We decided to examine the effect of macro-level factors (i.e., team attributes) and applied hierarchical linear modeling analysis to a sample of data collected from 96 individuals nested[r]
Trang 1Should I e-collaborate with this group? A multilevel model of usage intentions Ofir Turel * , Yi (Jenny) Zhang1
Sreven G Mihaylo College of Business and Economics, California State University, Fullerton, P.O Box 6848, Fullerton, CA 92834-6848, USA
1 Introduction
The use of ad hoc teams has become increasingly common in
today’s organizations because it allows flexible and efficient use of
expert employees with varying backgrounds[27] Furthermore,
recently, organizations have increasingly relied on electronic
collaboration tools, such as email and discussion boards, for
facilitating communications among ad hoc team members The
combination of these trends has led to the creation of virtual teams
or technology-mediated teams[5]consisting of employees who are
not necessarily co-located and who rely on technology for most of
their communication Furthermore, these individuals work on
interdependent tasks and share responsibility for the outcomes[29]
Because of the increased popularity it is important to
understand what drives their use and performance [18] It is
especially important to understand what drives individuals to use
e-collaboration tools when assigned to ad hoc teams, because it
can have important consequences, such as increased efficiency and
time saving[23] In many cases individuals assigned to a project
team can choose among several means of communication,
including electronic tools, face-to-face meetings, or a combination
For such decision, team members must consider a range of factors
and tradeoffs: online collaboration tools may benefit them for example by reducing peer pressure, but their use may also present challenges to effective communications due to, time delays in response, lack of social cues, and lack of assurance of participation
[20] Several studies have examined e-collaboration models [14], although mostly by taking either an individual- or team-level perspective, but not a joint perspective; i.e., one that takes into account the individual decision makers in the broader team context in which the decision is made While contextual team-level variables may influence individual team member behaviors
[13], little empirical research in the IS field has studied such cross-level effects[3]
The focus of our study was on level-spanning effects in electronic collaboration Specifically, we argued that users employed mental accounting processes [25] when deciding to use a specific electronic collaboration tool; and that the decision depended on both the properties of the tool and on the qualities of the team with which one is working These two elements are intertwined through Media Richness Theory; e-collaboration systems that often rely on lean-media (e.g., text based commu-nications) should be perceived as more suitable for interactions within competent teams due to their probable lower equivocality and task difficulty That is, team attributes may alter one’s general usage intentions beyond the mere effect of system-referenced perceptions
Examination of cross-level relationships required multilevel empirical perspectives The focal team attributes on which we
A R T I C L E I N F O
Article history:
Received 1 August 2009
Received in revised form 5 October 2010
Accepted 28 December 2010
Available online 13 January 2011
Keywords:
Online collaboration
Virtual teams
Social loafing
Hierarchical linear modeling
Technology use
Team potency
Multilevel theory
Mental accounting
Media richness
A B S T R A C T The use of online collaboration tools for virtual teamwork has been studied extensively, but mainly at the individual-level We decided to examine the effect of macro-level factors (i.e., team attributes) and applied hierarchical linear modeling analysis to a sample of data collected from 96 individuals nested in
34 virtual teams Our results suggested that the development of behavioral e-collaboration intentions by individual virtual team members was affected by their perceptions about the system, as described by individual-level IT use theories, and macro-level factors pertaining to the team The collaboration technology was perceived to be less useful when employed to communicate with social loafers; and collective social loafing negatively influenced the teams’ potency assessments After controlling for individual-level perceptions of system usefulness, team potency augmented team members’ intentions
to use the online collaboration technology with similar teams It also improved team performance
ß2011 Elsevier B.V All rights reserved
* Corresponding author Tel.: +1 657 278 5613; fax: +1 657 278 5940.
E-mail addresses: oturel@fullerton.edu (O Turel),
1
Tel.: +1 657 278 4851; fax: +1 657 278 5940.
Contents lists available atScienceDirect
Information & Management
j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / i m
0378-7206/$ – see front matter ß 2011 Elsevier B.V All rights reserved.
