International Journal of Information Management 19 (1999) 157—172

The analysis and study of the impact of technology on groups: a conceptual framework Narender K. Ramarapu, * Mark G. Simkin,
Mike Raisinghani ¼intec Software Corporation, Sunnyvale, CA, ºSA
Department of ACC/CIS-026, College of Business Administration, Reno, NV 89557, USA  University of Dallas, Dallas, TX 76019, USA

Abstract This paper examines how various aspects of technology, group and task interplay, affect group processing and performance. Efforts to provide groups with technological support are driven by three basic ideas: improving group task performance, overcoming time and space constraints on group collaborative efforts, and increasing the range and speed of access to information. These are discussed together with three conceptual frameworks within which groupware (defined as computer-based systems that support the efforts of groups engaged in common tasks) can be developed. A conceptual framework is developed within which information systems for assisting groups electronically can be designed and measured. The paper concludes with an agenda for further research.  1999 Published by Elsevier Science Ltd. All rights reserved.

1. Introduction Technological advances in computing and communications are changing the way that people meet and make group decisions. The most common form of computerized communication is electronic mail, but there are many other kinds of computerized systems that foster or support group interaction — for example, internet distribution lists, bulletin boards, on-line discussion group software, and computer-assisted conferencing systems. These technologies help people cross geographic, social, and psychological boundaries, and have such secondary or indirect effects as expanding the number of participants in collaborative decision making. What is most obvious about the dynamics of computer-assisted group decision making is how much they differ from what takes place in face-to-face meetings. But there is very limited empirical research that assesses the potential of computer-assisted groups, or evaluates the impact of computer-mediated interactions on groups. The purpose of this paper is to examine how various * Corresponding author. Tel.: 408 530 8036; fax: 408 530 8036; e-mail: [email protected]

0268-4012/99/$ - see front matter  1999 Published by Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 4 0 1 2 ( 9 9 ) 0 0 0 0 6 - 7

158

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

aspects of technology, group, and task interplay, affect group processing and performance, and to develop a conceptual framework within which information systems for assisting groups electronically can be designed and measured. While this research involves electronic communication in groups, it does not address communication technologies such as video conferences, telephone conferences, and facsimile that can be used for group decision making. The reason is that these technologies typically lack the power of computers for storing, editing, and channeling information, and for creating and using databases. Instead, this research studies electronic communication in groups, not because it is more or less efficient than other communication devices, but because it provides a larger scope for investigating basic group processes than many other technologies do.

2. Groupware for supporting computer-mediated work The term ‘‘groupware’’ was originally coined by Peter and Trudy Johnson-Lenz in 1978, and now describes a large number of computer-based systems that support the work efforts of groups engaged in achieving common tasks or goals (Ellis et al., 1991). Technically, ‘‘groupware’’ also includes software for collaborative games and other recreational activities, and is thus not a synonym for computer-supported cooperative work (CSCW) software. As suggested by Fig. 1, however, even CSCW groupware encompasses a wide range of products, including software that supports face-to-face meetings (such as electronic meeting room and presentation support products), systems that support electronic meetings (such as on-line or chat systems), and products that support activities between meetings (such as electronic calendars). Efforts to provide groups with technological support are driven by three basic ideas: improving group task performance, overcoming time and space constraints on group collaborative efforts, and increasing the range and speed of access to information. These are explained in a little more detail below. 2.1. Improving task performance Efforts to improve group productivity rest upon the idea that computers can reduce the time to perform mechanical chores (for example, revise manuscript drafts), sift through data (for example, large amounts of data in databases), and so forth. Improving task performance can also happen through improved communication channels. Efforts in this area often take the form of interventions, in which the person managing the situation (group facilitators, group supervisors, researchers) structures the communication within the group, the form and sequences of task responses required by the group, and/or the task information available to the group. Early intervention efforts of this kind are reflected in work identified under labels such as brainstorming (Osborn, 1957; Diehl & Stroebe, 1987), Delphi method (Dalkey, 1969), nominal group technique (NGT) (Van de Ven & Delbecq, 1974), and multiatttribute utility analysis (MAUA) (Eils and John, 1980). All of these systems used in the past decades are considered ‘‘manual’’ technologies today since they lacked computer-based group support systems. However, it is not clear that computers necessarily help. For example, despite the presence of ‘‘sponsors’’, ‘‘moderators’’, and other types of

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

Fig. 1. Examples of groupware applications.

