You are on page 1of 32

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/313592169

Trust Development in Globally Distributed Collaboration: A Case of U.S. and


Chinese Mixed Teams

Article  in  Journal of Management Information Systems · February 2017


DOI: 10.1080/07421222.2016.1267521

CITATIONS READS

49 824

3 authors:

Xusen Cheng Shixuan Fu


Renmin University of China Beijing International Studies University
96 PUBLICATIONS   1,180 CITATIONS    23 PUBLICATIONS   501 CITATIONS   

SEE PROFILE SEE PROFILE

Douglas A. Druckenmiller
Western Illinois University
24 PUBLICATIONS   195 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Future of Collaboration View project

Collaboration Engineering Definition View project

All content following this page was uploaded by Xusen Cheng on 23 January 2018.

The user has requested enhancement of the downloaded file.


Journal of Management Information Systems

ISSN: 0742-1222 (Print) 1557-928X (Online) Journal homepage: http://www.tandfonline.com/loi/mmis20

Trust Development in Globally Distributed


Collaboration: A Case of U.S. and Chinese Mixed
Teams

Xusen Cheng, Shixuan Fu & Douglas Druckenmiller

To cite this article: Xusen Cheng, Shixuan Fu & Douglas Druckenmiller (2016) Trust Development
in Globally Distributed Collaboration: A Case of U.S. and Chinese Mixed Teams, Journal of
Management Information Systems, 33:4, 978-1007, DOI: 10.1080/07421222.2016.1267521

To link to this article: http://dx.doi.org/10.1080/07421222.2016.1267521

Published online: 10 Feb 2017.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=mmis20

Download by: [University of South Florida] Date: 10 February 2017, At: 13:03
Trust Development in Globally Distributed
Collaboration: A Case of U.S. and Chinese
Mixed Teams
XUSEN CHENG, SHIXUAN FU, AND DOUGLAS DRUCKENMILLER

XUSEN CHENG (xusen.cheng@gmail.com; corresponding author) is an associate professor


and chair of the E-Business Department in the School of Information Technology and
Management at the University of International Business and Economics, Beijing, China.
He received his Ph.D. in informatics at the University of Manchester. His research focuses
on trust development in virtual teams and online communities, collaboration processes and
system design, and the integration of behavior and design issues in information systems.
His articles have appeared in Information Technology and People, Group Decision and
Negotiation, Computers in Human Behavior, and others. His work has also been presented
at leading conferences, including the International Conference of Information Systems
(ICIS) and Hawaii International Conference on System Sciences (HICSS).

SHIXUAN FU (shixuanfu0612@gmail.com) is a Ph.D. candidate in the School of


Information Technology and Management at the University of International
Business and Economics. Her interests focus on user behaviors in online commu-
nities, sharing economy, team collaboration and trust. Her publications have
appeared in Computers in Human Behavior, Information Technology and People,
and several conference proceedings, such as HICSS.

DOUGLAS DRUCKENMILLER (DA-Druckenmiller@wiu.edu) is a professor of informa-


tion systems in the School of Computer Sciences at Western Illinois University. He
received his Ph.D. in management information systems from Kent State University.
His research focuses on collaborative systems and virtual teaming technologies. He
has published in the Journal of the Association of Information Systems, International
Journal of Intelligence, Technology and Planning, FUTURES, and others. His
research has also been presented at leading conferences, such as Group Decision
and Negotiation (GDN) and HICSS.

ABSTRACT: Trust is frequently investigated as an indicator of a mutual relationship. Trust


is especially important for globally distributed collaboration in light of the lack of face-to-
face interactions. As the perception of trust is a dynamic process, however, little research is
conducted measuring trust development. Whether facilitation intervention is beneficial for
trust development is also unknown. In order to fill the research gaps, we followed a design
science approach and incorporated collaboration engineering into the design of the
treatment. Data were collected in a series of experiments with Chinese and U.S. mixed
teams, including a longitudinal survey, interviews, and documentation. Through the
comparison of the treatment group with the control group, we found that trust was
significantly improved in the treatment group. In addition, several trust antecedents
were found to explain the development. The power of facilitated collaboration is also

Journal of Management Information Systems / 2016, Vol. 33, No. 4, pp. 978–1007.
Copyright © Taylor & Francis Group, LLC
ISSN 0742–1222 (print) / ISSN 1557–928X (online)
DOI: 10.1080/07421222.2016.1267521
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 979

validated as helpful for trust development. This research makes several implications, such
as proposing a series of trust antecedents, a treatment design of a collaboration engineering
(CE) approach for trust improvement, and a new context application of CE. This research
could also help the practitioners in globally distributed collaboration who want to improve
trust over time.

KEY WORDS AND PHRASES: collaboration engineering, design science research,


globally distributed collaboration, thinkLets, trust development.

With the rapid development of information technology (IT) for communication and
online collaboration, many organizations, including schools and companies, choose
collaborative work in distributed teams to improve work efficiency [4]. Trust is widely
defined as the willingness to be vulnerable and having confident expectations of the
other party [45, 48, 62]. As one of the factors influencing collaboration, trust has played
an increasingly significant role in teams. Distributed teams need the support of informa-
tion technology [13], and also the level of trust in the team can be key to the team’s
success. If team members lack trust in collaboration, they will waste more time and have
to make a greater effort to collaborate [42, 58]. On the other hand, global virtual teams
are geographically distributed and culturally diversified, which calls for a higher level of
trust because of the absence of face to face interaction [27, 32].
Trust has been discussed frequently in distributed research settings, where a high level
of trust is positively correlated with team performance [17, 19, 65], group work
efficiency and effectiveness [56], team member satisfaction [25], and knowledge-shar-
ing behavior [71]. However, all this research treats trust as a state of mind and ignores
the process nature of trust dynamics. To follow the dynamics of collaboration behavior
through interactions, trust development is a great indicator.
Some of the studies that treat trust as a process focus on the initial stage of trust
relationships, such as trust-building mechanisms and antecedents [28, 37, 44], but they
do not address the measurement of dynamic trust, especially the longitudinal study of
trust development. Some report on trust dynamics over time in virtual teams and face-to-
face teams [13, 75], but that still leaves open the question of whether facilitation
intervention can improve trust, and the issue of how intervention principles help trust
improvement is still unsolved [20, 21, 51]. It would be useful to know more about trust
development over time (e.g., the role of facilitation intervention on trust development),
since trust is proposed to be intertwined with the relationship development process [32].
Facilitation intervention with a collaboration engineering (CE) approach provides
comprehensive guidelines covering six collaboration patterns [6]. Collaboration engi-
neering can be considered as a combination of facilitation, models, and repeatable
process deployment that aims to design a process for recurring high-value collaborative
tasks using facilitation techniques and technology [5, 40]. This research is motivated by
the intervention of a collaboration engineering process, and aims to explore the influ-
ence of facilitated collaboration on trust in globally distributed teams. This study
980 CHENG, FU, AND DRUCKENMILLER

contributes toward a complete understanding of trust development by comparing


differences in trust dynamics in facilitated collaboration teams and traditional teams.
Thus, we set out to investigate the following questions: (1) Does the development
of trust behave differently in conventional collaborative teams and process-guided
teams (control group and treatment group)? (2) What are the trust antecedents in
longitudinal globally distributed collaboration? (3) How has trust development been
influenced by the use of facilitated collaboration in globally distributed collabora-
tion? In order to answer these three questions, we followed a design science research
approach to design the facilitated collaboration process. Then we conducted experi-
ments in U.S. and Chinese mixed distributed teams to test the role of facilitation
intervention in trust development.

Research Background
Understanding Trust
Trust is a frequently investigated construct in the issue of mutual relationships.
Especially in online settings, lack of trust is often blamed for individuals’ lack of interest
in distributed interactions [37]. A factor-analytic study proposed by Hoy and
Tschannen-Moran [31] demonstrated that different aspects converge to form a coherent
construct of trust. A considerable amount of literature has been published on trust
dimensions [9, 15, 44, 52]. However, due to the diverse research perspectives and
disciplines, there is no unified agreement on the definition of trust, trust antecedents,
and consequences. Generally, there are two widely accepted definitions of trust from
different perspectives. On the one hand, trust is regarded as the willingness to take risks
and the acceptance of vulnerability [48]. On the other hand, trust is mostly related to
confidence and positive expectation of another party [62]. That is to say, perceived risk
and perceived benefit are two important streams that constitute trust [34, 49, 64].
Perceived risk is the recognized possibility of hazard or harm [18, 69]. In the context
of online teamwork, risk is an intentional interaction undertaken with uncertainty [53].
The mostly widely accepted definition of trust also refers to vulnerability to taking a risk
[48]. Perceptions of risk likelihood and the impact of risk play a role in the observed
variations in trust. While trust refers to an assessment of the likelihood of desirable
behaviors in another party, risk mostly shows the assessed probability of not achieving
satisfactory results. Therefore, the lower the risk perception, the higher the trust [9].
Trust is regarded as an expectation of a positive (or, at least, nonnegative) outcome
in uncertain circumstances [62]. Previous studies have also highlighted ability,
confidence, assurance and integrity as antecedents of trust [3, 33, 44, 59, 63]. In
teamwork, trust is the reciprocal outcome of trust propensity that is related to reliable
behaviors [74]. We incorporate these findings into the concept of perceived benefit.
Perceived benefit is defined as belief in the extent to which a person will be better
off as a result of interactions [26, 36], and can be regarded as the positive expecta-
tion and mutual confidence in the outcome due to other parties’ ability, benevolence,
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 981

and integrity [10]. The higher the value of perceived benefit, the higher the trust
perception [9].
From the analysis above, the dimensions of trust come down to a lower level of
risk and a higher level of perceived benefit. In other words, a low level of perceived
risk and a high level of estimated benefits contribute to the improvement of trust [38,
45, 50]. In our understanding, in distributed team collaboration, trust can be defined
as team members’ perceived likelihood of risk and the assessment of benefit in a
collaborative relationship. Therefore, based on the trust dimensions and relative trust
studies, we aim to investigate trust from the perspective of perceived risk and
perceived benefit.

