A comparison of
face-to-face and
virtual software
development teams
Hayward P. Andres
Virtual teams, within and across
organizations, are a recent phenomenon (Geber, 1995). Virtual teams are groups of individuals collaborating in the execution of a specific project while located at multiple individual sites or multiple group sites. Virtual teams have been brought about by the need for organizations to get projects done as quickly as possible while utilizing the skills of project team members that are geographically dispersed. The formation of virtual teams facilitate the incorporation of a wide range of knowledge and expertise possessed by individual members into a collective body of knowledge needed to conduct effective group problem-solving activities (Laughlin, 1980). Communication technologies used to support dispersed groups include electronic mail, computerized conferencing, and
videoconferencing (McDonough et al., 1999; O’Conaillet al., 1993; Ocker et al., 1996). Group process support provided by communication technologies include idea generation, problem-solving information exchange, clarification of efforts, conflict resolution, negotiation, and decision making. These communication technologies allow organizations to form virtual teams by dissolving the boundaries that separate groups. In addition, these virtual teams can be instantly dissolved. Finally, communication technologies give rise to potential
collaborations among workgroups and organizations that would otherwise not be considered.
Most of the investigations of the effects of electronic communication systems in the support of work groups have focused on the efficacy of group decision support systems (GDSSs) and distributed group decision support systems (DGSSs). These studies addressed the support of local and distributed group decision making (Chidambaram and Jones, 1993; Hiltzet al., 1986; Nunamaker et al., 1991). These studies have suggested that significant outcomes related to the use of electronic communication support include better distribution of participation across decision makers, increased number of decision alternatives, final decision quality, and satisfaction with the decision-making process. There are only a few studies that have begun to consider the impact of electronic communication technology on the The author
Hayward P. Andres is an Assistant Professor of the Information Systems Area in the School of Business Administration, Portland State University, Portland, Oregon, USA.
Keywords
Teams, Group dynamics, Interaction,
Communications technology, Videoconferencing Abstract
Looks at new communications technologies, such as videoconferencing systems, which have enabled the creation of ``virtual organizations’’ and ``virtual teams’’. Investigates the hypotheses that both ``social presence’’ and ``media richness’’ associated with a communication medium used to support geographically-disperse d software development teams, will have a significant impact on team productivity, perceived interaction quality, and group process satisfaction. Results supported the predicted superiority of the face-to-face setting over the videoconferencing setting with regard to team
productivity. They also indicated that a communication medium characterized as high in both ``media richness’’ and ``social presence’’ can engender a greater sense of interaction quality. There were no significant differences between the face-to-face and videoconferencing settings for group process satisfaction.
Electronic access
The research register for this journal is available at http://www.emeraldinsight.com/researchregisters The current issue and full text archive of this journal is available at
http://www.emeraldinsight.com/1352-7592.htm
Team Performance Management: An International Journal Volume 8 . Number 1/2 . 2002 . pp. 39±48
efficiency of production-task execution and product quality in distributed work group contexts (McGrath, 1990; Straus, 1996; Straus and McGrath, 1994). In an examination of where videoconferencing systems (VCS) fit into the media richness scale (Daft and Lengel, 1986), Kydd and Ferry (1994) noted that VCS is rated between teleconferencing and face-to-face meetings with regard to media richness and social presence.
Software development has been described as a collaborative problem-solving activity where success is dependent upon knowledge acquisition, information sharing and
integration, and the minimization of communication breakdowns (Walzet al., 1993). The purpose of this study was to investigate the impact of videoconferencing technology on software development task and affective outcomes of dispersed software development teams. Social presence, media-richness, time, interactions, and performance (TIP) theories and the software development literature are used to provide a framework for the comparison of the impact of face-to-face and videoconferencing collaboration on software project success. This study will address the media richness attribute of a communication medium and its ability to facilitate adequate information transfer needed for productive collaborative problem solving among geographically distributed team members. Social presence and TIP theories offer a theoretical basis for the examination of a communication medium’ s influence on group productivity, interaction quality, and satisfaction with the task execution process.