Trang 2focused included the overall level of social loafing in the team and
its potency (or shared belief in its effectiveness) These attributes
have been cited as key predictors of teamwork behaviors and
performance[4,10] The key system-referenced predictor we used
was perceived usefulness: the focal antecedent of IT usage intention
[15] Our research model is shown asFig 1
2 Conceptual background
While technology acceptance has been extensively studied,
little is known about such processes in virtual teams A Technology
Acceptance by Groups (TAG) framework addresses this issue It
posits that teams develop positive feeling towards a technology
based on behavioral attitudes of team members towards the
technology These attitudes are then shaped by technological
factors that the group considers important (e.g., complexity), and
psycho-social factors (e.g., majority support) [22] The TAG
framework has received some empirical support but not much
is known about the formation of behavioral intentions in the team
This was the focus of our study
While task considerations are applicable to user decisions
about any use of technology (e.g., task-technology fit
consider-ations), team characteristics are factors that should be considered
by users of electronic collaboration tools We conceptualized the
formation of technology use decisions on electronic collaboration
tools as a process that spanned levels of measurement Usage
decisions were therefore seen as being developed by individuals
based on the individual-level system assessment of perceived
usefulness and attributes of the teams to which they belonged (its
potency)
2.1 Social loafing
Social loafing is the lack of, or reduction in, motivation and thus
of effort when individuals work together as opposed to how they
work individually It is a negative phenomenon that results in
productivity loss when working in groups Social loafing is thus
important to investigate because it can negatively affect team
performance in both offline and virtual contexts
Individuals may loaf in team settings for many reasons First,
team members who do not feel their contributions are essential to
the final product, tend to loaf Second, team members loaf when
there is lack of evaluation of the individual Third, social loafing is
affected by the perceived fairness of a team’s decision process[21]
When people work collectively, they tend to match their co-workers efforts
At the individual level, perceived loafing refers to the perception of any loafing behavior of the team by an individual
At the group level, social loafing refers to the consensus of members about collective loafing in the group[19]
2.2 Team potency
Team potency is the shared belief of a group about its general effectiveness It is a group-level form of a general efficacy assessment Team potency obviously affects team outcomes Accordingly, IS studies have been focusing on this concept[1] While team potency increases team performance[6], it had not previously been seen as a predictor of technology use behavior Thus, we decided to examine its effect in a virtual team context, i.e., as a predictor of team performance and individual-level technology usage decisions We also examined a group-level antecedent for team potency: the level of within-team social loafing
3 Theoretical model and hypothesis development
3.1 Individual-level model
When deciding whether a technology is useful, individuals normally refer to their experiences with it and assess its ability to achieve their objectives In the context of virtual teams, it is argued that the level of social loafing is an external factor that can affect perceptions of system usefulness
Social loafing in virtual teams may be executed through silence (e.g., a team member who waits for others to generate their inputs, hoping that they will be sufficient) or result in undelivered promises (when a team member promises to do something but does not) Silence and text-based promises may
be difficult to interpret (e.g., ‘‘Are the teammates all working on the task?’’) Such messages (or lack thereof) would be classified
as equivocal According to Media Richness Theory, lean media (i.e., communications media that is deficient in transmitting information such as facial expressions, voice, etc.) is not effective in dealing with messages containing less (or no) information Therefore, with social loafers the underlying lean technology should be perceived as less efficient and useful Thus:
Fig 1 The multilevel research model.