Trust Development
The understanding of trust is a dynamic process grounded in long-term interactions. Many
studies have focused on the development of trust types as ongoing collaboration—for
example, some research has found that cognition-based trust is salient in the initial stage
and knowledge-based trust plays a dominant role in the later stages [41]. Others have
investigated the dynamics of trust from the perspective of trust-type transformation, from
individual-level trust to interpersonal trust and organizational trust [35, 66]. Moreover, the
dynamic account of trust is also investigated from the perspective of different approaches,
such as the behavioral approach and unidimensional, two-dimensional, and transforma-
tional approaches [43, 75].
Specifically, in online distributed collaborations, trust also evolves with the dynamics
of mutual relationships. Some researchers find trust development of importance and
and use it as a foundation for their research [12, 43, 54]. Generally, these literatures are
composed of three directions: first, the comparison of trust development trends
between computer-mediated teams and face-to-face teams [76]; second, approaches
to manage dynamic trust relationships in implementation projects [61]; and third, the
development of certain trust factors in distributed virtual collaboration [13].
Together, these studies provide important insights into the issue of trust develop-
ment. However, few studies have been conducted on the influence of facilitation
intervention, especially the IT-enabled collaboration process of trust development. In
team collaboration studies, teams using collaboration technologies and process
guidance may produce higher-quality ideas than individuals who do not use such
support [21, 55]. Therefore, we aim to investigate the influence after careful review
of the facilitated collaboration interventions.

Facilitated Collaboration Using thinkLets


Collaboration engineering (CE) is an approach that has a practical focus on task-
specific work practices. Generally, three roles exist in the theory of collaboration
engineering: the collaboration engineer, the facilitator, and the practitioner [39].
Practitioners are the task executors who accomplish the predictable collaboration
982 CHENG, FU, AND DRUCKENMILLER

deliverables with their special skills. The facilitators generally enable efficient group
collaboration, and skilled facilitators can often be very expensive [1], while colla-
boration engineers design the collaboration process to enable in-group members to
collaborate efficiently through facilitation intervention. In other words, CE provides
the possibility for nonexperts to realize their potential to execute a well-designed
collaboration process after quick training [6]. For each collaboration group, one of
the in-group members can be selected as the facilitator based on the well-designed
collaboration process. The collaboration engineer should coordinate with practi-
tioners to design the facilitation intervention process that users find useful.
A collaboration process is composed of a series of thinkLets. ThinkLets are the
packaged facilitation interventions that serve as the building blocks for a repeatable
collaboration process [5], and specify the rules and technology affordances that group
members work with to execute their tasks [22]. Through the five iterated steps of
collaboration process design, task diagnosis, activity decomposition, task-thinkLet
choice, agenda building, and design validation, a task-related collaboration process is
designed that can be executed with the help of group support systems [39]. There are
several thinkLets, such as Brainstorming, PopcornSort, StrawPoll, and FastFocus. CE
identifies six collaboration patterns, including generation, reduction, clarification, orga-
nization, evaluation, and building a consensus [7]. A certain thinkLet corresponds to the
patterns of collaboration.
Consequently, thinkLets modules and collaboration processes are considered to be
beneficial for the improvement of trust antecedents over time, and the trust antecedents
in turn lead to trust development from the perspective of risk and benefit. For example,
it has been observed that improving the quality of collaborative work through
thinkLets leads to better knowledge sharing [2, 30], and knowledge-sharing behavior
is positively correlated with reciprocal benefit and trust [30]. Meanwhile, cognitive
costs can be reduced and, as a result, group productivity can be improved according to
an IT-enabled CE process [5]. An improvement in group productivity may lead to an
increase in benefit that finally facilitates trust [78]. In the practical project of risk
management, the CE process also plays a positive role in mitigating risks [40].
Moreover, other researchers have found that communication, coordination and moti-
vation are all improved after executing the idea of creating a framework of thinkLets
[46]. More specifically, members’ activeness and passion in participation may be
improved after the adoption of a CE process [14], while research has studied the
positive relationships among communication, motivation, participation, and trust [32].

Research Framework
Our literature review reveals that the CE intervention has been widely investigated in
terms of team collaboration outcomes. We synthesize a number of trust dimensions
that have been identified in previous literature from the perspective of risk and
benefit. Combining these two aspects, we propose a research framework that
illustrates the influencing path of facilitated collaboration to collaboration outcomes.
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 983

Longitudinal influence

Treatment

ThinkLets Risk
modules Trust consequences
Trust
Trust
antecedents Satisfaction
Collaboration Benefit
process

Figure 1. Research Framework

Figure 1 presents our research framework. In summary, we argue that repeatable


facilitated processes using thinkLets are beneficial for the improvement of several
trust antecedents that help minimize risk and maximize benefit. As the collaboration
progresses, the development of trust will ultimately have a positive effect on the
collaborative performance of the distributed team.

Research Method and Data Collection


Our study serves as a design science research initiative founded on the logic of
exploratory research. Following the guidelines for design science in information
systems research [73], a facilitated-intervention process was designed to help with
trust development in team collaboration, satisfying the deliverables. To measure the
influence of facilitated collaboration principles on trust development in global virtual
collaboration, this study conducted a series of experiments targeting the case of U.S.
and Chinese mixed teams. Students in a Chinese university and a U.S. university
students took part in the experiments. Overall, we conducted a longitudinal experi-
ment with repeated investigations of the same participants over three stages of
collaboration. During the experiment, one group is the treatment group, and the
other is the control group. The treatment group was instructed to use well-designed
software based on principles of collaboration engineering. The control group
received no guidance in any IT-enabled collaborative process.

Participants
A total of 172 undergraduate students participated in our case experiment. They
received course credits for participating. The students were divided into two separate
case groups, 84 students in the control group and 88 students in the treatment group.
Both groups’ courses lasted nine weeks. The participants’ average age was 21.9 in
the control group and 21.8 in the treatment group. There was no significant
difference between the two groups in terms of age, educational background, or
gender composition.
984 CHENG, FU, AND DRUCKENMILLER

In the control group, 46 students were from a university in China and 38 students
from a university in the United States in the form of a distance education course.
They were divided randomly into nine hybrid subgroups to complete a team project
over nine weeks. Each subgroup consisted of four or five Chinese students and three
or four U.S. students. For the treatment group, a total of 88 students took part in the
experiment. Among those, 39 were from a U.S. university and 49 in a Chinese
university. Also, we randomly divided these students into nine subgroups. The
principle when grouping the students was to ensure the same percentage of U.S.
students and Chinese students within each group.

Procedures
Over the nine weeks, the students (both the control group and the treatment group)
were asked to collaborate to work on a cross-border e-business website analysis. The
task was exactly the same for the control group and the treatment group. Students
collaborated in the form of subgroups in both cases to simulate consulting teams.
Group decision making was required to explore international logistics management
problems in the e-business and finally to reach decisions about solutions and
recommendations for solving the problems. The two cases (treatment group and
control group) were conducted in two consecutive semesters in sequence. We first
conducted the experiment with the control group, then conducted another case
experiment with the treatment group in the following semester. Despite the different
course time, the two groups shared certain similarities in terms of the teachers,
course content, and collaboration task.
In the experiment, in both groups, Chinese students were allowed to interact face-
to-face with their Chinese teammates but students in the U.S. university were not
able to collaborate with them face-to-face due to the restrictions of the online course
environment. Instead, distributed students used several software tools to support
their collaboration, such as WeChat (group chatting software), e-mail, and Skype.
The difference between the treatment group and the control group lay in the process
control and collaboration tools. The control group received no process guidance: the
collaboration process was disorganized, with the teams using only WeChat and
Skype. For the treatment group in our case experiment, one student was selected
from each of the nine collaboration subgroups to be the facilitator of the teamwork.
We demonstrated the facilitated collaboration process to the facilitators and
instructed them to use the Discussion system, a self-developed professional colla-
boration tool designed according to the principles of CE. The facilitator was
responsible for being the timekeeper and group leader of the collaboration. The
participants used their own computers to register the Discussion system. Once
logged in, guidelines were provided on use of the system. The distributed team
members in the treatment group were asked to use the system to collaborate and
work toward their deliverables to solve the problems the company faced.
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 985

Design of Treatment
In this section, we aim to elaborate the CE process design for this study. We
developed a collaboration process to support participants’ collaborative tasks. The
process design was based on the general principles proposed by Kolfschoten and
Vreede [40]. Based on the tasks, an example of the collaboration process can be seen
in Figure 2. We labeled the sequence number, pattern of collaboration, activity name,
and time setting for each thinkLet. The facilitation process model (FPM) links the
collaboration agenda and thinkLets, and contains the step number, the activity, the
thinkLet name, the pattern of collaboration, and the duration of each activity [6, 22].
In Figure 2, the FPM sequence demonstrates the collaboration process. For each
FPM, the sequence number is the order in the collaboration process, which is indicated
in the left-upper field, while the pattern of collaboration is given in the lower-left
corner. In the lower-right corner, we explain the activity name of the particular
thinkLet that conveys what is supposed to be achieved in this step. In the upper-
right corner, we add the required time for each thinkLet module. During the execution
of the collaboration, the countdown function was used to ensure the duration of each
thinkLet module. Because the process can be used repeatedly, we merely designed the
duration instead of the exact clock time. Facilitators and group members could set the
starting time for each task themselves.