As new communication technologies make their way into organizations, it is suggested that there will be some induced changes in organizational forms (e.g. virtual teams, virtual organizations). Consequently, it is prudent to investigate the manner in which these technologies impact group processes during collaborative problem solving and distributed group decision making. This study reports the results of a laboratory experiment in which groups either used face-to-face meetings or videoconferencing technology to conduct communication needed to execute the assigned task.
The following section presents a review of empirical research on social presence, media richness, TIP, and software development.
Next, using relevant research findings, hypotheses regarding the role of videoconferencing’ s impact on task and affective outcomes of a typical software design activity are presented. This is followed by a discussion of the findings and suggestions for future research on assessing communication media and the support of distributed work groups.
Theoretical background
Social presenceSocial presence refers to the ability of a communication medium to allow a group member to feel the presence of the other group members and the feeling that the group is jointly involved in communicative
interaction (Shortet al., 1976). The extent of social presence associated with a
communication medium is a function of the amount of communication channels available to transmit rich information. These channels include verbal cues, facial expressions, gaze, gestures, posture, physical proximity, and back-channeling cues. Facial expressions and gestures can augment the content of a message by expressing relevancy or
importance, urgency, validity, or the extent of agreement with, information that has been transmitted or received. Mutual gaze and back-channeling cues (e.g. utterances such as ’’mm’’ , ’’uhu’’ , ’’right’’ , ’’okay’’ , or their associated gestures) indicate attention, support, or acceptance of a speaker’ s
message. According to Shortet al. (1976), as the variety of these channels decrease in a communication medium, less attention is paid to the presence of other participants engaged in a communication session. A number of researchers have suggested that there is a continuum whereby communication media can be placed according to the degree to which they transmit the ’’social context cues’’ discussed above (Daft and Lengel, 1986; McGrath, 1990). Past research has identified a continuum ranging from e-mail, teleconferencing, computer-desktop
videoconferencing, and face-to-face meetings. E-mail is lowest in social presence in that it restricts physical presence and the
nonverbal cues, back-channeling cues, and physical presence. Social presence has been used to account for differences in task orientation, depersonalization,
communicative tone, and participation experienced by groups supported by electronic communication mediums. The lack of social context cues in
computer-mediated communication has been found to lead to increased negative
communicative tone that included assertive and hostile language and an increased sense of depersonalization (Siegelet al., 1986; Sproull and Kiesler, 1986). Fulk et al. (1987) noted that e-mail and computer-conferencing were perceived to be less ’’ warm’’ than face-to-face communication. In contrast, Hiltzet al. (1986) found that communication quantity and consensus was higher in face-to-face groups than in the computer-mediated communication groups. However, when comparing imposed synchronous
communication (i.e. face-to-face meetings) versus asynchronous communication (i.e. e-mail and/or computer-conferencing), the asynchronous groups experienced greater productivity (Ockeret al., 1996; Turoff et al., 1993). The asynchronous group members were able to engage in ’’ anytime, anywhere’’ communication. In other words, they were not temporally or spatially constrained. Finally, under time constraints and increased task interdependence, face-to-face groups experienced superior productivity as compared to computer-mediated groups (Straus and McGrath, 1994). These findings suggest that context dictates whether
technology-supported communication can enhance or diminish team task
accomplishment.
Media richness
Daft and Lengel (1986) noted that organizational members engage in
communication activities in order to reduce uncertainty and equivocality associated with the information requirements of their assigned tasks. Uncertainty reduction refers to the elimination of the lack of information needed to complete tasks. Under such conditions, facts are elicited to answer any questions regarding the specification of the task to be performed and the procedure appropriate for task execution. Equivocality reduction refers to reducing ambiguity associated with a task. Ambiguity could be the
result of multiple interpretations and
conflicting assessments of task requirements. Equivocality reduction requires a higher degree of information richness as compared to uncertainty reduction, in that, rather than being limited to the acquisition of facts, the required communication must facilitate an evaluation of alternatives and must foster a shared understanding and consensus regarding the problem statement and an appropriate solution. Kydd and Ferry (1994) identified a continuum on which a
communication medium can be placed based on the ’’richness’’ of information it was capable of transmitting. Information richness content was rated on the availability of immediate feedback, the number of verbal and nonverbal cues, back-channeling cues, socio-emotional communication, and interpersonal interaction associated with a communication medium. Uncertainty reduction is handled most efficiently utilizing ’’lean’’ (e.g. e-mail and computer-mediated communication) media where there are minimal socio-emotional cues present thereby limiting the transmission to factual
information. Equivocality reduction is best addressed utilizing ’’rich’’ media
(e.g. videoconferencing, face-to-face) where there can be the immediate exchange of information and supporting nonverbal and back-channeling cues. Media differ in media richness and consequently differ in the extent to which they could adequately support team collaborative problem solving and decision making.
