In the present section, our focus turns to some social processes that are likely to occur within small crews. We shall consider the specific topics of leadership, cohesiveness, conformity, group performance, and group change over time.
Leadership is a social influence process. According to Hollander (1978), in the course of this process, leaders organize, direct, and coordinate followers. They also exert influence to (1 ) help the group maintain harmony and stability, (2) interpret the conditions that confront it, (3) set goals, and (4) meet challenges posed from without. The leader's right to exert this influence may be conferred through appointment by a higher authority, or by the group itself.
Although one group member may be appointed or elected leader and assigned distinguishing tokens of status and rank, the leader/follower distinction is oftentimes blurred (Hollander, 1978). First, leadership is a relational concept, with the result that the person who is leader from one perspective is a follower from another perspective. In any multilevel organizational hierarchy, most people will fill both leader and follower roles. Second, leadership involves a two-way influence process. Although the focus of discussions of leadership is likely to be on the leader's influence over his or her followers, the followers influence leaders in return. Leaders, for example, need followers' inputs to reach informed decisions. The appointed leader who fails to respond to followers' influence attempts may find it increasingly difficult to lead, and the nonresponsive elected leader may be booted out of office. Finally, despite the fact that there is an identifiable leader, certain leadership functions may be dispersed among various group members. As we shall  see, one person may organize and direct the group while another attempts to satisfy the group members' human needs.
The course and outcome of the leadership process depends on three general factors (Hollander, 1978). The first is the characteristics of the leaders themselves, including legitimacy, ability, experience, motivation, interests, expectations, and so forth. The second is the characteristics of the followers, again including such considerations as ability, motivation, interests, and expectations. The third is the characteristics of the situation, including the nature of the physical and social setting, the available physical and social resources, and the structure of the group's task. Prescriptions for good leadership often dwell upon the selection and training of leaders. However, such prescriptions could also involve the selection and training of followers, and the structuring of situations and tasks.
Heavy demands will be placed upon people performing leadership functions in space-capsule microsocieties. These demands are expected to become increasingly burdensome as the mission continues.
First there will be the stringent technical requirements associated with operating safely in a hostile environment. Although there will be advance preparation and some degree of communication with resource people at mission control, supplies will steadily decrease and as distance increases it will become increasingly difficult to maintain good communications with Earth. To a great extent, problems will have to be solved using the highly limited resources available in the closed environment of the space capsule.
Second, the demands on leaders' interpersonal skills are likely to be formidable. As noted earlier, it has been hypothesized that isolation, confinement, risk, and other conditions associated with spaceflight are conducive to deteriorating interpersonal relationships (George Washington University, 1974). Because of this possibility, people in leadership roles will have to be shrewd judges of human nature and display superior interpersonal skills.
A failure to fulfill the requirements of leadership can lead to severe penalties for an isolated and confined group. In the 1959-1960 fallout-shelter studies, a deliberately passive role on the part of the shelter commander was credited with a general lowering of standards of behavior and a loss of interest in matters of civil defense (Strope  et al., 1960, 1961). The Georgia fallout-shelter studies also revealed that a failure in management led to increased friction and decreased morale (Hammes et al., 1965; Hammes and Osborne,1965; Hammes and Watson, 1965). Weak or incompetent leadership has spelled disaster for a number of polar expeditions (Leonov and Lebedev, 1975). Competent leaders, on the other hand, may serve as models whose enthusiasm and even temper are emulated by the crew. Good leaders can prevent factionalism and ease group members through troubled relationships.
Task and socioemotional leadership activities- Repeatedly, distinctions have been made between task activities (also known as initiation of structure and concern for production) that help the group get the job done or move toward its goals, and socioemotional activities (also known as showing consideration and concern for people) that promote harmonious relations within the group. Group functioning requires people who take the initiative in each of these areas. Socioemotional leadership is at least as important as task leadership, and perhaps more so, judging by some of the research.