Trang 3H1 Social loafing, as perceived by team members, will negatively
affect the usefulness they attribute to the electronic collaboration
system
Consistent with technology use research, the individual-level
model assumes that the usefulness attributed to the electronic tool
will positively affect users’ intentions to use it in future similar
contexts
H2 Perceived usefulness will positively affect behavioral
inten-tion to use the electronic collaborainten-tion system in a similar context
in the future
3.2 Group-level model
While both team potency and social loafing have been studied in
team contexts, little work has examined the relationship between
them The literature has provided anecdotal evidence that they are
correlated [30] Yet no studies have apparently been made to
develop the relationship between them It is suggested that the
expectancy-value theory can explain the effect of social loafing on
team potency; an individual’s behavioral motivation is affected by
the expectation that the behavior will have a particular
conse-quence, and on the degree of affect towards the outcome Given the
contingency between effort and performance, lower effort results in
lower performance expectancy which can lead to further effort
reduction by other group members Thus, when some virtual team
members loaf, others may also loaf in order to adjust their feelings of
inequity; thus the collective effort will be reduced This will lead to a
collective reduction in the team’s shared belief in its competency
H3 The collective social loafing in a team will negatively affect
team potency
Highly competent teams can be expected to perform better
than others [12] Highly potent teams try to solve challenging
problems and work diligently towards this end, which results in
improved performance
H4 Team potency will positively affect team performance
3.3 Cross-level effects
Macro-level attributes (e.g., organizational climate) often
influence individual-level phenomena (e.g., individual
perfor-mance) Such cross-level effects have been well established in
traditional team contexts[24] The underlying mechanism for such
effects involves two steps First, team members assess the
macro-environment (the team in which they work) based on the observed
behaviors of their teammates Through a process of information
exchange, collection, and assimilation they develop a team
assessment (e.g., is my team competent?) In the second step, this
assessment is used as a partial basis for decision making; attributes
of the macro-environment considered and they can encourage (or
deter) the individuals to act in a certain fashion Arguably, the same
process applies to virtual teams: team attributes are assessed based
on team member behaviors and influence decision making
Specifically, team potency, a shared team-level assessment of
general ability, influences individual team members’ behavioral
intentions to e-collaborate Based on mental accounting theory the
decisions one makes are informed by a range of perceived utility
gains and losses In the virtual team context, one first considers the
value of the IT tool in performing the task Second, users should
consider the team members who use the technology for
communi-cation In essence, the individuals consider the general value of the
IT artifact for the task and then they consider how the particular
team will collaborate using the tool
Potent teams are likely to better utilize lean media in the e-collaboration environment, try harder to resolve any lean-media issues (such as lack of trust, conflict, etc.) and thus overcome the potential deficiencies of the online environment [17] We therefore argued that members of potent teams are more likely
to develop stronger future usage intentions towards an e-collaboration tool After judging that their team is highly competent, team members should expect lower ambiguity, higher quality of submissions, fast responses from peers, and low levels
of social loafing; all of which result in stronger willingness to e-collaborate Thus:
H5 After controlling for individual level perceptions of usefulness, team potency at the group level will incrementally and positively affect team members’ intentions to use online collaboration tech-nology with a similar team in the future
4 Research method
4.1 Participants
Students in an introductory MIS course presented at a US university were asked to complete an online collaboration assignment (worth 10% of their final grade) after which they were asked to voluntarily complete a survey Survey completion was encouraged by adding small grade incentives Business students were selected for this study because they often used online collaboration tools for completing group assignments; thus they were familiar with the technology and the collaborative setup with which our study was concerned
Special attention was given to group size and setting as these are pertinent to the phenomena we were investigating According to social impact theory, there are two macro-explanations for social loafing First, there is a dilution effect which is related to the size of the group: individual members have less motivation to contribute to group effort in larger groups Second, there is an immediacy gap; as members of a group become more isolated (e.g., in virtual teams), they contribute less to the group effort Both drivers existed in our setup In e-collaboration research it is common to use teams of three