End
Begin

1 Free
15 min. 6 Crowbar 5 min. 6 PointCounterpoint 5 min.
brainstorming
To find common ground
consensus

consensus
Generate

Team members To output a prioritized


Build

Build

between polarized
brainstorm their list of items and help
factions in a debatable
ideas freely. members know others.
situation.
When the group has
some unshared In a badly
opinions conflicted group
2 FastFocus 5 min. 5 StrawPoll 5 min.
To remove some To measure the
Evaluate
Reduce

invaluable ideas consensus within a


and pick up some group and rank the
reasonable ones. ideas.

3 PopcornSort 5 min. 4 Bucketwalk 5 min.

To validate the
Organize

Evaluate

convergence and To validate the results


quickly categorize of a PopcornSort.
ideas.

Figure 2. An Example of a Facilitated Collaboration Process Using thinkLets


986 CHENG, FU, AND DRUCKENMILLER

The participants in the treatment group were asked to use the collaboration tool,
the Discussion platform, designed according to the thinkLets–enabled collaboration
process. The Discussion platform was self-developed to be available in Windows,
Apache, MySQL, and PHP (WAMP) environments so that each participant could use
his or her personal computer as the server [11]. During the collaboration of every
subgroup in the treatment group, each team selected one member to be the facilitator.
The facilitators were trained in terms of what to monitor and what instructions to
give in terms of collaboration. Moreover, the collaboration process is intended to be
used frequently, as CE provides the possibility of designing a repeatable process that
can be used for recurring tasks [5, 7] or within the task.

Data Collection
We conducted the study by having participants in different countries interact and
collaborate with one another and then measuring the data at the individual level over
time. Quantitative data collection (a survey) showed some findings concerning the
general trust development trend and the comparison between different groups.
However, statistical results among variables alone cannot uncover the evidence of causal
relationships. Furthermore, the mechanism of why trust develops cannot be solved with
survey data alone. Qualitative data (e.g., interviews and documentation) are better able
to describe a phenomenon and identify the inner structures. Therefore, we adopted a
mixed data collection method.
Specifically, in the data collection for the experiment, we used three approaches:
survey, interview, and documentation. Triangulation of data sources provides greater
accuracy and improves the robustness of the research results [72, 77].

Survey
We used two separate questionnaires (one to measure trust, the other to measure the
collaboration process) designed by prior researchers to conduct our survey investigation
[7, 12, 33]. For the questionnaire of trust adapted from Cheng et al. [12] and Jarvenpaa
and Leidner [33], we measured three constructs related to trust: risk, benefit, and overall
level of trust. For the questionnaire on the facilitated collaboration process adopted from
Briggs et al. [7], four constructs were measured, including satisfaction with process
(SP), satisfaction with outcome (SO), tool difficulty (TOOLDIF), and process difficulty
(PROCDIF). All constructs were measured through multiple items. The items were
presented in the form of several statements. For each item, a five-point Likert scale was
used with anchors from 1 to 5. Among them, 1 represented the strongest disagreement
and 5 the strongest agreement.
For both groups, we administered the trust questionnaire three times to track the
development of trust over the whole project: the initial stage, middle stage, and final
stage. At the initial stage, we delivered the questionnaires to the participants as soon
as we divided them into several collaboration groups. The majority of the students
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 987

Table 1. Testing for Nonresponse Bias

Test for equality of two Collaboration


groups of respondents Gender experience
Respondents who completed Wilks’s lambda = 0.98 0.45 2.43
questionnaires at all three
stages
Respondents who did not 0.54 2.29
complete the latter two survey
stages
F = 0.51 t = –0.83 t = 0.72

had no prior collaboration experience. We allocated the collaboration task to the


participants and collected data when the participants finished their initial collabora-
tion interaction. We then asked the students to continually collaborate within the
group and, in the fifth week of the case—namely, the second stage—we asked the
students to complete the same questionnaires based on their four-week collaboration
experience. Later, in the final stage, we conducted follow-up surveys with the same
participants in the ninth week before the course finished.
We delivered trust questionnaires three times in the control group in total. For the
treatment group, since a facilitated collaboration process was used, we not only
delivered trust questionnaires three times but also delivered the collaboration process
questionnaire in the final stage of their team collaboration.
All the participants completed the questionnaires in the initial stage. However, in
the second stage, the response rate decreased to 94 percent in the control group and
91 percent in the treatment group. Later, in the final stage, the response rate was
even lower, with 88 percent responding in the control group and 87 percent in the
treatment group. In order to test the nonresponse bias in our research, we conducted
a comparison of participant characteristics between those who failed to deliver
questionnaires in the latter two stages and those who finished all the questionnaires
at the three stages. The results indicated no significant differences in terms of
demographic variables of gender and collaboration experience: thus nonresponse
bias does not exist in this study (see Table 1).

Interview
At the end of the project, we successfully interviewed 22 students in the control
group and 25 students in the treatment group individually. Interview protocols were
designed based on our research models, and are pertinent themes in trust develop-
ment and collaboration literature [5, 36]. Specifically, we mainly asked the infor-
mants about the dynamic mechanisms of several interesting findings from the survey
results. The interview questions took the form of four parts. First, we inquired about
the demographic features of the participants and their roles in the team collaboration.
988 CHENG, FU, AND DRUCKENMILLER

Second, we asked about their collaboration experiences and how they conducted the
deliverables. Third, we asked about their perceived trust dynamics, perceived risk
and benefits through team collaboration. Fourth, we conducted semistructured inter-
views with open-ended questions. The interviews were audio recorded with the
agreement of the interview participants.
For each collaboration group in both cases, two or three members from each group
were asked to participate in the interviews. Each interview took between half an hour
and an hour, depending on the participants’ responses. The interviews were con-
ducted by two of our authors together. One interviewer asked the general questions
according to the interview transcripts while the other raised questions about the
incidental findings. Once recorded, these interviews were transformed into text
transcripts as soon as possible. Clarifications and follow-up questions were asked
if any of the text transcripts were thought to be ambiguous.
We addressed potential informant bias in several ways. First, participants in each
collaboration group were involved in the interviews, including the facilitator in the
group collaboration and other team members. Second, open-ended questions were
involved in the interview transcripts, which help limit recall bias [47]. Third,
participants only received course credits for taking part. Their names and group
numbers were coded and we gave the interviewees anonymity so that they could
speak freely. Fourth, the interviewers (two of our authors) were trained profession-
ally to prompt interviewees to provide more details about their distributed group
collaboration but to avoid leading questions and apparent answers.

Documentation
Documentation was gathered mostly from the U.S. students concerning their overall
feelings about the distributed collaboration of culturally mixed groups. The docu-
mentation was about lessons learned from the mixed distributed collaboration and
best practices in accomplishing the project. We collected the documentation at the
end of the project in the form of a home assignment. The documentation serves as a
supplement to the qualitative interview data [23].

Data Analysis
This study used a multimethod approach to collect data. Through statistical analysis
of longitudinal surveys, we evaluated the general changing trend in trust develop-
ment over time. Through qualitative analysis of interview data, we further investi-
gated the in-depth mechanisms from the perspective of influencing factors. In this
study, the multimethod approach minimized the threat of monomethod variance,
enabling us to investigate the mechanism of the trust development trend and the
effect of facilitated collaboration on trust dynamics and group performance.
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 989

Table 2. Reliability and Validity Results of the Survey

Spearman–
Brown Composite
Variables coefficient reliability (CR) AVE
Control Risk (RA) 0.866 0.782 0.701
group Benefit (BA) 0.831 0.731 0.694
Overall trust (TA) 0.745 0.883 0.675
Treatment Risk (RB) 0.855 0.737 0.604
group Benefit (BB) 0.842 0.807 0.660
Overall trust (TB) 0.730 0.732 0.596
Evaluation Satisfaction with process (SP) 0.874 0.912 0.691
of Satisfaction with outcome (SO) 0.853 0.835 0.762
process Tool difficulty (TOOLDIF) 0.886 0.841 0.729
Process difficulty (PROCDIF) 0.821 0.882 0.689

Quantitative Results
Reliability and Validity of the Instrument

We adopted reliability tests to measure the inter-item consistency of both questionnaires.