It is not clear how often the same individual can satisfactorily fill both task and socioemotional leadership roles. The pioneering research by Bales and his associates (Bales, 1950, 1953, 1958, 1970) found that some group members engaged in more task and socioemotional activities than others, and as a result were offered leadership status. But it was also found that the person who engaged in the most task activities was not the same person who performed the most socioemotional activities. There were, in effect, two leaders: the task leader, who was rated as having the best ideas, offering the most guidance, and being most influential in forming the group's opinions; and the socioemotional leader, who was the best liked. The usual explanation for the emergence of the second leader is that a task leader's sense of purpose gives rise to activities (unpopular orders, sharp criticism, etc.) that hurt group members' feelings. The second leader emerges to smooth things over and restore harmony to the group.
However, it should be noted that the initial studies involved emergent leadership. That is, unacquainted individuals joined in a discussion, and social structure emerged as interaction progressed. The task leader took a role of power and influence, and it may have been his presumptuousness that caused the internal conflicts. According to Burke (1972), when a leader is designated by a higher authority and is hence perceived as "legitimate," group members are  more accepting of heavy-handed task acts and the need for the second leader diminishes. Certainly, Thor Heyerdahl and many of the other expeditionary leaders discussed by Leonov and Lebedev (1975) appear to have had heroic capacities to perform both task and socioemotional leadership roles. Yet, the evidence reviewed by Katz and Kahn (1978) suggests that only under rare conditions are task and socioemotional leadership roles best filled by the same individual.
Researchers, planners, and managers are thus confronted with the problem of understanding the optimal distribution of leadership behaviors within the crew. Specifically, to what extent should various task and socioemotional leadership behaviors be concentrated in the hands of a specific leader, rather than distributed across two or more people within the crew? Of particular interest in light of Burke's arguments is determining the extent to which a given crewmember should attempt to manage both task and socioemotional leadership roles. To begin with, Leonov and Lebedev's and Katz and Kahn's conclusions need to be reconciled. One possibility, suggested largely by Burke's research, is that structural factors, or possibly certain kinds of leader-follower compatibilities, make it possible for people such as Thor Heyerdahl to perform both roles. Another, definitely counterintuitive but intriguing possibility is that isolation, confinement, and risk make it relatively easy for one individual to perform both task and socioemotional leadership roles. Yet another alternative is that under conditions of isolation, confinement, and risk, followers have a strong defensive need to perceive their leaders as highly competent on both task and socioemotional dimensions. Evaluation of these alternatives awaits the results of further research.
Personality, situation, and leadership- A prevalent theme is that certain personality traits can be identified with effective leadership. Summarizing the results of scores of studies, Mann (1959) reported that intelligence, adjustment, and extroversion are moderately related to leadership, whereas dominance, masculinity, and interpersonal sensitivity are somewhat less closely related. Observations of Sealab II led Radloff and Helmreich (1968) to suggest that people under stress in isolation and confinement may not need a young, action-oriented leader as much as a mature individual who inspires identification and provides reassurance. Citing work by Misumi and Shirakashi (1966) and Cooper (1966), Kubis (1972, p. 55) derived the following composite picture of the effective spacecrew leader:
A person who is a highly competent leader under one set of conditions may prove to be a marginal or incompetent leader under another set of conditions; thus, both personality variables and situational variables influence quality of leadership (Fiedler, 1967, 1971, 1978; Hollander, 1978; Katz and Kahn, 1978; Mann, 1959). Perhaps the most promising theory which simultaneously considers personality and situational factors is Fiedler's (1967, 1971, 1978) contingency theory of leadership. Concerned with predicting performance rather than satisfaction or morale, the theory has been tested successfully in many military and civilian settings, and deserves close attention from space mission planners. The independent variables are situational favorableness and leadership style, and the dependent variable is leadership effectiveness.
Situational favorableness refers to the structural and social climate variables which make a group easy (high favorableness) or difficult (low favorableness) to lead. These include (1) the extent to which the leader is accepted and respected by the group, (2) the extent to which the group's goals are clear and structured, and (3) the extent to which the leader has been invested with the power to reward and punish group members.
Leadership style refers to the leader's orientation toward tasks and people. This can be measured by asking the leader to evaluate the least preferred co-worker (LPC) with whom he or she has ever worked. High scorers, who tend to give favorable ratings to the least preferred co-worker, are relatively socioemotional in outlook. Low scorers, that is, people who assign harsh ratings to their least preferred co-workers, have more of a no-nonsense, task orientation. In Fiedler's theory, high scorers and low scorers are referred to as high LPC leaders and low LPC leaders, respectively. Leadership effectiveness, the dependent variable, can be measured by any objective measure of task accomplishment.
