In this . . . [paper] we will consider differences in group composition and
structure and the effects of these differences on the group and its members.
Group composition refers to the properties represented by the aggregate of persons who are the members of a given group at a given time. We can consider group composition with respect to any of the many variables in terms of which we can characterize individuals: abilities, attitudes, personality characteristics, or personal attributes such as age, sex, educational background, and others.
Group structure refers to the relatively stable patterns of relationships that exist among members of groups. As the term is used here, all groups can be de scribed in terms of many aspects of their structure, even groups of relatively short duration. Group structure, like "personality," is not something that different groups have or do not have, or that different groups have more or less of.
Rather, groups differ in the kind or form of structure they have, just as individuals may differ in the kind or form of personality structure by which they can be characterized.
------------------
From Joseph E. McGrath, "Group Composition and Structure," Social Psychology: A Brief Introduction ( New York: Holt, Rinehart and Winston, Inc., (:) 1964), 72-86. Reproduced with permission of the author and publisher.
------------------
If we define groups as sets of interlocked roles, we cannot merely list a set of roles that form the "parts" for all groups. Rather, each group or type of group is likely to have a different set of particular roles. We can specify much of the role structure of a family in terms of the position and roles of father-husband, mother-wife, son-brother, and daughter-sister. These particular role relationships would be meaningless for describing the role structure of a work group or of an athletic team. It is possible, however, to describe at least three major dimensions in terms of which any set of roles may be differentiated from one another and related to one another. For each dimension, if we focus on the individual role as it is related to other roles, we are dealing with basic properties of roles; while if we focus on the total pattern of relationships among roles, we are dealing with different forms of group structure. The three dimensions of role relationships are:
1. Role differences in terms of task activities or responsibilities: the work structure
2. Role differences in terms of authority or influence: the power structure
3. Role differences in terms of communication channels: the communication structure One further dimension of group structure has to do with relations among members, rather than among roles:
4. The pattern of affect relations among members: the friendship structure Before we consider these patterns of group structure, we will discuss group com position, that is, the sets of abilities, attitudes, and other properties members bring with them to the group.
COMPOSITION OF THE GROUP
What a group is and what it does depends in part on characteristics of its members, independent of the role structure of the group. For example, a group of men with high ability can certainly do a task better than a group with lower ability. It does not always follow, however, that the group with higher ability will do the task better. If the task does not demand much skill, the additional ability of the better group will be superfluous and might even be detrimental if the task were so easy and unchallenging that the group had little motivation to perform it. Thus, a group does not necessarily need the highest possible levels of ability in all members. What it needs is a distribution of levels and types of abilities which best fits the requirements of its task.
Furthermore, the task, power, or communication structure of a group may operate to prevent the effective utilization of abilities of members. Torrance (1954) found that the effectiveness of group actions in a new and ambiguous situation (a physical-survival situation) was influenced adversely for groups whose prior power structure (the formal rank structure of air crews) remained unaltered. The formal structure of the crew, designed to perform a different task, tended to prevent the utilization of skills and resources of members who were low in that formal structure but had high levels of skills relevant to the new situation.
Fiedler and Meuwese (1963) found evidence that the utilization of member abilities depends on the type of formal leader and his style of leadership. Several studies of groups with restricted communication structures, which will be discussed later in this section, indicate that the degree to which the group gets the benefit of the skills of a particular member depends on the position of that member (how central he is) in the communication structure. Thus, while it is probably true that when all other things are equal groups of more able men perform tasks better than groups of less able men, there are many factors besides the task skills of members which affect group performance.