The Spearman–Brown coefficients of each construct are shown in Table 2. Moreover, we
used exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) to check
the convergent validity of the constructs measured in the questionnaires. The rotated factor
matrix results through SPSS showed that the measurement items for constructs in our
questionnaires were acceptable: all the items’ factor loadings are over 0.5 on the corre-
sponding constructs. Furthermore, we computed the average variance extracted (AVE) for
each construct. The CFA results showed that every indicator loads significantly (p < 0.01)
on the expected construct and that all loadings are above 0.5.
Overall, the results indicate that all the values of the constructs were greater than
the thresholds, indicating that the scales were acceptable for this research. For
detailed results, see Table 2.

Control Group

All the U.S. and Chinese participants were asked to complete questionnaires at three
stages during the project, the initial, the middle and the end. We calculated the arithmetic
mean value of all the variables of each group at the three stages, and then calculated mean
values of all the participants.
Table 3 shows the trust values at the three stages of the collaboration project in the
control group. We can see that the average level of risk increased from 2.03 to 2.24.
As for the mean values of benefit, we find that benefits decrease from Stage 1 to
Stage 2, and increase from Stage 2 to Stage 3. For the overall level of trust, we find
that the mean value of trust continued to decrease from Stage 1 to Stage 3.
990 CHENG, FU, AND DRUCKENMILLER

Table 3. Descriptive Results of Trust Development in Control Group

Median Mean SD
(Stage 1→2→3) (Stage 1→2→3) (Stage 1→2→3)
Risk 2.26→2.26→2.25 2.03→2.13→2.24 0.60→0.64→0.67
Benefit 4.00→4.00→4.17 4.08→3.92→4.01 0.61→0.69→0.67
Overall trust 4.00→4.00→4.00 3.98→3.79→3.77 0.85→0.85→0.76

Treatment Group

Similarly, we calculated the mean value, standard deviation, and the median
value from Stage 1 to Stage 3 of the treatment group in our experiment case.
The experiment group was instructed to use the facilitated collaboration tools to
help with their teamwork.
Table 4 presents an overview of the descriptive results of the treatment group.
We first analyzed the mean value results. It is apparent that risk decreases over
time, while benefit and overall level of trust continues to increase from Stage 1
to Stage 3. Moreover, the median value of each construct shows the same trend.
Table 5 shows the results obtained from the preliminary analysis of students’
perception of the facilitated collaboration process. From the results, the score for
participants’ satisfaction with the process (SP) ranges from 1 to 5, with a mean
value of 3.89. The mean value of SO is 4, which is the same as its median value.
For TOOLDIF, the mean value is 3.87, almost the same as the median value of
3.83. Moreover, participants’ perception of the ease or difficulty of the collabora-
tion process ranged from 1 (very difficult) to 5 (very easy), with a mean of 3.85.
Overall, the result of the process survey is satisfying.

Table 4. Descriptive Results of Trust Development in Treatment Group

Median (Stage 1→2→3) Mean (Stage 1→2→3) SD (Stage 1→2→3)


Risk 2.11→2.00→1.79 2.01→1.98→1.82 0.65→0.63→0.69
Benefit 4.33→4.33→4.50 4.17→4.24→4.36 0.64→0.66→0.72
Overall trust 4.00→4.00→4.00 4.03→4.08→4.13 0.86→0.82→0.89

Table 5. Results of Collaboration Outcome in Treatment Group

Median Mean SD
Satisfaction with process (SP) 4.00 3.89 0.77
Satisfaction with outcome (SO) 4.00 4.00 0.74
Perceived ease or difficulty of tools (TOOLDIF) 3.83 3.87 0.81
Perceived ease or difficulty of work process (PROCDIF) 4.00 3.85 0.68
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 991

Comparison of the Two Groups

Comparing the statistical results in the two case groups, one interesting ques-
tion arose—whether applying the CE process to team collaboration is more
useful to improve trust (with a lower level of risk and a higher level of benefit).
Since the results of the two groups separately indicate the general development
trend, we aim to compare the data further to ascertain the significance of the
changes.
First, repeated measures of analysis of variance (ANOVA) were conducted to
evaluate the longitudinal differences between the two groups. The resulting
design is 2 (type of groups) by 3 (stages). The facilitated process served as a
between-groups factor while the others were within-subject or repeated mea-
sures factors. As for risk, the result for Mauchly’s test indicated that the
assumption of sphericity was met, χ2(2) = 0.469, p = 0.291. The results showed
no significant effect of stages on risk perception, F(2,308) = 2.269, p = 0.105.
However, the interaction effect of stage and intervention (treatment group vs.
control group) had statistical significance on risk perception, F(2,308) = 6.279,
p = 0.002. In the meantime, the effect of intervention on risk was also
significant, with F(1,154) = 18.042, p = 0.000. For benefit, the results of
Mauchly’s test suggested that sphericity was also met, χ2(2) = 1.949, p =
0.377. There was a significant effect of intervention on benefit perception, F
(1,154) = 16.645, p = 0.000. However, the other two effects were not signifi-
cant. For trust, Mauchly’s test indicated that the assumption of sphericity was
also met, χ2(2) = 3.238, p = 0.198. Both the main linear effect of group (F
[1,154] = 10.942, p = 0.001) and the interaction effect of group and stage (F
[2,308] = 2.584, p = 0.077) had statistical significance on trust perception (see
Table 6).
In order to display the trust development trend in the two research groups
visually, we drew a column diagram using mean value (Figure 3). The results
of trust development in the control group are shown on the left, while the
results of the treatment group are shown on the right. The graph shows a
gradual decline in the value of risk and steady increases in the values of benefit
and overall trust in the treatment group. Notably, the comparison between two
groups demonstrates changing trends in trust development.
Since the main effect of group on all three indicators (risk, benefit, and trust) showed
significant differences, pairwise comparisons were conducted to explore which pairs of
variables were significantly different through difference stages. In Table 7, we only picked
those pairs that were significant among stages.
The results of the survey data reveal several interesting findings:
● From the perspective of risk, risk continues to decrease in the treatment group and
increase in the control group over time. The fluctuation of risk is the most obvious
whether in the control group or treatment group. Statistical evidence shows that the
longitudinal development of risk was significant.
992 CHENG, FU, AND DRUCKENMILLER

Table 6. Repeated-Measures ANOVA for Stages and Facilitated Intervention

Type III sum Mean


Source of squares df square F Sig.
Risk: Sphericity Stages 1.362 2 .681 2.269 .105
assumed (p = .291) Stages*Group 3.771 2 1.885 6.279** .002
Error(Stages) 92.491 308 .300
Group 8.660 1 8.660 18.042*** .000
Error(Group) 73.915 154 3.480
Benefit: Sphericity Stages .891 2 .446 1.310 .271
assumed (p = .377) Stages*Group 1.516 2 .758 2.229 .109
Error(Stages) 104.747 308 .340
Group 7.597 1 7.597 16.645*** .000
Error(Group) 70.291 154 .456
Trust: Sphericity Stages .424 2 .212 .422 .656
assumed (p = .198) Stages*Group 2.597 2 1.298 2.584* .077
Error(Stages) 154.773 308 .503
Group 7.383 1 7.383 10.942*** .001
Error(Group) 103.907 154 .675
Notes: ns: not significant; *p < 0.1; **p < 0.05; ***p < 0.01.

Figure 3. Trust Development Column Diagram (Left: Control Group; Right: Treatment Group)

● From the perspective of benefit, benefit constantly increases from Stage 1 to Stage
3 in the treatment group. The role of facilitation intervention in benefit development
showed statistical significance.
● From the perspective of overall trust, trust significantly decreased in the
control group. However, the trend of trust development in the treatment
group is not significant.
● The differences between groups (treatment group vs. control group) for risk,
benefit, and trust were all significant.
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 993

Table 7. Pairwise Comparisons

95% confidence
interval for
difference
Indicators (Group): Mean Std. Lower Upper
(I–J) Stages difference (I–J) error Sig. bound bound
Risk(Treatment group): 1–3 0.195** 0.072 .014 .041 .349
Benefit(Treatment group): 1–3 –.186** 0.075 .015 –.334 –.037
Trust(Control group): 1–3 .212* .123 .089 –.034 .458
Risk(Control group): 1–2 –.294** .091 .002 –.476 –.112
Risk(Control group): 1–3 –.208** .098 .037 –.404 –.013
Notes: ns: not significant; *p < 0.1; **p < 0.05; ***p < 0.01.

Qualitative Analysis
In order to find the in-depth mechanism of the aforementioned quantitative results, we also
conducted interviews and documentation studies. We analyzed the qualitative data con-
currently as we collected them [23]. Following the data analysis principles, we carefully
analyzed the qualitative data based on the theoretical lens and multisourced data. The main
purpose of the qualitative data analysis was to find the detailed influencing mechanism
that led to trust development over time. We assessed the influencing mechanism in trust
development from informants’ interview statements. These statements emerged in two
ways. The first was in response to our interview protocol, where we directly asked about
each trust dimension—for example, “Were there any problems in your team, and if so,
what were they? How were they resolved?” The second was in response to our open-
ended questions, where respondents expressed their overall feelings over the whole period
of team collaboration.
The qualitative data were transcribed and coded manually. Two of our authors partici-
pated in the coding process. Each coder separately categorized themes from the interview
transcripts and documentation. Review and discussion between the two authors continued
until they agreed on the key words to include in each category. Since our research purpose
was to investigate the trust antecedents that accounted for trust development over time, we
focused on the key words and sentences that might illustrate trust dynamics while coding.
The whole coding process was iterative to improve insights and generalities [70].