According to contingency theory, different degrees of situational favorableness require different leadership styles. Under conditions of very high or very low situational favorableness, the  task-oriented, low LPC leader is likely to prove most effective. As Jacobs (1971) so aptly puts it, the leader can afford to be firm when accepted by the group, pursuing clear goals, and invested with power to reward and punish. He or she must be firm when rejected by the group, grappling with ambiguous goals, and lacking the power to reward or punish. Under conditions of intermediate situational favorableness, the interpersonal sensitivity of the high LPC leader is likely to be of use in working through the moderately troubled relations within the group, thereby freeing the group to continue toward its goal.
Careful planning may create and maintain a high degree of situational favorableness on short-term missions, but such conditions may be difficult to sustain on prolonged flights. For example, it may be relatively easy to link the on-board leader's evaluation of crewmembers to the latter's continuation and advancement within the space program. However, as noted in chapter VIII, as the link with Earth becomes tenuous, traditional bribes and threats may lose force. Thus, we hypothesize that whereas task oriented, low LPC leaders may do best on carefully planned short flights, socioemotionally oriented, high LPC leaders may have an edge on extended flights.
Contingency theory posits that leadership style is fairly firmly ingrained. This implies that whereas leaders may be selected for missions of varying degrees of situational favorableness, training programs intended to change their styles could be ineffective. However, Fiedler and his associates have developed a self-instructional program called LEADER MATCH, which helps leaders self-select and gain control over such variables as those which determine situational favorableness (Fiedler, Chemers, and Mahar, 1976; Fiedler, 1978). This program, as it stands, or with some modifications, may prove useful for space-mission leader preparation.
From Fiedler's contingency theory of leadership flow many lines for future research. Efforts must extend beyond assessing the situational favorableness of a given mission and then choosing the leader with the most promising style. First, it is necessary to understand fluctuations in situational favorableness over time. Second, procedures must be developed to help spacecrew leaders learn to identify shifts in situational favorableness, and change their behavior accordingly.
Autocratic and participative procedures- Leaders who make decisions without soliciting subordinates' inputs are said to use  autocratic procedures. Leaders who solicit subordinates' inputs are said to use participative procedures. These latter procedures include (1) consultive decision making in which the leader seeks opinions of informal leaders and of rank and file, (2) representative democracies, and (3) full democracies.
Early "leadership climate" research undertaken on the eve of World War I I suggested many advantages to the democratic approach (Lewin, Lippitt, and White, 1939). Most reviewers seem to believe that modal group members can offer very useful information, and conclude that, more often than not, the quality of a decision will be enhanced by membership participation ( Kleinhans and Taylor, 1976; Steiner, 1972, 1976). In addition, it has been found that organizational members are likely to feel a more personal commitment to decisions they have helped make than to decisions imposed from above (Coch and French, 1948; Hollander, 1978). However, the overall picture contains many complexities, and the optimum point along the autocratic-democratic continuum depends upon such variables as the personalities of group members, the distribution of knowledge and skills within the group, the group's size and organization, and the degree of structure of the problem (Hollander, 1978; Vroom, 1976; Vroom and Yetton, 1976).
Mission managers may find it neither necessary nor desirable to establish procedures such that all decisions can be traced to one point on the autocratic-democratic continuum. For example, the Russians have reported success with procedures whereby mission decisions were made by the commander and crew issues were decided by democratic votes (Leonov and Lebedev, 1975).
An analysis of situations, personnel, and likely scenarios would seem essential for maximizing the advantages and minimizing the disadvantages associated with autocratic and democratic decision making procedures. Perhaps a useful starting point for such research is the Vroom-Yetton normative model of participative decision making (Vroom, 1976, Vroom and Yetton, 1976). In essence, this model is applied by first answering seven questions (regarding such issues as the availability of information, the degree of conflict among subordinates, and the need for subordinate acceptance of decisions). On the basis of the pattern of answers, one of five decision-making procedures is prescribed.