to six members To retain power we used groups of 3 or 4 members from different sections of the course
The assignment involved 103 individuals who were randomly divided into 33 groups of 3 members plus one group with 4 members Out of these, 96 individuals completed the post-exercise survey (a 93% response rate) They were nested in 34 groups: 26 with 3 completed surveys, 7 with 2 completed surveys, and one with 4 completed surveys The sample consisted of 51 men (53%) and 45 women with ages ranging from 18 to 56 (with an average age of 23.4) Many participants had worked full-time (from zero to
30 years; with an average of 3 years)
4.2 IT artifact and procedure
Individuals from different class sections were randomly pre-assigned to groups, and asked to collaborate online using only the assigned e-collaboration work-space when producing a report The work-space was a bulletin board through which individuals could exchange ideas and drafts, and develop their final submission by posting messages and files to others in their group Communica-tions were asynchronous, and did not allow the exchange of large media files or the use of streaming media Based on informal discussions with participants, the system was used for three tasks:
(1) Coordinating, timing, and dividing the work;
(2) Sending partial or draft submissions and exchanging ideas; and (3) Reviewing and integrating the partial submissions
Trang 4The assignment asked all teams to produce a short report on a
case study pertaining to privacy issues imposed by the Apple
iPhone It built on materials covered in class and relied on their
personal interpretation, analysis, and opinion
In order to minimize face-to-face collaboration, participants
from different sections were randomly assigned to groups and
asked to collaborate only via the assigned technology
Time-stamped messages were checked to ensure sufficient
collabora-tion Participants were given four weeks to complete
the assignment, after which they were asked to complete a
questionnaire
The respondents had a second group-assignment (given several
weeks after the first); it was similar to the first, in which they chose
between using the collaboration tool or other means of
collabora-tion (e.g., face-to-face meetings); the second assignment was
intended to measure their behavioral intentions
4.3 Measures
The survey instrument included two sections In the first, demographic information, such as age, gender, and work experi-ence was solicited; in the second, items pertaining to the research model were solicited These items were based on existing validated instruments All survey questions were measured by using a one to seven point Likert scale anchored on ‘‘strongly disagree’’ (1) and
‘‘strongly agree’’ (7) The grades that groups got for their team assignment (a maximum of 55 points; were assigned by an instructor external to this study) were used as a proxy for Team Performance This reduced the risk of common method variance because data were recorded from two separate and independent sources Several individuals were asked to review the survey and,
as a result, some minor changes were made The measures are shown inTable 1
Table 1
The measurement instrument.
Team potency (TP) [11] TP1 Our team has confidence in itself
TP2 Our team believes it can become unusually good by producing high-quality work TP3 Our team expects to be known as a high performing team
TP4 Our team feels it can solve any problem it encounters TP5 Our team believes it can be very productive TP6 Our team can get a lot done when it works hard TP7 No task is too tough for our team
Perceived social loafing (SL) [9] SL1 Individuals in this group deferred responsibilities they should assume to other group members
SL2 Individuals in this group put forth less effort when other group members were able to do the work SL3 Individuals in this group did not do their share of the work
SL4 Individuals in this group spent less time working on the task if other group members were available SL5 Individuals in this group put forth less effort than other members of the work group
SL6 Individuals in this group avoided performing undesirable tasks as much as possible SL7 Individuals in this group left work for others to do, which they should really complete SL8 Individuals in this group were less likely to volunteer to do a task if another group member was available SL9 Individuals in this group took it easy if other group members were willing to do the work
SL10 Individuals in this group deferred work to other group members if they were available Perceived usefulness (PU) [28] PU1 Using this e-collaboration tool improves group members’ ability to satisfactorily complete the group project
PU2 Using this e-collaboration tool saves group members’ time in completing the project PU3 Using this e-collaboration tool for completing the project enhances the group’s effectiveness PU4 I find this e-collaboration tool to be useful for completing group projects
Behavioral usage intentions (BI) BI1 Assuming I have access to this e-collaboration tool, and I can choose to use this tool or not, I intend to use
it for the second group project BI2 Given that I have access to this e-collaboration tool, and I can choose to use this tool or not, I predict that
I would use it
Table 2
Descriptive statistics for items and constructs.
