A prevalent image of the astronaut is that of the rugged individualist who is prepared to set forth to open new frontiers. Kubis and McLaughlin (1967, p. 361) offer the following summary of early NASA selection programs:
The search was for an individual with a high degree of intelligence, preferably characterized by mathematical and spatial aptitude. He was to be sufficiently creative to contribute not only to the development of test and space hardware, but also to the planning necessary for the success of the space program. With an ability to work closely with others, he was expected to tolerate extreme isolation without anxiety. Though reliable and consistent in his behavior, he was to possess the necessary flexibility and adaptability to meet any emergency without psychological disintegration. Deliberate rather than impulsive, and with outstanding capacity to tolerate stress, his motivation for volunteering in the space program was to be mission-oriented rather than based on personal need for achievement.
Given the nature of early pioneering efforts in space, the rugged individualists chosen as astronauts appear to have been very appropriate. However, there are important differences between the missions thus far accomplished and those envisioned for the future. Given that future missions will require increasing levels of cooperative functioning, selection and training procedures must not only yield effective individuals, they must yield effective groups.
Voluminous literature addresses the problem of selecting, as crewmembers, people whose behaviors are compatible with one another as well as with the environmental systems (Altman and Haythorn, 1965, 1967a, 1967b; Haythorn, 1968, 1970, 1973; Haythorn and Altman, 1967; Haythorn, Altman, and Myers, 1966; Haythorn et al., 1972; Kanas and Fedderson, 1971; Kubis, 1972; Natani, 1980; Sells and Gunderson, 1972). Crewmembers may be considered compatible to the extent that each member shows qualities and emits behaviors that the other crewmembers consider desirable and appropriate under spaceflight conditions. The problem is not merely finding people with the right kinds of behaviors, but finding people whose behaviors intermesh in a good or positive way.  The issue is exceedingly complex, because so many variables are suspected of entering in (Kubis, 1972).
According to Natani (1980), leaders who are experienced veterans of the isolation and confinement experience can assemble an effective crew if they have a large pool of applicants and the time to conduct thorough interviews. Selection based on interviews by psychiatrists, psychologists, and other mental health professionals has generally been less successful, and selection based on the results of psychological tests has been the least successful of all (Natani, 1980). Given that future space programs will require increasingly large numbers of astronauts, it seems worthwhile to perfect psychological tests that can be simultaneously administered to large groups of candidates and then scored by computer (Natani, 1980). Ideally, future testing programs will screen out candidates with undesirable characteristics, select in candidates with desirable characteristics, and take into account group as well as individual considerations.
The limited evidence available confirms the expectation that a lack of social compatibility has an adverse impact on morale and performance. Social compatibility emerged as the foremost factor in analyses of supervisory ratings and peer nominations at polar stations, and compatibility has been related to whether or not the Antarctic adventurers had a "good year" or a "bad year" (Gunderson, 1963, 1968; Gunderson and Mahan, 1966; Gunderson and Nelson, 1963, 1965, 1966; Nelson, 1965). In simulation research by Altman and Haythorn and their colleagues, members of isolated and confined groups who were incompatible showed increased stress, withdrawal, and territorial behaviors. In addition, they made more attempts to withdraw from the study (Altman and Haythorn, 1965, 1967a, 1967b; Altman and Taylor, 1973; Haythorn, 1968, 1970, 1973; Haythorn and Altman, 1967; Haythorn et al., 1966, 1972). Russian observors have reported associations between social compatibility on the one hand and morale and performance on the other ( Leonov and Lebedev,1975).
Space travel has been a male-dominated enterprise, but at least eight women astronauts are in training and it is recognized that in the long run some sort of sexual parity is likely to be achieved (Shurley, Natani, and Sengel, 1977). Extremely little is known about women in space. Women have visited polar stations, lived in underwater habitats, and participated in fallout-shelter studies, but the vast bulk of  the data come from all-male preserves. Those few studies which have involved women have not focused sharply on sex or gender variables.