In summary, future research on small-crew leadership should include an expanded range of leadership alternatives and options.  Particularly pressing problems include identifying the optimal distribution of task and socioemotional activities, and achieving a good fit between structural characteristics and leader characteristics. In addition, we need to know more about the consequences of various autocratic and participative decision-making procedures under conditions of isolation, confinement, and risk.
Cohesiveness refers to the strength, solidarity, or unity of a group. It is roughly synonymous to elan or morale. Although originally intended to be a unitary construct, cohesiveness sometimes designates a group with energy and a strong sense of purpose, and other times a group whose members have a strong sense of liking for and involvement with one another and who express positive attitudes about the group. Whereas drive and amiability often covary, it is possible for a group to be characterized by one of these attributes but not by the other (Lott and Lott,1965).
In some cases, adverse conditions and suffering seem to increase cohesiveness (Aronson and Mills, 1959; Gerard and Mathewson, 1966). In effect, undergoing a trying initiation encourages people to rationalize the discomfort by telling themselves that membership in the group is extremely desirable. However, most discussions focus on the rewards or satisfactions of group membership as the major cause of cohesiveness. Cartwright (1968), for example, has defined cohesiveness as the sum of the satisfactions which membership accords all of the members of the group. Satisfaction is likely to be high to the extent that the group (1) engages in activities that the members find intrinsically satisfying, (2) pursues goals of importance to the members, (3) provides social support and emotional gratifications, and (4) serves ulterior motives. Thus, a crew might be expected to be cohesive when the crewmembers (1) enjoy flight and adventure, (2) subscribe to the mission's overall goals, (3) encourage each other, and (4) help fulfill each others' needs in areas tangential to or unrelated to the mission.
The establishment and manipulation of group goals is one commonly prescribed way to promote cohesiveness (Sherif and Sherif, 1969). Goals refer to objectives or end states which have motivational properties in that they instigate, strengthen, and impart direction to behavior. Group goals refer to objectives or end states which are endorsed or found compelling by the different people who comprise the group. Group goals encourage people to coordinate their  activities for mutual gain, and hence are likely to affect the tone of interpersonal relations within the group. The isolation and confinement literature, for example, suggests that individuals may be able to suppress their differences in the interest of group goals. In Sealab II, some aquanauts commented that teammates who didn't always see eye to eye were able to get along for the period of the mission (Radloff and Helmreich, 1968). In a field experiment involving preadolescents who were camping in the wilderness, Sherif, Harvey, White, Hood, and Sherif (1961) found that incompatible or separatist goals promoted prejudice and open conflict, whereas goals that were of importance to all of the members of the group promoted amiability and unity.
First, mission planners and managers should take steps to help spacecrews identify group goals. Such goals should be superordinate in the sense that they override individual or subgroup goals which, if pursued, might encourage behavior that is detrimental to the overall mission. Crewmembers must feel personally committed to these goals; it will not suffice to simply impose them from above. Additionally, these superordinate goals must require a high degree of cooperative activity (Sherif and Sherif, 1969).
Second, steps should be taken to ensure that group goals are clear and well understood. Discussing Antarctic groups, Natani and Shurley (1974) have noted that scientists are given a brief introduction to the "big picture" at an orientation conference, but that their goals remain basically individualistic. Navy personnel are given only a minimal understanding of their science support role, with the result that they find it difficult to become firmly committed to the overall mission. Under such conditions, unwelcome forms of socioemotional behavior are likely to take precedence over welcome forms of task behavior.
Third, a means must be found to maintain astronauts' interest in distant goals over prolonged periods of time. It may thus be desirable to establish a number of interim goals which can be pursued and savored. Perhaps this has been best expressed by Sells and Gunderson (1972, p. 82):
Another way to promote cohesiveness would be to help minimize interpersonal conflicts. Presumably, initial social compatibility (as discussed earlier in this chapter) will be a major factor. Both task and socioemotional training can also be expected to help minimize or contain interpersonal conflicts. People who do not know what to expect and do not know how to do their jobs are likely to frustrate and annoy one another. In addition, people who are unskilled may respond to a poor overall level of performance by acting toward one another in negative ways (Shurley et al., 1977). Both Kubis (1972) and Berry (1973a) have advocated direct training in human relations. Such training may involve the entire crew, or, if this is impossible, crewmembers who are in managerial roles or whose assignment is to work with other people. Training in interpersonal relations was considered valuable by subjects in the Douglas simulation study (George Washington University, 1974).
