RESEARCH NOTE

The International Journal of Conflict Management

1996, Vol. 7, No. 4 (October), pp. 361-376

THE EFFECTS OF CONFLICT, TRUST, AND

TASK COMMITMENT ON PROJECT
TEAM PERFORMANCE

Thomas W. Porter
Bryan S. Lilly
Indiana University

Whereas recent research linking conflict to performance has focused on

strategic or executive teams, this paper examines task performing pro-
ject teams. The authors present an overall model for team performance
which includes relationship characteristics such as commitment, trust,
conflict, and task processes. The authors propose that conflict, which
may be quite beneficial for strategic teams, is more likely to hinder than
help performance in project teams. The structural model is tested using
data from 464 individuals in 80 student teams working on a new product
introduction case project. The empirical findings support the view that
(1) commitment and trust have only an indirect relationship with team
performance, (2) and conflict and task processes are key explanatory
variables directly related to team performance.
Recent interest in group conflict (e.g., Amason, 1996; Jehn, 1994) has refo-
cused attention on the vital relationship between conflict and performance. Ama­
son (1996) studied top management teams and found that task related or cognitive
conflict was positively related to decision quality in strategic decision making
teams whereas affective conflict was negatively related to decision quality. Thus,
according to Amason (1996), cognitive conflict is functional whereas affective
conflict is dysfunctional. However, these findings were studied within the context
of strategic decision making groups. The relationship between conflict and perfor­
mance is less clear across other types of groups. Therefore, a key conflict-related
issue concerns whether or not task related conflict can more generally be viewed as
functional.
Research attempting to better understand the performance of groups has a
long and distinguished history. Small group theorists have developed integrative
362 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

theories of groups (Hackman & Morris, 1975; Hackman, 1987; McGrath, 1964,
1991) and social interaction within groups (Bales, Cohen, & Williamson, 1979) to
explain why some groups are more effective than others. One finding of this
research has been that the relationship between group processes and effectiveness
is not constant, but rather varies with the nature of the task (Hackman & Morris,
1975). Therefore, it is important for researchers to specify the nature or type of
task faced by groups being studied.
Group tasks differ along several key dimensions (Steincr, 1972). Important
task dimensions include whether (1) the task is routine versus complex; (2) the task
is maximizing versus optimizing; and (3) creativity demands are high (versus low).
In this paper we limit our discussion to group tasks which are complex, optimizing,
and require limited creativity. Organizational projects such as TQM or new prod­
uct development are highly complex and require a great deal of information shar­
ing and integration as well as multiple stages to complete the task. Also, these
projects are optimizing rather than maximizing, where performance is assessed by
how well the product reaches some ideal, rather than how fast the task was com­
pleted or how much was produced. Finally, a large number of organizational pro­
jects have limited creativity demands. An example of a project requiring limited
creativity is the development of a line extension or me-too product. In these cases
development teams have a clear definition of the product being developed and are
able to utilize established new product development procedures. Such projects are
quite common in business, and therefore it is of practical significance to identify
antecedents and consequences of conflict in these types of project team activities.
We use the term project team to refer to small groups of individuals working
outside of traditional hierarchical lines of authority on a temporary basis on the
types of tasks discussed above. The purpose of this paper is to examine the role of
task conflict in such teams. We develop a model which links task commitment and
trust to conflict and ultimately to performance. Next, we discuss the measures used
in testing the model and interpret the empirical findings of model validation
efforts. Finally, we provide conclusions and suggestions for future research.

Prior Research on Group Interaction

Various aspects of the group's process and interactions among project mem­
bers may play an important role in understanding the relationship between conflict
and performance. Characteristics of the group interaction such as cohesion, task
commitment, and task process have previously been linked to conflict and team
performance. In the following section we outline the central team interaction
chacteristics in our model and briefly explain why we opted not to include cohe­
sion.
Cohesion has received much attention with mixed results. Goodman, Ravlin,
and Schminke (1987) argue that the evidence is decidedly mixed in regards to the
link between cohesiveness and performance. Goodman et al. suggested that cohe-
siveness leads to the acceptance of groups norms, although these norms may or