Trust Antecedents Related to Risk


In this section, we display mainly the qualitative results from the control group data. We
found three main factors of risk (antecedents of trust) through qualitative analysis—time
zone difference, language and cultural difference, and communication.
994 CHENG, FU, AND DRUCKENMILLER

1. Time Zone Difference


Time zone difference is a natural problem when many people work across
continents [24]. Teams may face greater complexity when members work
across multiple time zones, largely because of the difficulty of instant com-
munication [67]. By analyzing participants’ interviews and documentation,
we identified three negative effects (subfactors) of time zone difference that
correlate with risk and are negatively associated with trust and collaboration.

Subfactor 1: Difficulty in Communicating and Scheduling Work. Time


zone difference has a relationship with the difficulty in communicating and
scheduling work, thus correlating with the level of risk. Most participants had
at least ten hours’ time difference, so the time difference was obvious and
difficult to coordinate. For example, there is a comment in the documenta-
tion: “The biggest problem was time difference. Most of the discussions
among the Chinese students are after 11pm because it was daytime for
them. The main issue we had was the time difference since they would talk
to us so late in the day here, we would just answer when we could.”

Subfactor 2: Information Delay. Acquiring information was postponed, so


information delay was another negative effect correlated with risk. Due to
different time zones, there are always problems in acquiring information
promptly, which is gradually related to the level of risk over time. Talking
about this issue, an interviewee said:

Inevitably there are questions that don’t get answered until the following day
because of the time difference. Sometimes I feared whether or not they really
devoted themselves to our collaborative task as a result of the delayed
response. After several weeks’ interaction, my motivation to the task is lower
than the very beginning, because of the existence of a sense of cold-shoulder.
Subfactor 3: No In-depth Discussion. The time zone difference is a barrier
to conducting in-depth discussion. Choosing an exact time when everyone in
the team is free is usually difficult, but especially in globally distributed
teams. As one interviewee put it: “The time difference did make it a bit
difficult to have real lengthy discussions, however, so often we would post
and they would answer much later and vice versa.”

Most collaboration discussions could not be organized for a long time, such
as over two hours, and too many short collaboration sessions will reduce
efficiency and slow the project schedule. As problems arise, time zone
difference has a positive relationship with risk in distributed team collabora-
tions, thus correlating to collaboration trust.

2. Language and Culture Difference


Virtual team settings are more likely to include members who represent
different cultures than teams with collocated members [10]. Culture is a set
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 995

of values shared by a group of people and can be used to distinguish one


group from another [68]. Cultural values act as a filter, reflecting the sur-
rounding environment and guiding people’s behavior in decision making and
social interaction [54]. Language, as one cultural characteristic, is particularly
relevant to distributed teams [16]. After data analysis, we identified some
problems (subfactors) in language and culture differences that positively
correlate with risk.

Subfactor 1: Language Proficiency. Differences in language proficiency


among distributed team members created barriers to effective communication
and collaboration. Generally, team members with less proficient language
skills may suffer a confidence deficiency in their communication. They may
voice their opinions infrequently and sometimes misunderstand the meanings
expressed by others. A Chinese individual stated: “It was difficult at times to
get an idea across to them, possibly only because we weren’t sure how to
express the idea in a manner they would understand.”

Subfactor 2: Habitual Behavior. Moreover, habitual behavior is different


because of the diverse national contexts, which also has a positive relation-
ship with the level of risk. Hofstede and Bond [29] identified five dimensions
of culture. In our case, individualism versus collectivism is the most
obviously relevant dimension. Students in a U.S. university value personal
freedom and achievement more and are more task-oriented, while Chinese
students may believe the group is more important than the individual and
ascribe greater value to good relationships with others. Comments include:
“We try to keep friendly relationships with foreign students,” and “They
[United States] always issue their ideas individually, but we [China] always
discuss within the Chinese group and have a common idea to propose.”

3. Communication
Virtual team environments are made possible by information and commu-
nication technologies (ICT), so communication is a key factor [32].
Distributed teams depend heavily on computer-mediated communication
(CMC), which allows team members to engage in collaborative activities
[33]. The reliance on CMC in distributed collaboration relates to members’
perceived risk, which ultimately has a negative association with the devel-
opment of trust [60].

Subfactor 1: Collaboration Media. Even in terms of collaboration media,


participants in different countries have different habitual communication
tools. For example, some communication tools, such as Google+ and
Facebook, are not available in China, but U.S. students are not accustomed
to using WeChat. One interviewee reported: “While something [WeChat]
here is simple and easy to use, they found it inconvenient and not very
useful. I had hardly heard of WeChat before collaborating with the Chinese.”
996 CHENG, FU, AND DRUCKENMILLER

Subfactor 2: Information Integration. Another problem is that it is not


easy for users to integrate information. Everyone in the team presents
information in separate parcels and it is difficult to capture key points.
Especially for later information integration, this creates extra work. There
are some comments such as: “I made lots of efforts to integrate information
and it is too scattered. I have to download all the chatting records of our
group and read the records thoroughly to find the valuable information. It’s
time consuming and lowers our collaboration efficiency.”

Trust Antecedents Related to Benefit

We also display the qualitative data analysis results in the control group. We focused
on the factors related to benefit in this section. Generally, we found that task
accomplishment and leadership serve as two main factors influencing benefit during
collaboration.
1. Task Accomplishment
Subfactor 1: Job Finished on Time. When the group collaboration tasks
were allocated to each member, the team members were required to report
periodical deliverables to others. If all the participants within a group attach
importance to this collaboration task and accomplish their work on time, the
perceived benefit will change correspondingly. As one interviewee said,
“When a team member stated they would do or complete a task, if they
followed through with it that is what allowed me to trust.”

Subfactor 2: Effort on the Tasks. In the middle stage of team collaboration,


some team members failed to hand in their work on time: their strong focus
was on other issues, so the overall level of benefit was relatively low in the
middle stage of team collaboration. As one interviewee in the United States
put it: “We started meeting up and talking about the project early on, but
then I think we all got busy with other things and did very little work for quite
a while. So then we had to put a lot of effort in at the end to finish the
project.”

2. Leadership
The role of leader and leadership ability serve as contributors to the improve-
ment of benefit. We further discussed two aspects of leadership from the
qualitative data in this research.

Subfactor 1: Task Allocation. We divided each research case into several


collaborative subgroups, the leadership in different groups varied. For some
of the subgroups, one member was selected to be the team leader: responsible
leaders normally allocated tasks clearly, and guided other members to discuss
the tasks efficiently. However, for other groups, the team collaboration was
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 997

relatively disorganized, the perceived benefit was thus relatively low. As one
interviewee put it:
I think that we also had a bit of trouble with not having a “leader.” I
think it took a long time before somebody stepped into the leader role and
started pushing us to get the project done. Once the leader stepped up,
then it only took us a couple of days to pull everything together.
Subfactor 2: Powerful Leaders. On the other hand, powerful leaders have
the ability to gather team members together and facilitate within-group inter-
actions, especially the cross-cultural collaboration that broadens the vision of
the participants. Because the cross-cultural collaboration experience serves as
an important source of benefit, the leader who enables members to know each
other is very important. For example, one participant said:
I have never claimed to have superior knowledge or ability to influence others
and am always grateful for the opportunity to work with others and learn from
others, thanks to our team leader who gathered us together. New experiences
are always a plus in my book, this was definitely a new experience.

Risk Reduction through CE

In the previous two sections, we mainly discussed five trust factors that may
influence risk or benefit in the case of the control group. For each trust factor, we
elaborated several second-tier subfactors to illustrate the influencing mechanisms. In
the case of the treatment group, we mainly investigated the effect of a thinkLets–
enabled collaboration process on trust development. Following the trust factors and
the corresponding subfactors, we emphasized the influence of the CE process on
trust improvement.
In this section, we begin by discussing the first aspect of trust, risk, based on the
qualitative results of the control group.

1. After the adoption of the CE process, members have greater potential to


conduct in-depth discussions. Group productivity is improved using the
experiment process, even though time zone difference is an inherent problem
in globally distributed collaboration and becomes a barrier to conducting in-
depth discussion. Through the use of a facilitated collaboration process in the
treatment group, team members gathered together within a fixed period. The
countdown function embedded in the Discussion platform enables members
to concentrate on the collaboration while working together, which facilitates
in-depth discussion and further leads to the improvement of group produc-
tivity. As one interviewee put it: “After longitudinal collaboration and the
use of the Discussion platform, I don’t think the time zone difference is a
major problem. The facilitated collaboration guided by our team leader
enables our collaboration to be accomplished efficiently.”
998 CHENG, FU, AND DRUCKENMILLER

2. Habitual behavior caused by cultural difference is minimized through a well-


designed process. Although individualism versus collectivism exists in dif-
ferent cultural backgrounds, our experiment helped weaken the influence of
cultural differences. This can be further explained through the thinkLets
modules of Crowbar and PointCounterpoint. These thinkLets help with the
process of building a consensus. That is to say, no matter what their culture’s
tradition, team members should all vote for the solutions and discuss with
one another until conclusions are finally reached and agreed with other
members. For example, one U.S. student commented: “We have to follow
the collaboration process, no matter how we collaborated with native team
members before.”