To devise adequate training programs, more must be learned about the on-board diagnosis and management of interpersonal frictions and conflicts. Such research would be aimed at uncovering techniques that astronauts might use to identify and combat the underlying sources of interpersonal stress. It might address ways of recognizing and managing one's own rising tensions and tensions in other people. Satisfactory experiential training in interpersonal relations may require an authentic setting characterized by isolation, confinement, and stress.
Another technique for increasing spacecrew solidarity is derived from manning theory as developed by Barker (1968) and Wicker (1979). An implication of this theory is that the size of a spacecrew relative to the amount of work that needs to be done will affect the level of energy that the crew will apply to the task, the favorability of the attitudes that crewmembers maintain toward one another, and the overall level of involvement of the crewmembers in the group.
Adequate staffing occurs when there is a good match between the number of people available, on the one hand, and the situation's technical and social demands, on the other. Under conditions of adequate staffing, people are neither rushed nor are their abilities  allowed to languish. The group's interests and energies cover the immediate demands, but do not extend much beyond those demands.
Understaffing occurs when the number of people available is less than typically hoped for or expected, given the task demands. Wicker (1979) hypothesized that understaffing encourages people to (1) work close to the limits of their abilities and skills; (2) assume difficult, important, and varied tasks; (3) feel important, responsible, and versatile; (4) monitor and improve each others' performance; (5) display tolerance of the other people's occasional lapses; and (6) ignore personal characteristics (such as race, sex, and personality) that are irrelevant to performance. Thus, members of a "short" or understaffed spacecrew may feel a bit harried and pressured, but they may also work hard, feel good about themselves, and be supportive of one another.
Overstaffing occurs when the number of people available is greater than typically hoped for or expected, given the situations' demands. Wicker (1979) hypothesized that overstaffing (1) encourages perfunctory, lackadaisical performances; (2) fosters a high degree of specialization accompanied by a jealous guarding of all activities which fall within the associated narrow domain; (3) undermines one's sense of self-esteem; (4) encourages little concern of the crew for each others' progress; (5) develops cynical attitudes of the crew toward one another; and (6) encourages the crew to dwell upon race, sex, personality characteristics, and other task-irrelevant individual differences. Working in an overstaffed setting may be "easy time," but it may also be psychologically and socially unhealthy.
The theory of manning grew from studies of people engaging in voluntary activities (e.g., extracurricular school activities and church affairs) and in service professions, and most of its important tenets are yet to be tested in varied organizational settings. Nonetheless, this theory raises the interesting possibility that, subject to the limitations imposed by safety requirements, spreading a spacecrew just a little bit "thin" may enhance the crew's functioning. It must be emphasized, however, that manning theory needs to be evaluated m isolated and confined settings, that optimal understaffing is a question of degree, and that understaffing might have disastrous consequences when a group must continue its activities uninterrupted over a prolonged period of time.
 Some ambiguity surrounds the relationship between drive and amiability, on the one hand, and actual performance, on the other. Cohesive groups often are efficient and effective (Cartwright, 1968; Lott and Lott, 1965; Shaw, 1976). However, this does not necessarily mean that enhancing cohesiveness improves performance. First, successful performance can be a cause, rather than an effect, of cohesiveness. This is most clearly shown in Bakeman and Helmreich's (1975) study of divers aboard Tektite. It was found that whereas measures of cohesiveness obtained early in the mission were unrelated to measures of performance obtained later in the mission, measures of performance early in the mission were predictive of subsequent crew cohesiveness. Second, group standards or norms mediate the relationship between cohesiveness and performance. If norms support performance-related activities, then cohesiveness is likely to improve performance. If, on the other hand, norms support limited output or engagement in irrelevant tasks, cohesiveness may undermine performance (Berkowitz, 1954).
In summary, cohesiveness is typically viewed as a function of the rewards and satisfactions of group membership. Major research and planning issues are associated with promoting a high degree of cohesiveness within tomorrow's spacecrews.