A group's effectiveness depends also on the personalities of its members. But just what kinds of personalities make for good groups? "Folklore" on this question supports either one of two opposite premises: (1) that persons with similar personalities make good groups, as in "Birds of a feather flock together"; or (2) that persons with different personalities make good groups, as in "Opposites at tract." Recently, Schutz (1958) tested the hypothesis that persons with different but complementary patterns on certain personality characteristics are most compatible, hence make the most effective groups. He combined pairs of people in terms of the strength of their needs to give and to receive inclusion (belonging-ness), control, and affection. The compatible pairs were those in which member A had a high need to give inclusion, control, and/or affection to others, while member B had a high need to receive inclusion, control, and/or affection. He compared such compatible groups with pairs in which the two members had similar patterns of these needs (for example, both members had high needs to give affection or inclusion or control), and with other pairs whose members had different but non-complementary patterns of needs (e.g., member A had a high need to give control while member B had a high need to receive affection). He found the compatible pairs to be more effective on relatively complex group problem-solving tasks, although the different types of groups were about equal on easy tasks.
Other research on homogeneity of personality does not provide a very clear picture. Cohen (1956) studied the performance of two-person groups whose members were either similar or dissimilar in their preferred defense mechanism (projection, aggression, repression, and so on). He found that groups in which both members tended to use projection as a preferred defense mechanism per formed less effectively, but he did not find any consistent differences for groups that were either similar or dissimilar in the use of other defense mechanisms.
Haythorn et al. (1956) and Altman and McGinnies (1960) found that there was less intra-group hostility in groups whose members were homogeneous in attitudes toward authority. On the other hand, Hoffman (1959) has shown that task performance was better for groups whose members were heterogeneous in over all personality patterns.
The conflicting evidence makes it probable that neither homogeneity nor het erogeneity of members per se is desirable. Rather, it is likely that homogeneity on some characteristics and heterogeneity on others make for effective groups.
Homogeneity on certain personality characteristics such as dominance or the use of projection will almost certainly be disruptive in a group; homogeneity on other characteristics, such as sociability, may lead to greater member satisfactions and smoother cooperation on the task. Homogeneity or heterogeneity on other personality variables may make no difference at all.
Regarding member abilities, performance is likely to be better if a group has a heterogeneity of skills, or at least a distribution of skills that is of sufficient range to deal with all required tasks, than if it has a narrower range of skills common to all members. On the other hand, groups whose members have similar backgrounds and hence similar attitudes and values are likely to have better communication and better affect relations. For example, Fiedler, Meuwese, and Oonk (1961) found less friction and better performance in groups homogeneous in religious background. Here again, however, there are probably many specific back ground factors and attitudes for which the degree of member homogeneity or heterogeneity does not matter.
Another crucial aspect of group composition has to do with the level and diversity of the motivations of group members. Groups differ from one another in terms of how strongly motivated their members are, collectively, with respect to attaining the group's goals. Furthermore, it is unlikely that all members of any given group will be equally strongly motivated toward achievement of the group's goals. Both levels of over-all member motivation and differences in intensity of motivation among group members will affect how well the group can accomplish its objectives. As with levels of ability, adequate levels of motivation of group members are a necessary but not a sufficient condition for group success. A poorly motivated group will almost certainly not be highly successful on relatively difficult tasks. A highly motivated group, however, may or may not be successful, depending on the level and distribution of member abilities, the pat tern of member personality characteristics, and whether the group's task, power, communication, and friendship structures aid or hinder task performance.
Perhaps a more crucial aspect of group-member motivation has to do with whether or not group members are motivated toward group goals rather than individual goals. In any group situation, the individual is to some extent engaged in the business of satisfying his own motivations. Sometimes the individual motivations are such that all members want the same collective outcomes, that is, they share the same goals for the group. For example, business partners committed to a 50-50 split of costs and profits presumably share the goal of group profit.
What is in the best (economic) interests of the group is also in their own individual best interests. In other situations, the individual motivations of group members may not converge on the same goals and may in fact be incompatible.
In large measure, the motivations of individual group members depend on the nature of the rewards members hope to attain through group membership and group activity, the manner in which rewards are related to task performance, and the way in which rewards for group success are distributed among group members. We can distinguish at least three patterns in which individual and group performance and individual and group rewards can be related.
