Team Processes in Virtual Knowledge Teams: The Effects of Reputation Signals and Network Density

Abstract

Virtual knowledge teams (VKTs) depend on team processes that facilitate expertise coordination. VKTs use technology to map expertise, and thereby address the lack of familiarity among members. Despite the considerable interest in studying expertise coordination in teams, expertise coordination in VKTs is less understood. Moreover, technology’s role in expertise coordination in VKTs and the team processes, including expertise coordination, has received limited attention. This study argues and shows—through an online experiment—that technology enables VKTs by: connecting individuals through network ties; helping individuals to locate expertise, providing reputation signals; improving interpersonal processes through enhanced ties and signals; and enabling better performance through improved ties, signals, and processes. It contributes to theory by providing insights into how IT enhances expertise coordination and performance of VKTs. It contributes to practice by providing insights into the reputation signals that broadcast team members’ expertise and the effects of technology on team processes.

Key words and phrases:

• event sequence analysis
• network ties
• process research
• reputation signals
• team expertise
• team performance
• team processes
• virtual knowledge teams
• virtual teams

Notes

1 Appendix A summarizes the extensive literature, including under the banner of transactive memory systems.

2 More specifically, Marks et al. [50] use “temporal” order. Both “temporal” order and “sequential” order focus on the chronological order of events. We do not explicitly examine “time,” and hence use the term “sequential order.”

3 To avoid tautology, process quality and team performance should have distinct measures [34], as in this study.

4 The association between team quality and team performance has been previously proposed and found (e.g., [34]). But prior studies have not viewed team quality as a team’s following events in a specific order. Most prior studies find a positive association between following certain events (e.g., communication) and team performance, but overlook the order of events. We thank an anonymous reviewer for highlighting this aspect.

5 Amazon’s Mechanical Turk is a popular market where subjects participate in online surveys and are compensated financially. It has been used in prior IS studies (e.g., [31, 33, 59]). Subjects recruited from Mechanical Turk have been found to outperform student subjects and produce results comparable to real-world professionals [73].

6 Table E1 in Appendix E presents the correlation coefficients of individual scores in each of these subsections.

7 We tried to maintain the level of subject competence equal across teams, but the VKTs differed in performance of their members on the first task. So, we control for the overall performance of team members in the first task. Also, the deliberate assignment of subjects to teams based on task competence makes the study a quasi-experiment.

8 Execute tasks “involve the actual performance or execution of operations to achieve a group goal” ([43], p. 1258).

9 Marks et al. [50] contend that transition processes occur at the start of a collective entity’s formation and are followed by action processes, with interpersonal processes in both phases.

10 We refer to an activity in a team as an event.

11 As an exception, disagreement (an information action) and ignoring reputation signal (a signal response action) were considered distant (with substitution costs of 0.9) from other actions in their own broad categories, because they are both negative, in contrast to the positive nature of all other actions within their categories.

12 The centroid sequence in a situation is an actual sequence with the lowest mean distance from others in that situation, typical sequence is a hypothetical sequence that best represents the sequences in that situation, and ideal sequence is the typical sequence for the top 10 percent of performing sequences.

13 We estimated SUR including only Models 4 and 7; the significance levels and signs of coefficients, including mediational paths, were unchanged. The results are available from the authors upon request.

14 Coordination events also happen in teams with aggregated signals, but not commonly enough to be reflected in the typical sequence.

15 To empirically evaluate this assessment, we used a four-item scale (see Table E2 in Appendix E) to measure cognition-based trust at both start and end of the task, and computed the proportional increase in it from the start of the task to the end of the task. This proportional increase had a strong negative correlation (−0.748, p < 0.001) with the ratio of the number of teammate monitoring to the number of total events in a sequence, indicating that greater teammate monitoring relates to lack of trust in teammates. We thank an anonymous reviewer for suggesting a test that ensures teammate monitoring is indeed a close correlate of lack of trust.

16 The studies are collected from Web of Science, through search for the keywords “expertise coordination” OR “transactive memory system” AND “team”. Only English journal articles were selected to create this review table.

17 The Cronbach alpha of the items is 0.92. We averaged the scores for each item for each individual and calculated individual scores of trust. Then, we averaged the individual scores of trust in each team to estimate the team’s baseline trust. Table F2 in Appendix F presents the items adopted from Kanawattanachi and Yoo [4].

Additional information

Notes on contributors

Taha Havakhor

Taha Havakhor ([email protected]; corresponding author) is an assistant professor of management science and information systems in the Spears School of Business at Oklahoma State University. His research interests include social media platforms, value creation with information technology (IT), and IT governance. He serves on the editorial board of IEEE Transactions on Engineering Management, and his work has been published in venues such as Information Systems Journal.

Rajiv Sabherwal

Rajiv Sabherwal ([email protected]) is department chair (Information Systems) and Edwin and Karlee Bradberry Chair in the Walton College of Business at the University of Arkansas. His research interests include information systems planning and success, and utilization of emergent technologies to manage information and knowledge. He has published over 50 refereed papers in journals such as MIS Quarterly, Information Systems Research, and Management Science. He is editor in chief of IEEE Transactions on Engineering Management, senior editor of Journal of the Association for Information Systems, and a fellow of the Association for Information Systems.