Item Mean Std dev Factor loading Residual variance Item-total correlation Cronbach’s alpha Internal consistency Convergent validity (AVE)
Trang 55 Data analysis and results
A combination of data analysis methods was used PLS-Graph
was used for estimating the individual and group level models
(H1–H2 and H3–H4 respectively) Each model was estimated
separately using the relevant dataset (individual- or group-level)
Cross level effects (H5) were estimated using the hierarchical
linear modeling approach with HLM6.04, which allowed
decom-posing variance to within- and between-team components
5.1 Individual-level model estimation
The dataset with 96 individual-level responses was used as
input for this analysis First, common method bias was assessed
through Harman’s single factor test; the possibility was ruled out
because two major factors with opposite signs and reasonably
similar degrees of explained variance (48% and 30%) had resulted
Second, a controlled model that included age, sex, and working
experience was computed None of these control variables were
found to be significant, and thus they were removed from further
analyses Third, a table of descriptive statistics was constructed
(Table 2) to assess construct validity All factor loadings exceeded a
threshold value of 0.7, and all item-to-total correlation values
exceeded 0.35 with relatively low residual variance Construct
reliabilities were further supported by Cronbach’s alphas over 0.80,
measures of internal consistency over 0.7, and measures of
convergent validity over 0.5
A matrix of loadings and cross-loadings was constructed
together with a matrix of inter-construct correlations for assessing
the convergent and discriminant validities of constructs (Tables 3
and 4, respectively) The first demonstrated that all items loaded
on their respective constructs but did not load on other constructs
The second strengthened the validity of the measurement model;
the square root of the average variance extracted (AVE) for each
construct (on the diagonal) was larger than the corresponding
inter-construct correlations (below the diagonal) Overall, the
measurement model was therefore assessed as valid
T-statistics for the structural relationships were obtained using
a bootstrapping procedure with one hundred re-samples Both
coefficients were significant (p < 0.001), which provided support
to the two individual-level hypotheses Individuals who believed that their team members engaged in social loafing perceived the collaboration tool as less useful Perceptions of usefulness, however, augmented users’ behavioral intention to use the online collaboration tool in the future (the second assignment of our study) Perceived social loafing explained 13.4% of the variance of perceived usefulness, and it then explained almost 70% of the variance in behavioral future usage intent
5.2 Group-level model estimation
Team potency and perceived social loafing were assessed by individuals and aggregated to capture group-level feelings The aggregation was based on the assumption that the group-level unit was represented by individual ratings pertaining to the same attribute Thus the average score for a team would be appropriate, given an acceptable level of within-group consensus
Within-group inter-rater agreement (rwg) is often used for assessing consensus, and as the basis upon which aggregation decisions are made It captures the extent to which ratings from different team members are interchangeable While there are other measures of reliability (e.g., ICC), the strengths of rwgare that
it can deal with multi-item scales, and it is not based on between-team variance[7] In our study, rwgwas calculated for the social loafing and team potency scales following the James et al procedure Scores of 0.86 and 0.87 respectively, exceeded the suggested cutoff of 0.7 These scores indicate that individuals who belonged to the same group provided reasonably similar assess-ments of both social loafing and team potency of their groups Furthermore, the two scales were highly reliable at the individual level, with Cronabach’s alpha of 0.97 and 0.96, respectively Given the high within-scale reliability and the acceptable within-group reliability, perceptions of social loafing and team potency were aggregated into group level concepts by taking the mean across scales and team-members (i.e., the team potency of group 1 was calculated as the average of all team potency scale items as reported by all members of group 1) Group construct scores were further normalized (team potency and social loafing were measured on a one to seven scale, whereas performance was measured on a 0–55 scale)
Overall, a dataset with 34 normalized group level observations was used for group-level model estimation Because single indicators were used for capturing group level constructs, and their reliability had been established, the measurement model was irrelevant Descriptive statistics for the non-normalized group-level constructs and intra-construct correlations are shown in
Table 5 Using a bootstrapping procedure with 100 re-samples, the structural model demonstrated that all hypothesized relationships
Table 3
Loadings and cross-loadings.
Perceived social loafing Perceived usefulness Behavioral intentions
Table 4 Inter-construct correlations and square roots of AVE.
Perceived social loafing Perceived usefulness Behavioral intentions
Table 5
Descriptive statistics for group-level constructs.