First, group members can receive rewards on the basis of their individual performances, without regard to the task performance of other members. In such a situation, group members are independent, rather than interdependent, at least with regard to their motivation-task performance-reward patterns.
Second, group members can be striving for rewards under conditions where the more one member gains the less others can gain. Here, the members are interdependent but are in competition for rewards. This condition has been termed contriant interdependence (Thomas, 1957). Third, group members can be striving for rewards under conditions where the more one member gains the more all others gain as well. Here, members are interdependent but can gain rewards best if they cooperate with one another.
This condition has been called facilitative interdependence (Thomas, 1957). Deutsch (1949), Mintz (1951), Thomas (1957), and others have shown that group task performance, as well as interpersonal relations among group members, are better under conditions of facilitative interdependence, where the motivational conditions tend to foster cooperative behavior, than under conditions of contriant interdependence, where the motivational conditions tend to foster competition among group members.
A fourth pattern can also occur. Group members can be striving for rewards that stem from interpersonal relationships as such and that are not contingent on the group's task success. Under these conditions, task performance of group members is unrelated to their rewards and the rewards of others, but group members are highly interdependent for the rewards that stem from pleasant interpersonal relations.
GROUP STRUCTURE: PATTERNS OF ROLE RELATIONSHIPS WORK STRUCTURE AND THE TASK
The most obvious way to describe a set of roles is in terms of the part each role plays in the group's activity. When we consider the group's division of labor, or division of task responsibilities, we may focus on the nature of the separate roles (such as task specialization) or on the pattern by which the set of task roles are integrated (for example, the bases for sub-groupings of roles; the mechanisms for coordination between roles or sets of roles). This differentiation and integration of roles with respect to task activities is one major basis for de scribing the structure of a group.
The subject of division of labor and its consequences has been of concern in our society for many years. Much has been said about the advantages and disadvantages of task specialization. The main argument in favor of specialization has usually been based on the assumption that it makes for efficiency of task performance. If each man performs a limited set of activities for which he is highly trained, a set of specialists working together can accomplish a total job which involves so many diverse skills that no one man could have them all. This is one basic premise on which our modern, highly specialized, technologically sophisticated civilization has been built.
The main argument against task specialization has been in terms of its effects on individuals as human beings. The basic premise of this argument is that the satisfaction obtained from task accomplishment is an important factor in human adjustment and that such satisfaction is greatly reduced when the individual performs only a single, specific task operation and thus does not experience a feeling of accomplishment. This side effect of our industrial revolution has been considered a major cause of problems of individual adjustment in our modern world.
In spite of all the concern and discussion of the effects of task specialization, we still do not have much scientific knowledge (as opposed to opinion) about those effects. There is little evidence that task specialization as such affects individual morale or adjustment, although there is convincing evidence that being relatively isolated in the communication pattern of a work group does have such negative effects. We also have evidence that participation in decisions affecting one's job yields more favorable attitudes and better performance (see Coch and French, 1948; Mulder, 1959). But communication isolation and lack of opportunity to make job decisions are not necessary parts of task specialization.
There is little evidence about the task performance effects of specialization.
It is doubtless true that large and complex tasks can be accomplished better if they are divided among several people and that a task will not be done at all if no one who is available has the ability to do it. We do not know, however, how best to divide up any particular task into specialized jobs. Such a division must reckon with at least three facets of the problem. First, the division of jobs must relate closely to the distribution of skills among the people who do the jobs. For example, if all elementary school teachers were trained as specialists in only one subject, we would have to use a team-teaching system rather than a homeroom.
teacher system. Secondly, the division of jobs must be made so that the jobs tie in well together; we must be sure that completion of task A is all that is required before task B can be carried out. Third, we must divide tasks so that they can be efficiently coordinated in time and space. The efficiency of task specialization is lost, for example, if an automobile manufacturer ends up with 30,000 motors and only 10,000 bodies.