159

160

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

on-line facilitators, on-line decision making has proven difficult. If there is no formal process for decision making, for example, electronic groups tend to avoid closure. Electronic polling is one solution since many participants who are reluctant to contribute in other ways will respond to a poll. Although the more recent systems have raised the sophistication of technology by several orders of magnitude, they lack the depth of exposure to the theoretical and practical issues of dealing with groups. 2.2. Overcoming time and space constraints When groups are geographically dispersed and cannot meet face-to-face, decisions can be made electronically. When properly planned and implemented, electronic group communications can radically increase the number of people who can participate in decision-making processes — for example, over separate time zones or across organizational boundaries. Electronic communication, just because of its dependence on writing, can allow for more thoughtful input. It offers an ongoing record of communications that greatly improves the efficiency of groups working together in complex projects that have much detail and little time to rework issues and problems. Researchers have not paid much attention to the concept of time. Yet, there are a myriad of temporal issues central to the work of task groups. These include the temporal context in which the group’s work is being carried out (e.g., deadlines set by agencies outside the group), as well as the temporal patterning of the group’s behavior (e.g., the synchronization of related acts by different group members). With the advent of video-conferencing systems, all the members interacting in a group do not have to be in the same place at the same time. There has been some research directly testing the effects of physical separation (as distinct from reduction of communication modalities) on group interaction and task performance (Rice, 1980a, b, 1982; Kiesler & Sproull, 1992; Campion et al., 1996; Parker, 1996). However, getting the right information, just in time will continue to gain importance as work groups are increasingly likely to function across time and space. Groups will seek to interrelate time and task activities and interact with one another through highly visual displays so that the information users view and the activities they are likely to perform correspond closely to their underlying cognitive models of the tasks at hand. 2.3. Increasing information access With the internet and World Wide Web, access to global information has never been easier. The conceptual issues of information acquisition, the theoretical issues about the antecedents and consequences of different patterns of information distribution within work groups, and the conditions under which information can be and is easily shared among group members is of key interest in increasing information access. In many aspects of their work, groups use computers to increase their information access. For example, a group uses a computer as a tool to retrieve consistent information from a database to provide a structured format within which task performance takes place. Also, each group member can retrieve vast bodies of information via computer databases for their information access and processing, performance structuring and communication — all three of these functions may or may not go together. The use of a computer-aided communication system by a group is likely to lead to

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

161

a pattern of participation that overall is less in amount but more equally distributed among members (McGrath & Hollingshead, 1994). However, this conclusion does not take into account the temporal aspects of distribution of a group process or the group’s functions, modes, and tasks.

3. Prior research findings The theoretical merits of groupware are compelling. For example, it is easy to hope that computer-aided communication systems in a work group will increase interaction or participation for all participants, lead to less arguments (although there may be more uninhibited communication — such as ‘‘flaming’’), result in more positive socioemotional communication, and expose conflicts more effectively (Guzzo et al., 1995). The truth, however, has been something less than this. Researchers did find that groups reached a higher degree of consensus if they did make a decision, and tended to produce more ideas of higher quality on idea-generation tasks (although this can be attributed to task structure which may include procedures that simplify the handling of complex information). In a surprising number of cases, however, groups using computers also took longer to carry out a given task than did face-to-face groups, and, additionally, were less likely to finally reach a form of consensus. As explained in greater detail below, other interesting cases of dysfunction include productivity losses when brainstorming, group size problems, dissatisfaction/satisfaction with decision process and several secondary effects. 3.1. The productivity gap when brainstorming in groups Without computers, the empirical evidence suggests that groups are only about half as productive as when an equal number work alone. Current research suggests at least four reasons for this: (1) ‘‘free-riding’’, (2) ‘‘evaluation apprehension’’, (3) ‘‘production blocking’’, and (4) ‘‘matching’’. The first two of these — ‘‘free riding’’ and ‘‘evaluation apprehension’’ — suggest that groups depress performance because of motivational or emotional effects (Diehl & Stroebe, 1987). For example, ‘‘free-riding’’ happens when individuals reduce their own efforts because the others in the group are performing at high levels. The free rider may feel that his or her efforts are dispensable or not needed for group success. ‘‘Evaluation apprehension’’ occurs when individuals fear they will be judged by others in their groups. Individuals may also be uncomfortable or socially anxious in groups because of concern about potential negative reactions of other group members. However, groupware that allows users to contribute ideas anonymously appears to at least partially overcome the problems of evaluation apprehension. The behavioral and cognitive interference resulting from people attempting to communicate simultaneously also appears to account for part of the production deficit in brainstorming groups. The two major factors here are ‘‘production blocking’’ and ‘‘matching’’, both of which relate to the social activity rate of the group members. ‘‘Production blocking’’ or simply ‘‘blocking’’ occurs when an individual’s ideas differ substantially from what the other group members are thinking, thus creating a barrier to the initial line of thought. It can also happen when budding ideas are forgotten while the individual awaits his or her turn to talk in a group discussion.

