's life history when "a major jump" in progress occurs. It is the time when team-building interventions and related forms of feedback are likely to have significant effects on performance. Knowing just when these turning points are likely to occur is a major challenge to both analysts and organizational consultants.
This work has major implications for training. First, knowing which variables have strong effects on performance suggest where to focus the training effort. Second, being able to distinguish between those variables over which teams members have some control from those over which they have little control further defines a trainer's focus. Third, knowing when important changes in the life cycle of a team may occur helps to identify points of entry into the process. And, fourth, knowing about some consequences of member identification with teams can alert trainers to sources of conflict between teams.
Team-building interventions may boost morale and enhance team cohesion, but they have less effect on team performance, which is influenced more by contextual and organizational factors. Although further research is needed, these findings suggest caution be used in claiming that the techniques benefit performance. The research should follow a multimethod approach in which both quantitative and qualitative methods are used.
The enhanced cohesion and morale resulting from team-building activities can increase intra-organizational conflicts between teams. By strengthening the ties between members within teams, the interventions can weaken relationships with members of other teams. This effect is heightened to the extent that team-building exercises include strategy formation as part of the procedure. The impact of team building on interteam relationships requires further examination.
Team-building interventions can be improved in three ways: timing the intervention to occur during transition periods, defined usually at the temporal
midpoint of a team's calendar, when a major jump in progress occurs; after team members engage in a self-diagnosis of their problems; and when shared mental models are developed during the team's preparation sessions. The development of shared mental models among team members deserves further study.
Team variables exert stronger influences on the performance of individuals in teams than do member characteristics. However, the social processes of interaction, coordination, and cohesion are critical determinants of the way that team inputs (task characteristics and team resources) influence team outputs (quantity, accuracy, and proficiency of performance). For example, the effects of task fidelity on the accuracy or quantity of a team's output depends on the way that fidelity affects coordination. These findings derive from the path models developed from a meta-analysis of the team performance literature. Models of this type deserve greater attention in the development of theories of team performance.
Educational games are more effective than other methods in instilling positive attitudes toward a subject and in enhancing interest in the material, but they are not more effective than other methods in teaching subject matter. They may, however, aid in the retention of material already learned.
Further research should be designed to provide answers to questions about the reasons for effects shown to occur in gaming environments. Most evaluations reported to date have been demonstration studies focused on whether the technique works. In fact, both team-building and gaming exercises need to be "unpacked" in order to determine what works. In addition, research is needed to establish the value of games in team training.
Training in Teams
Team training can be distinguished from team-building or cooperative learning. Training in teams can be considered in a four-part framework: inputs, the resources and tasks used for training; the individual processes influenced by the training tasks; mediators, the social processes that must be taken into account for effective team training; and outcomes, the kinds of individual and team changes that result from training. The clear conclusion is that training in teams has many benefits. To realize these benefits, however, entails an understanding of the way that the process works. A framework that recognizes the process provides the ingredients for effective training: what to provide (inputs), what to focus on (processes), how to structure the training (mediators), and what to assess at the end (outcomes). Training effectiveness can be improved at each phase. With regard to
inputs, trainee motivation can be enhanced by a supportive atmosphere, by knowing what to expect, by encouraging voluntary participation, and by the realization that what is learned in training will be used on the job. The importance of ensuring that the resources needed for training are provided should not be underestimated. Especially for complex tasks, it is important to give learners a conceptual understanding of the problems, ensure active involvement in learning, provide process feedback from other team members, and strengthen critical attitudesnamely, commitment to quality work and self-efficacyduring training.
Teams have been found to be effective vehicles for enhancing these processes. But they are more likely to be effective if the teams are structured on the basis of positive interdependence and face-to-face promotive interaction. Equally important are the mediating variables of individual accountability for performance, interpersonal skills, and team processing. The idea of a team that monitors its own progress and reacts to feedback provided by observers is central. Such team processing has been shown to contribute to effective performance on tasks that require coordination for success. Results from many studies on team performance can be explained in terms of the development of shared mental models (e.g., Orasanu and Salas, 1993). Team outcomes are multifaceted: they refer to more than proficiency and productivity. Positive relationships, psychological health, self-esteem, and social competencies are important contributors to performance in the workplace. So, too, are the various changes that occur at the level of teams. The relative emphasis to be placed on each of these individual and team outputs depends on how tasks are accomplished on the job.
The committee also concludes that intact work teams are likely to gain more from training than ad hoc assemblies of people assigned to teams only for training sessions. Long-term performance gains for teams are more likely if trainees function in similar team configurations in the workplace and are supported for doing so.
MENTAL AND EMOTIONAL STATES
Chapters 8 and 9 deal with the role of mental states in performance. Of particular interest is the way that altered states may affect performance. We cover processes related to building self-confidence, hypnosis, restricted environmental stimulation, meditation, and sleep learning.
