3
Distributed Work and Group Processes

Distributed work has implications for many of the activities that organizations, groups, and individuals need to accomplish to be effective; its wider adoption will present both organizational and personal challenges. Technology is likely to solve only a subset of these challenges, but nonetheless should be developed and applied when appropriate.

The impact of distributed work on group processes and performance is worthy of examination because group performance encompasses individual performance, because so much of the work in organizations is done by groups, and because organizations themselves have functions that parallel those of groups (Katz and Kahn, 1979). Reviewing the voluminous research literature on groups and organizations is beyond the scope of this report. Rather, the intent in this chapter is to point out group processes most likely to be affected by the distribution of work across space and time and to discuss basic functions that must be supported for distributed work to be conducted effectively.

Such functions go beyond tasks with a single focus such as writing a report, selling a product, or handling a customer's transaction. Individuals in organizations are typically members of multiple groups, each of which is involved in multiple projects. As members of groups, people perform their assigned work tasks, as well as many ancillary activities that keep them coordinated, motivated, well trained, and



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 35
Research Recommendations: To Facilitate Distributed Work 3 Distributed Work and Group Processes Distributed work has implications for many of the activities that organizations, groups, and individuals need to accomplish to be effective; its wider adoption will present both organizational and personal challenges. Technology is likely to solve only a subset of these challenges, but nonetheless should be developed and applied when appropriate. The impact of distributed work on group processes and performance is worthy of examination because group performance encompasses individual performance, because so much of the work in organizations is done by groups, and because organizations themselves have functions that parallel those of groups (Katz and Kahn, 1979). Reviewing the voluminous research literature on groups and organizations is beyond the scope of this report. Rather, the intent in this chapter is to point out group processes most likely to be affected by the distribution of work across space and time and to discuss basic functions that must be supported for distributed work to be conducted effectively. Such functions go beyond tasks with a single focus such as writing a report, selling a product, or handling a customer's transaction. Individuals in organizations are typically members of multiple groups, each of which is involved in multiple projects. As members of groups, people perform their assigned work tasks, as well as many ancillary activities that keep them coordinated, motivated, well trained, and

OCR for page 35
Research Recommendations: To Facilitate Distributed Work adapted to their organizational context. Moreover, individuals may well play different roles in different groups: leader in some, member in others; expert in some, novice in others; and so on. DISTRIBUTED WORK IN THE CONTEXT OF BUSINESS ORGANIZATIONS Even though businesses are not identically structured, they share similarities that can be discussed usefully in terms of the effects of distributed work. Every organization has two primary, albeit competing, needs: to divide labor into the various tasks that need to be performed and to coordinate these tasks so that the overall goals of the organization can be met. As management theorist Henry Mintzberg (1979) has written, the ''structure of an organization can be defined simply as the sum total of the ways in which it divides its labor into distinct tasks and then achieves coordination among them." According to Mintzberg, each relatively large business organization consists of three basic parts: (1) an operating core that has as its primary function the fundamental work of producing goods and/or services; (2) an administrative component that provides strategic management, middle or line management, and technical services and leadership; and (3) a support staff that provides ancillary services such as legal counsel and personnel, finance, communications, and building services. Any introduction of technology into an organization must take into account both the nature of the technology and where in the organization the technology is introduced. Introducing advanced robotics on the shop floor will yield a very different relationship between technology and organizational culture than will increasing the availability of laptop computers, cellular telephones, and modems in the ranks of middle management or at the strategic apex of the corporation. The introduction of new technologies to facilitate distributed work is simply a particular case of the broad interaction between organizations and technology. While distributed work in almost all cases means increased independence of place for the worker, how it plays out precisely will vary according to the kind of work, the intention of the worker, and the area within the organization in which it occurs. In some cases, "knowledge" workers can work outside of the centralized office—at home or in a satellite office—so as to achieve greater mental concentration or a more desirable lifestyle. Much of the contemporary discussion regarding telecommuting concerns this aspect of distributed work. In other cases, distributed work may mean that employees will travel