3. Information integration became easier after the adoption of the CE


process. Lack of ease in integrating information was an indicator of
the communication barrier proposed by the control group, while in the
eyes of the treatment group members, the designed process helped
remove invaluable ideas and the final decision was the outcome agreed
by most team members. Therefore, communication was improved
through the CE process, leading to the decrease of risk. As one infor-
mant noted: “I think it [the Discussion platform] is great! Although we
[the Americans and Chinese] all [are] not familiar with the tool, some
of the functions are useful and effective for our virtual communication.”

Benefit Promotion through CE


As for the benefit factor, we also found several advantages in the experiment process
compared with the control group of using regular collaboration tools and processes.
1. The leaders became more powerful and organized due to the CE process.
On the one hand, the facilitators serve as the team leaders who guide the
collaboration process, reserve time for the group discussion, and ensure
members’ participation. The experiment process does not require a profes-
sional facilitator because the tool is easy to use and it is easy to grasp how
to be a facilitator in the Discussion platform. One member who served as
a facilitator said: “After the training that included a demonstration and
practice of each thinkLet by a professional facilitator, I think I am able to
lead our collaboration process using the Discussion platform.”

2. The motivation to collaborate changed due to the CE process, and so had a


positive relationship on the effort put into the tasks. Although the experiment
clearly may not have had a direct influence on task accomplishment, team
members valued the whole collaboration highly and their motivation in the
task was relatively high as a result of the ordered collaboration process,
which may in turn promote task accomplishment. When talking of
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 999

motivation, a U.S. student commented: “The members of the Chinese team


we talked to on a regular basis seemed to be reliable and gave us great
brainstorming ideas for the project. We are motivated to get the work done.”

Discussion
Summary of the Results
In order to answer the three research questions proposed in the introduction section,
we conceptualized and empirically tested a research model relating to the influencing
mechanism from trust antecedents to consequences. Specifically, there are four
major findings in our present study. We present the findings according to each of
the research questions.
Research Question 1: Does the development of trust behave differently in
conventional collaborative teams and process-guided collaborative teams?
Trust development in the control group and treatment group shows different
trends. Specifically, risk significantly increases in the control group and decreases
in the treatment group, while benefit significantly increases in the treatment group,
trust significantly decreases in the control group. Among these, the effects of group
types on trust development were all significant on all three indicators (risk, benefit,
and trust). Of all the trust dimensions, the change of risk was the most significant
whether in the control group or the treatment group. Therefore, these findings
through the analysis of longitudinal data answer RQ1.
The trust development trends are different in conventional collaborative teams and
process-guided collaborative teams. Compared with the ideal development trend of
lower risk and higher benefit [38], the process-guided collaborative teams (i.e., the
treatment group) have more satisfying collaboration results from the perspective of
trust dynamic.
Research Question 2: What are the trust antecedents in longitudinal globally
distributed collaboration?
Five factors were found to be the antecedents of trust in longitudinal globally
distributed collaboration: these factors are time zone difference, language and
cultural difference, communication, task accomplishment, and leadership. In sum,
these five antecedents help answer our second research question. The former three
factors serve as the influencing factors of risk, while the latter two factors were
found to have an influence on the perceived benefit. Furthermore, for each trust
antecedent, we also see several second-tier subantecedents that clarify the in-depth
reason that a certain factor is regarded as a trust antecedent in our study. The
antecedents found in this research are all in line with previous studies [24, 32,
54], although the influencing mechanisms are slightly different in various research
settings. The relationships of the antecedents are shown in Figure 4.
1000 CHENG, FU, AND DRUCKENMILLER

Information No in-depth Job finished on


delay discussion time

Time zone
difference Leadership
Collaboration Difficulty in
media communicating

Effort on the
tasks
Communication Risk Benefit

Powerful
leaders
Information
integration Trust
Language
Task
and cultural
accomplishment
difference

Language Habitual
proficiency behavior Task allocation

Figure 4. Proposed Relationships of Trust Antecedents

Research Question 3: How has trust development been influenced by using


facilitated collaboration?

The well-designed CE process enabled trust improvement in the treatment group.


In order to answer the third research question, we deliberately reviewed the inter-
view transcripts and the documentation to find the influence of CE on trust devel-
opment. Several thinkLets modules were found to contribute to the improvement of
trust. For example, the designated duration of each activity causes team members to
concentrate on the collaboration and improves group productivity, thus helping to
weaken the influence of time zone difference. Also, some thinkLets modules helped
to weaken the influence of cultural diversity and communication barriers, such as
Crowbar to create a prioritized list and PopcornSort to concentrate ideas. As CE-
enabled collaboration provides the possibility of repeating the collaboration process
for recurring tasks [5], the influence of the thinkLets can be a longitudinal process.
Just using the process once may not influence the level of trust significantly, but the
cumulative influence over time serves as a catalyst for trust improvement in globally
distributed collaboration. Moreover, from the perspective of the collaboration pro-
cess, the designed process is organized and enables team members to reach a
consensus within the scheduled time. The facilitator role helps to improve leadership
skills through guided scripts. The improvement in leadership also facilitates an
improvement in trust after long-term team collaboration. In total, we have found
five attractions of CE theory that help with risk reduction and benefit improvement
that ultimately contribute to trust development over time. Each attraction is asso-
ciated with a single trust subfactor, as discussed above.
Finally, through the measurement of trust consequences, the average value of SO,
PO, TOOLDIF, and PROCDIF are all around 4. This reflects a relatively satisfying
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 1001

collaboration outcome through the improvement of trust using the CE process. The
Discussion platform is relatively easy to use (mean value of TOOLDIF = 3.87) and
the designed process is easy to follow (mean value of PROCDIF = 3.85). The
participants’ outcome expectations are generally met (mean value of SO = 4), and
they are generally satisfied with the collaboration process we designed (mean value
of PO = 3.89).

Contribution to Research
This research fills gaps by making the following theoretical contributions.
First, based on the perspectives of risk and benefit, we found several trust ante-
cedents and their corresponding subfactors. Although previous studies have exam-
ined several antecedents of trust (e.g., [9, 15, 24, 44, 54]), few were conducted from
the perspective of subfactors. The present study contributes to the theory of trust
antecedents in the context of globally distributed collaboration [33, 52], and elabo-
rates the influencing mechanism through the detailed description of trust subfactors
from the perspectives of risk and benefit, respectively [36, 48, 63]. Generally, the
trust antecedents found in our study are all in line with previous studies that have
identified the relationships of leadership, communication, cultural difference, time
zone difference, and task accomplishment with trust [24, 32, 54, 60]. Moreover, in
our research, several corresponding subfactors help clarify various aspects of influ-
encing mechanisms of the antecedents to trust, providing more detailed explanations
to further understand trust antecedents in the context of globally distributed
collaboration.
Second, this study contributes to the literature on trust development research [13,
76] and the literature on trust dynamics by applying CE principle to help with trust
improvement. While prior trust research has often regarded trust as a dynamic
construct, few investigations have been conducted from the development perspective
[48]. Although some focus on the dynamics influencing antecedents of trust, few
have been conducted based on the facilitation of intervention in trust development.
Among various facilitation interventions, CE provides a repeatable collaboration
process that can be used in recurring tasks [7, 39, 40]. In turn, the repeatable use of
the process contributes to successful trust improvement over time. Therefore, com-
pared with previous literature, such as that on trust dynamics of trust-type transfor-
mation [35], trust development in implementation projects [61], trust development in
computer-mediated local teams [12], and face-to-face collaboration [76], this
research attempts another new approach that not only describes trust development
but also improves trust development through in-depth investigations on the applica-
tion of CE [5, 39]. In doing so, this study complements work on the dynamic view
of trust [10, 76] from the perspective of facilitation intervention [8].
Finally, this study conducted CE application in the context of globally distributed
collaboration. The CE principle was validated as being helpful for shared under-
standing [4], idea generation [39], and risk assessment [22] in the context of
heterogeneous work groups, technology-supported groups [8], and project
1002 CHENG, FU, AND DRUCKENMILLER

management teams [40]. In this study, the new research context of globally dis-
tributed collaboration has broadened the research domain of CE [7]. Moreover,
through the longitudinal experiment cases of both the control group and the treat-
ment group, our study highlighted the feasibility of introducing the CE principle to
help improve trust in cross-cultural research settings involving participants from
China and the United States. The cross-cultural dimension also serves as a future
research direction for further applying CE to more complex research settings.

Practical Implications
This research has practical implications for globally distributed collaboration, espe-
cially for those who need to improve their trust longitudinally. The research could be
used by team leaders/managers to improve their skills and manage their team with
better understanding of the trust antecedents, as well as facilitating collaboration
intervention in order to help improve trust development in globally distributed
teams.
Moreover, global team members always have various norms and habitual cultural
traditions, and mutual communication also faces problems compared with traditional
face-to-face collaboration. It is suggested that practitioners design a collaboration
process that helps minimize cultural and linguistic influences—for example, con-
ducting an iterative democratic vote that avoids the contradiction of the individual-
ism versus collectivism culture. Furthermore, for recurring and complex tasks,
practical workers are also advised to adopt a wider range of thinkLets according
to the requirements of the tasks. In addition, we have discussed in depth the reason
why time zone difference is a problem, and found that lack of instant interaction is
the major contributor. Global collaboration practitioners are encouraged to reserve a
fixed time with countdown functions to facilitate effective collaboration. Well-
organized collaboration process guidance can help save time and reach satisfying
collaboration outcomes through effective and in-depth discussions.