Trang 6at the group level were supported Groups with high social loafing
developed lower assessments of team potency (b= 0.66, p < 0.01)
Team potency, in turn, positively affected team performance
(b= 0.28, p < 0.05) The collective social loafing in groups explained
44% of the variance in team-level team potency, and team potency
explained 8% of the variation in team performance
5.3 Estimating cross-level effects
Because the dependent variable inH5was at the individual level
and the predictor at the group level, HLM 6 was used First, a null
model which had no predictors at either level was estimated This
was used for assessing the within-group and between-group
variance components, and testing whether there was sufficient
between-group variation for further analysis The results
demon-strated that 78% of the variance in behavioral intentions resided
within teams (individual level), and the rest (ICC = 22%) resided
teams (group level) A chi-square test for the
between-groups variance component (p < 0.05) indicated that it significantly
differed from zero, and a multilevel analysis was thus plausible
Second, a model that includes only individual level effects
(perceived usefulness, as a control variable for us) was constructed
and estimated Centering was performed around group means for
individual-level variables, because group-mean centering is less
biased, and leads to less ambiguous interpretation than other
centering approaches[8] The results demonstrated that perceived
usefulness (centered on group means) had a significant effect on
behavioral intentions (b= 0.87, p < 0.001) After controlling for
perceived usefulness, 66% of the variance was between-teams and
this is significant (p < 0.001) Thus, a multilevel model should be
examined for estimating cross-level effects Therefore a two-level
model was specified and tested In this, perceived usefulness
predicted behavioral intentions at the individual level, and team
potency, at the group level, predicted the intercept of the
behavioral intentions regression equation at the individual level
The results provided support for H5 (g= 0.69, p < 0.001), after
controlling for perceived usefulness (b= 0.87, p < 0.001)
The HLM procedure allowed us to decompose the variance
explained by behavioral intention into within-group and
between-group parts Perceived usefulness explained 56% of the
within-group variance in behavioral intentions, whereas team potency
explained 33% of the between-group variance These components
were combined using the formula of Bryk and Raudenbush[2] The
total R2 showed that the full model explained 51% of the total
variance in behavioral intentions This result was more accurate
than the individual-level PLS analysis because PLS analyses assume
independence of individual records, whereas data from individuals
within a team may not be fully independent Therefore, the final
model reported the HLM results when applicable (seeFig 1)
6 Discussion
We conceptualized and validated a multilevel theory of
technology usage in electronic collaboration settings, linking the
negative effects of social loafing behaviors to future usage intentions
by individuals, and to team performance Our study further covers
important aspects of virtual teams, and distinguished between
within- and between-group variance components
Social loafing emerged as a key concept in e-collaboration
settings Individuals who perceived their group as engaging in it
believed that electronic collaborating means were less useful A
group’s consensus on the social loafing behaviors within the group
affected its potency evaluation Groups with high levels of social
loafing were less likely to believe in their ability to excel across
tasks Teams that believed they could perform well across tasks
outperformed less potent teams
Separating group-level from individual-level effects apparently was warranted, because after controlling for perceived usefulness, 66% of the variance in behavioral intentions was found to be between-teams This further demonstrated that team potency was a macro-level factor that affected team members’ future use intentions, after controlling for individual level assessments of usefulness
Note that an individual-level only model may be flawed in the presence of moderate to high within-team correlations[31] In such models, the interdependence between responses from individuals who belong to the same virtual team is ignored This
is typically done for keeping the model simple, but it can generate inaccuracies[26]
Ultimately, our study demonstrated the importance of group level attributes, through measures taken from individual team members, as well as the potential importance of cross-level effects
in IS research
6.1 Managerial implications