OCR for page 35
Research Recommendations: To Facilitate Distributed Work less to the centralized office—because they now have access to information or services that once would have required them to make trips into the office—and travel more freely and frequently to clients or to other operating plants. For example, advanced technology currently allows more time in the field for General Electric Company service staff. A GE service technician returns home at the end of each day and uses a laptop computer and modem to obtain assignments for the next day, order parts necessary to replenish inventory, and order whatever special equipment may be needed for the next day's calls. While the technician sleeps, the van, still sitting in his or her driveway, is restocked, saving a trip to a parts depot and allowing for more repair time the following day. Such a work arrangement requires fundamental changes in core work concepts and practices: inventory is delivered to technicians rather than having to be picked up at parts depots, and assignments are obtained at home rather than at the office. The technician's data become part of a larger performance management system so that management can track trends in the nature of service calls, anticipate preventive maintenance needs, and so on. Distributed work can enhance productivity and locational independence for a variety of work purposes. The various approaches chosen by individuals and organizations to fulfill their particular purposes are almost certain to gain increased attention and importance over the coming years. Distributed work also raises questions regarding the impacts of various work arrangements on the individual and on the culture and functioning of organizations. In addition to the technological research outlined in this report, extensive sociological and organizational research will be required to ensure that distributed work can be carried out to serve the needs of individuals and organizations effectively. STRATEGIES FOR ACCOMMODATING DISTRIBUTED WORK According to estimates by the U.S. Department of Transportation (1993), more than 2 million people currently telecommute in the United States, and more than 10 times as many work at home under other arrangements, including corporate after-hours work or various forms of self-employment. In addition, large numbers of people travel as a routine part of their jobs: sales people travel to customer sites, maintenance personnel travel to immobile equipment, and executives travel to meetings, for example. Yet currently available technology for distributed work places substantial constraints on the types of work and

OCR for page 35
Research Recommendations: To Facilitate Distributed Work learning that can be done remotely from a central site, on the road, or removed from peers and supervisors. When a work group is co-located and physically embedded within a larger organization, many of the processes necessary for success are tacit or informal. People learn the norms of an organization and many job-relevant behaviors by observing the people around them. Supervisors manage groups by monitoring employee effort, recognizing and averting incipient failure, helping resolve personality conflicts, appraising performance, and providing advice and instruction. Workers may call over to a colleague at the next desk for help. They share budgets, designs, or machinery by simply gathering in a common physical place with the appropriate facilities, such as a white board or overhead projector, or with the item under discussion. Substantial research suggests, however, that while these informal and tacit processes are common, they are not invariably the best ones for ensuring organizational effectiveness. Workers may, for example, learn norms of minimal work effort by being able to ask a convenient local for advice while ignoring an expert farther away. And, as organizations become larger and more specialized, these informal processes often become increasingly impractical. When members of a work group are physically separated from each other or from a supporting organizational structure, tacit and informal processes no longer suffice to support group work and organizational effectiveness. Groups need different strategies for handling group formations, operations, external relations, and reorganization. The approaches that can be taken are of three major types. Organizations can attempt to reduce the interdependence among distributed group members. Studies of telecommuting and home-based employment suggest that reducing interdependence is a common mode of coping with a distributed work force in the face of inadequate communications facilities. For example, companies occasionally reduce the interdependence between supervisor and employee by paying piece rates or by creating independent contractor arrangements. Consultants may tend to recommend only stand-alone jobs such as writing manuals as being suitable for telecommuting. Scientists working remotely from each other may engage in a more formal division of labor and may communicate less among themselves. Firms can attempt to reproduce at a distance the informality of the processes typically supported by physical proximity and, in so doing, may reproduce both the adequacies and the inadequacies of the informal processes. For example, some organizations have deployed electronic mail and computer conferencing to support informal

OCR for page 35
Research Recommendations: To Facilitate Distributed Work communication, discussion, and decision making. In several organizational experiments, long-duration video connections between remote work sites have also been used to support informal communication (Abel et al., 1990; Fish et al., 1992; Bly et al., 1993). Alternatively, organizations can attempt to use managerial and technological mechanisms to formalize at a distance the processes normally supported by proximity-based informal communication. For example, tools currently exist to facilitate the successful development of software programs by geographically distributed teams. Typically, a highly specialized database is used to track the programmers' progress and coordinate the program segments that have been placed within a globally distributed file system. BASIC GROUP PROCESSES TO BE SUPPORTED FOR EFFECTIVE DISTRIBUTED WORK Despite the use of strategies for handling decentralized work, several group and organizational processes are likely to be disrupted by the distribution of work over space and time. Among these are the selection of goals, recruitment of group members, management of distributed groups, and information retrieval. Sketched below for each is the way the process is generally performed in a centralized organization, how distributed work might disrupt the process, technological solutions that might mitigate such disruption, and, if applicable, application and infrastructure research needed to implement these solutions. Selecting Goals The process of selecting initial goals depends on the type of group being formed. For some groups, such as a crisis task force, the group goal is largely predefined. However, in other groups, like senior management teams, quality improvement circles, or teams of collaborating scientists, specific goals often emerge from informal communication. This communication can occur in distributed communication environments through the use of electronic bulletin boards or electronic mail distribution lists, or through the use of real-time audio-or videoconferences. The technical challenge is to support both casual and formal small-group communication to facilitate the formulation, evaluation, and selection of goals in distributed work groups. Today, the traditional written medium is generally too slow for the give and take through which new, shared ideas emerge and goals are