Self-Confidence and Performance
It has become apparent that perceptions of self-confidence play an important role in performance. It is also apparent that these perceptions can be manipulated to enhance performance. However, such manipulation is not a
simple matter. Perceived self-confidence involves both cognitive and motivational processes. It is rooted in beliefs about what contributes to performance. It also reflects a desire to perform in the sense of being "psyched" for the challenge. Thus, information about accomplishments and feelings of readiness to perform contribute to people's sense of effectiveness. These findings provide a basis for suggesting programs designed to enhance performance through increasing a person's perception of his or her efficacy.
One suggestion is that the speed and quality of skill acquisition can be improved by emphasizing the learnability of the skill to be taught. Another is to provide opportunities for learners to observe people of widely different backgrounds succeeding at the task. A third suggestion is to help learners to attribute their success to skill improvement or hard work and their sub-par performances to a lack of effort, a lack of sufficient practice time, or the use of inappropriate strategies. And a fourth suggestion is to make rewards contingent on performance rather than given simply for participation or distributed in the context of competitions that heighten social comparisons.
Four kinds of programs provide alternative ways to strengthen perceptions of self-confidence. Programs that emphasize performance accomplishment consist of instructional strategies that use goal setting and feedback to bolster the process-related goals of effort, form, and strategy, all of which are under a performer's control. Programs that focus on modeling techniques use model performers who demonstrate how a task is done, along with coping strategies. Included also in such programs is the use of videotapes to enable a performer to edit out mistakes. The third kind of programs use persuasion techniques, which are especially useful during the early stages of skill acquisition. They consist of "sandwiching" skill instructions between words of encouragement, such as compliments for achievements, and goads to keep trying.
The fourth program type is designed to reduce anxiety in order to increase a performer's perceived coping efficacy. As concluded in the committee's first report (Druckman and Swets, 1988), anxiety is effectively reduced when individuals feel that they have control over an uncertain future and the potential threats or risks that exist in those unknown situations. By concentrating on short-term goals, learners' feelings of uncertainty are likely to be reduced. Participation in simulation exercises can serve to reduce uncertainty, especially when the experience is viewed as being successful. Although those four types of programs are often presented as alternative strategies, they can be used in combination.
Altering States of Consciousness
Chapter 9 discusses some implications of altered states of consciousness for performance: hypnosis, transcendental meditation, restricted environmental stimulation, and sleep learning.
A long history of research and practice clearly shows that hypnosis is effective in reducing the experience of pain. Therefore, it enhances performance to the extent that performance is impaired by feelings of pain and fatigue. Although not everyone is hypnotizable enough to experience this effect, some people may still receive some benefit from the placebo component of the hypnotic procedure or from training in nonhypnotic stress inoculation. The effectiveness of hypnosis in increasing muscular strength and endurance, sensory thresholds, learning, and remembering has not been demonstrated. Hypnotized subjects may believe that they are doing better, and this belief may have positive motivational implications for performance; however, the subjective experience of performance enhancement due to hypnosis appears to be illusory.
Transcendental meditation has been promoted as a technique for enhancing performance, largely by reducing the deleterious effects of stress. The evidence to date does not support that claim. The committee's analysis focused on physiological arousal, relaxation and anxiety, and psychological health. Although there is a voluminous body of research on transcendental meditation, the studies reported to date suffer from a variety of methodological flaws that prevent making firm conclusions. For example, it is not clear whether the positive effects reported for transcendental meditation (versus other forms of relaxation) are due to the unique features of the technique, or to the frequency and discipline with which transcendental meditation is practiced.
Restricted Environmental Stimulation
Restricted environmental stimulation (REST) is a set of techniques designed to reduce the level of environmental stimulation to a practicable minimum. It has also been promoted as a technique for enhancing performance. The evidence supporting this claim is based on the therapeutic effects of REST in habit control. Despite examples of the performanceenhancing effects of REST and published reports of demonstration experiments, the evidence to date does not lead to firm conclusions about the effects and their underlying mechanisms.
In the committee's first report (Druckman and Swets, 1988), the possibility was raised that sleep learning may be possible, that material learned
during light sleep could be expressed as implicit memory in the absence of explicit recollection. Recent evidence, however, does not support this possibility. Although some degree of sleep learning may be possible, it is likely to be inefficient and to have detrimental effects on a person's subsequent waking performance.
Chapters 10 and 11 present new directions for future research and application on two topics that have recently been considered more closely in relation to enhancing human performance: socially induced affect and thought suppression. We cannot draw implications for techniques from the tentative conclusions now possible, but they are a basis for a research agenda proposed by the committee.