First, managers should be aware of the negative impact that social loafing has on both electronic collaboration technology adoption and team performance They should further devise ways for reducing the actual and perceived social loafing The methods may include reducing team size, increasing team cohesiveness, and emphasizing the importance of teams’ missions [16] Managers should have a face-to-face start-up meeting for their collaboration teams, in which they help team members know each other and to internalize the importance of their mission Alternatively, they should restructure tasks so that interdependencies are reduced, and individuals, with their contribution, are identifiable Finally, managers can implement control mechanisms and milestones to increase oversight and minimize social loafing
6.2 Limitations
Potential limitations of this study included:
1 Our use of student participants may have limited the generalizability of our findings Teams in organizational contexts may face different issues (e.g., budget constraints) Nevertheless, the motivational mechanism (grades as opposed
to appraisals by a supervisor), together with the behavioral future intentions objective, added realism to our findings By using a controlled experimental setting with a well defined task,
we controlled for many external factors, thus focusing on the variables of interest
2 We did not capture changes in behaviors and assessments over time Longitudinal studies would further benefit our under-standing of e-collaboration phenomena
3 We used only one level of context (team level) for explaining usage intentions and performance However, there are higher-levels (division, company, etc.) that may also affect the results
4 Team performance may have been affected by the capabilities
of a single outstanding team member and not the collective To mitigate this concern, the correlation between subject-matter knowledge, as measured by the midterm score of the highest performing team member, and team performance was assessed A correlation of 0.24 (p < 0.17) indicated that there was no significant relationship
7 Conclusion
Our study has presented a multilevel theory of the development
of technology usage intentions by individual members of virtual teams It showed that users of e-collaboration tools consider both system attributes (micro-level factors) and team characteristics
Trang 7(macro-level factors) when developing e-collaboration intent This
seems to be anchored in media richness and mental accounting
theories Taken together, this study shows that social loafing
behaviors within teams can diminish team potency assessments,
perceptions of technology usefulness, and thus, behavioral usage
intentions and team performance As such, ways for reducing
social loafing and increasing team potency in virtual teams should
be explored
References
[1] A.E Akgun, H Keskin, J Byrne, S.Z Imamoglu, Antecedents and consequences of
team potency in software development projects, Information & Management 44
(7), 2007, pp 646–656.
[2] A.S Bryk, S.W Raudenbush, Hierarchical Linear Models: Applications and Data
Analysis Methods, SAGE, Newbury Park, CA, 1992.
[3] A Burton-Jones, M.J Gallivan, Toward a deeper understanding of system usage in
organizations: a multilevel perspective, MIS Quarterly 31 (4), 2007, pp 657–679.
[4] L Chidambaram, L.L Tung, Is out of sight, out of mind? An empirical study of
social loafing in technology-supported groups Information Systems Research 16
(2), 2005, pp 149–168.
[5] J.A Colquitt, J.R Hollenbeck, D.R Ilgen, J.A LePine, L Sheppard, Computer-assisted
communication and team decision-making performance: the moderating effect of
openness to experience, Journal of Applied Psychology 87 (2), 2002, pp 402–410.
[6] A de Jong, K de Ruyter, M Wetzels, Antecedents and consequences of group
potency: a study of self-managing service teams, Management Science 51 (11),
2005, pp 1610–1625.
[7] M.A Dixon, G.B Cunningham, Data aggregation in multilevel analysis: a review of
conceptual and statistical issues, Measurement in Physical Education and Exercise
Science 10 (2), 2006, pp 85–107.
[8] C.K Enders, D Tofighi, Centering predictor variables in cross-sectional multilevel
models: a new look at an old issue, Psychological Methods 12 (2), 2007, pp 121–
138.
[9] J.M George, Extrinsic and intrinsic origins of perceived social loafing in
organiza-tions, Academy of Management Journal 35 (1), 1992, pp 191–202.
[10] S.M Gully, K.A Incalcaterra, A Joshi, J.M Beaubien, A meta-analysis of
team-efficacy, potency, and performance: interdependence and level of analysis as
moderators of observed relationships, Journal of Applied Psychology 87 (5), 2002,
pp 819–832.
[11] R.A Guzzo, P.R Yost, R.J Campbell, G.P Shea, Potency in groups: articulating a
construct, British Journal of Social Psychology 32 (1), 1993, pp 87–106.
[12] T.D Hecht, N.J Allen, J.D Klammer, E.C Kelly, Group beliefs, ability, and
perfor-mance: the potency of group potency, Group Dynamics – Theory Research and
Practice 6 (2), 2002, pp 143–152.
[13] G Johns, The essential impact of context on organizational behavior, Academy of
Management Review 31 (2), 2006, pp 386–408.
[14] P Kanawattanachai, Y Yoo, The impact of knowledge coordination on virtual
team performance over time, MIS Quarterly 31 (4), 2007, pp 783–808.
[15] W.R King, J He, A meta-analysis of the technology acceptance model, Information
& Management 43 (6), 2006, pp 740–755.