A great deal of "applied" research has been done on task differentiation and coordination in industrial and military settings. Early efforts on these problems were called time and motion studies and grew out of F. W. Taylor's (1911) crusade for scientific management. Later efforts have been carried out under the headings of operations research and recently systems research. Unfortunately, however, most of this research is particularistic, that is, geared to determine the best division and coordination of activity for a particular complex task (such as the launching of a particular missile). Hence, it does not add much to our general knowledge of the effects of different work-structure patterns.
Research on the more general problem of work structure is fairly scarce.
Lanzetta and Roby (1956) have compared two general types of work structure: a horizontal division of labor, in which each man handles the several different parts of a group task but does so for only one portion of the group's area of responsibility; and a vertical division of labor, in which each man handles only one part of the group's task but does so for the group's entire area of responsibility.
They found the former to be slightly more efficient, especially for reasonably light work loads, but this difference was reversed for heavier work loads. While differences were not strong, they suggest that the best division of labor even for a particular task depends partly on work load.
The Lanzetta and Roby study is based on the condition that every man has the necessary skills to perform all tasks involved. This is often not the case, especially for groups performing highly specialized tasks requiring extensive training.
The nature of the group's task interacts with the group's work structure. The kind of task the group is performing affects the kinds of activity and coordination required, and this in turn affects the choice of the most appropriate division of labor. A wide variety of tasks have been studied in small group research, but relatively little systematic investigation has been done of the properties of the tasks themselves.
Carter et al. (1950) studied group performance on a number of tasks and concluded that from the point of view of effective leadership there are at least two distinct kinds of group tasks: intellectual tasks and manual, or "do with the hands," tasks. Many further distinctions can be made within the global category of intellectual tasks. For example, many researchers have distinguished between problem-solving and creative tasks (for example, Fiedler, 1962). Problem-solving tasks require the group to integrate various pieces of information into "the" solution, or the best solution. Creative tasks require the group to generate a number of original ideas about a topic. Some tasks, of course, require both the generation of new ideas and their synthesis into a coherent solution. We can distinguish also another type of intellectual task, negotiation tasks, which require members to integrate divergent attitudes or values rather than generate or integrate information.
It would appear that such task differences, which require different patterns of task activity and interaction among members, may also require different pat terns of role differentiation and coordination. Recently, Shaw (1963) proposed a set of ten dimensions for describing and differentiating a wide range of group tasks. These include such properties as single versus multiple paths to goal; clarity of the relationship between path and goal; specificity versus generality of the solution. These dimensions offer a substantial refinement of the gross categories of intellectual versus manual tasks and problem-solving versus creative tasks.
Such a schema, when fully elaborated, may make it possible to compare results obtained on various group tasks and to explore relationships between type of task and the group's division of labor.
Analytically, we can specify the connection between task roles of any two people as one of several types: (1) A and B can do the same task, while working independently (for example, Lanzetta and Roby's horizontal structure, or tellers at adjacent windows in a bank) ; (2) A and B can work on separate, sequential steps of the same task, so that A's output provides B's input (Lanzetta and Roby's vertical structure, or workers at adjacent stations on an assembly line) ; (3) A and B can work on complementary aspects of the same task simultaneously (a blocker and a ball carrier on a football team, or the members of a land-survey team) ; (4) B may direct or supervise the activities of A. These types of task connections have different implications for the degree to which A and B are dependent on each other for their activities. In type 1, A and B are unrelated; their task performances are independent. In type 2, B is dependent on A, because an error in performance by A disrupts B's task performance.
In type 3, A and B are interdependent; the performance of each affects the requirements placed on the other. In type 4, A is dependent on B for direction of his work and in terms of its evaluation, but B is dependent on A in another sense, because A must do part of the job for which B is responsible. Such task relations between group members also affect the influence relations and the communication relations that are likely to develop between them.
POWER STRUCTURE
The set of roles within a group may be described in terms of the relative authority or power of each role. In a formal group, each role can be designated as to its position in an authority or power hierarchy, and as to its super-ordinate or subordinate relation to other roles. In informal groups, differentiations in power are likely to be equally prominent though less explicitly defined.