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

 T. W. PORTER AND B. S. LILLY 363

may not facilitate group performance. For example, in groups where the norm is to
not "show up" slow-moving workers, cohesiveness might tend to reduce perfor­
mance. Mullen and Copper (1994) traced the cohesiveness literature back to Fes-
tinger (1950) who identified three main dimensions of the cohesiveness construct:
commitment to task, group attractiveness, and group pride. Mullen and Copper's
major Finding was that commitment to the task had a significant and positive rela­
tionship with performance, while group attractiveness and group pride were not
significantly related to performance. This led the authors to concur with Goodman
et al. (1987) that commitment to task is the cohesion element that should be stud­
ied. In light of these recommendations, we have elected to focus on task commit­
ment in our conceptualization.
Trust is another concept that also has played an important role in many orga­
nizational theories. For example, the group's trust in its leader has frequently been
studied (Bennis & Nanus, 1985). In the group literature, trust plays a central role in
theories of resource scarcity (see Kramer, 1990, for a review of this literature).
More recently, trust in an exchange partner has been linked to a variety of out­
comes including: relationship survival, cooperation, and the use of marketing
research (Morgan & Hunt, 1994; Moorman, Deshpande, & Zaltman, 1993).
Finally, although practitioner literature (e.g., Katzenbach & Smith, 1993) often
points to trust as a key characteristic of high performing teams, empirical valida­
tion of this important concept is lacking.
Conflict is often included in models of group effectiveness. Traditionally,
conflict has carried a pejorative connotation that often includes hostility, tension,
or adversarial positioning (Saavedra, Barley, & Van Dyne, 1993). On the other
hand, there is another type of conflict, often referred to as functional (Anderson &
Narus, 1990; Deutsch, 1969) that may positively influence team effectiveness.
Recent work (Jehn 1994; Amason & Schweiger, 1994; Amason, 1996) may hold
the key to clearing up the conflict paradox. These researchers distinguish between
affective conflict and cognitive conflict. According to Amason and Schweiger,
affective conflict tends to be emotional and focused on personal incompatibilities
or disputes. In contrast, cognitive conflict is generally task oriented and focused on
judgmental objectives about how to best achieve common objectives. Contention
exists, however, on whether affect should be treated as a conflict dimension. Wall
and Nolan (1986) and Sessa (1996) proposed that the key conflict dimensions are
people oriented and task oriented. Furthermore, Sessa (1996) suggests that affect is
a consequence of conflict, characterized by tone and arousal. Given the lack of
consensus on a uniform definition or even meaning of conflict, in this study we
elected to focus on task-related conflict.
Task processes is a final team interaction characteristic that has been linked
to performance. Task processes are behaviors aimed at organizing members to get
work done (Ancona & Caldwell, 1992; Philip & Dunphy, 1959). Here the focus is
on activities such as setting goals, prioritizing work, and developing workable
plans that facilitate task accomplishment. Task processes may be particularly rele­
vant to project teams. Katzenbach and Smith (1993) argue that team effectiveness

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

364 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

depends upon the group's ability to' effectively decide upon objectives, goals, and
common approach. Teams which can effectively manage their task processes may
be more able to achieve consensus and avoid the negative effects of unresolved
conflict.
In the following section, we present a model to aid us in our attempt to better
understand team interaction antecedents and performance consequences of task
related conflict in project teams (see Figure 1). Towards this objective, we review
literature linking these important constructs and hypothesize how these variables
interrelate to influence group performance.