Several issues are involved when we consider women entering space. Perhaps the least of these is whether or not women are equipped for the rigors of life in space. Early doubts are giving way to a conviction that women can do the job and have the right to be there. This conviction reflects, in part, an increasing recognition of women's capabilities outside of the traditionally feminine sphere, and a growing recognition that technical systems are as easily engineered to meet women's as men's requirements.
There is evidence of attitudinal barriers against attaining sexual parity m space. Some space veterans, for example, have expressed the view that space tasks are men's tasks, and that women are unlikely to be able to meet the challenges (e.g., see Cunningham and Herskowitz, 1977; Oberg, 1981). Cosmonaut Valentina Tereshekova's orbital flight in 1963 has been dismissed as a political stunt, and her contributions to space exploration discounted, ostensibly because unlike the U.S. spacecraft of the era, the Russian craft were almost entirely controlled from the ground (Cunningham and Herskowitz, 1977; erg, 1981). The same critique has not been applied with equal force to Yuri Gargarin and other of Tereshekova's male contemporaries. In recent years, secondary "feminine" roles such as physician or stewardess have been deemed appropriate for female cosmonauts (Oberg, 1981); the second female cosmonaut, who joined a Salyut space station in 1982, was in fact a physician.
There are many complex issues regarding mixed-sex crews. Inclusion of crewmembers of both sexes can increase social diversity and provide the opportunity to exercise role options which might otherwise be forfeited for the duration of the mission. On the other and, there may be a certain lack of ease in dealing with members of the opposite sex under conditions of isolation and confinement, and the attitudes that astronauts of each sex hold about the other could prove problematical. For example, difficulties could arise if crewmembers of either sex have strong prejudices about the capacities and/or appropriate roles of the other. A review by Adams (1980) suggests that sexually mixed groups within the military are fraught with conflicts, often centering around male officers' expectations concerning female personnel. Other spectres include the failure of male subordinates to respond appropriately to female leaders' commands stress-induced strengthening of sexual stereotypes (leading, or example, to men taking unnecessary risks to impress women or  women faking helplessness in the presence of men), and clique formation along sexual lines. Additional research is required to identify ways to select crewmembers who are flexible and tolerant in their attitudes and responses toward members of the opposite sex.
Intimate sexual relationships could form within same-sex or mixed-sex crews, but concerns about such relationships are heightened in the case of mixed-sex crews. The reason for this is that two powerful factors are likely to prevent the formation of intimate sexual bonds within same-sex crews: the clear preponderance within the general population of heterosexual preferences, and strong cultural taboos that prevent the expression of homosexual impulses. Sexual bonds of any kind are potentially disruptive because Jealousies may arise as the result of other crewmembers "pairing off". In addition, a terminated intimate relationship which proves merely painful under normal conditions could prove devastating under conditions of isolation and confinement.
There is little basis for predicting sexual behavior in space. On the one hand, the periods of deprivation are likely to be so long that otherwise effective internal restraints may lose their effectiveness. Also, space travel can be very exciting, and a high state of excitement or arousal can fan sexual passions (Berscheid and Walster,1978). On the other hand, weightlessness and a rigorous program of technical activities may detract from sexual interest. In addition, there is at least some evidence that, within some small social systems, there develops social norms (i.e., shared expectations regarding appropriate attitudes and behaviors) which discourage members from choosing one another for sexual liaisons. Members of these systems seem to recognize that endogamous choices can fan jealousies and reduce privacy to a dangerously low level. The findings are tentative, however, and come from kibbutzim (Talmon, 1964) and residential colleges (DeLamater, 1974), which maintain relatively permeable boundaries and thereby make exogamous choices possible. Such expectations do not develop within all small social systems, for within the Navy, shipboard pregnancies have become a problem (Adams, 1980).