Limitations and Future Research


As with all research, there are some limitations. First, the participants in our special
case context are all university students who lack work experience: because of
differences between students and practical workers, global collaboration may be
more complex in a real business,. Moreover, we tested globally distributed colla-
boration only in U.S. and Chinese universities, so the tested population is limited.
Therefore, it would be useful to generalize our study to additional experiments with
various subject demographics in different research settings in future studies.
Another future research direction relates to differences in trust development in
terms of multicultural participants in globally distributed collaboration. Though
cultural diversity is widely investigated, such research is seldom conducted in
terms of facilitation intervention, especially the evaluation of CE theory in
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 1003

longitudinal trust improvement in different cultural settings. Whether the influence


of CE on the collaboration outcome has the same results in different cultural back-
grounds is an interesting question. Extending the principle of collaboration engineer-
ing and integrating it into a comparison of culturally diversified team collaboration
will be of theoretical and practical significance.

Conclusions
This study investigated the role of facilitated intervention in trust development.
Following the design science research approach, collaboration engineering principles
were used to design a collaboration process for the treatment group. Survey results
showed that the changing trust trend in the treatment group was more positive than
in the control group. Qualitative data explained the in-depth mechanism of trust
development in the two groups. From the results, time zone difference, language and
cultural difference, communication, leadership, and task accomplishment all serve as
contributors to trust development. Moreover, the treatment group that adopted the
CE process mitigated the influence of the aforementioned factors through several
thinkLets modules and the collaboration process, and thus positively correlated with
trust over time.

Acknowledgments: We would like to thank all the participants. We gratefully acknowledge


constructive feedback and suggestions from the anonymous reviewers and editors, which have
helped us to strengthen the study. We also thank the National Natural Science Foundation of
China (Grant no. 71571045) and the Beijing Higher Education and Teaching Reform Project
(2015-ms080) for providing funding for part of this research.

REFERENCES
1. Anson, R.; Bostrom, R.; and Wynne, B. An experiment assessing group support system
and facilitator effects on meeting outcomes. Management Science, 41, 2 (1995), 189–208.
2. Arduin, P.E.; Grundstein, M.; and Rosenthal-Sabroux, C. From knowledge sharing to
collaborative decision making. International Journal of Information and Decision Sciences, 5,
3 (2013), 295–311.
3. Awad, N.F., and Ragowsky, A. Establishing trust in electronic commerce through online
word of mouth: An examination across genders. Journal of Management Information Systems,
24, 4 (2008), 101–121.
4. Bittner, E.A.C., and Leimeister, J.M. Creating shared understanding in heterogeneous
work groups: Why it matters and how to achieve it. Journal of Management Information
Systems, 31, 1 (2014), 111–144.
5. Briggs, R.O.; Vreede, G.J. de.; and Nunamaker, J.F. Jr. Collaboration engineering with
thinkLets to pursue sustained success with group support systems. Journal of Management
Information Systems, 19, 4 (2003), 31–64.
6. Briggs, R.; Kolfschoten, G.; Vreede, G.J. de.; and Douglas, D. Defining key concepts
for collaboration engineering. Proceedings of the Twelfth Americas Conference on
Information Systems. Acapulco, 2006, pp. 121–128.
7. Briggs, R.O.; Kolfschoten, G.L.; Vreede, G.J. de.; Lukosch, S.; and Albrecht, C.
Facilitator-in-a-box: Process support system to help practitioners realize the potential of colla-
boration technology. Journal of Management Information Systems, 29, 4 (2013), 159–199.
1004 CHENG, FU, AND DRUCKENMILLER

8. Briggs, R.O.; Reinig, B.A.; and Vreede, G.J. de. Meeting satisfaction for technology-
supported groups: An empirical validation of a goal-attainment model. Small Group Research,
37, 6 (2006), 585–611.
9. Cheng, X.; Fu, S.; Sun, J.; Han, Y.; Shen, J.; and Zarifis, A. Investigating individual
trust in semi-virtual collaboration of multicultural and unicultural teams. Computers in Human
Behavior, 62 (2016), 267–276.
10. Cheng, X.; Li, Y.; Sun, J.; and Huang, J. Application of a novel collaboration engineer-
ing method for learning design: A case study. British Journal of Educational Technology, 47,
4 (2016), 803–818.
11. Cheng, X.; Li, Y.; Sun, J.; and Zhu, X. Easy collaboration process support system
design for student collaborative group work: A case study. Proceedings of the 47th Hawaii
International Conference on System Science (HICSS). Waikoloa, 2014, pp. 453–462.
12. Cheng, X.; Macaulay, L.; and Zarifis, A. Modeling individual trust development in
computer mediated collaboration: A comparison of approaches. Computers in Human
Behavior, 29, 4 (2013), 1733–1741.
13. Cheng, X.; Nolan, T.; and Macaulay, L. Don’t give up the community: A viewpoint of trust
development in online collaboration. Information Technology and People, 26, 3 (2013), 298–318.
14. Cheng, X.; Yin, G.; Azadegan, A.; and Kolfschoten, G. Trust evolvement in hybrid team
collaboration: A longitudinal case study. Group Decision and Negotiation, 25, 2 (2016), 267–288.
15. Chiu, C.M.; Hsu, M.H.; Lai, H.; and Chang, C.M. Re-examining the influence of trust
on online repeat purchase intention: The moderating role of habit and its antecedents. Decision
Support Systems, 53, 4 (2012), 835–845.
16. Chudoba, K.M.; Wynn, E.; Lu, M.; and Watson‐Manheim, M.B. How virtual are we?
Measuring virtuality and understanding its impact in a global organization. Information
Systems Journal, 15, 4 (2005), 279–306.
17. Colquitt, J.A.; LePine, J.A.; Piccolo, R.F.; Zapata, C.P.; and Rich, B.L. Explaining the
justice–performance relationship: Trust as exchange deepener or trust as uncertainty reducer?
Journal of Applied Psychology, 97, 1 (2012), 1–15.
18. Das, T.L., and Teng, B.S. The risk-based view of trust: A conceptual framework.
Journal of Business and Psychology, 19, 1 (2004), 85–116.
19. De Jong, B.A., and Elfring, T. How does trust affect the performance of ongoing teams?
The mediating role of reflexivity, monitoring, and effort. Academy of Management Journal,
53, 3 (2010), 535–549.
20. Dean, D.L.; Hender, J.M.; Rodgers, T.L.; and Santanen, E. Identifying good ideas:
Constructs and scales for idea evaluation. Journal of Association for Information Systems, 7,
10 (2006), 646–699.
21. Dennis, A.R.; Minas, R.K.; and Bhagwatwar, A.P. Sparking creativity: Improving
electronic brainstorming with individual cognitive priming. Journal of Management
Information Systems, 29, 4 (2013), 195–216.
22. Vreede, G.J. de.; Briggs, R.O.; and Massey, A.P. Collaboration engineering:
Foundations and opportunities. Editorial to the special issue of the Journal of the
Association for Information Systems, 10, 3 (2009), 121–137.
23. Eisenhardt, K.M. Building theories from case study research. Academy of Management
Review, 14, 4 (1989), 532–550.
24. Espinosa, J.A.; Nan, N.; and Carmel, E. Temporal distance, communication patterns, and
task performance in teams. Journal of Management Information Systems, 32, 1 (2015), 151–191.
25. Fang, Y.; Qureshi, I.; Sun, H.; McCole, P.; Ramsey, E.; and Lim, K.H. Trust, satisfac-
tion, and online repurchase intention: The moderating role of perceived effectiveness of
e-commerce institutional mechanisms. MIS Quarterly, 38, 2 (2014), 407–427.
26. Gefen, D.; Karahanna, E.; and Straub, D.W. Trust and TAM in online shopping: An
integrated model. MIS Quarterly, 27, 1 (2003), 51–90.
27. Gulati, R. Does familiarity breed trust? The implications of repeated ties for contractual
choice in alliances. Academy of Management Journal, 38, 1 (1995), 85–112.
28. Hardwick, J.; Anderson, A.R.; and Cruickshank, D. Trust formation processes in
innovative collaborations: Networking as knowledge building practices. European Journal
of Innovation Management, 16, 1 (2013), 4–21.
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 1005