Compliance, Conformity, and Independence
A certain amount of social activity aimed at ensuring adherence to group norms is generally regarded as beneficial, because it promotes coordination of efforts and a sharing of values within the group. However, such influence processes have certain potentially adverse effects which may become pronounced under conditions of isolation and confinement. Strong social pressures can inhibit the flow of creative ideas, particularly in a cohesive group. Individuals may fear that unorthodox suggestions will incur the leader's displeasure, undermine morale, or lead to rejection. This fear is justified, for it has long been known that a person who violates group norms is likely to trigger a specific series of events (Schachter, 1951). The initial reaction is an increase in communications intended to bring that person back into line. If these attempts are unsuccessful, communication ceases and the deviant is ignored. Under normal conditions, such ostracism may simply result in the deviant leaving the group, but under conditions of isolation and confinement, the deviant cannot leave the group.
 Prolonged rejection may lead to the pathological "long eye" syndrome (Haggard, 1964; Rohrer, 1961). Noted primarily in polar camps, this syndrome may involve hallucinations, tears, loss of appetite, silence, suspiciousness, and sloth. This is not only extremely punishing to the rejected individual, but it penalizes the group by robbing it of the services of one of its members. This can be a major problem in small crews that begin the mission only minimally staffed. Rohrer (1961) documented at least 10 cases of "long eye" in the Antarctic. Fortunately, the effects of "long eye" are temporary and vanish quickly when the individual is reaccepted by the group. Further research is required to better understand the causes and consequences of ostracism under conditions of prolonged isolation and confinement. It is essential to find ways to soften the blow of social rejection without entirely eliminating it as a mechanism of social control. Finally, it is important to find ways to reintegrate the deviant into the crew once the episode is over.
Strong conformity pressures may reflect attempts to maintain group harmony. Such harmony may be purchased at the expense of performance. Janis has coined the term "groupthink" to refer to conditions under which efforts to maintain group harmony undermine critical thought and lead to poor decisions (Janis,1971,1974; Janis and Mann, 1977). Spacecrews may be quite vulnerable, since groupthink becomes likely when the group is concerned with maintaining amiability, when there is little or no communication with people outside of the group, and when the group is confronted with a threatening situation. Among the most important characteristics of groupthink are (1) false optimism and lack of caution, (2) direct pressures on nonconformers, (3) a fear of disapproval for expressing new alternatives, (4) an illusion of unanimity, (5) the emergence of "mind guards" who protect the leader from criticism, and (6) efforts to deny or rationalize all ill-omens.
Safeguards against groupthink include soliciting external inputs during the decision-making process, appointing a devil's advocate to challenge majority views, and reconsidering decisions before action is taken. It is not clear that all of these safeguards are effective; for example, Bennis (1976) argues that a devil's advocate is ignored because the group recognizes that he or she is merely playing a social role. Those safeguards which are effective under "normal" conditions may or may not be workable under conditions of isolation and confinement. Specifically, some of the remedies proposed for groupthink presuppose a social system with a relatively permeable  boundary. More research is needed on the emergence and control of groupthink under conditions of isolation and confinement.
Special problems arise in emergency situations in which leaders are likely to exert authority to gain prompt acceptance of a course of action. Ideally, leaders would invariably make appropriate decisions which would then be gracefully accepted and implemented by the crew. Actually, leaders may make faulty decisions because they may lack certain information or fail to process information correctly. There are cases on record when aviation accidents or near accidents could have been averted if crewmembers had forcefully drawn the pilot's attention to unnoticed conditions or had openly questioned the pilot's instructions (Yanowitch, 1977; Murphy, 1980). That is, correctable pilot errors have gone uncorrected because of unquestioning attitudes, a lack of assertiveness, or deficient communication skills.
Certainly, we may expect selection and training procedures to yield leaders who have a higher degree of competence than the average member of the crew. Furthermore, we do not expect a leader's performance to be enhanced if he or she is constantly bombarded with wrong-headed suggestions. However, even as a leader's inability to induce compliance can be regarded as a failure in leadership, so can an inability to attend to relevant subordinate input. It is necessary to find ways to help crewmembers identify the conditions under which they should speak up, and to devise training procedures to ensure that they have sufficient assertiveness to do so. Moreover, it is necessary to find ways to help leaders respond appropriately to unsolicited follower inputs. In summary, it is necessary to identify and achieve a healthy balance between crewmembers' acceptance of social influence and a willingness to engage in appropriate independent action.