In summary, there are many unanswered questions regarding interaction within mixed-sex crews. Intolerance toward members of the opposite sex; counterproductive sexual stereotypes, attitudes, and activities; the formation of cliques on the basis of sex; and the generation of rivalries and jealousies are among the problems that need to be understood.
Although polar camps, subaquatic dwellings, and space simulators have tended to be male preserves, inhabitants have ranged in age from their late teens to their mid-forties or beyond. Group members who have deviated noticeably from the group's mean age have, like other members, been physically and mentally fit to stand the environmental rigors and make positive contributions to the group. Within the ranges studied, age has not emerged as an appreciable source of friction. Indeed, there are special advantages to including mature individuals. Research in subaquatic environments suggests that a mature individual may serve as a parent-surrogate who satisfies important needs of younger crewmembers (Radloff and Helmreich, 1968). It is significant that both members of the highly successful pioneering Space Shuttle crew were over 40.
Both relatively young and relatively old candidates have something to offer a spacecrew (e.g., peak energy and physical fitness in the former instance, and experience and perspective in the latter instance), but it will remain necessary to identify people who are biologically too young or too old to satisfy mission requirements. Computations of upper bounds, of course, must take mission duration into account; a person with sufficient youth and vitality to complete a 6-month mission may not be able to satisfactorily complete a 2-yr mission. It is also necessary to identify calendar ages at which people are likely to be perceived as too young or too old by other crewmembers. It is hypothesized that the inclusion of a person who is (Justly or unjustly) defined as unfit by other crewmembers will generate negative attitudes which undermine the person's self-esteem and disrupt group functioning.
Another area of concern is the changes that astronauts may undergo during truly extended missions. Thus far, even the longest studies of isolated and confined individuals have involved but a very small segment of the participants' lifespans. But developmental changes which are undetectable on short missions may become prominent during missions measured in years. Work in the emerging field of adult developmental psychology suggests that people undergo fairly pronounced changes at several points during their adulthood (Kimmel, 1980). The major changes of interests and goals which sometimes accompany these transitions could reduce the astronauts' fitness for the technical side of the mission, and also their social compatibility. At present, missions are not measured in years, our knowledge of adult development is modest, and it may well be that  commitment to a mission may prevent major changes of interests and identity. However, potentially significant developmental changes require consideration when planning a truly extended mission.
In summary, the evidence available suggests that tomorrow's astronauts will not have to be selected from a narrow range of ages. Additional research is required to identify means for establishing an appropriate age range and mix for a mission of given specifications. Eventually, some attention will have to be directed toward understanding adult development in space, and how such age- related changes are likely to affect fitness for and commitment to a mission.
The Americans who have visited space thus far have primarily been exemplars of the nation's dominant caucasian ethnic group; this is rapidly changing. Future missions will include a larger proportion of Black Americans, Latinos, and representatives of other minority groups. There is a possibility that, despite careful selection and training, prolonged isolation and confinement will bring long-standing prejudices to the fore. Although Kanas and Fedderson (1971) have discussed some of the implications of ethnically mixed missions, racial and cultural backgrounds have not been major variables in studies of isolated and confined groups. The literature, in fact, provides some basis for optimism. Specifically, three conditions associated with life in space may minimize certain ethnic prejudices.
First, some prejudice appears to be the result of an assumption that people from other ethnic groups maintain attitudes that are different from one's own (Stein, Hardyck, and Smith, 1965). In fact, astronauts are likely to discover that they have many interests and values in common (e.g., those centering around the mission). Such similarities should militate against prejudice.
Second, some prejudice flows from the perception of low social status rather than the perception of race or ethnicity per se (Allport, 1954; Amir, 1969). Since space voyagers are likely to come from a highly select population in terms of ability, education, and health, premission social status is unlikely to contribute to prejudice. However, care has to be taken not to somehow create invidious status distinctions along ethnic or cultural lines during premission activities or during the mission itself. Such distinctions could arise, for example, if minority astronauts were not fully trained, or if they were limited to relatively menial roles during the flight. In the  Russian program, guest cosmonauts from Eastern bloc nations received less training than their Russian counterparts, a discrepancy which seems to have occasionally led to some frictions (Oberg, 1981).