162

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

Finally, the social process of ‘‘matching’’ means comparing rates of performance when individuals are generating ideas within a group. In particular, matching involves the concept of a group productivity norm, and the belief that it is not desirable to vary substantially from it. Thus, when brainstorming a task, individual group members may not want to perform at a rate that differs too much from those of other group members. As a result, high-output members may decrease their output, while other group members free ride. On the other hand, low-output members may attempt to increase their output to match higher output members so as not to appear lazy or unintelligent. The interactive effects of free riding, evaluation apprehension, blocking, and matching are also noteworthy. Free riding, evaluation apprehension, and blocking initially lower the rate of production of ideas for the group as a whole, and then, with the rate lower than it would have been, matching processes kick in to maintain that rate. However, blocking and matching are inversely proportional. For example, when individuals are exposed to high rates of performance by co-workers, the tendency to match should lead to increased performance but the blocking effects of productive co-workers should decrease performance. If the impact of these counteracting forces is similar, it may be difficult to observe variations in performance related to the presence of highly productive co-workers. Paulus et al. (1996), succeeded in demonstrating separate effects of social comparison and blocking in a modified electronic brainstorming paradigm. The evidence suggests that computers can help solve the problems of lowered productivity in group environments. With electronic brainstorming, individuals generate ideas on computers in groups of varying sizes — perhaps anonymously. In such group formats, the ideas generated by others can be seen on the top half of a monitor screen (Valacich et al., 1992). Part of the advantage of computer mediation may stem from the limited feedback about how much others, both individually and collectively, have done. This type of electronic ‘‘group’’ is not associated with the production losses related to conventional group brainstorming (Gallupe et al., 1992; Valacich et al., 1994), thus making it well-suited for directly examining the role of free riding, evaluation apprehension, and matching in groups. The paradigm of electronic brainstorming (e.g., Nunamaker et al., 1991) facilitates the study of cognitive simulation and memory effects of idea exchange in groups and allows for greater control in laboratory brainstorming studies. The output of a computer-simulated brainstormer can be controlled precisely according to the parameters of the brainstorming model (e.g., Connolly et al., 1993). The ability to simulate a specific type of group makeup (e.g., participants with low sensitivity to matching) would allow individual brainstormers to experience what it is like to participate in a variety of group situations. Yet, thus far, there is no evidence that electronic brainstorming is more effective than solitary oral brainstorming (Paulus et al., 1996). 3.2. The effects of group size in computer-mediated work groups Can the size of the work group affect performance? Several researchers have examined the issues of numerical group size (ranging from 5 to 12 members) and the heterogeneity of task information among group members to determine the extent to which numerical group size enhances performance when each member possesses unique, task-relevant information (Valacich et al., 1995; Connolly et al., 1993; Dennis & Valacich, 1993). In addition, logical size was varied by the amount of task-relevant information given to a particular group member. ¸arger logical group sizes

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

163

(heterogeneous groups) were induced by distributing unique task-relevant information among groups members; smaller logical group sizes (homogeneous groups) were induced by providing all task information to all members. The researchers found that larger groups addressing an ideageneration task using computer-mediated communication outperformed smaller groups. Numerical group size interacted with logical group size resulting in greater performance gains for increased numerical group size within heterogeneous groups. Finally, the average contributions per group member diminished with increased numerical group size for homogeneous groups and increased for heterogeneous groups. Dennis and Valacich (1993) found that 12-member groups interacting through computer mediation outperformed 12-member nominal groups, but found no performance differences between 6-member computer-mediated groups and 6-member nominal groups. They state that since larger groups were able to generate more ideas than smaller groups, smaller groups may not overcome the overhead associated with interacting within the task time allotted. They found that the performance benefits from interacting came through the ability of groups to build on the ideas of others and avoid ideas already proposed. However, interacting computer-mediated groups have an overhead associated with their processing that nominals do not (i.e., reading the ideas of others). Thus although small interacting groups benefited from synergy and redundancy avoidance, these benefits did not overcome the overhead associated with interacting. Thus more task processing time may have helped overcome the overhead, as this overhead has been mathematically modeled to be a ‘‘fixed’’ process loss (Valacich & Dennis, 1994), at least in some idea generation situations. The conclusion is that the optimal group size in computer-mediated groups addressing an idea generation task is not a fixed size such as five members. Most likely the optimal size is influenced by the task and the task-relevant resources of the available group members. In addition, the optimal group size for heterogeneous groups addressing a relatively complex task will be larger than the optimal size for homogeneous groups. Implications for practice suggest that the optimal size of a group is determined by the job requirements (i.e., task skills and task-relevant knowledge): if members are relatively homogeneous, then the optimal size may be smaller; if members are relatively heterogeneous, then the optimal size may be larger. 3.3. Satisfaction Research also shows that computer-mediated groups were more satisfied with the decision process than face-to-face groups while employing the expert, devil’s advocacy, or dialectical inquiry decision-making approach (Valacich & Schwenk, 1995). Since these results run counter to a substantial amount of prior research (e.g., Benbasat & Lim, 1993; McLeod, 1992), it may be that either the groups perceived the computer-mediated environment to provide a more efficient process, or that the groups may simply have been influenced by the novelty of the computer-mediated environment. However, if novelty did influence process satisfaction, it did not influence outcome process satisfaction, as no satisfaction differences for the group outcome were found by the authors. 3.4. Secondary effects The essence of group dynamics is the interplay between each individual member of the group and the group itself. Using groupware often shifts this relationship in delicate and often unanticipated