A Model of Group Characteristics and Team Performance

For this paper, we adopt the conceptual definition of task commitment as the
group's determination to perform to a degree that is superior to acceptable stan­
dards of performance. This definition differs from other task commitment concep­
tualizations in that it focuses on team members' perceptions that the group perfor­
mance norm is superior performance rather than simply doing the job. We expect
that actual task commitment is manifested by behaviors (Salancik, 1977), and that
these behaviors drive perceptions of task commitment.
Trust is defined as confidence in the group's reliability and expertise
(Moorman, Deshpande, & Zaltman, 1993). One situation in which trust may be
low, even if task commitment is high, is when team members have different goals
and objectives for the joint activity. Conflicting goals or objectives may be partic­
ularly relevant in teams composed of people from different departments or func­
tions. In general, however, commitment and trust have been viewed as related con­
structs with trust influencing commitment (Morgan & Hunt, 1994). The basic
argument is that if trust exists between exchange partners, they will tend to be
more committed to the relationship. While we concur with the notion that trust and
commitment are related variables, temporally, it is more likely that trust is a con­
sequence of task commitment. As group members perceive that others in the group
are highly determined to excel, their trust in the group increases. For example,
behavior on the part of team members which indicate a lack of commitment to the
task is likely to result in a reduction in the confidence in the team's reliability.
Alternately, where team member behaviors signal high task commitment, confi­
dence is likely to increase.
Hypothesis 1: The group's task commitment has a positive influence on
group trust.
Ancona and Caldwell (1992) view task processes as behaviors aimed at orga­
nizing members to get work done. In this paper, quality of task processes refer to
the group's effectiveness at organizing work related activities. Three types of task
processes relevant to project groups are: identifying task objectives, prioritizing
work, and developing workable plans. The quality of task processes provide an
indication of how well the group is able to work together. In a study of new prod­
uct team performance, Ancona and Caldwell (1992) hypothesized and found a sig-
The International Journal of Conflict Management, Vol. 7, No. 4, October 1996
 T. W. PORTER AND B. S. LILLY 365

nificant and positive relationship between task processes and performance as rated
by the team, but not as rated by managers. Ancona and Caldwell reasoned that
team members may have schema that link task processes to performance; however,
managers, who have less access to the group, rely on more quantitative perfor­
mance data such as budgets and schedules. Because prior research has failed to
consistently find a significant relationship between task processes and manager
ratings of performance, we do not hypothesize a direct link between quality of task
processes and performance.
To properly fit quality of task processes in the model, we first turn to trust.
Task processes may be influenced by trust because high trust should lead to greater
ability to achieve consensus on important task-related issues which increases the
quality of task processes. For example, group members that have high confidence
in the group's ability may be more likely to accept group decisions. When trust is
present, the team members may be more willing to accept ideas and contributions
of other team members rather than adopt an adversarial position. This analysis
suggests that under conditions of high trust, team members are likely to view the
team's task processes favorably. Therefore, we hypothesize:
Hypothesis 2: Group trust has a positive influence on the quality of group
task processes.
In this paper conflict is defined as the group members' perceptions of the
amount of opposing work-related views articulated within the group. This defini­
tion differs from the way that conflict is treated in decision making groups. Rather
than focusing on a particular conflict episode, we examine the perceived level of
task conflict that occurs during the project. What, then, is the effect of conflict on
project team performance? The answer may depend on the importance of creativity
in the team's problem solving process. Creative problem solving activities such as
problem definition, alternative generation, and alternative evaluation may well be
improved through productive disagreements and debates. However, when the
activity requires effectively implementing a preferred alternative, the value of
content related debates likely diminishes. In this case, increased task conflict may
distract the team from its focal task. In the type of group tasks considered here,
conflict may lead to consensus building activities that may sap time and energy
which could be more fruitfully utilized moving the project ahead. We expect that,
on balance, in project activities with low creativity requirements, conflict has a
negative impact on performance.
Hypothesis 3: Group conflict has a negative influence on team performance.
Even when creativity demands may be low, the team is responsible for many
task-related decisions that can effect the overall success of the project. If the group
is able to hash out disagreements over task-related issues and arrive at a decision
acceptable to the whole group, then the impact of conflict on performance should
be less negative or even positive. This notion is consistent with arguments made by
Amason (1996), who notes that conflict may positively impact performance if dealt
with functionally. On the other hand, an inability to achieve consensus on impor-
TheInternational Journal of Conflict Management, Vol. 7, No. 4, October 1996
366 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

tant task-related issues is likely to adversely effect performance, through reduced