OCR for page 35
Research Recommendations: To Facilitate Distributed Work selected. Computer-facilitated communication is faster but often reduces social context cues, thereby blurring the link between what is being said and the person saying it and thus reducing the likelihood that consensual goals will emerge. Recruiting Group Members In addition to determining an initial project goal, the most important processes in forming a project group are identifying, investigating, and recruiting potential members. Substantial research literature suggests that physical proximity plays a large role in forming teams. As Hagstrom (1965) noted over 25 years ago, "since collaboration often begins through informal contacts, anything that increases the frequency of such contacts increases the likelihood of collaboration. …Spatial propinquity often leads to collaboration since it is likely to lead to informal communication." Of course, more recent experiences with electronic networks suggest new options for stimulating informal communication, although it may be too early to tell how productive they are. During the initiation phase of a collaboration, potential collaborators must establish an intellectual and interpersonal relationship based on shared interests. Selecting people and ideas are both important and often intertwined. Physical proximity creates opportunities for potential collaborators to become acquainted, to assess and develop interpersonal compatibility, to identify common interests, to explore new ideas, and to accomplish rudimentary planning before they become committed to working together. In this way, individuals determine whether potential partners are smart enough to help think through problems, responsible enough to do their share of the work, and humble enough to accept only their share of the credit, as well as whether they are sympathetic enough with one's perspective and compatible enough in work style and personality to make working together pleasant. Sometimes collaborations come about simply because the potential collaborators like each other, before they have any shared ideas to work on. While research collaborations and other types of ad hoc groups can and do form at a distance, the evidence indicates they are much less likely to do so than if the potential members are or have been co-located. Nevertheless, despite the barriers, some amount of scientific research has always been carried out by individuals who share common or complementary interests but happen to be located at different institutions. Notably, scientists have been among the early adopters

OCR for page 35
Research Recommendations: To Facilitate Distributed Work of network, data communication, and computing technologies, the most basic tools and technologies needed for collaboration at a distance. In addition, supervision is generally less of an issue in scientific research teams than in project teams in other areas. Although members of a team are often identified and evaluated informally, database and information retrieval technology could aid in the selection process for distributed workers on at least some dimensions. For example, it is theoretically possible to apply the statistical techniques used in information retrieval to a database of research personnel. Potential candidates for a work group could be represented by the documents they have written or textual descriptions of the projects they have worked on. Queries against such a database could return information on the relevance of each candidate to the query, subject of course to the known inaccuracies of information retrieval systems. In the case of seeking a partner for scientific collaboration, a query of the published literature could return a list of researchers who have, for example, applied factor analysis to the problem of second language learning. Such techniques, however, do not currently suffice for identifying people with the personal attributes and skills necessary for a particular work group. The insufficiency is not intrinsic to an information retrieval approach to personnel recruiting, but occurs because personal attributes are seldom articulated explicitly by people seeking to form groups and, more importantly, because this information is now rarely recorded in organizational archives about projects or people. For example, effective group work may require individuals with a sense of humor, creative or leadership abilities, and a willingness to do their share of the work; knowledge of such attributes as they apply to individuals is most often uncovered through informal conversations with people who have worked with these individuals before. Of course, this phenomenon is also reflected in conventional hiring procedures, in which substantial screening is done based on resumes or comparable documentation but final decisions tend to be based on interviews and consultations with references. The ability to assess personal, work-relevant attributes from databases of public-record materials or, more likely, to interact informally with individuals who could attest to such attributes could promote more effective, efficient formation of distributed work groups (assuming that privacy, liability, and other ancillary social and legal considerations can be accommodated). Social scientists and computer scientists should consider joint research aimed at providing a database structure that could aid in the formation of distributed groups.