There are many forms or bases of power. French and Raven (1959) define power as the potential influence of person 0 on person P in a particular area.
They distinguish five bases of power: (1) reward power; (2) coercive power; (3) legitimate power, such as the "right" of a formally chosen leader to direct the actions of members; (4) referent power, or power based on P's liking for or identification with 0; and (5) power based on expert knowledge. Even in informal and transient groups, members tend to differ in their relative power (potential influence). Degree of power tends to correlate with the holding of a position of formal leadership, but is by no means entirely a matter of such formal authority. The degree of power 0 has over P is also related to his friendship or affect relations with P. A group I or one of its members) can have power to influence a particular member P to the extent that P is attracted to that group, that is, to the extent that the group can provide or withhold rewards which are valuable to P. Sometimes the power of a role derives from the status of the person who happens to occupy that role in the group. By "status" we mean the prestige the individual possesses as a person, independent of his position in a particular group. Such status can arise from the person's general social status (wealth or position of prestige) or from specific features of his background such as age, sex, race, education. (The latter are the institutionalized positions he occupies in the general society.) Strodtbeck and his associates (1957) have shown that both males and persons of higher socioeconomic status have more influence in jury deliberations than females or persons of lower socioeconomic status. . . . Hurwitz and his associates 11960) found that persons representing higher status professions (such as medicine) in an interdisciplinary conference on mental health had more influence than those representing lower-status groups (such as nurses, teachers, clergymen), independent of their actual contributions to the conference deliberations. . . . In children's groups, influence may be related to physical size, athletic skill, or fighting ability (Lippett et al., 1952). Thus, even though a group explicitly assumes that all its members are equal in power (and both the jury and the interdisciplinary conference are predicated on that assumption), members may actually wield different degrees of power based on their general or specific statuses in the broader community within which the group exists.
FIGURE 1. Diagram of some communication patterns used in studies of groups
with experimentally restricted communication networks. Boxes labeled A, B,
C, D, and E represent positions; lines represent possible communication channels
between persons in designated positions.
------- The circle, a decentralized communication network in which al members are equally central; The wheel a centralized communication network in which position C is central and all others are peripheral; The chain, a moderately centralized communication network in which position C is central, positions B and D are intermediate, and positions A and E are peripheral. -----
COMMUNICATION STRUCTURE
A third basis for describing the structure of a group is in terms of the pat tern of communication linkages among its members. A group's communication structure may be considered as the set of possible or permissible communication links or as the pattern of communication channels actually utilized during group activity. Here again, it is possible to focus on the communication properties of a particular role (for example, how central it is in the group's communication net work; how heavy the load of messages which arise at or are sent to that position) ; or on the over-all pattern of the group's communication net (how much it is centralized around one person; how many total communication links there are).
A number of experiments have shown that the communication structure of a group has important consequences for the group's task performance and for the members' morale or satisfaction. Most of these studies were conducted with small groups that had experimentally imposed restrictions on the communication links among members. (For further descriptions of studies of restricted communication nets, see Leavitt, 1951; Bavelas, 1950; Guetzkow and Simon, 1955; Shaw, 1954; and a review of Glanzer and Glaser, 1961). The logic of these communication-net studies is as follows. If we restrict the potential communication channels within a group (if we permit each member to communicate only with certain other members), the actual flow of communications will necessarily be restricted.
How will differences in flow of communication affect the group's task performance and the reactions of the members to the situation? Most communication-net studies have been done with four- or five-man groups. Only certain pairs of members were permitted to communicate with each other, some of the potential channels being eliminated by experimental arrangements. The method of communication was restricted either to written notes or to non-face-to-face verbal communication over an intercom system. The open channels were usually two-way, permitting feedback. Sometimes limits were placed on the rate and the content of messages. The problems used always required an interchange of information; they sometimes required some problem-solving activities; they usually required feedback of information or solutions to other group members. The communication nets varied primarily in degree of centralization, that is, in how much they forced communication to be routed to one central per son. Some typical communication nets studied in these experiments are shown in Figure 1.