Third, under certain conditions, interaction is likely to lead to a reduction of prejudice (Allport,1954; Amir, 1969). Two of the most important conditions - cooperation and the pursuit of common goals-are likely to be found in space missions.
Prejudicial attitudes are but one factor potentially affecting the compatibility of ethnically heterogeneous crews. First, as noted in our chapter on communication, ethnically mixed crews, especially international crews, may experience language problems. Second, it may prove difficult to identify equipment, provisions, and schedules of activities that are acceptable to representatives of different cultural groups. For example, individuals from different nations may have different dietary preferences and aversions, and seek different types of leisure activities. Such within-group similarities and between-group differences may encourage the formation of cliques along national lines. Once such cliques have been formed, it may be but a small step to international disputes. Yet another problem is that representatives of different cultures may have conflicting views regarding appropriate astronaut behaviors. For example, Americans have relied upon the image of the assertive, competitive, achieving pioneer (Wolfe, 1979). American astronauts may thus find it difficult to accept an astronaut from a culture that stresses mild manners, decorum, and the tight control of hostility.
Thus, there are many important questions concerning the effects of cultural variability on the compatibility of tomorrow's spacecrews. Although little data are presently available, there are some interesting prospects for future research. The mixed-nationality crews of the large oil supertankers (Mostert, 1975) may provide a useful analog of international spacecrews. Increased interaction between the European Space Agency and NASA, and the inclusion of international crews aboard the Space Shuttle may also prove to be revealing.
Rawls, Hopper, and Rawls (1969) instructed college students to "List as many things as you can possibly think of that would determine how closely you would be willing to interact with another  individual." The other person's attractiveness in terms of such considerations as cleanliness, appearance, dress, and general demeanor emerged as a major consideration.
The search for complex bases for social compatibility should not cause personal attractiveness to be overlooked. Given that the need to staff large missions or to simultaneously staff a number of different missions will necessarily result in decreased selectivity and increased crew heterogeneity, it is necessary to learn more about the attributes which make an individual personally appealing in light of spacecrew norms. It might be useful, in this regard, to devise and validate an instrument for identifying personal characteristics which crewmembers are likely to find distasteful or annoying. This might involve a listing of physical characteristics, personality traits, and mannerisms (unkempt hair, dirty fingernails, stubbornness, and so forth) to be rated in terms of irritation value. Once perfected, such an "annoyance questionnaire" could be used in two ways. First, norms could be established to provide a basis for eliminating "unattractive" spacecrew candidates. Second, the instrument could be used for weeding out "finicky" individuals who find too many human frailties aversive.
A highly emotional or uncontrolled individual poses an unacceptable threat in any hazardous environment. Accordingly, it has been noted that Antarctic personnel place high premium on having calm, even-tempered, emotionally mature companions (Doll and Gunderson, 1971; Law, 1960; Lugg, 1973).
Many of the research questions surrounding emotional stability are questions of selection. Much more is known about how to exclude people who are liable to react badly than how to choose people of exceptional psychological health (Perry, 1965, 1967; Natani, 1980). Whatever the ultimate screening procedures, there is no getting around the fact that as more and more people are chosen for space missions, a few "high risk" individuals will inadvertently be included. Mission planners and managers need to know more about the kinds of supports or props that can be used to help people preserve or restore their emotional stability under conditions of isolation and confinement.
In some cases, crew perceptions of emotional stability may be more important than the facts. Behaviors which result in the  inference that the astronaut is emotionally unstable (whether or not that inference is correct) may demoralize that astronaut and undermine the confidence of the remainder of the crew. Of particular interest is identifying those conditions under which undue significance is read into an outburst or other act, with the result than an effectively functioning crewmember is considered no longer a member of the team..