The low immediacy of communication and lack of sufficient socio-emotional cues present in computer-mediated communication, as compared to face-to-face meetings, has been attributed to longer time to reach decisions and failure to reach consensus within the allotted time (Hiltzet al., 1986; Siegel et al., 1986). The inability of ’’ lean’’ media, such as computer-mediated communication, to convey socio-emotional content in messages was found to engender lower satisfaction with the problem-solving process, when compared to face-to-face groups. Lower productivity has also been associated with ’’lean’’ media support of work groups (Straus, 1996; Straus and McGrath, 1994).
Time, interaction, and performance
functions: production, group well-being, and member support. The production function refers to the purposeful activity performed by the group. Work groups are typically assigned a project that can be decomposed into tasks, which are further decomposed into sub-tasks. In execution of the production function, a project team must:
identify goals and select an initial performance strategy for project implementation;
devise techniques, procedures or algorithms to carry out the project; resolve conflicting preferences, values, and interests within the group; and engage in task performance.
Group well-being refers to activities aimed at maintaining an intact and continuing social structure. Activities associated with the well-being function include:
use of appropriate communication channels;
acquisition of role clarity; conflict resolution; and
conducting interpersonal interactions needed for task performance.
The member support function refers to inclusion of and cooperation with project team members. In the maintenance of member support, project team members will:
display acceptance and support of another team member’ s contribution; allow participation and status attainment of others; and
conduct conflict resolution through negotiation.
In summary, TIP theory suggests that project teams engage in both task-oriented activities and socio-emotional activities. McGrath (1990) noted that communication media could have both positive and negative effects on a work group’ s production, group well-being, and member support functions. As noted above, social presence and media richness of a communication medium can have a significant impact on uncertainty and equivocality reduction associated with a specific task production function. In addition to impacting task production, this study suggests that a communication medium’ s ability to reduce task uncertainty and equivocality may also affect a work group’ s ability to carry out its group well-being and member support functions.
Software development and project
success
Software development context
Managing large-scale software development projects requires the ability to adequately allocate technical and human resources in relatively complex organizational settings (Zmud, 1980). The division of labor on large software development efforts typically require the creation of specialist groups such as technical programming teams, configuration management, quality assurance, technical writing, systems engineering, hardware engineering, and integration/test (Crowston, 1997; Kraut and Streeter, 1995). These software development team members differ considerably in formal structure, training, cognitive orientation of members, career paths, and departmental missions (Rasch and Tosi, 1992). In addition, Symon (1998), noted that all of the activities performed during the software development process are interdependent (both task and goal
interdependent). These characteristics of software development project efforts give rise to problematic situations such as task and role ambiguity and inter-group and intra-group conflict. It has been suggested that conflict in software development may arise from
differing interests, goals and cultural aspects of:
software developers and users; different user departments; and individual departments and the organization as a whole (Tripp, 1991). Krasneret al. (1987) noted that large projects experienced communication breakdowns that were attributable to changes in people, conflicting cognitive orientation and goals, conflicting assessments of the problem space, technology and so forth. In fact, the three most salient problems were:
(1) the thin spread of application domain knowledge;
(2) fluctuating and conflicting requirements; and
(3) communication and coordination breakdowns.
adequately facilitate uncertainty reduction and, more importantly, equivocality
reduction. This would be essential in order to reach a consensus regarding the multiple interpretations of the software requirements held by members of the design team and clarify any existing instances of role ambiguity. Finally, the communication medium used must also allow the transmission of socio-emotional content needed to support factual information along with personal interaction needed to
implement conflict resolution when addressing differing interests, goals, and cognitive orientations.
