Content uploaded by Hans Georg Gemuenden
Author content
All content in this area was uploaded by Hans Georg Gemuenden
Content may be subject to copyright.
1047-7039/01/1204/0435/$05.00
1526-5455 electronic ISSN
O
RGANIZATION
S
CIENCE
,2001 INFORMS
Vol. 12, No. 4, July–August 2001, pp. 435–449
Teamwork Quality and the Success of
Innovative Projects: A Theoretical Concept
and Empirical Evidence
Martin Hoegl • Hans Georg Gemuenden
Washington State University, Department of Management and Decision Science, 601 W. First Avenue,
Spokane, Washington 99201–3899
Technical University of Berlin, Chair for Technology and Innovation Management, Hardenbergstr. 4–5,
HAD 29, 10623 Berlin, Germany
hoegl@wsu.edu • hans.gemuenden@tim.tu-berlin.de
Abstract
An extensive body of literature indicates the importance of
teamwork to the success of innovative projects. This growing
awareness, that “good teamwork” increases the success of in-
novative projects, raises new questions: What is teamwork, and
how can it be measured? Why and how is teamwork related to
the success of innovative projects? How strong is the relation-
ship between teamwork and various measures of project success
such as performance or team member satisfaction? This article
develops a comprehensive concept of the collaboration in
teams, called Teamwork Quality (TWQ). The six facets of the
TWQ construct, i.e., communication, coordination, balance of
member contributions, mutual support, effort, and cohesion, are
specified. Hypotheses regarding the relationship between TWQ
and project success are tested using data from 575 team mem-
bers, team leaders, and managers of 145 German software
teams. The results of the structural equation models estimated
show that TWQ (as rated by team members) is significantly
associated with team performance as rated by team members,
team leaders, and team-external managers. However, the mag-
nitude of the relationship between TWQ and team performance
varies by the perspective of the performance rater, i.e., manager
vs. team leader vs. team members. Furthermore, TWQ shows a
strong association with team members’ personal success (i.e.,
work satisfaction and learning).
(Teamwork;Innovation;Software Development)
The importance of teams to the success of innovation pro-
cesses is well documented in the theoretical literature.
Popular approaches to new product development (Johne
and Snelson 1990, Clark and Fujimoto 1991, Cooper
1993) such as project management (Pinto et al. 1993,
Fleming and Koppelman 1996, Gemuenden and Lechler
1997), speed and cycle time management (Gemuenden
1990, Eisenhardt and Tabrizi 1995), as well as total qual-
ity management and continuous improvement (Griffin
and Hauser 1992, Lawler et al. 1995, Mohrman et al.
1995) consider teamwork as a crucial success factor.
Empirical evidence regarding the influence of team-
work on the success of teams with innovative tasks, how-
ever, is lacking in two ways. First, past research has often
employed rather crude measures of teamwork, without
adequately addressing the multifaceted nature of the func-
tioning of teams (Denison et al. 1996). While many studies
demonstrate the link between the mere existence of a team-
based organization and innovative performance (Gupta et
al. 1987, Hise et al. 1990, Cooper and Kleinschmidt 1995,
Gupta and Wilemon 1996), they fail to specify and mea-
sure the collaborative work of teams and to illustrate how
this relates to various aspects of the success of innovative
projects. Second, existing empirical evidence regarding
the impact of teamwork on the success of innovative proj-
ects is conflicting. As past studies report influences of
varying magnitude on team members’ ratings of perfor-
mance, managerial ratings or other measures of innova-
tive performance (e.g., number of ideas) seem less af-
fected by the quality of teamwork (Thamhain and Kamm
1993, Campion et al. 1993, Campion et al. 1996, Cohen
et al. 1996).
The present study aims to contribute to the clarification
of these critical issues. In particular, it addresses the fol-
lowing questions: (1) What is “teamwork,” and how can
it be measured? (2) Why and how is teamwork related to
the success of innovative projects? (3) How strong is the
relationship between teamwork and different measures of
project success? Thus, the primary contribution of this
research is to provide a comprehensive understanding of
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
436 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
teamwork quality, to develop a valid and reliable measure
of this construct, and to test its influence on different as-
pects of project success using data from 575 interviews
with members, leaders, and managers of 145 software
development teams in Germany.
Teamwork Quality: A Measure of
Collaboration in Teams
Following the literature, a team can be defined as a social
system of three or more people, which is embedded in an
organization (context), whose members perceive them-
selves as such and are perceived as members by others
(identity), and who collaborate on a common task (team-
work) (Alderfer 1987, Hackman 1987, Wiendieck 1992,
Guzzo and Shea 1992).
The many experienced managers who emphasize the
merits of “good teamwork” have numerous behavioral
requirements in mind. Their claims remain rather vague
and meaningless, however, as long as the essence of a
team, the quality of its collaborative working, is neither
precisely defined nor validly and reliably measured. We
advocate first developing theories that explain which as-
pects of teamwork are relevant to team performance and
then testing these propositions to make distinctions that
are useful for practical purposes (i.e., distinguishing be-
tween teams that achieve results and those that do not).
Following the work of Homans (1974) on the elemen-
tary forms of social behavior, we can conceptualize hu-
man behavior in teams as activities, interactions, and sen-
timents. Activities are observable actions of individuals
that can be measured by quantity (e.g., the production
output of a factory worker) as well as by the correctness
of their execution (e.g., the effectiveness of an action).
Interaction refers to the connectedness or the “being in
contact” of two or more people regardless of the activities
that bring them into contact. According to Homans
(1974), interaction can be studied in terms of frequency
and intensity. The third element of social behavior is sen-
timent, which refers to human emotions, motivations, or
attitudes. Sentiments cannot be directly observed, but
nevertheless influence interactions and activities and are,
in turn, influenced by them.
The focus of this research is solely on the quality of
interactions within teams rather than team members’
(task) activities. Starting from the widespread fundamen-
tal proposition that the success of work conducted in
teams depends (beyond the quantity and correctness of
the task activities) on how well team members collabo-
rate, or interact, we propose the construct teamwork qual-
ity (TWQ) as a comprehensive concept of the quality of
interactions in teams. To capture the nature of team mem-
bers’ working together, we specify six facets of the col-
laborative team process that integrate to the concept of
TWQ. These facets capture both task-related and social
interaction within teams. The quality of interactions with
external parties (e.g., management, other teams, etc.) is
not part of the TWQ construct (Gladstein 1984, Ancona
and Caldwell 1988, Ancona and Caldwell 1990). The six
TWQ facets—communication, coordination, balance of
member contributions, mutual support, effort, and cohe-
sion—comprise performance-relevant measures of team
internal interaction. Thus, the focal point is the quality
of a team’s collaborative work rather than the content of
its tasks and activities (e.g., we are concerned with how
well team members communicate task-relevant informa-
tion, not with what their communication is about). There-
fore, measures of the team-task process, such as the ap-
propriateness of the task strategy employed (Hackman
1987, Denison et al. 1996) or the quality with which task
activities are carried out by team members, are not the
subject of the TWQ construct as we propose it. Further-
more, leadership processes including activities such as
goal setting, task planning, resource acquisition and dis-
tribution, and task controlling, as well as feedback and
performance appraisal, are not within the scope of the
TWQ concept. These variables refer to the content of task
activities rather than to the quality of interactions within
teams. We do assume that there is a connection between
many of the above-mentioned team-based task content
and leadership activities such that high TWQ would fa-
cilitate the efficiency and effectiveness of the execution
of these activities in a team setting; however, these pos-
sible relationships are not investigated in the present re-
port.
Based on this understanding of interaction in teams, our
review of the relevant literature and various exploratory
empirical case studies conducted on this subject has re-
sulted in six concepts that are descriptive of the quality
of collaborative working in teams. The table below gives
an overview of the six concepts included in the TWQ
construct and the principal questions they address.