Spacecrews are likely to have a strong work orientation. Like inhabitants of other dangerous environments, they will no doubt recognize that poor or incompetent performance on one person's part can jeopardize everyone's welfare. Thus, they are likely to exhibit a strong preference for colleagues who are able and willing to get the job done. Gunderson and Nelson (1965) found that "task motivation" related to "good years" and "bad years" in the Antarctic, and Shears and Gunderson (1966) reported that both personal motivation and perceptions of the group's achievements were related to satisfaction with the Antarctic assignment. Studies undertaken by the Alaskan Air Command also suggest that marginal performance is correlated with poor adjustment and dissatisfaction (George Washington University, 1974), and Day (1969) has offered a fascinating account of adverse reactions generated by crewmembers who failed to fulfill their performance requirements in the days of sailing ships.
Generally, attention has focused on technical or task competence at the expense of interpersonal or social competence. Because selection procedures are such that people who lack essential physical and intellectual skills are likely to be eliminated at the outset, the success of a group in an exotic environment may very well depend upon the group members' abilities to cope with one another. Along with adding interpersonal performance measures to research on competence, we need additional data regarding peer perceptions and ratings of competence. For example, as in the case of emotional stability, a competent crewmember who is not seen as such may have more of an adverse effect on performance and morale than an incompetent crewmember whose inadequacies are not correctly identified. Alternatively, anyone who is able to convey an impression of knowledge and skill may have a calming effect on the rest of the crew. A high degree of perceived competence is particularly crucial for those who lead isolated and confined groups (Leonov and Lebedev, 1975).
Space voyagers embark on highly interdependent ventures which require the utmost cooperation for success. According to McClintock (1972, 1978), people vary in terms of their interests in coordinating their efforts for mutual gain. Individuals who are governed by own-gain motivation tend to be interested in achieving what they can for themselves, and to be relatively uninfluenced by the anticipated effects of their actions on other people. Individuals who are governed by relative-gain motivation tend to be interested in doing better than other people; they view the absolute level of reward that they get as unimportant relative to whether or not they "come out on top" in the competition. Individuals who are governed by joint-gain motivation tend to prefer courses of action which yield benefits for other people as well as for themselves; thus, they show a sensitivity to other people's needs and a concern for their welfare.
McClintock and his associates hypothesize that each person is consistently governed by one of these three motives (Maki, Thorngate, and McClintock, 1979; McClintock, 1972, 1978). Each motive is believed to stem from early childhood socialization and reflects both familial and cultural values. A better understanding of these motives may prove useful in the flight-personnel selection process, or in establishing the most effective reward structures in the space-capsule microsociety.
Relevant to both competence and cooperativeness is Helmreich's work on the achievement orientation, or need achievement (Helmreich, et al., 1980). Classically, need achievement has been defined as a persistent preference for engaging in success-related activities (Atkinson, 1958). People with high need achievement have many admirable qualities, but problems may arise aboard a space vehicle if attaining standards of excellence involves "prima donna" behaviors or a put-down of other members of the crew. According to Helmreich, need achievement can be divided into three independent factors. Work orientation refers to the motivation to work hard because work is a valuable activity in and of itself. Mastery orientation refers to a desire to continually improve one's own performance. Competition refers to an attempt to do better than other people. Helmreich et al. hypothesize that the combined interests of task accomplishment and social compatibility will be best served if crewmembers show a strong work and mastery orientation but relatively little competitiveness. The hypothesis is provocative, given the competitive orientation of the early astronauts (Cunningham and  Herskowitz, 1977; Wolfe, 1979). The rationale is that competitive individuals are likely to generate interpersonal stress and hostility, adversely affecting collaborative performance and further undermining the quality of a social environment whose quality has already been heavily sapped by the conditions of life in space (Helmreich et al, 1980). The available evidence suggests that work orientation and mastery orientation positively correlate with performance, and competition negatively correlates with performance (Helmreich et al., 1980). Additional research is required to test this hypothesis under conditions analogous to extended-duration spaceflight. In addition, it would be of interest to explore the possibility that the extent of the frictions caused by one crewmember's competitiveness may depend on the orientations of the other crewmembers.