Project success
Project success has been defined as a combination of two types of implementation outcomes; task outcomes and perceived psychosocial outcomes (Pinto and Pinto, 1990). Software project task outcomes typically refer to adherence to the estimated schedule and budget and optimal productivity in terms of delivered source code instructions per man-hour (Henderson and Lee, 1992). Dimensions of software quality such as user-friendliness and adherence to end-user specifications are also associated with software development task outcomes (Yeh, 1993). Team psychosocial outcomes refer to the evaluation of the degree of experienced friendliness and support, positive feelings associated with interactions, acquired knowledge and skills, enjoyment of participation, and sense of pride and value resulting from participation in project implementation process (Pinto and Pinto, 1990).
Research hypotheses
Empirical research on technological support for group work indicate mixed results regarding the hypothesis that communication technology support facilitates group process through enhancing group problem-solving capabilities, group interaction, and
satisfaction with the task process (Alaviet al., 1995; Chidabaram and Jones, 1993;
Nunamakeret al., 1991; Straus, 1996). However, in the majority of cases, face-to-face settings have been found to be superior because adequate transfer of rich information was needed for effective task completion and subsequent satisfaction with the task
execution and group interaction process. It is
suggested that software development, when characterized as a collaborative problem-solving process, is certainly dependent upon the group process experienced by work groups engaged in that activity. An adequate software development process must have communication mechanisms that facilitate uncertainty and equivocality reduction and socio-emotional communication. Figure 1 graphically depicts the theoretical framework used in the formulation of the research hypotheses.
Walzet al., (1993) noted that an effective software development group process involves knowledge acquisition and the sharing and integration of that knowledge. Software design teams must acquire knowledge regarding the problem domain, user requirements, and design approaches (Watersonet al., 1997). Consequently, the communication mechanisms used during systems analysis and design can impact software project outcome (i.e. task and psychosocial outcomes) depending on the extent of social presence and media richness associated with the communication medium.
Social presence theory suggests that social context cues such as mutual gaze, back-channel cues such as nods and utterances should facilitate: effective turn-taking or smooth information exchange; clarification of task assignments/efforts; and immediate feedback. This, in turn, leads to better team coordination and optimal productivity. In addition, the member support and group well-being functions, delineated in TIP theory, are more easily performed when there is reduced opportunity for the expression of assertive and hostile language and perceived depersonalization (normally associated with minimal social presence) during task
execution. This should give rise to increased satisfaction with interactions and process satisfaction. Alternatively, low social presence should give rise to disruptions in
Consequently it is hypothesized that: H1. Groups working in the face-to-face
setting will experience greater team productivity than in the
videoconferencing supported setting. H2. Groups working in the face-to-face
setting will experience greater perceived interaction quality than in the videoconferencing supported setting.
H3. Groups working in the face-to-face setting will experience greater group process satisfaction than in the videoconferencing supported setting.
Research methodology
Research designThree separate one-way analysis of variance (ANOVA) designs were used to test the hypotheses presented earlier (see Figure 1). As depicted in Figure 1, the single factor was comprised of two levels of communication media: face-to-face and videoconferencing. Each of the three dependent variables (i.e. team productivity, interaction quality, and process satisfaction) was analyzed separately. Although satisfaction with the software development process and perceived interaction quality were significantly
correlated (r = 0.417, p < 0.01) they were treated separately because they are conceptually distinct (see Table I). In addition, neither of the dependent variables were significantly correlated to the covariate, basic program design ability.
Subjects
In this study, subjects were drawn from a population of management information systems undergraduate students familiar with the Systems Development Life Cycle
approach to software design and knowledge of
the C++ structured programming language. The students were randomly assigned to a four-person design team. Each design team was configured into two dyads. For their participation, each design team was eligible to receive a $100 award for the highest team productivity score under each of the experimental conditions (face-to-face and videoconferencing).