We conceptualize TWQ as a multifaceted higher order
construct. The underlying proposition of this construct is
that highly collaborative teams display behaviors related
to all six TWQ facets. Thus, these six facets are indicators
of the collaborative work process in teams and combine
to the TWQ construct. This conceptualization as a higher
order (latent) construct is comparable to Hackman’s
(1987) “process criteria of effectiveness” in that several
critical indicators (or subconstructs) are combined in the
specification of the team task process. The following is a
discussion of the TWQ facets that provides a detailed
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 437
Table 1 The Teamwork Quality Construct
•Communication
Is there sufficiently frequent, informal, direct, and open
communication?
•Coordination
Are individual efforts well structured and synchronized
within the team?
•Balance of Member Contributions
Are all team members able to bring in their expertise to
their full potential?
•Mutual Support
Do team members help and support each other in carrying
out their tasks?
•Effort
Do team members exert all efforts to the team’s tasks?
•Cohesion
Are team members motivated to maintain the team? Is
there team spirit?
description of these concepts as they relate to the TWQ
construct as a whole.
Communication
The most elementary component of TWQ is the com-
munication within a team. Communication provides a
means for the exchange of information among team mem-
bers (Pinto and Pinto 1990). The quality of communica-
tion within a team can be described in terms of the fre-
quency, formalization, structure, and openness of the
information exchange. While frequency refers to how ex-
tensively team members communicate (i.e., time spent
communicating), the degree of formalization describes
how spontaneously team members are able to converse
with each other. Communication that requires a large
amount of preparation and planning before it can occur
(e.g., scheduled meetings, written status reports) is con-
sidered more formal, whereas spontaneously initiated
contacts (e.g., talks in the hallway, quick phone calls,
short e-mails) constitute informal communication. It is
this informal, spontaneous communication that has been
shown to be crucial to the work of teams with innovative
projects because ideas and contributions can be shared,
discussed, and evaluated with other team members more
quickly and efficiently (Katz 1982, Pinto and Pinto 1990,
Brodbeck 1994, Domsch and Gerpott 1995). In addition,
it is important to the quality of collaboration in teams that
team members be able to communicate directly with all
other team members (communication structure) because
the exchange of information through mediators (e.g.,
team leader) is time consuming and a possible cause of
faulty transmission. Apart from frequency, formalization,
and structure, it is critical to TWQ that members share
their information openly with each other (Gladstein 1984,
Pinto and Pinto 1990). A lack of openness within a team
(i.e., holding back important information) hinders the
most fundamental function of teamwork, namely the in-
tegration of team members’ knowledge and experience
on their common task.
Coordination
The degree of common understanding regarding the
interrelatedness and current status of individual contri-
butions also determines the quality of teamwork per-
formed. While teams must work together on fundamen-
tal aspects of a common task, many activities in the task
process should be delegated to individual members
working on parallel subtasks. One important component
of the quality of collaboration in teams is the harmoni-
zation and synchronization of these individual contribu-
tions (Tannenbaum et al. 1992, Larson and Schaumann
1993, Brannick et al. 1995). To do this effectively and
efficiently, teams need to agree on common work-down
structures, schedules, budgets, and deliverables. Thus, co-
ordination means that the teams have to develop and
agree upon a common task-related goal structure that has
sufficiently clear subgoals for each team member, free of
gaps and overlaps.
Balance of Member Contributions
It is important to the quality of teamwork that every team
member is able to contribute all task-relevant knowledge
and experience to the team (Hackman 1987, Seers et al.
1995). This is especially critical for teams with innovative
tasks because they often consist of members whose ex-
pertise is in different functional areas (e.g., R&D, mar-
keting, finance, etc.). It would defeat the purpose of such
cross-functional teams if some team members could not
bring in their views and ideas because others were domi-
nating discussions and decision-making processes. There-
fore, it is considered essential to TWQ that contributions
to the team task are balanced with respect to each mem-
ber’s specific knowledge and experience. While not ev-
eryone must bring in, for instance, the exact same number
of ideas, no one should be limited in presenting and con-
tributing relevant knowledge to the team.
Mutual Support
Building on the work of Tjosvold (1984, 1995), we find
that mutual support among team members is an essential
component of TWQ. The intensive collaboration of in-
dividuals depends upon a cooperative rather than a com-
petitive frame of mind. Without questioning the motiva-
tional potential of competition in the case of independent
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
438 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
individual tasks, Tjosvold (1995) argues that, for inter-
dependent tasks, mutual support is more productive than
are the forces of competition. Thus, team members work-
ing on a common goal should display mutual respect,
grant assistance when needed, and develop other team
members’ ideas and contributions rather than trying to
outdo each other. Competitive behaviors in a team lead
to distrust and frustration, whereas mutual support fosters
the integration of team members’ expertise and is, there-
fore, a critical aspect of the quality of collaboration in
teams.
Effort
Norms are defined as shared expectations regarding the
behavior of team members (Levine and Moreland 1990,
Goodman et al. 1987, Helfert 1998). While such shared
expectations can exist for every kind of observable be-
havior in teams, norms regarding the effort of team mem-
bers are of particular importance to TWQ. Workload
sharing and prioritizing of the team’s task over other ob-
ligations are indicators for the effort team members exert
on the common task (Hackman 1987, Pinto and Pinto
1990, Campion et al. 1993). To achieve high TWQ and
avoid conflict among team members, it is important for
everyone in the team to know and accept the work norms
concerning sufficient effort. A uniformly high level of
effort by all team members is primary to the quality of
collaboration.
Cohesion
Team cohesion refers to the degree to which team mem-
bers desire to remain on the team (Cartwright 1968). Sev-
eral forces play a part in a person’s desire to stick with
the team. In their meta-analysis (including 49 empirical
studies) Mullen and Copper (1994) distinguish between
three forces of cohesion: (1) interpersonal attraction of
team members, (2) commitment to the team task, and
(3) group pride-team spirit. High TWQ can hardly be
achieved without an adequate level of cohesion. If team
members lack a sense of togetherness and belonging, if
there is little desire to keep the team going, then intensive
collaboration seems unlikely. An adequate level of co-
hesion is necessary to maintain a team, to engage in col-
laboration, and thus to build the basis for high TWQ.
Success of Innovative Projects
The conceptualization of project success as a multi-
variable construct is widely acknowledged in the litera-
ture (Gemuenden 1990, Pinto et al. 1993, Denison et al.
1996, Gemuenden and Lechler 1997, Hauschildt 1997).
Many authors distinguish between task-related outcomes
(e.g., quality, adherence to budget and schedule) and
people-related outcomes (e.g., team member satisfaction,
viability of the team). This distinction was adopted for
the present study because the outcome categories team
performance and the personal success of team members
are considered separately.
Thus, our conceptualization of project team success (as
well as the measurement model built on it) is following
a consistent stream of team-success modeling, including
Gladstein (1984), Hackman (1987), and models published
since then (e.g. Sundstrom et al. 1990, Tannenbaum et al.
1992, Denison et al. 1996). The following table illustrates
the relations between the conceptualization of project
team success as applied in the present study with the ap-
proaches taken by Gladstein (1984), Hackman (1987),
and Denison et al. (1996). As shown, our concept builds
and expands on these three models.
Team Performance
Team performance can be defined as the extent to which
a team is able to meet established quality and cost and
time objectives (Gemuenden 1990, Schrader and Goepfert
1996, Gemuenden and Lechler 1997). The perception of
project success depends, in part, on the perspective of the
evaluator. Thus, it is important to include multiple views
(e.g., of the company, the customer, the team) when rat-
ing team performance. It must also be acknowledged that
setting clear and precise performance objectives at the
outset of a project is particularly difficult in the case of
innovations because the subject matter is often highly
complex and uncertain (Gemuenden 1995, Hauschildt
1997).