Socially Induced Affect
Social influences on performance are evident in many of the chapters of this volume. Communication and influence are central processes in cooperative learning, team training, and motivational programs. The social dimensions of performance were also evident in earlier topics discussed by the committee; for example, influence through mimicry (Druckman and Swets, 1988:Ch. 8), group cohesion (Druckman and Swets, 1988:Ch. 9), and deception detection (Druckman and Bjork, 1991:Ch. 9). In each case, the communication process contains both verbal and nonverbal elementsmessage content and expressions that convey emotion. Socially induced affect considers the possible effects of emotions transmitted from one person to another during social interactions.
Considerable evidence supports the idea of socially induced affect, that is, that one person's expressed feelings can influence another's feelings. The transmission is especially strong for influencing emotions of the same valence (happy to happy or sad to sad feelings) than for influencing emotions of the opposite valence (happiness to envy or vice versa). Less clear, however, is the role of socially induced affect as a contributor to one's overall emotional experience, the frequency and strength of its occurrence, the way it is manifest in real-life situations, and the variables that influence it. Also unclear are the mechanisms that cause it to occur: some alternative theories concern the role of cognition in attributing intentions to the model, classical conditioning of the association between events and emotions, and mimicry of facial expressions that precede the experienced emotions. It may be that each of these mechanisms plays a role in the transmission process, but that one or another is more important in any particular instance. There are some interesting implications for performance of socially induced affect, but they are still subject to further
documentation by research, as are the more basic questions about the fundamental phenomenon.
Thought suppression is the intentional avoidance of a thought or category of thoughts. Although everyone has reasons to avoid thoughts about painful experiences or uncertain future events, recent research suggests that the desire to suppress thoughts may not be effective. Paradoxically, when people use suppression to free themselves of unwanted thoughts, they may actually increase the emotional power of the very thoughts they are trying to avoid. This paradox was recognized by Freud and it is now being explored in laboratory experiments. The early research suggests that an attempt not to think about an unwanted thought is likely to fail if it is the only strategy a person adopts for dealing with that thought.
Alternatives to thought suppression exist that are likely to be more effective. In cases of anxiety-producing or obsessive thoughts, successful avoidance of the unwanted thought may occur when one faces the thought and even concentrates on it. Encouraging people to talk about their unwanted thoughts may enhance their ability to cope with the events. It is not known whether this strategy is useful in all cases, and there are important exceptions. For example, encouraging depressed people to dwell on their problems is a technique that has not received enough research attention to allow any evaluation. In the cases of unwanted thoughts about fears, traumas, or worries, however, the approach of confronting them may be more beneficial than the approach of trying to suppress them.
Further research is needed on the circumstances in which dwelling on an uncomfortable thought can be detrimental. Research is also needed on the implications of forgetting old, unneeded, information that may interfere with performing a new task, such as reaching for a control that was in a different location on an old plane. Continued sensitivity to the old information may result from earlier attempts to suppress that information. But if there are new replacements for old information that allows a person to update his or her memory, the person may not suffer from the undesired retrieval of irrelevant old items. The research challenge is to discover ways to forget unneeded information without suffering the troubling effects of thought suppression.
IMPEDIMENTS TO EFFECTIVE TRAINING
In the epilogue the committee discusses institutional impediments to effective training. Many of the conclusions reached in this report, and in the committee's previous two volumes, suggest training procedures for improving
performance, but whether organizations actually implement such suggestions is at issue. On the basis of more than two dozen site visits, the committee has observed a number of organizational values, attitudes, and structures that appear to impede effective training.
Training is often not highly valued within organizations. This is due in part to financial constraints, but it is also the result of an attitude that attributes differences in performance not to levels of training, but to "aptitude." This attitudethe innate-ability fallacyleads organizations to emphasize selection more than training. It ignores the large body of research findings that suggest that practice, not innate ability, is the critical factor in determining performance.
This attitude toward the primacy of selection also leads to a misunderstanding of the training process in several ways. First, the tendency to avoid errors during training contradicts the findings that making mistakes is a necessary part of the training process. Second, the tendency to view tests as assessment devices negates their value as learning devices. And, third, the tendency to use evaluations of performance immediately after training as indicators of success misses the goal of training, which is to transfer learned skills to the settings in which trainees will work. These attitudes create a "catch 22" that impedes progress: training programs are not as effective as they might be because training is not highly valued, and training programs are not highly valued because they are seldom as effective as they might be.
Organizational attitudes and structures also effect the extent to which a trainer can maximize his or her effectiveness. The tendency to view teaching as a talent rather than a learned skill impedes the development of the skills needed to be an effective teacher. Similarly, the failure to view teaching itself as a difficult skill to be learned leads organizations to recruit experts in a given domain without regard to their experience as teachers. In addition, there is the tendency of organizations to design administrative structures that isolate instructors or put them in competition with each other. These structures impede the kind of communication and cooperation needed to share knowledge, innovations, and solutions that enable teachers and trainers to be as effective as possible.