164

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

Fig. 2. Classification of communication category by space/time dependence.

ways, and may be especially noticeable in business environments that are themselves experiencing dramatic changes. Examples of such effects include empowering heretofore geographically remote users who were formerly ‘‘out of the loop’’, changing the social status of interpersonal organizational relationships (e.g., when a lower-level employee can now email the corporate president), or replacing telephone operators with automated telecommunications equipment. Early adopters of groupware tend to emphasize the planned uses of the systems because the known advantages are both better understood and so technologically compelling (Daft & Lengel, 1986). Yet, two of the most important characteristics of secondary effects is that they are usually difficult to foresee and easy to underestimate. The fact that such effects tend to emerge slowly and are so integrally connected to the social, cultural, and political fabric of an organization also means that they will continue to challenge would-be implementers for some time to come.

4. The need for a conceptual model The advantages of groupware have been heavily debated in the literature, with opinions ranging from deep skepticism to enthusiastic acclaim. Part of the problem has been the fact that groupware products cover such a wide spectrum of applications, with uses ranging from conventional timesharing systems to real-time, multi-user software and tools. Thus, one need for a conceptual framework for group processes is to formalize the different types of space/time dependencies. An example is shown Fig. 2. Groupware that is space or time independent allows groups to meet without being in the same place and/or time. This gives the individual enormous freedom and flexibility because it permits ‘‘virtual’’ meetings by allowing group members to participate via their computers. Computer-mediated groups represent a paradigmatic shift. The traditional computing paradigm sees the computer as a personal tool for manipulating and exchanging data. In contrast, the computer-mediated group paradigm views the computer as a ‘‘shared space in which people collaborate’’ — a clear shift in the relationship between people and information. Computer-mediated group designers are not only concerned with the information processing issues, but also with the human

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

165

relationships computers can support. This section discusses the conceptual foundations that allow us to see the varied ways that people organize their work, and to describe the complex social situations in which computer-mediated groups must operate. If researchers are to make sense of computer-mediated group interaction, they will need models with which to evaluate the various dimensions of group interaction. At present, there are at least three conceptual frameworks within which groupware can be developed: (1) augmentation, (2) language, and (3) coordination. Each of these is discussed briefly below. 4.1. Augmentation Engelbart (1984) asserts that computers and people must evolve together, and that computers should be designed to augment (rather than to replace) human capabilities. In fact, augmentation is really a simultaneous growth for people and computers. The computer side comprises tools, methods, skills, knowledge, language, training, and organization. The human side adds highperformance work, specific roles in the workplace, and the ability to be engaged in work while not physically being present at the office. Engelbart sees the computer as a tool supporting wide-scale augmentation, which requires software designed for interoperability and openness. Interoperability allows a person to access the materials of any other person, with software tools that operate correctly in work contexts and vary over time and distance. Openness comes from an underlying hypertext system, and provides a common facility for creating, transporting, storing, and manipulating work materials. Interoperability and openness are key concerns for the computer industry today. For example, groupware designs must address both the transparency of information access, regardless of its locale, and the evolution of the system to suit users’ needs. These requirements call for thoughtful deliberation during design, and an architecture that separates user interface and user profiles from application and network functionality. 4.2. Language Bodker et al. (1988) suggest developing computer applications for cooperative work by ensuring a cooperative process design. From this viewpoint, design is a process that creates computer artifacts in participatory ways. Participation is a collective approach stressing industrial democracy, quality of work and product, and human-centered design. Bodker et al. suggest structuring practice through language games. In a language game, users teach designers the ‘‘game of work’’ and designers teach users the ‘‘game of design’’. This method of joint learning equalizes the contributions people make and creates an open learning atmosphere. Language games help users and designers to share their different backgrounds and understandings of terms and participate in the totality of the work to aid the development of an interdisciplinary design that addresses the whole work situation (‘‘the big picture’’). The experience of Bodker et al. with language games suggests that computers are used in multilingual environments, evidenced by the variations in terminology used by different groups within the same company. Sensitivity to the fact that users cannot anticipate all future situations and that users tend to be traditional in their view of both the work and the computer’s potential is required by the designers.