For the purpose of the present study, team performance
is described in terms of the variables effectiveness and
efficiency. Effectiveness refers to the degree to which the
team meets expectations regarding the quality of the out-
come. In the case of innovative projects, an effective per-
formance regularly entails adherence to predefined qual-
itative properties of the product, service, or process to be
developed, e.g., functionality, robustness, reliability, per-
formance, etc. The team’s efficiency is assessed in terms
of adherence to schedules, e.g., starting the manufactur-
ing and/or marketing on the target date, and budgets, e.g.,
staying within target costs with both the project and the
finished product. Thus, effectiveness reflects a compari-
son of actual versus intended outcomes, whereas effi-
ciency ratings are based on a comparison of actual versus
intended inputs.
Personal Success of Team Members
In addition to achieving performance objectives, teams
must also work in a way that increases members’ moti-
vation and ability to engage in future teamwork
(Hackman 1987, Sundstrom et al. 1990, Denison et al.
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 439
Table 2 Comparison of Team Success Conceptualizations
Present Study Gladstein 1984 Hackman 1987 Denison et al. 1996
Effectiveness (Quality)
Efficiency (Schedule and Budget)
Performance Acceptability of output to those who
receive/review it.
Information Creation
Time Compression
Overall Effectiveness
Work Satisfaction
(personal satisfaction of team members;
desire to work in teams in the future)
Satisfaction Members’ needs are more satisfied
than frustrated by the group experience.
Capability of members to work together
in the future is strengthened.
Growth Satisfaction
Learning
(acquisition of knowledge and skills)
— — Learning
Capability Development
Figure 1 Teamwork Quality and the Success of Innovative
Projects
1996). For the purpose of this investigation, the two vari-
ables satisfaction and learning build the category per-
sonal success of team members. Satisfaction with working
in teams leads to increased motivation for participating
in future team projects. Also, collaborating with other
people provides the opportunity for learning social, proj-
ect management, technical, and creative skills. Such ac-
quisition of new skills relates to team members’ desire
for personal and professional growth (Denison et al.
1996) as well as increasing the potential of the team mem-
bers for future teamwork.
Hypotheses
It is the central proposition of this article that TWQ is
positively related to the success of innovative projects.
Figure 1 depicts the two categories of project success
(i.e., team performance and personal success of team
members) as well as the proposed positive relationship
with TWQ.
This conceptual framework constitutes a process-
outcome model. In comparison to comprehensive input-
process-outcome models of team effectiveness (Gladstein
1984, Hackman 1987, Tannenbaum et al. 1992), our
model focuses on the collaborative work process (i.e.,
TWQ) and illustrates how this affects various team out-
comes. As such, our research framework specifically ad-
dresses the “black box” of input-outcome models of team
effectiveness (Guzzo and Shea 1992) and offers a com-
prehensive concept of team members’ collaboration on
their common task. Given this focus, however, possible
antecedents of TWQ (i.e., inputs) are not investigated in
this report.
Teamwork Quality and Team Performance
TWQ is a measure for the quality of collaboration in
teams and consists of six facets: communication, coor-
dination, balance of member contributions, mutual sup-
port, effort, and cohesion. Building on the discussion of
the TWQ construct above, we will present below the
theoretical rationale as well as empirical evidence for the
proposed positive relationship between TWQ and team
performance on the basis of every TWQ facet.
Communication. It is widely agreed upon in the liter-
ature that the flow of communication within teams influ-
ences the success of innovative projects (Griffin and
Hauser 1992). Research by Katz and Allen (1988), in-
volving 50 R&D teams, demonstrates a strong positive
impact of within-team communication on project success.
More recent studies confirm this fundamental finding.
Hauptman and Hirji (1996), in their investigation on 50
cross-functional project teams, show that frequent two-
way communication within teams exerts a positive influ-
ence on team performance. In addition, a large-scale em-
pirical study in Germany on the success factors of project
management finds communication and information flow
to be a direct prerequisite of project success (Gemuenden
and Lechler 1997).
Coordination. Coordinating contributions from differ-
ent functional areas efficiently is of critical importance to
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
440 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
the successful design and implementation of innovations.
Teams can be seen as a mechanism for integrating various
skills needed to perform complex and uncertain tasks suc-
cessfully. From this perspective, teams provide a form of
organization in which individuals with different expertise
can collaborate directly on a task and, thus, achieve a very
high degree of coordination among each other (Lawrence
and Lorsch 1967, Nadler and Tushman 1988, Adler 1995).
A study by Gupta et al. (1987), which includes 167 high
tech companies in the United States, confirms this widely
held opinion. Highly integrated companies use mainly
teams (venture teams and temporary task forces) in their
new product development.
Balance of Member Contributions. It is essential to the
success of innovative projects that all team members feel
free to bring in their task-relevant expertise. If discussions
and decision making are dominated by some team mem-
bers, and others in the team are unable to contribute their
views and ideas, then this will have negative conse-
quences on the team’s performance (i.e., quality, costs,
time) because avoidable mistakes are likely to be made.
In his empirical study, Seers (1989) demonstrates that the
balance of member contributions is significantly related
to both task performance and team-member satisfaction.
Mutual Support. According to Tjosvold (1984, 1995),
it is important for team members not to compete (e.g., for
resources or prestige), but to cooperate to achieve a com-
mon goal. A laboratory study by Cooke and Szumal
(1994), involving 64 groups of students, demonstrates
that constructive-cooperative behaviors within teams (as
opposed to aggressive interaction or passivity) increase
both the quality and the acceptance of the solutions de-
veloped. It is further reasonable to assume that the level
of mutual support impacts team performance through its
influence on communication and coordination within
teams.
Effort. The effort that team members exert on their
common task influences the success of the project
(Hackman 1987). This proposition reflects the fundamen-
tal assumption that, independent of other factors such as
task-relevant knowledge and skills, the level of effort
brought to bear on a task influences performance. A study
by Weingart (1992) provides support for this proposition
at the team level of analysis. The results from data of 56
student groups indicate that effort, among other variables
such as planning and coordinating of tasks, has a signifi-
cant influence on team performance.
Cohesion. Even though the empirical evidence of the
influence of team cohesion on performance is not entirely
conclusive (see Janis 1995 for the detrimental effects of
the groupthink phenomenon in the case of very high lev-
els of cohesion, particularly for longstanding groups), a
number of recent reviews and empirical studies underline
a positive relationship between the two. As one main re-
sult of their meta-analysis, including 49 empirical studies,
Mullen and Copper (1994) report that it is chiefly the
commitment to the task (as an indicator of cohesion) that
shows a significant impact on team performance. Gully
et al. (1995) include 51 effects of 46 empirical investi-
gations in their meta-analysis and conclude that cohesion
influences performance, particularly if the team task re-
quires coordination and communication (e.g., innovative
tasks). These results indicate that an adequate level of
cohesion impacts the performance of innovation teams
through its positive influence on communication and co-
ordination.
The theoretical considerations and the evidence pre-
sented on the basis of the six TWQ facets result in the
following hypothesis:
H
YPOTHESIS
1. TWQ is positively related to the per-
formance of teams with innovative projects.
Teamwork Quality and the Personal Success of
Team Members
In addition to the positive relationship with team perfor-
mance, TWQ is assumed to influence the personal success
of team members. High TWQ leads to team members’
satisfaction with their work situation and provides an op-
portunity for team members to acquire knowledge and
skills. Recent empirical evidence provides support for this
assumption. Research by Pinto et al. (1993), involving 72
task forces, shows that within-team cooperation influ-
ences both performance and team member satisfaction.
These results are confirmed by the studies of Campion et
al. (1993) and in a later replication by Campion et al.
(1996). Both empirical investigations demonstrate strong
relationships between measures of within-team coopera-
tion and communication and the satisfaction of team
members. In light of these theoretical considerations as
well as of the empirical evidence, the following hypoth-
esis is proposed:
H
YPOTHESIS
2. TWQ is positively related to the per-
sonal success of team members.