These studies show two major results:
a. For simple problems, centralized nets (such as the wheel and to a lesser degree the chain in Figure 1) produce faster and more accurate problem solving than less centralized nets (such as the circle in Figure 1). The leader (position C in the centralized nets in Figure 1) is happy in his central position but other group members are less satisfied. Individual satisfaction is directly related to the individual's centrality in the group (how near he is to the central person in terms of communication links).
b. For more complex problems, performance differences among nets tend to diminish. The effectiveness of highly centralized nets depends largely on the leader's ability and his utilization of members' skills. The relationship between member satisfaction and member centrality, however, tends to persist for the more complex problems.
Laboratory studies by Kelley (1951) and Thibaut (1950) have shown also that the frequency, direction, and content of actual communications in groups which had no externally imposed communication restrictions are related to the relative status of members and the likelihood that their status could change. In general, high-status members talked more than low-status members, and both high- and low-status members directed more communications to those with high status. When low-status members had no chance for upward status movement, they directed more communications to high-status persons and engaged in more task-irrelevant communications than did low-status members who had some possibility of an upward change in status. Presumably, upward communication in this instance served as a substitute for upward status movement . . . .
FRIENDSHIP OR AFFECT STRUCTURE
There is at least one other general basis for description of group structure, namely, the friendship or affect pattern of a group. The affect structure has to do with patterns of relationships among group members as persons, rather than the relationships among roles. The member structure patterns, however, are highly interdependent with various dimensions of the group's role structure.
For one thing, there are fairly strong role expectations about what kind of affect relations ought to exist between persons in certain role relationships. For example, mothers "should" love their children; bosses "should" be friendly but retain a certain aloofness or businesslike attitude in relation to subordinates; soldiers "should" hate enemy soldiers; in some situations the worker is expected to be hostile toward the boss. For the most part, such expected affect relations be tween occupants of different positions actually become a part of the role relation ship between them.
The communication pattern also influences (and is influenced by) the affect relations that develop between occupants of different positions. We seldom develop either positive or negative attitudes toward persons with whom we do not interact (except, of course, toward persons in positions for which our own position gives us an expected affect relation, for example, toward "an enemy"). We may, of course, develop strong attitudes toward public figures, such as political leaders. When we do, though, we often feel that we know them, even though we are aware that they do not have reciprocal attitudes.
In fact, Festinger and his associates (1950) have shown that mere proximity, which provides the opportunity to interact, is the most important factor in determining what friendship patterns arise between families within a housing development. That is, we develop friendships with those with whom we have occasion to interact. Conversely, Newcomb (1961) has shown that once a friendship pattern has been established, we tend to seek opportunities to interact with those we like and avoid interaction with those we dislike.
Affect relations show the same kind of interdependence with power relations.
In all the studies of effects of power or status which we have cited (Strodtbeck et al., 1957; Hurwitz el al., 1960; Lippett et al., 1952; Kelley, 1951; Thibaut, 1950), persons with high status not only received more deference but also were more popular. We tend to be attracted to people who occupy positions of high power. Of course, this tendency depends on the form or basis of that power; it probably does not hold for power based on coercion, for example. Conversely, we tend to influence and be influenced by those we like. Back (1951) has shown that two-person groups whose members like one another exhibit more attempted influence and more often succeed in these attempts, than groups whose members do not like one another. The tendency to be influenced by those we like may account for the rule in many work organizations which forbids nepotism (employment of two or more persons who are related by blood or marriage). Role relations on the job are expected to be affectively neutral; while role relations among kinfolk are expected to include positive affect. We therefore expect (and indeed often find) that the juxtaposition of these two kinds of role relations within the same role system may have disruptive effects.