Experimental manipulation
The experimental manipulations involved the use of a two-way audio-video conferencing system to link the two dyads that made up the design team (i.e. a ’’virtual’’ team) or a face-to-face setting to connect the two dyads. The two dyads were required to collaborate with each other throughout the entire execution of the task. In other words, each dyad or sub-team was responsible for the complete development and documentation of the software design. This restriction was posed in order to prevent each team from sub-dividing the tasks and working separately. Another intent of this restriction was to force communication among all group members in order to assure that their design decisions and subsequently developed design documents were consistent. Information acquisition, sharing, and integration was limited to verbal communication for both the
videoconferencing setting and the face-to-face setting; physical documents were not shared. In addition, to force knowledge sharing and Figure 1 Communication medium and software project success
Table I Pearson correlations
integration, each dyad was responsible for conveying a portion of the design
requirements intentionally not supplied to the other. In the face-to-face setting, a partition was used to eliminate visibility of the other dyad’ s design documentation. The partition used in the face-to-face setting did not obstruct eye contact or viewing of facial expressions.
Experimental task
The students were required to enhance the functionality of a hypothetical university information system. The design teams were required to construct design documentation that included: a hierarchy chart; list of function prototypes; and pseudocode for each function identified as part of a solution to the problem. The enhancement required the development of software modules that would determine athletic participation eligibility status and scholarship award amounts based on a student’ s classification, credit hours acquired, and grade point average. In addition, the system’ s enhancement was also required to provide queries regarding a student’ s athletic participation eligibility and scholarship award. The final enhancement requirement provided automation to parking fee calculations for a parking garage. The experimental task duration was three hours with a ten-minute break scheduled during the mid-point of the experimental session. During the break, there was no communication between the two dyads, only within the dyads.
Dependent measures
There were two classes of dependent measures: task outcome and psychosocial outcomes. The task outcome was a measure of team productivity. The team productivity measure was determined by assessing the completeness of the required design documentation. Team productivity was a combined score on the completeness of file design, specification of function prototypes, and pseudocode for each function. Credit was given only to design documentation that was consistent among both dyads. Any part of the solution not developed and documented by both dyads was ignored. The psychosocial measures were comprised of interaction quality and process satisfaction. These measures were assessed using questionnaire items adapted from Green and Taber (1980) (see Appendix). The interaction quality scale
elicited perceptions regarding the extent of negative opinions made by and received by individual group members and the degree of frustration with any of the other team members’ behavior. The process satisfaction scale obtained an assessment of the perceived fairness, understandibility, and satisfaction associated with the method in which task execution was conducted.
The reliabilities (Cronbach’ s alpha) for the interaction quality and process satisfaction scales were 0.65 and 0.67, respectively. A grade obtained from a typical management information systems course that presents basic skills in functional decomposition of a software design solution was used as a covariate. None of the dependent measures were significantly correlated with this covariate.
Results
The correlations presented in Table I indicate significant correlations among the three outcome variables interaction quality, process satisfaction, and team productivity. Multivariate analysis was not performed because these outcome variables do not share any common conceptual meaning (Stevens, 1992). Three separate ANOVAs were conducted in the test of the hypotheses. Table II shows the means and standard deviations for overall team productivity, interaction quality and process satisfaction. The results for the one-way ANOVAs for team
productivity, interaction quality, solution satisfaction, and process satisfaction are presented in Table III. The study results found support for bothH1 and H2. The results for H3 were not significant.
As predicted inH1, Table III indicates that there was a significant effect (F(1, 46) = 27.795, p < 0.01) for the impact of the communication medium on team
productivity. A rich communication medium (i.e. face-to-face) resulted in significantly higher team productivity (mean productivity = 15.50) than in the design teams supported by videoconferencing (mean productivity = 8.67). Table III indicates that face-to-face groups were, on average, nearly twice as productive as the videoconferencing groups.
quality H2. H2 predicted that the face-to-face groups would experience greater perceived interaction quality because of the opportunity to utilize increased verbal, nonverbal, and back-channeling cues to promote efficient turn-taking, immediate feedback, and confirmation of conceptual consensus. The face-to-face groups did experience a significantly higher degree of interaction quality (mean interaction quality = 12.83) as compared to the
videoconferencing supported groups (mean interaction quality = 10.88).