Methods
Sample and Data Collection
A total of 145 software development teams from four
German software development laboratories participated
in this research. All four laboratories were part of larger
organizations, with two of them being independent opera-
tions of the same U.S. parent company. The other two
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 441
laboratories belong to organizations headquartered in
Germany. Each laboratory employed between 100 and
500 software developers. Included in the study were soft-
ware development projects that were undertaken within
18 months prior to data collection and that were worked
on by teams that met the above-described definition (i.e.,
three or more members, common identity). The first au-
thor was provided a list of projects, including names and
contact information of team members, while the employ-
ees of the software laboratories were informed that a
study about team management was to be conducted. All
team leaders and team-external managers as well as ran-
domly chosen team members were contacted for inter-
view appointments. Respondents’ participation in this
study was strictly voluntary. All contacted respondents
were interviewed.
Data were gathered by the first author and research as-
sistants in individual interviews using a fully standardized
questionnaire (five-point answer scale). All interviews
were conducted on site in dedicated interview rooms
which assured similar conditions for every interview. The
interviews followed a very structured pattern. First, team
membership as stated on the list was confirmed with the
respondent to ensure that he or she was indeed a member
of the team in question and to ascertain that all respon-
dents of one team were referring to the same set of in-
dividuals as the team. Then the questionnaire was handed
over for the respondent to complete by reading it himself
or herself. This way, possible interviewer effects were
minimized, while there was still an interviewer present to
clarify questions if any occurred. The time consumption
of one interview was about 45 minutes. A total of 575
interviews with members, leaders, and (team external)
managers referring to 145 software development teams
was conducted. Multiple respondents from each team,
one of whom was the team leader, were interviewed to
obtain more reliable ratings of the team level constructs
in consideration. On average, three members of each team
were interviewed. Of the team members, 22% were fe-
male. Our sample contains 26% female respondents. The
teams in this sample have an average of 6.3 members
(median 6, standard deviation 3) and an average
age of all team members of 36.6 years (median 36,
standard deviation 5).
Measures
All constructs considered in this investigation refer to the
team as the unit of analysis. Accordingly, all measures
were specified on the team level. Thus, respondents were
asked to evaluate properties and behaviors of the team as
a whole. The questionnaire was administered in the Ger-
man language. All constructs investigated in this study
were assessed using multiple questionnaire items. Trans-
lations of the items used in the following analyses are
included in the appendix. A pretest was conducted, in-
cluding 23 members of product development teams, at a
machine tool manufacturing company. Following this
pretest, the wording of some items was refined for later
use in the present study.
The TWQ facets were measured using three to 10 items
each. The items were specifically generated for the pres-
ent study based on the discussions of these constructs in
the literature (see the construct descriptions above). Some
items used by Pinto et al. (1993) were translated to Ger-
man for inclusion in the communication and mutual sup-
port scales. The discussions by Seers (1989) and Seers et
al. (1995) gave guidance in generating the items for bal-
ance of member contributions.
The measurement scales for effectiveness and effi-
ciency were based partly on the scales used by
Gemuenden and Lechler (1997) in their large-scale study
of project management in Germany. Effectiveness, i.e.,
the technical quality of the software solution, including
the satisfaction with the software solution from different
perspectives (customer, team), was measured using 10
items. Another five items were used for measuring the
teams’ adherence to schedule and budget (i.e., efficiency).
Team members’ work satisfaction and learning were
assessed using three and five items, respectively. The items
for these scales were generated based on the discussions
of the constructs by Hackman (1987), Sundstrom et al.
(1990), and Denison et al. (1996). All measurement scales
used in this study demonstrated strong reliability (Cronbach’s
alpha coefficients between 0.72 and 0.97).
To ensure content validity and avoid a possible com-
mon source bias, data from different respondents were
used to measure the different variables. TWQ and team
members’ personal success were measured using aggre-
gated responses from multiple team members (excluding
team leaders). The team performance variables, however,
were assessed from three perspectives. Team members,
team leaders, and team-external managers were all asked
to rate effectiveness and efficiency using the same mea-
surement scales.
Explanatory statements made by managers during in-
terviews and after the study suggest that their judgements
were primarily reliant on customer feedback (once the
software was in operation) for assessing quality and on
project controlling reports for assessing budget and time
performance. Some managers explained to us that they
could offer a “bird’s-eye view” of project results because
they were not very familiar with how individual projects
progressed (i.e., task process). Team members, in con-
trast, often explained that they were less familiar with
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
442 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
Table 3 Number of Items, Means, Standard Deviations, Reliabilities, and Intercorrelations
Items Mean SD Alpha
b
IRR
c
12345678910111213
TWQ
a
(1) Communication 10 4.20 0.44 0.94 0.95
(2) Coordination 4 4.04 0.59 0.85 0.79 0.58
(3) Balance of
Member Cont. 3 4.08 0.50 0.72 0.84 0.73 0.54
(4) Mutual Support 7 4.13 0.55 0.93 0.93 0.80 0.55 0.75
(5) Effort 4 3.91 0.59 0.94 0.88 0.61 0.52 0.65 0.62
(6) Cohesion 10 3.89 0.56 0.97 0.94 0.69 0.49 0.74 0.75 0.78
Team Performance (Team Member Rating)
(7) Effectiveness-TM 10 3.91 0.59 0.91 0.96 0.41 0.39 0.45 0.44 0.36 0.42
(8) Efficiency-TM 5 3.76 0.77 0.86 0.91 0.36 0.28 0.38 0.42 0.34 0.38 0.64
Team Performance (Team Leader Rating)
(9) Effectiveness-TL 10 4.11 0.55 0.90 — 0.20 0.25 0.19 0.18 0.20 0.19 0.54 0.32
(10) Efficiency-TL 5 3.93 0.78 0.86 — 0.17 0.21 0.23 0.19 0.24 0.24 0.33 0.51 0.58
Team Performance (Manager Rating)
(11) Effectiveness-Mgr 10 4.11 0.62 0.87 — 0.22 0.22 0.15 0.18 0.14 0.15 0.37 0.20 0.40 0.24
(12) Efficiency-Mgr 5 3.96 0.85 0.85 — 0.17 0.10 0.16 0.17 0.18 0.19 0.17 0.37 0.30 0.47 0.61
Personal Success of Team Members
a
(13) Work Satisfaction 3 4.01 0.55 0.79 0.86 0.68 0.51 0.72 0.73 0.60 0.77 0.62 0.48 0.34 0.24 0.24 0.18
(14) Learning 5 4.06 0.46 0.76 0.92 0.48 0.34 0.49 0.52 0.51 0.65 0.38 0.31 0.13 0.08 0.07 0.01 0.64
a
team member ratings
b
cronbach’s alpha coefficient
c
coefficient of interrater reliability (James et al. 1984)
project controlling reports and often had less contact with
the customer after the development project was com-
pleted. The team leaders, when asked by the interviewers,
appeared well aware of both the “manager’s world,” i.e.,
project controlling reports and customer reactions, and
the “team members’ world,” i.e., operational details about
the task process. Thus, the three types of respondents (i.e.,
team members, team leaders, and managers) represent
three different perspectives drawing on varying sources
of information in their evaluations of team performance.
Prior to aggregating team members’ evaluations, inter-
rater agreement (James 1982, James et al. 1984, Campion
et al. 1993) was assessed using the multiple-item esti-
mator for within-group interrater reliability as proposed
by James et al. (1984). This test yielded results indicating
generally very strong agreement of ratings referring to the
same team. The average scores of this test across all teams
are included in the table below for the applicable con-
structs (i.e., team member ratings of TWQ, team perfor-
mance, and personal success). Given this homogeneity of
within-team ratings, data were aggregated by calculating
the arithmetic mean.