OCR for page 35
Research Recommendations: To Facilitate Distributed Work Managing Distributed Groups Typically employees in groups distributed at two or three different locations (usually in different time zones and often in different countries) are charged with completing a development task by working collaboratively with their remote colleagues. The development target could be a new piece of hardware, a new financial product, an advertising campaign, and so on. Each site contains several people, at least some of whom have worked together before. And at least some of those people know some of the people at the other sites. Typically the project manager is co-located with the employees at one of the sites. The first management problem is building one cohesive group out of distributed subgroups. How can technology be used to help people create shared mental models, shared goals, and trust? Technology such as videoconferences may even inadvertently subvert this process by setting up unintended "we/they" dynamics. Project reporting and tracking are more difficult across multiple sites and are exacerbated as project schedules become more complex. The integration of the work products of each site may become an issue if the project involves intellectual or creative efforts that are only marginally suited to detailed specification. Finally, it may be difficult to evaluate the relative contributions of the different groups and individuals on a distributed team. Obtaining Information To function successfully within modern organizations, individuals and groups must have increasing amounts of information. Given adequate data transmission links, conventional information systems can currently provide many workers the core information they need. Thus, reservation clerks, directory assistance operators, and even reference librarians working from home can easily obtain needed information from the same information systems they would use in centralized job settings. Text-based Electronic Sources Many library collections and other archives of books, articles, and text-oriented documents are becoming increasingly available over general-purpose computer networks such as the Internet. As more of this information becomes available, facilities to browse and search it become increasingly important. More than 30 years of research on

OCR for page 35
Research Recommendations: To Facilitate Distributed Work information retrieval, coming primarily from a library science perspective, can be applied to the problem of helping people find information when they have a relatively well defined informational need (Salton and McGill, 1979). However, recent experience shows that current query techniques are much less effective for very large databases, or collections of databases, especially if the user is unfamiliar with the underlying detailed structure of the information or database. Often, even the best retrieval systems return only a portion of the material that is relevant to a query, while at the same time returning other material that is irrelevant. Paper Sources Currently, much of the information workers receive and would like to have available in multiple locations or to share with others does not arrive in electronic form. While newspaper articles, consultants' reports, and company memos may exist electronically during some stage in their life cycles, they often come to one's desk in paper form. To share hard copy with a colleague electronically is a cumbersome process. A user needs tools to easily convert paper sources of information back into electronic form and to ship, store, and retrieve them. The individual parts for these tools are commercially available in the form of scanners, optical character recognition systems, facsimile machines, multimedia databases, and multimedia electronic mail. However, they are still separate and sizable hardware items, some are still largely outside mainstream use, and users must be familiar with the details of hardware and software compatibility, file formats, image resolution, and other technical details. Individual Sources Empirical studies of how managers, scientists, and other knowledge workers search for and use information shows that text-oriented documents are not necessarily the most sought after or the most valuable information in many circumstances. In science, for example, because of their timeliness and relevance to problems at hand, personal contacts and the "gray literature" in a discipline are often as valuable as the published literature (Garvey et al., 1970), which often involves long lead-times for publication. Thus, scientists often know about new findings in their field well before they appear in conference proceedings or archival journals. When attempting to

OCR for page 35
Research Recommendations: To Facilitate Distributed Work solve a problem, scientists, engineers, and sales personnel are as likely to get help from colleagues as they are to find information in any written source (Constant et al., 1994; Culnan, 1983; Kraut and Streeter, 1994; Tushman, 1977). For example, an individual needing to understand or apply a new statistical technique is as likely to seek guidance from a local expert as from books on statistics or journal articles that have used the technique. (Whether this underuse of the archival literature is desirable or is more a reflection of the poor tools for information search is a question worthy of research in its own right.) In many other circumstances, individuals' personal stores of information are more valuable than new information gleaned from a library. Assuming that the widespread use of individuals' knowledge stores and interpersonal queries is desirable, supporting their use in distributed organizations suggests several directions for research. Information Sharing A special problem in supporting distributed work groups and organizations is ensuring that people who have useful information will provide it to others. Within organizations, colleagues and experts often respond to the requests of others for information. The problem with respect to answering such questions is that self-interest can discourage actions that promote the public good. Thus, while many people benefit from having others in an organization contribute their expertise on various topics, it can also be in an individual's interest to take advantage of these databases without contributing to them (Thorn and Connolly, 1987). For example, with properly designed networks and software, a question about using a particular statistical technique might be posed on a university-wide network. Using new indexing mechanisms, the posting could be directed to a small group of campus members with relevant expertise instead of being broadcast to the entire campus or merely posted on a computer bulletin board. These experts could be encouraged to provide answers by systems that implement payment of a royalty for information, that archive particular requests for later use by others, or that highlight an expert's similarity to or identification with the person asking the questions. In many conventional organizations, information gatekeepers play a special role in spreading needed knowledge (Tushman, 1977; Allen, 1977). These people typically have knowledge of the information needs of other group members, sources of information both within and outside the group, and the communication skills and organizational

OCR for page 35
Research Recommendations: To Facilitate Distributed Work motivation to proactively connect the two. Research on this type of selective dissemination of information has been ongoing for almost 25 years (Foltz and Dumais, 1993; Gentile and Houseman, 1970), but it has been based on conventional information retrieval techniques. Future research should also consider the dissemination of information in electronic form.