We can consider the over-all pattern of affect relations in a group, as well as the affect ties of a particular member. Groups vary in terms of the extent to which the pattern of affect relationships among members tends to partition the group into separate subgroups or cliques, and in terms of whether there is active hostility or just lack of positive relations between members of one clique and an other. Interaction cliques tend to become affect cliques, and vice versa. Such divisions within a group may or may not correspond with task-based sub-groupings.
The presence of antagonistic cliques in a group detracts from the efficiency of group-task performance, at least when the task requires close coordination be tween persons in the different cliques.
Similarly, we can consider the effects of the over-all level of positive or negative affect relations in a group. Much has been written about the importance of good interpersonal relations in work groups, most of it on the assumption that groups whose members like one another perform their tasks better. Actually, there is little research evidence on the connection between interpersonal relations and task effectiveness, and what evidence does exist is not at all clear-cut. It does seem clear that negative affect relations within a group disrupt group performance; it does not necessarily follow, however, that positive affect relations facilitate group performance.
Too much friendship within a work group can result in the group's spending most of its time on social activities rather than on the task. Even when a group with highly positive interpersonal relations does concentrate on task activities, there is no guarantee that it will do so effectively. When a group does have positive interpersonal relations among its members, though, there is at least some assurance that the group's task performance will not be hindered by the disruptive effects of negative interpersonal relations, and the group's task performance will depend more closely on the skills and abilities of its members and on the extent to which its task and communication structures facilitate (or hinder) task performance.
Although positive interpersonal relations among members do not necessarily lead to group task success, task success does seem to lead to an increase in positive affect relations among group members. Thus, we tend to like those who have helped us gain rewards.
This discussion is related to a structure problem that can have serious consequences for members of formal work organizations; namely, the problem of the "fit" between the affect structure of the group and its formal work, authority, and communication structures. Formal organizations usually spell out task responsibilities and power relations among members. Usually they also specify preferred or required communication relations among members (and, by implication, those possible communication linkages that are not supposed to be used), at least regarding the communication of task-relevant and organization-relevant in formation. These structure patterns are often embodied in a formal organization chart. Within such formal organizations, informal friendship groups often develop. These are likely to have informal communication structures, utilized primarily for nontask communications. The informal communication pattern may or may not conflict with the formal pattern of communication as prescribed in the organization chart.
An individual member is likely to have a position and a set of roles in both the formal and informal structures. In certain situations role expectations for his behavior may be divergent and mutually exclusive. Thus, the individual may be placed in a role-conflict situation. Resolution of such role conflicts depends on the relative attraction of the member to the formal and informal groups, which in turn depends upon how well these groups satisfy his various needs and aspirations. The formal organization may largely control the satisfaction of certain of the individual's goals, such as economic gain or occupational status. The informal group is likely to be the main instrument for satisfaction of other goals, such as recognition, prestige, friendship. Resolution of such role conflicts often re quires the individual to choose between a number of important and separate goals or compromise ...
REFERENCES:
ALTMAN, I., & MCGINNIES, E. Interpersonal perception and communication in discussion groups of varied attitudinal composition. J. abnorm. soc. Psychol., 1960, 60, 390-395.
BACK, K. W. Influence through social communication. J. abnorm. soc. Psychol., 1951, 46, 9-23. BAVELAS, A. Communication patterns in task groups. J. accoust. Soc. Amer., 1950, 22, 725-730.
CARTER, L. F., et al. A further investigation of the criteria of leadership. J. abnorm. soc. Psychol., 1950, 45, 350-358.
Cocu, L., & FRENCH, J. R. P., Jr. Overcoming resistance to change. Hum. Relat., 1948, 1, 512-532.
COHEN, A. R. Experimental effects of ego defense preference on interpersonal relations. J. abnorm. soc. Psychol., 1956, 52, 19-27.
DEUTSCH, M. A theory of cooperation and competition. Hum. Relat., 1949, 2, 129-152.
FESTINCER, L.. et al. Social pressures in informal groups: A study of human factors in housing. New York: Harper & Row, Publishers, 1950.