commitment to the team's approach or by detracting attention from task issues in
order to deal with group maintenance issues (McGrath, 1991).
As stated previously, task processes may provide a behavioral indicant of the
group's ability to achieve consensus. When the group is able to effectively identify
objectives, prioritize work, and develop workable plans, then the negative impact
of conflict on performance may be reduced. Alternatively, when task processes are
poor, the group may be unable to achieve consensus on critical issues and perfor­
mance will likely suffer to a greater degree. This discussion suggests that task pro­
cesses moderate the relationship between conflict and performance.
Hypothesis 4: The effect of conflict on performance becomes less negative
as task processes increase.
Next we address the issue of how trust influences the level of task-oriented
conflict within the group. When group trust is high, group members may feel less
compelled to challenge the ideas and positions of fellow group members. Further­
more, when trust is high, groups may have more ability to exert normative pressure
on dissenting group members, thus reducing the overall amount of conflict occur­
ring within the group. This is essentially one of the central assertions of the
groupthink model (Janis, 1982). The basic premise of Janis' theory is that under
certain conditions, group cohesiveness may lead to poorer decision quality because
individuals with dissenting views will be pressured to accept the group's position.
Although distinct constructs, trust may be similar to cohesion in terms of its
effect on conflict. Research exists to suggest that for groups in the mature stage of
the group development process (Tuckman, 1965), cohesiveness may lead to
increased task-related conflict (Leana, 1985). This occurs because group members
are secure enough in their roles to challenge one another. A parallel explanation for
how trust influences conflict is that when group trust is high, group members may
be more willing to express opposing viewpoints, and challenge each other's ideas
without the fear that this "conflict" would be interpreted as a personal assault.
Members of project teams often work with individuals from other functional
areas, rather than people from their own work groups (Page, 1993). Furthermore,
team composition tends to be somewhat amorphous, with individuals sometimes
entering or leaving the team as particular skills are needed. These factors make it
less likely for the maturity effect described above to occur. These factors lead us to
believe that trust reduces the level of task conflict within the group. Therefore, we
suggest:
Hypothesis 5: Group trust is negatively related to conflict.
Finally, we address the issue of how task commitment influences the team's
performance. In our model, task commitment is proposed to have indirect effects
on performance as mediated through other variables in the model (trust, conflict,
and quality of task processes). Thus, the key issue is whether or not task commit­
ment has a direct effect on the team's performance, above and beyond any indirect
effects it may have. It is possible that task commitment operates primarily as a sig-
The International Journal of Conflict Management, Vol. 7, No. 4, October 1996
 T. W. PORTER AND B. S. LILLY 367

nal which fosters trust but is not directly related to performance. However, given
the fairly consistent support for a small but positive relationship between task
commitment and performance (see Mullen & Copper, 1994, for a review), we
expect that task commitment does indeed have a direct effect on performance.
Therefore, we hypothesize:
Hypothesis 6: Task commitment has a positive influence on the team's per­
formance.

Method
Research Setting
Empirical work was conducted at a large midwestern university using groups
of undergraduate students working on a multi-course project. All students in their
first semester of the business program simultaneously take a set of courses in mar­
keting, finance, and operations management. A key feature of this curriculum is
that in the last four weeks of the course students are broken into project groups of
five or six students each in order to work full time on a demanding integrative
case. The case requires students to work together as a team, applying concepts they
have learned throughout the semester.
The project itself is a new product commercialization case and requires the
group to carry out all the steps necessary to bring an innovative new product to
market. Strategic decisions such as the new product design and the market segment
to be targeted are provided in the case and, hence, do not need to be made. Typical
marketing issues include items such as deciding on pricing and advertising. Opera­
tions issues include production scheduling and inventory management. Financial
issues include identifying a cost of capital and estimating the net present value
(NPV) of the project under various scenarios. An optimal solution exists based
largely on analysis of spreadsheets, work flow diagrams, and multiple equations.
Each project is assigned a grade according to (1) how closely the project matches
this optimal solution, and (2) the quality of the supporting discussion. Thus, the
grades assigned to these projects provide a fairly objective performance measure.
Finally, group projects are a key component of the business curriculum and team
members realize that they may have to work together on future projects. Therefore,
just as in organizational settings, team members have expectations of future inter­
actions which may influence their behavior.
Sample
Eighty five teams participated and data were collected from all 464 members
of eighty teams for a 94% response rate. Data from five project groups were not
returned by the grader. Furthermore, surveys for three individuals were discarded
because of patterned responses (i.e., 4's marked for all items). Constructs were
measured with a questionnaire which was included in the case packet distributed to
each student. Students were required to complete and return the questionnaire and
were told that the survey would not affect their grade, rather that it would be used
The International Journal of Conflict Management, Vol. 7, No. 4, October 1996
368 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