The database for this empirical study includes teams
from four different software development laboratories.
Prior to combining the data, the samples were tested for
possible differences in means. The one-way ANOVA
yielded significant results (p⬍0.01) only for effective-
ness as rated by the team leaders (Effectiveness-TL). Fur-
ther t-tests revealed that Laboratory 2 and 3 showed sig-
nificantly different means for this variable. Apart from
this, the four samples showed generally strong homoge-
neity in both means and variances for all variables inves-
tigated in this study.
Data Analysis
Teamwork Quality as a Higher Order Construct. Fac-
tor analyses were conducted to assess whether all six
specified TWQ facets pertain to the same latent construct.
First, we conducted a factor analysis (principle-
component method) on the team level using aggregated
team member responses. The results confirm the latent
construct assumption concerning the TWQ construct.
Following the Kaiser criterion, the one-factor solution can
be accepted because the eigenvalue of the second com-
ponent falls far below one. All six facets are loading high
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 443
Table 4 Teamwork Quality as a Higher Order Construct:
Factor Analysis, Regression Analysis, Reliability
Analysis
TWQ Facet
Factor
Loading
Std.
Regr.
Coefficients P-Value
Communication 0.88 0.16 0.00
Coordination 0.71 0.22 0.00
Balance of Member Contributions 0.89 0.19 0.00
Mutual Support 0.89 0.20 0.00
Effort 0.82 0.22 0.00
Cohesion 0.89 0.20 0.00
Eigenvalue 4.29 R
2
100%
Variance explained
(Factor TWQ) 71.51%
Cronbach’s alpha
coefficient 0.91
Note. N 145.
on one factor. The factor TWQ explains 71.51 % of the
variance of the six scales.
To ensure that the results obtained at the team level
(N145) are not the result of inflated correlations due
to data aggregation, we have conducted further factor
analyses at the individual level (N285). In doing so,
possible dependencies of observations within one team
must be dealt with. We have employed two procedures
to test the TWQ factor structure at the individual level.
First, we have randomly sampled 145 responses and con-
ducted factor analyses. This process was repeated 15
times. All individual-level factor analyses support the
team-level findings. All analyses resulted in the extraction
of one factor using the Kaiser criterion. Second, to test
the factor structure at the individual level including all
285 responses, we have regressed group on every TWQ
facet index and saved the standardized residuals for fur-
ther analysis. We have used the standardized residuals
from this procedure (“purified” from group effect) as in-
put for another factor analysis including all 285 team
member responses. The results again confirm the team-
level factor analysis. One factor was extracted using the
Kaiser criterion.
Furthermore, we followed the procedures suggested by
Henik and Tzelgov (1985) to control for a possible halo
effect whereby team members’ general sense of their
team’s success may have caused them to give correlated
ratings across the TWQ facets. We have regressed the first
item of the effectiveness scale as a general indicator of
project success (“Going by the results, this project can be
regarded as successful”) on the standardized residuals
from the group effect procedure above. The standardized
residuals (“purified” from halo effect) were saved and
used as input for another factor analysis at the individual
level (N285). Again, a one-factor solution for the six
TWQ facets emerged, indicating that the six TWQ facets
pertain to the same latent construct.
The following table shows the loadings of the team-
level factor analysis as well as the standardized regression
coefficients of a linear regression with the (aggregated)
TWQ construct as the dependent variable.
Structural Equation Modeling. Structural equation
modeling was the principal analytic technique employed
in this study. Using the statistical software package
AMOS, version 4.0, structural equation modeling (SEM)
allows simultaneous testing of the measurement and
structural models. Analyses were conducted using the
unweighted-least-squares method (ULS). To reduce the
number of free parameters in the model (to be appropriate
given the number of cases; N145 teams) (Bagozzi and
Yi 1988), the multiple-item scales were each aggregated
by calculating the arithmetic mean and treated as ob-
served indicators in the model. The error variances of the
scales were left unspecified in the SEMs. Detailed statis-
tics for these scales are provided in Table 3.
Results
Team Performance from Different Perspectives
Gathering data concerning team performance from dif-
ferent perspectives makes it possible to analyze whether
managers, team leaders, and team members experience
project success in the same way. While comparisons of
mean ratings do not yield any major differences between
these three perspectives (all three parties rate all teams
similarly on average; see Table 3), the correlations be-
tween the three ratings reveal agreement of varying mag-
nitude.
Although the differences in correlations are not statis-
tically significant, the results in Figure 2 display a com-
mon pattern underlying both effectiveness and efficiency
ratings. While there is considerable agreement (r⬎0.50)
between the people that worked together on the project
(team leader and team members), as well as reasonably
strong coherence (r⬎0.40) between team leaders and
managers, there is less of a relationship (r⬎0.35) be-
tween manager judgments and team members’ evalua-
tions of performance.
There are a number of possible reasons for these dis-
crepancies (Gemuenden 1990). According to Hauschildt
(1997), these differences in perceived success may be at-
tributed to certain properties of the rater as well as to
deviations in the reference criterion. As described above,
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
444 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
Figure 2 Correlations Between Performance Evaluations from Different Perspectives
Note. Correlations are significant at the .01 level (2-tailed).
managers and team members particularly seemed to have
quite different sources of knowledge concerning the out-
comes of individual projects. While team members were
primarily familiar with the details of the software product
and the course of the project, managers appeared to rely
on data from project controlling reports and information
from customer contacts in judging team performance. The
team leaders seem to stand right in between the managers
and the other team members, which could indicate that
they are to some degree familiar with the experiences and
realities of both parties.
Furthermore, managers’ more distant observation of in-
dividual projects and their outcomes (“bird’s-eye view”),
could suggest that some managers may have been lacking
detailed information about the relevant performance mea-
sures (quality, schedule, budget). Also, managers’ and
team leaders’ ratings may have been influenced by per-
ceptions of the overall success of the larger development
endeavor or customer relationship to which a particular
project team was contributing. Moreover, it is possible
that managers, perhaps lacking better knowledge of actual
performance, evaluate the outcomes based on their gen-
eral impression of the expertise of the team leader or other
team members, rather than solely considering actual per-
formance.
Teamwork Quality and the Success of Innovative
Projects
In testing our hypotheses we have estimated three struc-
tural equation models based on the three different project
success ratings. Model 1: TWQ (evaluated by team mem-
bers) predicts team members’ (self-)evaluations of team
performance and personal success. Model 2: TWQ (eval-
uated by team members) predicts team leaders’ evalua-
tions of team performance. Model 3: TWQ (evaluated by
team members) predicts managers’ evaluations of team
performance. Details of the three models, including factor
loadings, standardized coefficients, variance explained
(R-square), and goodness-of-fit measures (GFI, AGFI,
RMR, Chi-Square, Degrees of Freedom, P-Value)
(Bagozzi and Yi 1988), are documented in Table 5. The
results for the three models indicate strong overall model
fit with the adjusted goodness-of-fit indicators at 0.99. All
effects are significant on the 1% level.
The results of the three models estimated provide sup-
port to Hypothesis 1, proposing a positive influence of
TWQ on team performance. The path coefficients from
TWQ to team performance are significant on the 1% level
in all three models. However, there are considerable dif-
ferences in the magnitude of the relationship between
TWQ and team performance as rated by the team mem-
bers versus team leaders and managers. TWQ explains
about 41% of the variance in team performance as rated
by the team members, 11% as rated by the team leaders,
and 7% as rated by the managers.