As noted in our introductory chapter, the conditions of spaceflight will restrict the opportunity to engage in varied roles. Crewmembers who can easily perform a wide range of role-related behaviors in flight should prove valuable, for they should help reinstate lost social opportunities. We hypothesize that the value of selecting crewmembers on the basis of an ability to engage in a wide variety of social behaviors should be inversely proportional to crew size and directly proportional to mission duration.
In Western society, men are expected to engage in task-oriented instrumental activities, and women are expected to engage in socioemotional or expressive activities. Men are expected to be autonomous, independent, somewhat dominating and aggressive, and emotionally inhibited. Women are expected to be warm and nurturant and to openly display their feelings (Bem and Bem, 1970; Broverman, Vogel, Broverman, Clarkson, and Rosenkrantz, 1972). There are, of course, great individual differences in the extent to which members of each sex adopt and enjoy such traditional social roles (Bem, 1974,1975).
Research by Helmreich and Spence (Helmreich, Spence, and Holahan, 1979b; Helmreich et al., 1980; Spence and Helmreich, 1978; Spence, Helmreich, and Holahan, 1979) suggests that some people have the capacity for both goal-oriented instrumental behavior and interpersonal sensitivity. The extent to which these capacities are exercised depends largely upon situational demands. Such people, who are referred to as androgynous, appear to have positive self-concepts and build rewarding interpersonal relationships. Helmreich  et al. (1980) note that androgyny appears highly desirable for astronauts, for a strong instrumentality combined with interpersonal sensitivity should be associated with both task accomplishment and social harmony.
In effect, androgynous individuals are individuals of either sex who can perform multiple roles, or at least central components of traditional masculine and feminine roles. Androgynous crewmembers, then, may have the value of increasing social variety within a crew. Future research should explore additional dual traits. As an example, consider the Jungian introversion-extroversion dichotomy. Who might be more valued under conditions of isolation and confinement: the shy and self-contemplative introvert, or the sociable and outgoing extrovert? On the one hand, introverts should be self-sufficient and accommodating of others' privacy needs. They may be well-equipped to weather extended communication breaks with Earth. On the other hand, extroverts should provide social stimulation and help satisfy others' affiliative needs. They may also thrive on attention from Earth when communication with Earth is possible. The hypothesis that each type of person offers mutually exclusive qualifications for extended-duration spaceflight poses a dilemma for those involved in the selection process. This dilemma would be resolved if future research revealed a third type of person, a person who exhibited behaviors commonly associated with both extroversion and introversion. In theory, this relatively flexible "ambivert" might be better able to deal with the rigors of space than the relatively inflexible introvert or extrovert. There are many other dimensions in which one might find that the best crew candidate is the one who possesses a duality of traits.
Similarities and Complementarities
There are numerous personal qualities and attributes whose effects on group compatibility are determined by the qualities and attributes of the other people in the group. In some cases, it is people's similarities that make for social compatibility; in other cases, people's differences intermesh. Such factors include attitude and value homogeneity, skill complementarily, and need compatibility.