In contrast to the prediction inH3, Table III indicates that there was no significant effect (F(1, 46) = 1.369, p < 1.0) for the impact of the communication medium on group process satisfaction. However, a significance level of 0.120 does indicate an approach towards marginal significant differences between the videoconferencing and face-to-face groups. A rich and high social presence communication medium (i.e. face-to-face) resulted in a higher level of group process satisfaction (mean group process satisfaction = 11.50) than in the design teams supported by system videoconferencing (mean productivity = 10.42).
Discussion
Previous research on communications technology supported groups (e.g. computer-conferencing, videoconferencing) has shown increases in number of decision alternatives presented, greater equity in group member participation, and superior decision quality when compared to face-to-face groups. The common thread in these studies is the use of a decision-making task. The results from this study complements this research stream by
showing that task type (e.g. decision making or production task) may be a factor in
determining the impact of communication technology on distributed project teams or work groups. Software development has been characterized as a highly interdependent production-type task (i.e. creating and documenting a design using a hierarchy chart, writing function prototypes and psuedocode) (Olsonet al., 1992; Waterson et al., 1997). The results show that the videoconferencing-supported group performing a production-type task was less productive in a fixed period of time and in a distributed synchronous group setting, as compared to face-to-face groups. This suggests that productivity could decrease when computer-mediated communication is used to support dispersed project team members working on production-type tasks that require equivocality reduction and conceptual consensus. The superior team productivity experienced by the face-to-face groups suggest that multiple channels of communication (i.e. ’’rich’’ media) are needed to facilitate better information acquisition, sharing, and integration. This gives rise to a shared understanding of the software requirements and derivation of a consensual solution to the problem in a timely manner. Visual cues easily observable in face-to-face settings allow team members greater awareness of when others are attending to them and they subsequently use this feedback to determine appropriate behavior. System designers can often possess distinct models of a software systems’ structure and functionality (Waterson et al., 1997). Subgoals of these designers then result in competitive behaviors devised to influence the design according to their cognitive representation. It is concluded that the face-to-face group members were better able to immediately discuss conflicts and Table II Cell means and standard deviations
Team productivity Interaction quality Process satisfaction Communication medium Mean Std. dev. Mean Std. dev. Mean Std. dev. Videoconferencing 8.67 5.44 10.88 2.51 10.42 2.73 Face-to-face 15.50 3.28 12.83 2.06 11.50 1.93
Table III Communication medium ANOVA results
Dependent variable df MS F Sig. Observed power
problems directly with each other to make mutually beneficial agreements that
contributed to successful collaboration. The experimenter observed a tendency for the face-to-face groups to be more critical of
alternatives posed and fewer requests for clarification of ideas presented as compared to the videoconferencing groups. In addition, the videoconferencing groups engaged in less dialogue and appeared to have less confidence in the content and validity of information exchanged.
The greater interaction quality experienced by the to-face groups, suggests that face-to-face communication is better suited to the implementation of the group member support and group well-being functions delineated in McGrath’s TIP theory (1990). High social presence communication media apparently provided a setting where mutual eye gaze, nods, helped to reduce disruptions in communication and to engender feelings of personalization and message clarification that would encourage behaviors aimed at
maintaining a cooperative climate and feelings of group well-being. Finally, the experimenter observed a general tendency for the face-to-face groups to express disagreement with solution strategies posed, but yet these groups reported a greater sense of interaction quality than that of the videoconferencing group. This suggests that greater social presence reduce the tendency to openly display disagreements with, or negative opinions regarding, another’ s suggestions.