These differences may be attributed to the above-
discussed suboptimal agreement between managers, team
leaders, and team members on team performance. An-
other possible explanation is known as implicit theories
(Gladstein 1984). According to this concept, team mem-
bers, in light of the TWQ experienced, attribute good per-
formance to (in their view) good team processes (Model
1). Furthermore, the relationships in Model 1 may be in-
flated due to common source bias, given that both TWQ
and the project success constructs are based on team
member data. In contrast, manager evaluations of team
performance are much less likely to be influenced by im-
plicit theories, given their limited knowledge of the col-
laborative task processes of individual teams. Addition-
ally, the results of Models 2 and 3 are not subject to
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 445
Table 5 Structural Equation Models
Model 1. TWQ
Predicting Team Members’
Evaluations of Team
Performance and
Personal Success
Model 2. TWQ Predicting
Team Leaders’ Evaluations
of Team Performance
Model 3. TWQ Predicting
Managers’ Evaluations of
Team Performance
Stand.
Factor
Loading
Stand.
Coefficient/
R-Square
Stand.
Factor
Loading
Stand.
Coefficient/
R-Square
Stand.
Factor
Loading
Stand.
Coefficient/
R-Square
Teamwork Quality (TWQ)
Communication 0.82 0.83 0.85
Coordination 0.63 0.65 0.63
Balance of Member Contr. 0.84 0.85 0.84
Mutual Support 0.86 0.85 0.86
Effort 0.76 0.79 0.79
Cohesion 0.88 0.88 0.88
Team Performance 0.64/0.41 0.34/0.11 0.26/0.07
Effectiveness (Quality) 0.86 0.74 0.80
Efficiency (Schedule and Budget) 0.74 0.79 0.76
Personal Success of Team Members 0.93/0.87
Work Satisfaction 0.94
Learning 0.69
GFI (Goodness-of-Fit Index) 0.995 0.996 0.997
AGFI (Adjusted GFI) 0.992 0.993 0.994
RMR (Root Mean Square Residual) 0.013 0.012 0.010
Chi-Square 92.05 46.64 48.93
Degrees of Freedom 33 19 19
P-Value ⬍0.01 ⬍0.01 ⬍0.01
Note. N 145; TWQ is based on aggregated team member evaluations (excl. team leaders) in all three models. All path coefficients are
significant at the 0.01 level.
common source bias, as team leaders’ and managers’
evaluations of team performance are predicted by team
members’ evaluations of TWQ.
While recognizing the differences in the extent of the
relationships given the different performance evaluations,
Hypothesis 1 is supported because TWQ is significantly
associated with all ratings of team performance.
The results of Model 1 provide strong support for Hy-
pothesis 2. TWQ shows a very strong association with
team members’ personal success (standard coefficient
0.93, 87% variance explained). The results indicate that
the quality of collaboration in a team determines the con-
tentment of team members with their work situation to a
large degree. Also, high TWQ fosters learning on the part
of the team members, which accommodates the desire for
personal and professional growth of the generally highly
educated workforce of software development laborato-
ries.
Discussion
This research yields several findings. (1) The quality of
collaboration in teams can be captured through the six
facets of TWQ. (2) TWQ shows a relationship with the
success of innovative projects as measured by team per-
formance (effectiveness and efficiency) and by the per-
sonal success of team members (satisfaction and learn-
ing). (3) The magnitude of the relationship between TWQ
and team performance varies with the perspective of the
rater (team member versus team leader versus manager).
Our findings support the conceptualization of TWQ as
a higher order latent construct, as the six facets loaded on
one factor in both the team- and individual-level factor
analysis. The TWQ construct provides a comprehensive
measure of the collaborative team-task process focusing
on the quality of interactions rather than on activities in
teams (Homans 1974). Thus, TWQ complements other
elements of the team-task process, such as the task strategy
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
446 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
employed, (Hackman 1987) and the teams’ external re-
lations (Gladstein 1984; Ancona and Caldwell 1988,
1990).
The empirical results of this research document that
TWQ is significantly related to project success (including
team performance and team members’ personal success).
These findings on the basis of data from 145 software-
development teams in Germany confirm prior studies in
the United States, relating individual aspects of TWQ,
such as mutual support (Pinto et al. 1993), balance of
member contributions (Seers 1989), and team effort
(Campion et al. 1993, Campion et al. 1996) to task- and
people-related outcomes. Furthermore, our findings relate
to previous research on software-development teams in
Germany showing a positive relationship between infor-
mal communication and team performance (Brodbeck
1994). While our results do not allow us to draw infer-
ences across cultural contexts, the consistency of findings
of different studies in both countries seems to indicate
some cross-national robustness of these relationships.
This research contributes to clarifying how teamwork
relates to different measures of project success. Both team
performance (effectiveness and efficiency) and the per-
sonal success of team members (satisfaction and learning)
are significantly connected to TWQ. Thus, the TWQ con-
struct offers a way to assess the quality of collaboration
within teams and to actively influence this critical success
factor by focusing management activities on improving
the six TWQ facets (Hoegl 1998). It is worth noting that
all six TWQ facets, when considered individually, display
rather uniform relationships with the respective ratings of
team performance. The correlations of the six facets with
the evaluations of effectiveness and efficiency document
homogeneous associations (for each rater) without show-
ing any of the TWQ facets significantly deviating (see
Table 3).
TWQ explains about 41% of the variance in team per-
formance based on team member ratings. As this docu-
ments the explanatory power of TWQ, it leaves a bigger
portion of the variance in team performance unexplained.
Obviously, TWQ as a measure of the quality of collabo-
ration cannot account for team performance entirely
(Gemuenden and Hoegl 1998). Other aspects of project
management, such as correct planning and controlling of
the project, the existence and application of the necessary
skills to do the job (Hackman 1987), the interaction be-
tween the team and its main constituencies (e.g., man-
agement, customer, other projects) (Gladstein 1984,
Ancona and Caldwell 1988, 1990), and various influences
from the organizational context (e.g., resource scarcity,
organization structure) (Tannenbaum et al. 1992), may be
important predictors of performance as well.
The results of this study indicate that there are con-
siderable differences in the way in which team members,
team leaders, and managers experience and, thus, rate
team performance. This is an issue of great importance.
If consequences (career development, financial bonus,
etc.) are to be attached to certain performance levels, it
is necessary to reach consensus among the parties in-
volved regarding targeted and actual performance. It
seems possible that factors on the organizational level,
such as vertical and horizontal differentiation, could
help to explain discrepancies in performance ratings. In
an organization with a wide span of control, managers
might have less in-depth knowledge of any particular
project. Further inquiry into this phenomenon seems es-
sential if teams are to be managed by objectives as well
as rewards.
A few limitations of this study should be noted. First,
the data for this research are cross-sectional rather than
longitudinal. While our study demonstrates associations
between variables, it cannot establish causality. A lon-
gitudinal research design using multiple informants
would further our knowledge toward both causality of
relationships and the development of team collaboration
and team success perceptions over time. Second, the
scope of the empirical data gathered for this research al-
lows generalization of the results obtained chiefly to the
domain of teams with innovative tasks, such as R&D
teams, new venture teams, etc. Because innovative tasks
are highly complex and dynamic, they require that a wide
range of skills be closely integrated to achieve high ef-
fectiveness and efficiency. It is this integrative function
that teams with innovative tasks must fulfill, and TWQ
can be seen as a measure of just how well this is done.
Therefore, we believe that as tasks get more innovative,
TWQ becomes increasingly important to team perfor-
mance. Hence, when tasks are routine (and even R&D
teams sometimes have projects that pose modest levels of
uncertainty and complexity), the quality of teamwork
may have less effect on performance, while other factors
such as external team relations, the organizational context,
and process know-how become more important. Because
our study was aimed at developing concepts specific to a
defined domain of teams, we encourage empirical re-
search that further increases our understanding of possi-
ble moderating effects, such as the influence of task char-
acteristics on the relationship between TWQ and team
performance.