First, conflicts of social, moral, and ethical values have proven to be a problem in some of the fallout-shelter studies (George Washington University, 1974), and almost all reviewers have tended to accept the position that homogeneous attitudes, values, and  interests will militate against interpersonal conflict. The expectation that crews composed of individuals with shared attitudes and values will tend to be compatible is certainly supported by studies in other contexts. Results from the field and from the laboratory have been spectacularly consistent: attitudinal similarity is a powerful determinant of mutual attraction. It has been repeatedly found that the proportion of shared attitudes determines the extent to which people find each other attractive (Byrne, Gouaux, Griffitt, Lamberth Murakawa, Prasad, Prasad, and Ramirez, 1971). Group members do not have to be completely "like-minded," for attitudes vary in terms of their relevance to the group. A group that reacts in a highly spirited fashion to dissimilarities on issues that it feels are significant may allow considerable latitude for differences of opinion in areas unrelated to its purposes and tasks (Schachter, 1951).
Thus, attitude and value similarity on mission-related issues should be a powerful determinant of crew compatibility. Still, one would hope to find, within a given crew, sufficient attitudinal variability to generate interaction and provide new ideas during problem-solving sessions. Additional research is necessary to identify the appropriate balance between similarities and differences in attitudes opinions, and beliefs. We hypothesize that it is essential for crewmembers to share certain general values and desirable for them to exhibit variability in the ways that these values are expressed. Value homogeneity is likely to decline as crew size increases, but the development of new selection and indoctrination procedures should help to offset this decline.
Second, as noted by Haythorn and his associates, interlocking or complementary abilities should enhance group compatibility (Haythorn et al., 1972). One type is skill complementarily, which exists when one person is skilled in an area in which the other person is unskilled. Another is cognitive complementarily, which exists when people have nonoverlapping knowledge and must learn from or rely upon each other. Complementary abilities should allow each crewmember to contribute to the crew's welfare, sensitize each to the importance of the others' contributions, and in consequence promote solidarity and high morale. However, there is little or no research characterized by systematic efforts to relate complementary and overlapping abilities to compatibility within isolated and confined groups.
Third, people's needs may fit together in such a way as to affect group compatibility. Particularly important for present purposes is Haythorn s version (1968), which has been tested under conditions  of isolation and confinement. This theory involves three patterns of needs. Congruent needs are similar-appearing needs such that the satisfaction of one person's need results in the satisfaction of the other person's need. For example, two people who have needs to affiliate could find mutual satisfaction by affiliating with one another. Complementary needs are different-appearing needs such that the satisfaction of one person's need also satisfies the other person's need. For example, a person with a need to teach might establish a satisfying relationship with a person with a need to learn. Competitive needs are such that the satisfaction of one person's need results in the frustration or aggravation of the other person's need. This might occur, for example, when two people vie for dominance within a group.
Important tests of this theory have been performed by Altman and Haythorn (Altman and Haythorn, 1965, 1967a, 1967b; Altman and Taylor, 1973; Haythorn and Altman, 1967; Haythorn et al., 1966). In these studies, isolated and confined subjects were paired to form dyads varying in homogeneity or heterogeneity along dimensions of dogmatism, dominance, need for achievement, and need for affiliation. Adaptability in isolation was a direct function of need-compatibility. All of the isolated dyads that experienced serious difficulties had been composed in such a way as to be incompatible. None of the nonisolated control dyads composed of individuals with incompatible needs showed comparable levels of difficulty. These results suggest that need compatibility gains importance under conditions of isolation and confinement.
Research to date thus suggests that it would be both useful and desirable to mount a comprehensive effort aimed at a better understanding of need compatibility. Such a program should attempt to identify relevant needs, show how they fit together, and spell out the consequences of compatibility and incompatibility. Ultimately, screening procedures may be devised for weeding out crew candidates whose needs are too likely to conflict, or ways may be found for keeping competitive needs under control. Researchers involved in such a program should remain sensitive to the possibility that incompatibility may not be a problem if conditions conspire to prevent crewmembers from detecting their differences, and that incompatibilities that disrupt one group may not affect another.