Conclusion
The purpose of this study was to assess the relative effectiveness of the use of
videoconferencing technology to support a distributed software design team. The results indicate that managers must find innovative ways to promote smoother information exchange and instill confidence when using videoconferencing technology in collaborative problem solving by virtual project team members (e.g. use of task facilitator). Despite the superiority of face-to-face meetings, the use of videoconferencing may offset the cost of lower productivity through reductions in travel time and costs. Future research should address the potential of various communication technologies (e.g. desktop videoconferencing, videoconferencing, etc.) in the construction of
’’ virtual teams’’ and ’’virtual organizations’’ . Such successful applications could afford invaluable collaborations and reduce the time to complete joint ventures and the associated costs in engaging in joint ventures. Future research should also address other contextual factors (e.g. task interdependence, task type, project phase) that might mitigate the impact of the communication medium used to support distributed work groups. Upon the identification of relevant technical and group process issues in communication media use, task execution strategy can be modified in order to overcome any shortcomings of videoconferencing technology, as compared to face-to-face settings. The use of
videoconferencing technology can potentially act as a new business process ’’enabler’’ through the formation of new organizational forms and new software product developments teams resulting from inter-organizational ’’ virtual’’ teams. These organizational forms will be able to access highly-skilled
professionals that have been traditionally inaccessible because of both temporal and physical boundaries.
References
Alavi, M., Wheeler, B.C. and Valacich, J.S. (1995), ``Using IT to reengineer business education: an exploratory investigation of collaborative telelearning’’, MIS Quarterly, Vol. 19 No. 3, pp. 293-312. Chidambaram, L. and Jones, B. (1993), ``Impact of
communication medium and computer support on group perceptions and performance: a comparison of face-to-face and dispersed meetings’’, MIS Quarterly, Vol. 17 No. 4, pp. 465-91.
Crowston, K. (1997), ``A coordination theory approach to organizational process’’,Organization Science,
Vol. 8 No. 2, pp. 157-75.
Daft, R.L. and Lengel, R.H. (1986), ``Organizational information requirements , media richness, and structural design’’, Management Science, Vol. 32 No. 5, pp. 554-71.
Fulk, J., Steinfield, C., Schmitz, J. and Power, G. (1987), ``A social information processing model of media use in organizations’’, Communications Research, Vol. 14 No. 5, pp. 529-52.
Geber, B. (1995), ``Virtual teams’’, Training, April, pp. 36-40.
Green, S.G. and Taber, T.D. (1980), ``The effects of social decision schemes on decision group process’’, Organizational Behavior and Human Decision Processes, Vol. 25 No. 1, pp. 97-106.
Henderson, J.C. and Lee, S. (1992), ``Managing I/S design teams: a control theories perspective’’,
Management Science, Vol. 18 No. 6, pp. 757-77. Hiltz, S.R., Johnson, K. and Turoff, M. (1986),
``Experiments in group decision making:
Communications Research, Vol. 13 No. 2, pp. 225-52.
Krasner, H., Curtis, B. and Iscoe, N. (1987), ``Communication breakdowns and boundary spanning activities on large programming projects’’, in Olson, G.M., Shepard, S. and Solloway, E. (Eds), Empirical Studies of Programmers: Second Workshop, Ablex Publishing Corporation, Norwood, NJ, pp. 47-64.
Kraut, R.E. and Streeter, L. (1995), ``Coordination in software development’’ ,Communications of the ACM, Vol. 38 No. 3, pp. 69-81.
Kydd, C.T. and Ferry, D.L. (1994), ``Managerial use of videoconferencing’’, Information & Management, Vol. 27 No. 6, pp. 369-75.
Laughlin, P.R. (1980), ``Social combination processes of cooperative, problem solving groups as verbal intellective tasks’’, in Fishbein, M. (Ed.), Progress in Social Psychology, Lawrence Erlbaum, Hillsdale, NJ, pp. 127-55.
McDonough, E.F., Kahn, K.B. and Griffin, A. (1999), ``Managing communication in global product development teams’’, IEEE Transactions on Engineering Management, Vol. 46 No. 4, pp. 375-86.
McGrath, J.E. (1990), ``Time matters in groups’’, in Galegher, J., Kraut, R.E. and Egido, C. (Eds), Intellectual teamwork: Social and Technological Foundations of Cooperative Work, Erlbaum, Hillsdale, NJ, pp. 23-61.
Nunamaker, J.F., Dennis, A.R., Valacich, J.S., Vogel, D.R. and George, J.F. (1991), ``Electronic meeting systems to support group work’’,Communications of the ACM, Vol. 34 No. 7, pp. 40-61.