Perhaps the most important issue for further research
derives from the effect of TWQ on project success. Be-
cause this research provides empirical evidence that
TWQ is an important success factor of innovative proj-
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 447
ects, it seems necessary to ask about the major antece-
dents of TWQ. What can managers of innovative projects
do to encourage their teams to practice high TWQ? Ex-
isting theoretical models of team effectiveness (McGrath
1964, Gladstein 1984, Hackman 1987, Shea and Guzzo
1987, Sundstrom et al. 1990, Tannenbaum et al. 1992)
provide a valuable starting point for necessary empirical
research in this area.
Acknowledgments
The authors would like to thank the Editor as well as the three anonymous
reviewers for their helpful comments.
Appendix
Construct Variable (Items)
Teamwork Quality
(TWQ)
Communication. There was frequent communication within the team. The team members communicated often in spontaneous
meetings, phone conversations, etc. The team members communicated mostly directly and personally with each other. There
were mediators through whom much communication was conducted.
R
Project-relevant information was shared openly by all
team members. Important information was kept away from other team members in certain situations.
R
In our team there were
conflicts regarding the openness of the information flow.
R
The team members were happy with the timeliness in which they
received information from other team members. The team members were happy with the precision of the information received
from other team members. The team members were happy with the usefulness of the information received from other team
members.
Coordination. The work done on subtasks within the project was closely harmonized. There were clear and fully
comprehended goals for subtasks within our team. The goals for subtasks were accepted by all team members. There were
conflicting interests in our team regarding subtasks/subgoals.
R
Balance of Member Contributions. The team recognized the specific potentials (strengths and weaknesses) of individual team
members. The team members were contributing to the achievement of the team’s goals in accordance with their specific
potential. Imbalance of member contributions caused conflicts in our team.
R
Mutual Support. The team members helped and supported each other as best they could. If conflicts came up, they were easily
and quickly resolved. Discussions and controversies were conducted constructively. Suggestions and contributions of team
members were respected. Suggestions and contributions of team members were discussed and further developed. Our team was
able to reach consensus regarding important issues.
Effort. Every team member fully pushed the project. Every team member made the project their highest priority. Our team put
much effort into the project. There were conflicts regarding the effort that team members put into the project.
R
Cohesion. It was important to the members of our team to be part of this project. The team did not see anything special in this
project.
R
The team members were strongly attached to this project. The project was important to our team. All members were
fully integrated in our team. There were many personal conflicts in our team.
R
There was personal attraction between the
members of our team. Our team was sticking together. The members of our team felt proud to be part of the team. Every team
member felt responsible for maintaining and protecting the team.
Team Performance Effectiveness. Going by the results, this project can be regarded as successful. All demands of the customers have been
satisfied. From the company’s perspective, all project goals were achieved. The performance of our team advanced our image
to the customer. The project result was of high quality. The customer was satisfied with the quality of the project result. The
team was satisfied with the project result. The product required little rework. The product proved to be stable in operation. The
product proved to be robust in operation.
Efficiency. From the company’s perspective one could be satisfied with how the project progressed. Overall, the project was
done in a cost-efficient way. Overall, the project was done in a time-efficient way. The project was within schedule. The
project was within budget.
Personal Success Work Satisfaction. After this project, the team members could draw a positive balance for themselves overall. The team
members have gained from the collaborative project. The team members would like to do this type of collaborative work
again.
Learning. We were able to acquire important know-how through this project. We see this project as a technical success. Our
team learned important lessons from this project. Teamwork promotes one personally. Teamwork promotes one professionally.
R
reverse coded item
References
Adler, P. S. 1995. Interdepartmental interdependence andcoordination:
The case of the design / manufacturing interface. Organ. Sci. 6(2)
147–167.
Alderfer, C. P. 1987. An intergroup perspective on group dynamics.
J. W. Lorsch, ed. Handbook of Organizational Behavior. Prentice-
Hall, Englewood Cliffs, NJ, 190–222.
Ancona, D. G., D. F. Caldwell. 1988. Beyond task and maintenance:
Defining external functions in groups. Group and Organ. Stud.
13(4) 468–494.
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
448 O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001
——, ——. 1990. Beyond boundary spanning: Managing external de-
pendence in product development teams. J. High Tech. Manage-
ment Res. 1(2) 119–135.
Bagozzi, R. P., Y. Yi. 1988. On the evaluation of structural equation
models. J. Acad. Marketing Sci. 16(1) 74–94.
Brannick, M. T., A. Prince, C. Prince, E. Salas. 1995. The measurement
of team process. Human Factors 37(3) 641–651.
Brodbeck, F. C. 1994. Intensive kommunikation lohnt sich fuer SE-
Projekte. F. C. Brodbeck, M. Frese, eds. Produktivitaet und Qual-
itaet in Software-Projekten. R. Oldenburg Verlag, Muenchen,
Wien 51–67.
Campion, M. A., G. J. Medsker, A. C. Higgs. 1993. Relations between
work group characteristics and effectiveness: Implications for de-
signing effective work groups. Personnel Psych.46(4) 823–850.
——, E. M. Papper, G. J. Medsker. 1996. Relations between work team
characteristics and effectiveness: A replication and extension.
Personnel Psych. 49(2) 429–452.
Cartwright, D. 1968. The nature of group cohesiveness. D. Cartwright
and A. Zander, eds. Group Dynamics: Research and Theory, 3rd
ed. Tavistock Publications, London, U.K., 91–109.
Clark, K. B., T. Fujimoto. 1991. Product Development Performance.
Harvard Business School Press, Boston, MA.
Cohen, S. G., G. E. Ledford, G. M. Spreitzer. 1996. A predictive model
of self-managing work team effectiveness. Human Relations
49(5) 643–676.
Cooke, R. A., J. L. Szumal. 1994. The impact of group interaction
styles on problem-solving effectiveness. J. Applied Behavioral
Sci. 30(4) 415–437.
Cooper, R. G. 1993. Winning at New Products: Accelerating the Pro-
cess from Idea to Launch. Addison Wesley, Reading, MA.
——, E. J. Kleinschmidt. 1995. Benchmarking for firm’s critical suc-
cess factors in new product development. J. Product Innovation
Management 12 374–391.
Denison, D. R., S. L. Hart, J. A. Kahn. 1996. From chimneys to cross-
functional teams: Developing and validating a diagnostic model.
Acad. Management J. 39(4) 1005–1023.
Domsch, M. E., T. J. Gerpott. 1995. Fuehrung in forschung und en-
twicklung. A. Kieser, ed. Handwoerterbuch der Fuehrung (2 Au-
flage). Schaeffer-Poeschel, Stuttgart, Germany, 369–380.
Eisenhardt, K. M., B. N. Tabrizi. 1995. Accelerating adaptive pro-
cesses: Product innovation in the global computer industry. Ad-
min. Sci. Quart. 40 84–110.
Fleming, Q. W., J. M. Koppelman. 1996. Integrated project develop-
ment teams: Another fad . . . or a permanent change. Internat. J.
Project Management 14(3) 163–168.
Gemuenden, H. G. 1990. Erfolgsfaktoren des Projektmanagements—
eine kritische Bestandsaufnahme der empirischen Untersuchun-
gen. Projekt Management 90(1,2) 4–15.
——. 1995. Zielbildung. H. Corsten, M. Reiss, eds. Handbuch Unter-
nehmensfuehrung: Konzepte-Instrumente—Schnittstellen. Gabler,
Wiesbaden, Germany, 251–266.
——, M. Hoegl. 1998. Teamarbeit in innovativen Projekten: Eine kri-
tische Bestandsaufnahme der empirischen Forschung. Zeitschrift
fuer Personalforschung 12(3) 277–301.
——, T. Lechler. 1997. Success factors of project management: The
critical few. Reviewed paper, Portland Internat. Conf. Manage-
ment of Eng. Tech. Portland, Oregon July 27–31.
Gladstein, D. L. 1984. Groups in context: A model of task group ef-
fectiveness. Admin. Sci. Quart. 29 499–517.