FIEDLER, F. E. Leader attitudes, group climate, and group creativity. J. abnorm. soc. Psychol., 1962, 65, 308-318.
FIEDLER, F. E., & Meuwese, W. A. T. Leader's contribution to task performance in cohesive and uncohesive groups. J. abnorm. soc. Psychol., 1963, 66, 83-87.
FIEDLER, F. E., MEUWESE, W. A. T., & OONK, S. An exploratory study of group creativity in laboratory tasks. Acta Psychol., Amsterdam, 1961, 18, 100-119.
FRENCH, J. R. P., JR., & RAVEN, B. The bases of social power. In D. Cartwright (Ed.), Studies in social power. Ann Arbor, Mich.: University of Michigan, Institute for Social Research, 1959.
GLANZER, M., & GLASER, R. Tediniques in the study of group structure: II. Empirical studies of effects of structure in small groups. Psych. Bull. 58, 1961, 1-27.
GUETZKOW, H., & SIMON, H. A. The impact of certain communication nets upon organization and performance in task-oriented groups. Mgmt. Science, 1955, 1, 233-250.
HAYTHORN, W., et al. The behavior of authoritarian and equalitarian personalities in groups. Hum. Relat., 1956, 9, 57-74.
HOFFMAN, L R. Homogeneity of member personality and its effect on group problem solving J. abnorm. soc. Psychol., 1959, 58, 27-32.
HURWITZ, J. L., et al. Some effects of power on the relations among group members. In D. Cartwright, & A. Zander (Eds.), Group dynamics-research and theory, New York: Harper & Row, Publishers, 1960.
KELLEY, H. H. Communication in experimentally created hierarchies. Hum. Relat., 1951, 4, 39-56.
LANZETTA, J ., & ROBY, T. B. Effects of work group structure and certain task variables on group performance. J. Abnorm. Soc. Psychol., 1956, 53, 307-314.
LEAVITT, H. J. Some effects of certain communication patterns on group performance. J. Abnorm. Soc. Psychol., 1951, 46, 38-50.
LIPPITT, R., et al. The dynamics of power: A field study of social influence in groups of children. Hum. Relat., 1952, 5, 37-64.
MINTS, A. Non-adaptive group behavior. I. Abnorm. Soc. Psychol., 1951, 46, 150-159.
MULDER, M. Power and satisfaction in task-oriented groups. Acta Psychol., Amsterdam, 1959, 16, 178-225.
Newcoma, T. M. The Acquaintance Process. New York: Holt, Rinehart and Winston, Inc., 1961.
SCHUTZ, W. C. FIRO: A Three-Dimensional Theory of Interpersonal Behavior. New York: Holt, Rinehart and Winston, Inc., 1958.
SHAW, M. E. Some effects of problem complexity upon problem solution efficiency in different communication nets. J. Exp. Psychol., 1954, 48, 211-217.
SHAW, M. E. Scaling group Tasks: A Method for Dimensional Analysis. Tech. Rep. No. 1, ONR Contract NR 170-266, Nonr 580(11), July 1963.
STRODTBECK, F . L., et al. Social status in jury deliberations. Amer. Socio!. Rev., 1957, 22, 713-719.
TAYLOR, F . W. The Principles of Scientific Management. New York: Harper and Row, Publishers, 1911.
THIBAUT, J. W. An experimental study of the cohesiveness of underprivileged groups. Hum. Relat., 1950, 3, 251-278.
THOMAS, E. J. Effects of facilitative role interdependence on group functioning. Hum. Relat., 1957, 10, 347-366.
TORRANCE, E. P. Some consequences of power differences on decision making in permanent and temporary three-man groups. Res. Stud., Wash. State Coll., 1954, 22, 130-140.
Also in Part 5:
- Communication Theory: Social Context
- How People Interact in Conferences
- Group Composition and Structure
- Achieving Change in People: Some Applications of Group Dynamics Theory
- Interpersonal Networks: Communicating Within the Group
- Traditions of Research on the Diffusion of Innovation