166

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

4.3. Coordination Malone and Crowston (1990) define ‘‘coordination’’ as the act of working harmoniously, and thus define coordination theory as a body of principles about how actors can work together in harmony. The central theme of coordination theory is that the workplace centers around the interactions of actors (people) and agents (computerized procedures). Their framework for coordination theory is developed from the components of coordination and associated processes. They suggest the following four-tiered approach: (1) identify work goals, (2) map goals to activities, (3) assign activities to actors, and (4) manage the interdependencies among actors. The actors involved in a specific work situation may or may not distinguish all these components. For example, an entire marketing department may be viewed as a single actor at one time, while at other times each marketing person may be viewed as a distinct actor. Hence, Malone and Crowston (1990) focus specifically on those aspects of a situation that are unique to coordination, defining coordination as ‘‘the act of managing interdependencies between activities performed to achieve a goal’’. Managing interdependencies is a key to good coordination. The interdependencies among agents are classified as generic (for example, sequenced actions, a shared resource, or simultaneous actions), or domain-specific (for example, a part designed by an engineering team must be able to be manufactured in quantity). The authors see domain-specific coordination as an extremely common requirement of today’s systems, asserting that the effectiveness of a general theory of coordination is measured by its transferability to a variety of work domains.

5. A model for analyzing the effects of computers on work groups Research on group support systems has focused more on technical developments and applications than on the basic theoretical issues involving the functioning of work groups. Although there is a wealth of well-founded research in work psychology concerning group interaction, group dynamics (leadership, roles, norms, development, and communication), and the behavior of individuals in groups — e.g., by Festinger (convergence and divergence and groups), Ringelmann (model of group losses), Tuckmann, Forgas, McGrath and Hackman (group/task modeling) — there is a lack of an integrative, systematic conceptual framework in CSCW that can serve as a foundation for future research. In work environments, groupware reflects a change in emphasis from ‘‘using computers to solve problems’’ to ‘‘using computers to facilitate human interaction and communication’’. Thus, it is a technology that focuses directly on groups and group processes. At its core, groupware includes the following three elements: (1) electronic communication to link group members, (2) one or more databases and concomitant software to store group information, and (3) features that support, foster, and remember group activity. What is needed is a paradigm that enables researchers to not only measure the relative importance of these elements within a group decision-making environment, but also a model that identifies the major external factors that affect these elements and measures these effects in meaningful ways. Because technology profoundly affects the nature of group work (Huber, 1990; Nunamaker et al., 1991; DeSanctis & Gallupe, 1987; Simon, 1976), it is inappropriate to generalize the outcomes from

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

167

Fig. 3. Meta-level research model.

non-computer-supported work groups to the computerized environment. A better approach is to take a meta view of the research, as illustrated in Fig. 3 (Dennis et al., 1988; Nunamaker et al., 1991). Meeting outcomes (e.g., efficiency, effectiveness, satisfaction) depend upon the interaction within the meeting process of the group (e.g., group size, group proximity, group composition, group cohesiveness, etc.); task (e.g., idea generation, decision choice, task complexity, etc.); context (e.g., organizational culture, time pressure, reward structure, etc.) and technology factors that differ from situation to situation. To understand these interactions, an analysis of the group processes at a lower level of detail is required. This new framework is illustrated in Fig. 4. Like the models in McGrath and Hollingshead (1994) and Hackman (1969), this model views the groupware paradigm as the interaction of four basic sets of variables. The contextual variables in the first, left-hand panel of Fig. 4 are the basic constituents of work groups: technological support, group structure, personal factors, and task characteristics. Within each of these groups are the individual elements that define or operationalize these variables — for example, the type of decision software for ‘‘level of technological support’’ or the complexity of the problem for ‘‘type of task’’. The second panel of variables, termed ‘‘Operating Conditions’’ in Fig. 4, reflects the prevailing conditions under which a given group implements its tasks. Examples of such conditions are the degree to which a group is pressured by time, the degree of anonymity of the group members, or the situational factors that define group membership or cohesiveness outside the immediate work environment. The third panel of variables are ‘‘Group Process Variables’’ that define ongoing group activity. Examples of such variables are the amount of participation, the time required (or given) to reach a decision, or the level of cooperation among group members. These can either be dependent outcomes (for example, that measure the degree of social interaction or the clarity of communication among group members), or independent variables that form antecedents for yet other variables. Finally, the ‘‘Task/Group Related Outcomes’’ provide outcome measures of group performance. Examples include the quality of the decision making, the confidence of the users in the system, or

