Experiments in Organizational Embeddedness.

Although a great deal has been learned about the effects of communication structures upon group processes (Leavitt, 1951; Shaw, 1964; and Cohen, 1964a), the research so far has been limited to small groups (six members or fewer) working in isolated networks. Actually, however, small groups typically perform as subgroups which are parts of larger networks or organizations. The effectiveness of a group may be quite different when it is in isolation from its effectiveness when embedded in a larger organization. Furthermore, a group may organize and function differently when it is part of a larger organization, depending on both the kind of overall organization and the kinds of subgroups that interact with one another. Differences may also be expected in satisfactions, leadership, and other relevant variables. As an organization becomes larger and more complex, opportunities for interaction and communication become more important. As groups and subgroups become larger and more numerous, and as the relations within and among them become more complex, there is an increased need to understand the process of communication and the factors in group structure that influence communication. A general question that arises is: What kinds of communication networks will be appropriate for meeting the demands of large and complex organizations for increased speed, accuracy, and communication? Early work in small and isolated communication networks suggested that it may not be possible to maximize both morale and efficient communication at the same time, and that these maxima might have to; be sacrificed to achieve a balanced operation (Bavelas and Barrett, 1951). This problem can be expected to be magnified with in-