[16] R.C Liden, S.J Wayne, R.A Jaworski, N Bennett, Social loafing: a field
investiga-tion, Journal of Management 30 (2), 2004, pp 285–304.
[17] E.M Lira, P Ripoll, J.M Peiro, V Orengo, How do different types of intragroup
conflict affect group potency in virtual compared with face-to-face teams? A
longitudinal study Behaviour & Information Technology 27 (2), 2008, pp 107–
114.
[18] L.L Martins, L.L Gilson, M.T Maynard, Virtual teams: what do we know and
where do we go from here? Journal of Management 30 (6), 2004, pp 805–835.
[19] P.W Mulvey, H.J Klein, The impact of perceived loafing and collective efficacy on
group goal processes and group performance, Organizational Behavior and
Hu-man Decision Processes 74 (1), 1998, pp 62–87.
[20] A Powell, G Piccoli, B Ives, Virtual teams: a review of current literature and directions for future research, Database for Advances in Information Systems 35 (1), 2004, pp 6–36.
[21] K.H Price, D.A Harrison, J.H Gavin, Withholding inputs in team contexts: member composition, interaction processes, evaluation structure, and social loafing, Journal of Applied Psychology 91 (6), 2006, pp 1375–1384.
[22] S Sarker, J.S Valacich, S Sarker, Technology adoption by groups: a valence perspective, Journal of the Association for Information Systems 6 (2), 2005, pp 37–71.
[23] C Steinfield, Realizing the benefits of virtual teams, Computer 35 (3), 2002, pp 104–106.
[24] K Tasa, S Taggar, G.H Seijts, The development of collective efficacy in teams: a multilevel and longitudinal perspective, Journal of Applied Psychology 92 (1),
2007, pp 17–27.
[25] R.H Thaler, Mental accounting and consumer choice, Marketing Science 27 (1),
2008, pp 15–25.
[26] O Turel, Interdependence issues in analyzing negotiation data, Group Decision and Negotiation 19 (2), 2010, pp 111–125.
[27] O Turel, Y Zhang, Does virtual team composition matter? Trait and problem-solving configuration effects on team performance Behaviour & Information Technology 29 (4), 2010, pp 363–375.
[28] V Venkatesh, F Davis, A theoretical extension of the technology acceptance model: four longitudinal field studies, Management Science 46 (2), 2000, pp 186– 204.
[29] J Webster, S Staples, Comparing virtual teams to traditional teams: an identifi-cation of new research opportunities, in: J.J Martocchio (Ed.), Research in Personal and Human Resources Management, (25), Elsevier, Boston, 2006, pp 181–214.
[30] J.W Whiteoak, The relationship among group process perceptions, goal commit-ment and turnover intention in small committee groups, Journal of Business and Psychology 22 (1), 2007, pp 11–20.
[31] O Turel, Y.F Yuan, C.E Connelly, In justice we trust: Predicting user acceptance of e-customer services, Journal of Management Information Systems 24 (4), 2008,
pp 123–151.
Ofir Turel is a Professor of Information Systems and Decision Sciences at the College of Business and Economics, California State University, Fullerton Before joining the academia, he held senior positions in the information technology and telecommunications in-dustries His research interests include a broad range of behavioral and managerial issues in various informa-tion systems contexts His work received several national and international awards, and was presented
in many conferences He published over 30 articles in journals such as MIS Quarterly, Journal of MIS, Communications of the ACM, Information & Manage-ment, Journal of Information Systems, Behavior & Information Technology, Telecommunications Policy, Group Decision and Negotiation, and Communications
in Statistics.
Yi (Jenny) Zhang is an Associate Professor in the Department of Information Systems and Decision Sciences at California State University, Fullerton She holds a B.S in Electronic Engineering, a M.S and Ph.D in Information Systems Her current research interests include virtual teams, virtual communities, and busi-ness intelligence Her work has been published in Behavior & Information Technology, the International Journal of E-Business Research, the Journal of Informa-tion Privacy & Security, and the Journal of EducaInforma-tion for Business.