168

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

Fig. 4. A conceptual framework for studying the impact of technology on groups.

the level of satisfaction with the decision or with the process by which the decision was reached. These variables typically form the criteria for evaluating the effectiveness of the communication system or the other input factors that were modulated for testing purposes.

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

169

In keeping with the principles of general system theory, transfer of knowledge of these tools across various fields can help prevent the duplication of the same concepts in different fields that are isolated from each other. For instance, field researchers could use these tools when presenting qualitative investigations of CSCW effects on group process on different meeting situations and lab experimenters could develop contingency models to isolate and explain why certain CSCW features (i.e., types of process support, process structure, task support and task structure) are of value for certain groups, tasks and contexts.

6. An agenda for future research A critical review of the literature on computer-mediated groups indicates some deficiencies which obviously seriously impede the generalization of study findings. Some examples include the lack of attention paid to the composition of groups with respect to such characteristics as gender, task experience, or technical competence, or the length of organizational tenure of the individuals within the group. Also, researchers have not considered the arrangements and types of group task, the interaction of member and group attributes in relation to task, and technology factors. Future research needs to be more systematic and comprehensive with respect to the variables being studied and their interactions. It will also have to develop methodologies that allow comparisons of findings across multiple studies. In addition, longitudinal research on the incontext effects of the use of technology in groups will be more useful than studying short-run effects under context-stripped, relatively artificial operating conditions. Multiple-criteria assessment should give an in-depth view of the group performance. Here, one needs to consider variables such as quality, quantity, cost of group task performance, group member attributes (abilities, experience, etc.) and interrelate them to group attributes — for example, group size and the heterogeneity of members regarding a number of attributes (McGrath & Hollingshead, 1994; Lipnack and Stamps, 1997; Mankin et al., 1997). Another avenue will be to study the interaction of member and group attributes with task and technology factors, or comprehensively examine interactions of technology and task (or technology-task fit). In recent years, several research areas in the IS literature address the issue of technology-task fit regarding information richness, problem-solving, and/or decision making (Goodhue & Thompson, 1995; Ramarapu et al., 1994; Vessey, 1991; Vessey & Galleta, 1991). Technology-task fit is the degree to which a technology would assist group members in performing their task. This concept of technology-task fit has shown evidence that the impact of technology on performance seemed to depend on fit with the task (Jarvenpaa, 1989; Vessey, 1991; Goodhue & Thompson, 1995). This same concept can extend to the computer-mediated work group environment. Also, another avenue can be knowledge required for the given task should be tied with the technology. Task knowledge type (superficial vs. causal knowledge) can be introduced via the concept of fit to the technology factor to determine group performance. Research needs to be focused on finding tested models and methods for improving team effectiveness in any organizational context and show the empirical results of using information technology to help teams function and fundamentally reshape the way individuals work together. A more theory-guided, multivariate, and longitudinal approach that can explore all the variables discussed above in a systematic manner is required. Research needs to take a fresh look at the

170

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

management of controversy, crisis, and conflict in computer-mediated teams; examine team member diversity and its effect on the group process; and elevate to the team level existing models for understanding individual performance, decision-making processes, and the effects of stress.

7. Summary and conclusions A major trend is underway that shifts the computer framework of today’s modern organization from a mechanistic one to a contextual one, a shift that replaces ‘‘standard practice’’ with procedures that emphasize augmentation, language, coordination, and other conceptual foundations yet to be discovered. In addition, there is a rapidly growing need for cooperation among many large organizations — a need that depends heavily on groups to plan strategies and make decisions of increasing complexity. These groups will need to use an infrastructure that supports wide-scale cooperative work. ¹he mechanistic approach that underlies current computing practices is proving to be inadequate for building this infrastructure. For all the technology that is now available for these tasks, it is easy to hypothesize that computer-mediated work groups are still in their infancy. But because of their importance, we need to have a better understanding of how groups and organizations function and evolve than we now possess. At the same time, we also need to know more about individual differences in responding to technology if we are to develop systems that can support entire groups. However, smart organizations can get a headstart on the competition by gaining familiarity with groupware since the changes required in learning how to work in groups with technology are pervasive. The impact of technology on group process and performance operates in dynamic interdependence with key features of the group composition, task, and situation. Organizations can employ a combined approach, for example, by using widely available asynchronous electronic mail systems to surface and share solutions and face-to-face communication to select a final solution. It is becoming possible to merge today’s understanding of technology with an understanding of group dynamics. This foundation can be the basis for a new way of conducting business where information and time are focal points of business strategy. Computer support for the activities of individuals in their group and organizational contexts will unquestionably change the way people work in significant ways. As researchers and developers, we must develop a better understanding of our own decision-making processes and broaden our intuitions.