29. Hofstede, G., and Bond, M.H. Hofstede’s culture dimensions an independent validation
using Rokeach’s value survey. Journal of Cross-Cultural Psychology, 15, 4 (1984), 417–433.
30. Horwitz, S.K., and Santillan, C. Knowledge sharing in global virtual team collaboration:
Applications of CE and thinkLets. Knowledge Management Research and Practice, 10, 4
(2012), 342–353.
31. Hoy, W.K., and Tschannen-Moran, M. Five faces of trust: An empirical confirmation in
urban elementary schools. Journal of School Leadership, 9 (1999), 184–208.
32. Jarvenpaa, S.L.; Knoll, K.; and Leidner, D.E. Is anybody out there? Antecedents of trust
in global virtual teams. Journal of Management Information Systems, 14, 4 (1998), 29–64.
33. Jarvenpaa, S.L., and Leidner, D.E. Communication and trust in global virtual teams.
Journal of Computer‐Mediated Communication, 3, 4 (1998), 791–815.
34. Jarvenpaa, S.L.; Shaw, T.R.; and Staples, D.S. Toward contextualized theories of trust:
The role of trust in global virtual teams. Information Systems Research, 15, 3 (2004), 250–267.
35. Jones, G.R., and George, J.M. The experience and evolution of trust: Implications for
cooperation and teamwork. Academy of Management Review, 23, 3 (1998), 531–546.
36. Kim, D.J.; Ferrin, D.L.; and Rao, H.R. A trust-based consumer decision-making model
in electronic commerce: The role of trust, perceived risk, and their antecedents. Decision
Support Systems, 44, 2 (2008), 544–564.
37. Kim, H.W.; Xu, Y.; and Koh, J. A comparison of online trust building factors between
potential customers and repeat customers. Journal of the Association for Information Systems,
5, 10 (2004), 392–420.
38. Kramer, R.M. Trust and distrust in organizations: Emerging perspectives, enduring
questions. Annual Review of Psychology, 50, 1 (1999), 569–598.
39. Kolfschoten, G.L.; Briggs, R.O.; Vreede, G.J. de.; Jacobs, P.H.; and Appelman, J.H. A
conceptual foundation of the thinkLet concept for collaboration engineering. International
Journal of Human–Computer Studies, 64, 7 (2006), 611–621.
40. Kolfschoten, G.L., and Vreede, G.J. de. A design approach for collaboration processes:
A multimethod design science study in collaboration engineering. Journal of Management
Information Systems, 26, 1 (2009), 225–256.
41. Kuo, F.Y., and Yu, C.P. An exploratory study of trust dynamics in work‐oriented virtual
teams. Journal of Computer-Mediated Communication, 14, 4 (2009), 823–854.
42. Lee, D.; Moon, J.; Kim, Y.J.; and Mun, Y. Antecedents and consequences of mobile
phone usability: Linking simplicity and interactivity to satisfaction, trust, and brand loyalty.
Information and Management, 52, 3 (2015), 295–304.
43. Lewicki, R.J.; Tomlinson, E.C.; and Gillespie, N. Models of interpersonal trust devel-
opment: Theoretical approaches, empirical evidence, and future directions. Journal of
Management, 32, 6 (2006), 991–1022.
44. Lu, B.; Zhang, T.; Wang, L.; and Keller, L.R. Trust antecedents, trust and online
microsourcing adoption: An empirical study from the resource perspective. Decision
Support Systems, 85 (2016), 104–114.
45. Malhotra, D., and Lumineau, F. Trust and collaboration in the aftermath of conflict: The
effects of contract structure. Academy of Management Journal, 54, 5 (2011), 981–998.
46. Marques, M., and Ochoa, S.F. Improving teamwork in students’ software projects. IEEE
Proceedings of the 27th Conference on Software Engineering Education and Training
(CSEE&T). Klagenfurt, Austria, 2014, pp. 99–108.
47. Martin, J.A., and Eisenhardt, K.M. Rewiring: Cross-business-unit collaborations in
multibusiness organizations. Academy of Management Journal, 53, 2 (2010), 265–301.
48. Mayer, R.C.; Davis, J.H.; and Schoorman, F.D. An integrative model of organizational
trust. Academy of Management Review, 20, 3 (1995), 709–734.
49. Maznevski, M.L., and Chudoba, K.M. Bridging space over time: Global virtual team
dynamics and effectiveness. Organization Science, 11, 5 (2000), 473–492.
50. McKnight, D.H.; Cummings, L.L.; and Chervany, N.L. Initial trust formation in new
organizational relationships. Academy of Management Review, 23, 3 (1998), 473–490.
51. Michinov, N. Is electronic brainstorming or brainwriting the best way to improve
creative performance in groups? An overlooked comparison of two idea‐generation techni-
ques. Journal of Applied Social Psychology, 42, 1 (2012), 222–243.
1006 CHENG, FU, AND DRUCKENMILLER

52. Mortensen, M., and Neeley, T.B. Reflected knowledge and trust in global collaboration.
Management Science, 58, 12 (2012), 2207–2224.
53. Nidumolu, S. The effect of coordination and uncertainty on software project perfor-
mance: Residual performance risk as an intervening variable. Information Systems Research,
6, 3 (1995), 191–219.
54. O’Leary, M.B.; Wilson, J.M.; and Metiu, A. Beyond being there: The symbolic role of
communication and identification in perceptions of proximity to geographically dispersed
colleagues. MIS Quarterly, 38, 4 (2014), 1219–1243.
55. Paulus, P.B.; Kohn, N.W.; Arditti, L.E.; and Korde, R.M. Understanding the group size
effect in electronic brainstorming. Small Group Research, 44, 3 (2013), 332–352.
56. Pavlou, P.A., and Gefen, D. Building effective online marketplaces with institution-
based trust. Information Systems Research, 15, 1 (2004), 37–59.
57. Piccoli, G., and Ives, B. Trust and the unintended effects of behavior control in virtual
teams. MIS Quarterly, 27 (2003), 365–395.
58. Qu, W.G.; Pinsonneault, A.; Tomiuk, D.; Wang, S.; and Liu, Y. The impacts of social
trust on open and closed B2B e-commerce: A Europe-based study. Information and
Management, 52, 2 (2015), 151–159.
59. Ridings, C.M.; Gefen, D.; and Arinze, B. Some antecedents and effects of trust in
virtual communities. Journal of Strategic Information Systems, 11, 3 (2002), 271–295.
60. Robert, L.P.; Denis, A.R.; and Hung, Y.T.C. Individual swift trust and knowledge-based
trust in face-to-face and virtual team members. Journal of Management Information Systems,
26, 2 (2009), 241–279.
61. Rose, J., and Schlichter, B.R. Decoupling, re‐engaging: Managing trust relationships in
implementation projects. Information Systems Journal, 23, 1 (2013), 5–33.
62. Rotter, J.B. A new scale for the measurement of interpersonal trust. Journal of
Personality, 35, 4 (1967), 651–665.
63. Rusman, E.; Van Bruggen, J.; Sloep, P.; and Koper, R. Fostering trust in virtual project
teams: Towards a design framework grounded in a trustworthiness antecedents (TWAN)
schema. International Journal of Human–Computer Studies, 68, 11 (2010), 834–850.
64. Sarker, S.; Ahuja, M.; Sarker, S.; and Kirkeby, S. The role of communication and trust
in global virtual teams: A social network perspective. Journal of Management Information
Systems, 28, 1 (2011), 273–310.
65. Schaubroeck, J.; Lam, S.S.; and Peng, A.C. Cognition-based and affect-based trust as
mediators of leader behavior influences on team performance. Journal of Applied Psychology,
96, 4 (2011), 863–871.
66. Schilke, O., and Cook, K.S. A cross-level process theory of trust development in
interorganizational relationships. Strategic Organization, 11, 3 (2013), 281–303.
67. Schweitzer, L., and Duxbury, L. Conceptualizing and measuring the virtuality of teams.
Information Systems Journal, 20, 3 (2010), 267–295.
68. Shachaf, P. Cultural diversity and information and communication technology impacts on
global virtual teams: An exploratory study. Information and Management, 45, 2 (2008), 131–142.
69. Shrivastava P. Ecocentric management for a risk society. Academy of Management
Review, 20, 1 (1995), 118–137.
70. Smith, W.K. Dynamic decision making: A model of senior leaders managing strategic
paradoxes. Academy of Management Journal, 57, 6 (2014), 1592–1623.
71. Staples, D.S., and Webster, J. Exploring the effects of trust, task interdependence and
virtualness on knowledge sharing in teams. Information Systems Journal, 18, 6 (2008), 617–640.
72. Venkatesh, V.; Brown, S.A.; and Bala, H. Bridging the qualitative-quantitative divide:
Guidelines for conducting mixed methods research in information systems. MIS Quarterly, 37,
1 (2013), 21–54.
73. Von Alan, R.H.; March, S.T.; Park, J.; and Ram, S. Design science in information
systems research. MIS Quarterly, 28, 1 (2004), 75–105.
74. Walther, J.B., and Bunz, U. The rules of virtual groups: Trust, liking, and performance
in computer‐mediated communication. Journal of Communication, 55, 4 (2005), 828–846.
TRUST DEVELOPMENT IN GLOBALLY DISTRIBUTED COLLABORATION 1007

75. Wildman, J.L.; Shuffler, M.L.; Lazzara, E.H.; Fiore, S.M.; Burke, C.S.; Salas, E.; and
Garven, S. Trust development in swift starting action teams: A multilevel framework. Group
and Organization Management, 37 (2012), 137–170.
76. Wilson, J.M.; Straus, S.G.; and McEvily, B. All in due time: The development of trust in
computer-mediated and face-to-face teams. Organizational Behavior and Human Decision
Processes, 99, 1 (2006), 16–33.
77. Yakovleva, M.; Reilly, R.R.; and Werko, R. Why do we trust? Moving beyond
individual to dyadic perceptions. Journal of Applied Psychology, 95, 1 (2010), 79–91.
78. Zaheer, A.; McEvily, B.; and Perrone, V. Does trust matter? Exploring the effects of inter-
organizational and interpersonal trust on performance. Organization Science, 9, 2 (1998), 141–159.

View publication stats

You might also like