The determinants of group performance are many and complex. Some degree of order and coherence is imposed by the Hackman and Morris (1975, 1978) process model of group performance. The model's name is derived from its emphasis on the processes that occur as group members interact with one another.
According to the model, three factors or sets of variables affect group performance. The first factor is the knowledge and skills which group members are able to bring to bear. The second factor is the  amount of energy or effort that the group members apply. The third factor is the performance strategies or procedures that the group members follow while addressing their tasks. To the extent that conditions foster the application of skill and effort, and to the extent that groups choose appropriate performance strategies, the group will prove to be efficient and effective. To the extent that the application of skill or effort is dissuaded, and to the extent that groups follow inappropriate performance strategies, poor performance will result.
Hackman and Morris identify three levels of inputs which directly affect group processes and indirectly affect performance outcomes. Individual-level inputs include such variables as the group members' skills, attitudes, and personality characteristics. Grouplevel inputs include such variables as group size, morale or cohesiveness, communication patterns, and social norms. Environmental-level inputs include such variables as the nature and structure of the group's tasks, and the level of environmental stress. Through the manipulation of inputs, group processes can be altered and performance outcomes improved.
First, by manipulating group composition, changes can be effected in the knowledge and skills which group members bring to bear on the group's tasks and problems. For example, it might be possible for space crewmembers to be chosen, in part, on the basis of complementary skills and interests (Haythorn et al., 1972). A careful analysis of mission requirements and of the people who might satisfy them could result in a large crew with a good spread of abilities and skills.
Second, the model suggests that the manipulation of task characteristics can increase group members' commitment to work hard on the task. Drawing on the work of Hackman and Lawler (1971) and Hackman and Oldham (1974), Hackman and Morris (1975, 1978) suggest that a high degree of individual-effort expenditure becomes likely when individual assignments (1) provide the opportunity to use a variety of personal skills and abilities, (2) involve "whole and visible" pieces of work, (3) are related to the physical or psychological well-being of other people, (4) make some allowance for individual initiative and discretion, and (5) are structured in such a way as to allow the person to monitor his or her level of performance.
Finally, through manipulating social norms it should prove possible to influence task performance strategies. For example, as already noted, groups whose norms foster a high degree of  conformity may have difficulty solving problems that require innovative solutions. Pressures "not to be different" discourage group members from sharing new or unusual ideas. If new norms could be developed that encourage independence of thought and expression, the group could rapidly uncover a wide array of creative solutions to its problems. Hackman and Morris urge group members to continually ask themselves if their usual performance strategies are the most appropriate ways of proceeding under emerging conditions. They suggest that an open preperformance discussion of strategies can pay big performance dividends, particularly if the group finds itself confronted with a nonroutine problem or task.
This promising model of group performance has received only limited testing, and these tests have not involved isolated and confined groups. Certain of the issues raised by Hackman and Morris are prime research issues. These revolve around constituting crews in such a way as to ensure a desirable range of knowledge and skills, structuring tasks in such a way as to maximize motivation and commitment, and identifying and encouraging social norms which foster appropriate performance strategies under conditions of isolation, confinement, and risk. Some of these issues will be further addressed in our discussion of motivation in chapter VIII.
Interpersonal dynamics within a crew can be expected to change over time. Building on work by Bales (1950), a number of researchers including Bennis and Shepard (1956), Mann, Gibbard, and Hartman (1967), and Tuckman (1965) have identified developmental stages within small groups. Although these formulations differ in terms of their specifics, there is some agreement that early in the group's life, activities center around the feelings that members have about each other and their relationship to the leader. Only after these issues are resolved will the group become highly cohesive and gain the ability to effectively focus its energies on the job at hand. Illustrative is Tuckman's (1965) formulation based on studies of therapy groups. According to Tuckman, groups pass through four sequential stages: (1) testing and dependence or "forming," (2) intragroup conflict or "storming," (3) development of group cohesiveness or "norming," and (4) task functioning or "performing."