References Benbasat, I., & Lim, L.H. (1993). The effects of group, task, context, and technology variables on the usefulness of group support systems: A meta-analysis of experimental studies. Small Group Research, 24 (4), 430—462. Bodker, S., Knudsen, J.K., Morten, K., Ehn, P., & Madsen, K.H. (1988). Computer support for cooperative design. Conference on Computer-Supported Cooperative ¼ork (pp. 377—394). Campion, M.A., Papper, E.M., & Medsker, Gina, J. (1996). Relations between work team characteristics and effectiveness: A replication and extension. Personnel Psychology, 49 (2), 429—452. Connolly, T., Routhieaux, R.L., & Schneider, S.K. (1993). On the effectiveness of group brainstorming: Test of one underlying cognitive mechanism. Small Group Research, 24, 490—503.

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

171

Daft, R.L., & Lengel, R.H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32 (5), 554—571. Dalkey, N.C. (1969). ¹he Delphi method: An experimental study of group opinion. Santa Monica, CA: RAND. Dennis, A.R., George, A.F., Jessup, L.M., Nunamaker Jr., J.F., & Vogel, D.R. (1988). Information technology to support electronic meetings. MIS Quarterly, 12 (4), 591—624. Dennis, A.R., & Valacich, J.S. (1993). Computer brainstorms: More heads are better than one. Journal of Applied Psychology, 78 (4), 531—537. DeSanctis, G., & Gallupe, R.B. (1987). A foundation for the study of group decision support systems. Management Science, 33 (5), 589—609. Diehl, M., & Stroebe, W. (1987). Productivity loss in brainstorming groups: Toward the solution of a riddle. Journal of Personality and Social Psychology, 53, 497—509. Engelbart, D. (1984). Authorship provisions in Augment. Proceedings of the IEEE COMPCON, March. Eils, L.C., & John, R.S. (1980). A criterion validation of multiattribute utility analysis and of group communication strategy. Organizational Behavior and Human Performance, 25, 268—288. Ellis, C.A., Gibbs S.J., & Rein, G.L. (1991). Groupware: Some issues and experiences. Communications of the ACM, 34 (1), 38—58. Gallupe, R.B., Dennis, A.R., Cooper, W.H., Valacich, J.S., Bartianutti, L.M., & Nunamaker, J.R. Jr. (1992). Electronic brainstorming and group size. Academy of Management Journal, 35, 350—369. Goodhue, D.L., & Thompson, R.L. (1995). Task-technology fit and individual performance. MIS Quarterly, 19 (2), 213—236. Guzzo, R.A., Salas, E., & associates (1995). ¹eam effectiveness and decision making in organizations. Ch. 3: Computer assisted groups and Ch. 6: Modeling and Simulations (1st ed.), CA:Jossey-Bass Inc. Hackman, J.R., (1969). Toward understanding the role of task in behavioral research. Acta Psychological, 31, 97—128. Henry, J.E., & Hartzler, M. (1997). Virtual teams: Today’s reality, today’s challenge. Quality Progress, 30 (5), 108—109. Huber, G.P. (1990). A theory of the effects of advanced information technology on organizational design, intelligence, and decision making. Academy of Management Review, 15 (1) 47—71. Johansen, R. (1989). User approaches to comptuter-supported teams. In M.H. Olson (Ed.), ¹echnological support for work group collaboration (Chapter 1, pp. 1—31). Jarvenpaa, S. (1989). The effect of task and graphical format on information processing strategies. Management Science, 35 (3), 285—303. Kiesler, S. and Sproull, L. (1992). Group decision making and communication technology. Organizational Behavior and Human Decision Processes, 52, 96—123. Lipnack, J., & Stamps, J. (1997). »irtual teams: Reaching across space, time and organizations with technology. New York: Wiley. Malone, T., & Crowston, K. (1990). What is coordination theory and how can it help design cooperative work systems? Proceeding of the CSC¼’90, October. Mankin, D., Cohen S., & Bikson, T. (1997). ¹eams and technology: Fulfilling the promise of the new organization. Boston: Harvard Business School Press. McGrath, J.E., & Hollingshead, A.B. (1994). Groups interacting with technology (Vol. 194). Thousand Oaks, CA: Sage. McLeod, P.L. (1992). An assessment of the experimental literature on the electronic support of group work: Results of a meta-analysis. Human Computer Interaction, 7 (3), 257—280. Nunamaker, J.F., Dennis, A.R., Valacich, J.S., & Vogel, D.R. (1991). Information technology for negotiating groups: Generating options for mutual gain. Management Science, 37, 1325—1346. Osborn, A.F. (1957). Applied imagination (2nd ed.). New York: Charles Scribner’s Sons. Parker, G.M. (1996). ¹eam players and teamwork: ¹he new competitive business strategy. San Francisco: Josey- Bass Publishers. Paulus, P.B., Larey, T.S., Putman, V.L., Leggett, K.L., & Roland, E.J. (1996). Social influence processes in computer brainstorming. Basic and Applied Social Psychology, 18, 3—14. Ramarapu, N., Frolick, M., & Wilkes, R. (1994). The impact of sequential versus non-sequential access to information presentation in problem solving. Proceedings of the International Conference on Information Systems, December. Rice, R. (1980a). Computer conferencing. In B. Dervin, & M. Voigt (Eds.), Progress in communication sciences (Vol. 2). Norwood, NJ: Ablex.