O’Conaill, B., Whittaker, S. and Wilbur, S. (1993), ``Conversations over video conferences: an evaluation of the spoken aspects of video-mediated communication’’, Human-Computer Interaction, Vol. 8 No 4, pp. 389-428.
Ocker, R., Hiltz, S.R., Turoff, M. and Fjermestad, J. (1996), ``The effects of distributed group support and process structuring on software requirements development teams: results on creativity and quality’’,Journal of Management Information Systems, Vol. 12 No. 3, pp. 127-53.
Olson. G.M., Olson, J.S., Carter, M.R. and Storrosten, (1992), ``Small group design meetings: an analysis of collaboration’’,Human-Computer Interaction,
Vol. 7 No. 3, pp. 347-74.
Pinto, J.K. and Pinto, M.B. (1990), ``Project team communication and cross-functional cooperation in new program development’’, Journal of Product Innovation and Management, Vol. 7 No. 3, pp. 200-12.
Rasch, R. and Tosi, H. (1992), ``Factors affecting software developers’ performance: an integrated approach’’, MIS Quarterly, Vol. 16 No. 3, pp. 395-413. Siegel, J., Dubrovsky, V., Kiesler, S. and McGuire, T.W.
(1986), ``Group processes in computer-mediate d communication’’, Organizational Behavior and Human Decision Processes, Vol. 37 No. 2, pp. 157-87.
Short, J., Williams, E. and Christie, B. (1976), The Social Psychology of Telecommunication, John Wiley, New York, NY.
Sproull, L. and Kiesler, S. (1986), ``Reducing social context cues: electronic mail in organization
communication’’, Management Science, Vol. 32 No. 11, pp. 1492-512.
Stevens, J. (1992), Applied Multivariate Statistics for the Social Sciences, 2nd ed., Lawrence Erlbaum, Hillsdale, NJ, p. 151.
Straus, S.G. (1996), ``The effects of communication media and information distribution on participation and performance in computer-mediate d and face-to-face groups’’, Small Group Research, Vol. 27 No. 1, pp. 115-42.
Straus, S.G. and McGrath, J.E. (1994), ``Does medium matter? The interaction of task type and technology on group performance and member reactions’’,
Journal of Applied Psychology, Vol. 79 No. 1,
pp. 87-97.
Symon, G. (1998), ``The work of IT system developers in context: an organizational case study’’, Human-Computer Interaction, Vol. 13 No. 1, pp. 37-71. Tripp, R.S. (1991), ``Managing the political and cultural
aspects of large-scale MIS projects: a case study of participative systems development’’ , Information Resources Management Journal, Fall, pp. 2-13. Turoff, M., Hiltz, S.R., Bahgat, A.N.F. and Ajaz, R.R.,
(1993), ``Distributed group support systems’’, MIS Quarterly, Vol. 17 No. 4, pp. 399-417.
Walz, D.B., Elam, J.J. and Curtis, B. (1993), ``Inside a software design team: knowledge acquisition, sharing, and integration’’, Communications of the ACM, Vol. 36 No. 10, pp. 62-77.
Waterson, P.E., Clegg, C.W. and Axtell, C.M., (1997), ``The dynamics of work organization, knowledge, and technology during software development’’,Journal of Human-Computer Studies, Vol. 46 No. 1,
pp. 79-101.
Yeh, H. (1993), Software Process Quality, R.R. Donnelley & Sons.
Zmud, R.W. (1980), ``Management of large software development efforts’’, MIS Quarterly, Vol. 4 No. 2, pp. 45-55.
Appendix. Questionnaire items
Interaction qualityWhen working on this project, to what extent did you:
feel frustrated or tense about the other team members’ behavior;
express negative opinions about any project team member’ s behavior;
observe others express a negative opinion about your behavior.
Thegenuine participation items were rated on a five-point (1 5) Likert scale.
Group process satisfaction
How would you describe your team’ s software development process?
fair . . . unfair;
confusing . . . understandable; satisfying . . . unsatisfying.