The research on group development implies that using, as a crew, a well-established or mature group that has thrashed out norms, tested limits, come to grips with authority relations, and  reconciled individual differences would help minimize interpersonal conflicts within the spacecraft. A relatively mature group may have another advantage: It can be studied as a group prior to departure and this study could provide a backup to the initial selection process (Perry, 1965, 1967). The disadvantage of a well-established group is that the members may find that they have little new to offer each other socially. That is, as the natural processes of social penetration continue (Altman and Taylor, 1973), crewmembers will become increasingly familiar to one another, find increasingly less to discuss, and perhaps lose interest in one another. Future research is required for selecting and training crews that are highly coordinated but retain the capacity to provide appropriate levels of mutual stimulation.
As missions increase in duration, crews will lose the characteristics associated with short-term groups and gain the characteristics associated with long-term groups. Ziller (1977) has identified four important ways in which short-term groups and long-term groups differ. First, whereas short-term groups are locked into the here-and-now, long-term groups have a sense of continuity and view themselves in relationship to the past, the present, and the future. Second, whereas short-term groups tend to be dominated by immediate outcomes, long-term groups have images of the future and may be willing to forego short-term gains in order to achieve long-term goals. Third, relationships within short-term groups tend to be impersonal and reflective of social roles, whereas relationships within long-term groups tend to be personal and based on strong attachments among members. Finally, whereas short-term groups tend to be relatively static, long-term groups tend to be dynamic, or, in Ziller's words, always "in process." Most group dynamics research involves short-term groups, yet it is long-term groups that we must understand to plan extended-duration space missions.
Not all microsocieties in space will be closed systems. There is likely to be some turnover in orbiting laboratories or settlements. This raises the problem of introducing and assimilating newcomers into preformed groups. According to a review by Crandall (1978), because newcomers do not share the continuing members' knowledge and attitudes, they may unintentionally act in disruptive ways and hence come to be seen as disloyal to the group. Aware of this problem, newcomers themselves are likely to be anxious and prone to conform.
Crandall reviews several methods for easing the integration of newcomers into ongoing groups. Each of these procedures may  reduce conflict and attrition, and hence deserves careful consideration for incorporation into mission plans. First, there is preentry therapy, which is expected to encourage anxiety control and reduce the need to conform. Second, newcomers can become acquainted with current or former group members prior to their entry. Third, newcomers can be given candid and realistic (as compared with guarded and idealistic) information about the group. Finally, newcomers can be sponsored; that is, an established group member can introduce and tutor each newcomer.
In the area of personnel rotation, there are many topics for future research. One hypothesis is that there is an advantage to using, as newcomers, people selected and trained along with those who have already entered space. Another possibility is that there is an advantage to letting the crew help select its own new members. Still another possibility is that assimilation is enhanced as a result of telecommunication with the newcomer prior to the newcomer's departure from Earth. Finally, it would be useful to know more about the kinds of conditions that will result in newcomers being given more time to gain acceptance by the group.
Yet another issue is fixing the number or proportion of crewmembers that should be rotated or replaced at any one time. In the military, piecemeal replacement has not been particularly successful. On the other hand, introducing large groups of newcomers means that many people have to be socialized simultaneously and that old-timers will feel particularly threatened. It is also necessary to find means for selecting individuals who are to be placed during a given personnel exchange.
In the future, it should become increasingly likely to actually train crewmembers or entire crews in space. This might be done, for example, inside a well-established satellite or lunar station. Such bases could provide a relatively safe environment where new astronauts could learn to function under conditions of weightlessness and come to grips with the psychological and social realities of isolation, confinement, and risk. In addition, the promise of being in outer space at a relatively early point in one's career might help generate and sustain a high level of motivation. On-board training should ease a new crew's transition from training to operational status. Seasoned operational crews might be more willing to accept new crewmembers who have already gained experience in space than those whose prior experiences have been limited to Earth.
 In summary, crew dynamics will change over time. Future research is required to better understand these changes. In addition, future research is needed to ease the integration of newcomers into crews that will have rotating memberships. On-board training may serve a number of useful purposes, including the integration of newcomers into seasoned operational crews.