172

N.K. Ramarapu et al. / International Journal of Information Management 19 (1999) 157—172

Rice, R. (1980b). The impacts of computer-mediated organizational and interpersonal communication. In M. Williams (Ed.), Annual Review of Information Science and ¹echnology (Vol. 15, pp. 221—250). White Plains, NY: Knowledge Industry. Rice, R. (1982). Communication networking in computer-conferencing systems: A longitudinal study of group roles and system structure. In M. Burgoon (Ed.), Communication ½earbook (Vol. 6, pp. 925—944). Beverly Hills, CA: Sage. Simon, H.A. (1976). Administrative behavior (3rd ed.), New York: Free Press. Valacich, J.S., & Schwenk, C. (1995). Devil’s advocacy and dialectical inquiry effects on face-to-face and computermediated group decision making. Organizational Behavior and Human Decision Processes, 63 (2), 158—173. Valacich, J.S., Wheeler, B.C., Mennecke, B.E., & Wachter, R. (1995). The effects of numerical and logical group size on computer-mediated idea generation. Organizational Behavior and Human Decision Processes, 62 (3), 318—329. Valacich, J.S., Dennis, A.R., & Connolly, T. (1994). Idea generation in computer-based groups: A new ending to an old story. Organizational Behavior and Human Decision Processes, 57, 448—476. Valacich, J.S., & Dennis, A.R. (1994). A mathematical model of performance of computer-mediated groups during idea generation. Journal of Management Information Systems, 11 (1), 59—72. Valacich, J.S., Dennis, A.R., & Nunamaker, J.F. Jr. (1992). Group size and anonymity effects on computer mediated idea generation. Small Group Research, 2 (1), 49—73. Van de Ven, A.H., & Delbecq, A. (1974). The effectiveness of nominal, delphi, and interacting group decision making processes. Academy of Management Journal, 17, 605—621. Vessey, I. (1991). Cognitive fit: A theory-based analysis of the graph versus tables literature. Decision Sciences, 22, 219—239. Vessey, I., & Galletta, D. (1991). Cognitive fit: An empirical study of information acquisition. Information Systems Research, 2 (1), 63—86.

Dr. Narender K. Ramarapu is the Principal Director of Wintec Software Corporation in Sunnyvale, California. He leads a team of consultants who specialize in the area of ERP and Electronic Commerce applications. Prior to the current position, he used to be an assistant professor at University of Tennessee and Univeristy of Nevada. He has an MBA and a Ph.D. in MIS from the University of Memphis. His current research is in the areas of Emerging Technologies, Information Presentation and Business Reengineering. Mark G. Simkin is a professor of Computer Information Systems at the University of Nevada, Reno. He earned his MBA and Ph.D. from the University of California, Berkeley. His research in such area as end-user computing, computer education, and computer crime appears in over 75 academic journals articles and 12 books. Two of his most recent books are ‘‘Core Concepts of Accounting Information Systems’’ and ‘‘Applications Programming in Visual Basic 5.0’’. Mike Raisinghani is an Assistant Professor of Information Systems at the University of Dallas, TX. He has received his Ph.D. from The University of Texas at Arlington.