 Group Homeostasis
According to Russian researchers, people who can synchronize their activities in very minimal kinds of social situations tend to form highly compatible and effective groups (Leonov and Lebedev,1975). This research involves sitting groups of three or more candidates in front of a "homeostat," which consists of individual panels mounted with dials, pointers, and other controls. Each candidate is instructed to adjust his or her controls according to specifications; unknown to each candidate is the fact that as they proceed with the task they alter the settings of their colleagues' controls. The difficulty level of the task can be varied to study such variables as differences in group functioning under various levels of frustration. The ability to achieve the correct setting of one's own controls without preventing others from achieving their correct settings defines group homeostasis. As group members learn to coordinate their activities to operate the homeostat, their psychophysiological responses (heartbeat rate, respiration, certain brain waves, and so forth) also become synchronized. The groups that achieve homeostasis are seen as particularly suitable for isolated and confined settings and as highly capable of dealing with threatening or demanding conditions. The research on group homeostasis is intriguing, and needs to be replicated and extended m the West.
Crew Size and Social Compatibility
Increasing crew size increases the number of possible dyadic relationships within the crew according to the formula (n2 - n)/2, where n is the number of people in the crew. Thus whereas a three-person crew could generate 3 dyadic relationships, a six-person crew could generate 15 dyadic relationships and a twelve-person crew could generate 66 dyadic relationships. Increasing crew size increases the number of possible social relationships and, among other things, options for social stimulation, options for developing friendships, and options for exercising varied role behaviors.
The evidence from studies of isolated and confined groups is a bit sketchy, and is complicated by the problem that relatively large groups may be stationed at a relatively comfortable main base, whereas relatively small groups may be located in primitive quarters that offer few of the main base's amenities. However, S. Smith's (1969) review indicates fewer emotional and interpersonal problems in relatively large isolated and confined groups. In one study, Doll  and Gunderson (1971) found that Antarctic parties varying in size from 8 to 10 reported less in the way of compatibility and accomplishment than those varying in size from 20 to 30. They also report that military personnel stationed at small bases were more hostile than their counterparts at more heavily populated bases. Although cross-study comparisons are difficult, it is interesting to note that Georgia fallout-shelter studies (Hammes, Ahearn, and Keith, 1965; Hammes and Osborne, 1965; Hammes and Watson, 1965), which imposed very Spartan conditions on unselected but unusually large groups, had very low defection rates. S. Smith and Haythorn (1972) found triads more harmonious than dyads in a simulation study.
Because increasing crew size should increase the level of social stimulation, the number of friendship options available, and the opportunity to exercise role-related behaviors, it is expected that, socially, membership in a large crew should be more easily endured than membership in a small crew. However, the relationship between group size and interpersonal compatibility is not well understood, and (because few studies have involved varying group size while holding other variables constant) we lack the necessary bearings for making confident predictions. A basic research question is in identifying the functional form of the relationship between crew size (particularly over the range from about 2 to 30 crewmembers) and social compatibility.
Much of the literature on selection is based on the assumption that there are certain qualities that will affect a person's overall suitability for membership in a spacecrew. To some extent, this assumption is reasonable; for example, it is difficult to imagine conditions under which an incompetent, emotionally unstable astronaut would make a positive contribution to crew performance and morale. However, to some extent, people's strengths and weaknesses are to be found in the eyes of the beholder. For example, the quality of the contributions of female or minority astronauts might depend in part on the prevalence of racist or sexist attitudes among the rest of the crew. Furthermore, as we noted in our discussion of research on attitudinal similarities, complementary needs, and group homeostasis, compatibility often depends on the way that different people s qualities fit together. Consequently, it is necessary not only to consider each crewmember's personal qualities, but to relate these to the personal qualities of the rest of the crew.
 The end result is that future selection research may require a shift of focus from characteristics of isolated individuals to combined characteristics of different members of the crew. Associated with this shift are many complex issues. First, there are theoretical and empirical issues regarding the interplay of different people's characteristics. Second, there are technical issues regarding the assessment of these characteristics. Third, there are extremely knotty questions regarding appropriate statistical-analysis techniques.