Goodman, P. S., E. Ravlin, M. Schminke. 1987. Understanding groups
in organizations. B. W. Staw, L. L. Cummings, eds. Research in
Organizational Behavior 9JAI Press, Greenwich, CT, 121–173.
Griffin, A., J. R. Hauser. 1992. Patterns of communication among mar-
keting, engineering and manufacturing: A comparison between
two new product development teams. Management Sci. 38(3)
360–373.
Gully, S. M., D. J. Devine, D. J. Whitney. 1995. A meta-analysis of
cohesion and performance: Effects of level of analysis and task
interdependence. Small Group Res. 26(4) 497–520.
Gupta, A. K., S. P. Raj, D. Wilemon. 1987. Managing the R&D mar-
keting interface. Res. Management (March) 38–43.
——, D. Wilemon. 1996. Changing patterns in industrial R&D man-
agement. J. Product Innovation Management 13 497–511.
Guzzo, R. A., G. P. Shea. 1992. Group performance and intergroup
relations in organizations. M. D. Dunnette, L. M. Hough, eds.
Handbook of Industrial and Organizational Psychology, 3rd ed.
Consulting Psychologists Press, Palo Alto, CA, 269–313.
Hackman, J. R. 1987. The design of work teams. J. W. Lorsch, ed.
Handbook of Organizational Behavior. Prentice-Hall, Englewood
Cliffs, NJ, 67–102.
Hauptman, O., Hirji, K. K. 1996. The influence of process concurrency
on project outcomes in product development: An empirical study
of cross-functional teams. IEEE Trans. Engrg. Management 43(3)
153–164.
Hauschildt, J. 1997. Innovations management, 2nd ed. Franz Vahlen,
Muenchen.
Helfert, G. 1998. Team management of business relationships: A
framework for effectiveness. H. G. Gemuenden, T. Ritter, A.
Walter, eds. Relationships and Networks in International Mar-
kets. Elsevier (Pergamon Press), Oxford.
Henik, A., J. Tzelgov. 1985. Control of halo error: A multiple regres-
sion approach. J. Appl. Psych. 70(3) 577–580.
Hise, R. T., L. O’Neal, A. Parasuraman, J. U. McNeal. 1990. Market-
ing/R&D interaction in new product development: Implications
for new product success. J. Product Innovation Management 7
142–155.
Hoegl, M. 1998. Teamarbeit in innovativen Projekten: Einflussgroes-
sen und Wirkungen. Deutscher Universitaetsverlag, Gabler, Wies-
baden, Germany.
Homans, G. C. 1974. Social Behavior: Its Elementary Forms. Harcourt,
Brace, Jovanovich, New York.
James, L. 1982. Aggregation bias in estimates of perceptual agreement.
J. Appl. Psych. 67(2) 219–229.
——, R. Demaree, G. Wolf. 1984. Estimating within-group interrater
reliability with and without response bias. J. Appl. Psych. 69(1)
85–98.
Janis, I. 1995. Groupthink. Psychological Dimensions of Organiza-
tional Behavior, 2nd ed. Barry M. Staw, ed. Prentice-Hall, En-
glewood Cliffs, NJ, 391–399.
Johne, A., P. Snelson. 1990. Successful Product Development. Basil
Blackwell, Oxford, U.K.
Katz, R. 1982. The effects of group longevity on project communica-
tion and performance. Admin. Sci. Quart. 22 81–104.
MARTIN HOEGL AND HANS GEORG GEMUENDEN Teamwork Quality and the Success of Innovative Projects
O
RGANIZATION
S
CIENCE
/Vol. 12, No. 4, July–August 2001 449
——, T. J. Allen. 1988. Investigating the not invented here (NIH) syn-
drome: A look at the performance, tenure, and communication
patterns of 50 R&D project groups. M. L. Tushman, W. L. Moore,
eds. Readings in the Management of Innovations. Ballinger Pub-
lishing Company, Cambridge, MA, 293–309.
Larson, J. R., L. J. Schaumann. 1993. Group goals, group coordination,
and group member motivation. Human Performance 6(1) 49–69.
Lawler, E. E., S. A. Mohrman, G. E. Ledford. 1995. Creating High
Performance Organizations. Jossey-Bass, San Francisco, CA.
Lawrence, P. R., J. W. Lorsch. 1967. Differentiation and integration in
complex organisations. Admin. Sci. Quart. 12 1–47.
Levine, J. M., R. L. Moreland. 1990. Progress in small group research.
Ann. Rev. Psych. 41 585–634.
McGrath, J. E. 1964. Social Psychology: A Brief Introduction. Holt,
Rinehart and Winston, Inc., New York.
Mohrman, S. A., S. G. Cohen, A. M. Mohrman. 1995. Designing Team-
Based Organizations: New Forms for Knowledge Work. Jossey-
Bass, San Francisco, CA.
Mullen, B., C. Copper. 1994. The relation between group cohesiveness
and performance: An integration. Psych. Bull. 115(2) 210–227.
Nadler, D. A., M. L. Tushman. 1988. Strategic linking: Designing for-
mal coordination mechanisms. M. L. Tushman, W. L. Moore, eds.
Readings in the Management of Innovations. Ballinger Publishing
Company, Cambridge, MA, 469–486.
Pinto, M. B., J. K. Pinto. 1990. Project team communication and cross-
functional cooperation in new program development. J. Product
Innovation Management 7200–212.
——, ——, J. E. Prescott. 1993. Antecedents and consequences of
project team cross-functional cooperation. Management Sci.
39(10) 1281–1297.
Schrader, S., J. Goepfert. 1996. Structuring manufacturer-supplier in-
teraction in new product development teams: An empirical anal-
ysis. H. G. Gemuenden, T. Ritter, A. Walter, eds. Proc. 12th In-
ternat. Conf. Industrial Marketing and Purchasing (1).
Universitaet Karlsruhe, Karlsruhe, Germany, 557–598.
Seers, A. 1989. Team-member exchange quality: A new construct for
role-making research. Organ. Behavior and Human Decision
Process 43 118–135.
——, M. M. Petty, J. F. Cashman. 1995. Team-member exchange un-
der team and traditional management: A naturally occuring quasi-
experiment. Group & Organization Management 20 18–38.
Shea, G. P., R. A. Guzzo. 1987. Groups as human resources. Res.
Personnel and Human Resources Management 5323–356.
Sundstrom, E., K. P. De Meuse, D. Futrell. 1990. Work teams: Appli-
cations and effectiveness. Amer. Psych. 45(2) 120–133.
Tannenbaum, S. I., R. L. Beard, E. Salas. 1992. Team building and its
influence on team effectiveness: An examination of conceptual
and empirical developments. K. Kelley, ed. Issues, Theory, and
Research in Industrial/Organizational Psychology. Elsevier, Am-
sterdam, Holland, 117–153.
Thamhain, H. J., J. B. Kamm. 1993. Top-level managers and innovative
R&D performance. A. Cozijnsen, W. Vrakking, eds. Handbook
of Innovation Management, Blackwell Publishers, Oxford, U.K.,
42–53.
Tjosvold, D. 1984. Cooperation theory and organizations. Human Re-
lations 37(9) 743-767.
——. 1995. Cooperation theory, constructive controversy, and effec-
tiveness: Learning from crisis. R. A. Guzzo, E. Salas and Asso-
ciates, eds. Team Effectiveness and Decision Making in Organi-
zations. Jossey-Bass, San Francisco, CA, 79–112.
Weingart, L. R. 1992. Impact of group goals, task component com-
plexity, effort, and planning on group performance. J. Appl.
Psych. 77(5) 682–693.
Wiendieck, G. 1992. Teamarbeit. Frese, Erich, eds. Handwoerterbuch
der Organisation, C. E. Poeschel Verlag, Stuttgart, Germany.
Accepted by M. Scott Poole