to evaluate the use of teams on the integrative case. Instructions requested students
to fill out the survey after completion of the case and seal it in an envelope with
their peer evaluations. The measures were taken prior to the group receiving any
feedback on their performance, which helps avoid the problem of implicit theories
of group performance (Guzzo, Wagner, Maguire, Herr, & Hawley, 1986) affecting
the team member's perceptions of the various group characteristics. In order to
ensure confidentiality, groups were identified by number, and students were not
required to identify themselves.
Measures
An objective set of performance measures is distributed to all graders who
score group performance systematically. The score is a grade applied to the group's
written paper and may range between zero and ninety five. Performance scores
ranged between 55 and 94 with a mean of 81.53. Measures for the other variables
were either developed or adapted from previous research as follows (see Appendix
for final items).
Trust. Four items from the Moorman, Zaltman, and Deshpande (1992) scale
for "User Trust in Researcher" were adapted for the present study by replacing the
term "researcher" with "group." One of the items from that scale referred to trust­
ing someone outside the group. This item was not relevant for the current purpose
and was replaced with a new item. The Moorman et al. scales have been empiri­
cally shown to possess high reliability and discriminant validity (Moorman, Desh­
pande, & Zaltman, 1993).
Task Processes. Following work by Ancona and Caldwell (1992), three
items were used to assess member's perceptions of task-related group processes.
These items related to the team members' ability to identify task objectives, priori­
tize work, and develop workable plans, and have been empirically shown to have
high reliability.
Conflict and Task Commitment. For these constructs, scales were devel­
oped to be consistent with the present construct definitions. As a face validity
check, items and definitions were given to marketing researchers and feedback was
used to further revise the measures. Next, following the procedure advocated by
Churchill (1979), a list of construct definitions and revised measures was given to
eight academicians. Recipients were requested to assign each measure to the con­
struct they thought appropriate and to indicate if the measure represented more
than one construct.
These preliminary scale development efforts for the four latent independent
variables resulted in a survey instrument which included nineteen 7-point Likert
scale items which were then pretested. For pretesting, the survey was administered
to 80 undergraduate students representing 18 project groups from three courses.
Responses were factor analyzed. Although the hypothesized four factor solution
emerged, six items were dropped because of cross-loadings above .25 which
resulted in thirteen final items (see Appendix).

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

 T. W. PORTER AND B. S. LILLY 369

Analysis
The model shown in Figure 1 was estimated with path analysis using the
group scores as input. Path analysis differs from OLS regression in that all
hypotheses are estimated concurrently. Thus, four regressions are estimated
simultaneously in our model; trust is regressed on commitment, conflict and task
processes are regressed on trust, and performance is regressed on commitment,
conflict, and the task process-conflict interaction. Path model estimation is pre­
ferred to combined measurement-structural model estimation in our case because
of (1) the non-linear relation posited between conflict and performance; and (2)
because structural models require large sample sizes (see Bollen, 1995, for a
review and critique of non-linear latent variable modeling techniques and assump­
tions).
Variables in this study exist at the group level both conceptually and opera­
tionally. In order to determine if aggregating the individual responses to provide a
group score is justified, we examined inter-rater reliabilities as recommended by
James, Demaree, and Wolf (1984). Fifty-two of the eighty groups had reliability as
above .90, thirteen groups had as between .80 and .90, four groups had as between
.70 and .80, six groups had as between .60 and .70, and the remaining five groups
had as below .60. These a levels are favorable according to Nunnally (1978) and
indicate that group level aggregation is appropriate for our data. Data were ana­
lyzed including all eighty groups and again excluding the five groups with low
reliabilities. Excluding these five groups did not significantly change the results,
and we report results based on data from all eighty groups.
Responses were aggregated in two ways. First, group scores for Likert scale
items were created by averaging individual scores, which is consistent with
between-group variation analysis recommendations (Bryk & Raudenbush, 1992).
This resulted in group scores for each of the thirteen survey items. Next, the items
which correspond to each factor were averaged as recommended by Ping (1995).
The four trust items, for example, were averaged to create an overall trust score.
Thus, a group's score for trust is the average of all group member responses to all
four trust items. This aggregation method results in a data set with 80 cases and
five variables per case including the performance score (see Table 1 for summary
statistics of these variables). We then created the task process-conflict interaction
term by multiplying the overall group conflict score by the overall group task-pro­
cess score which again is recommended by Ping (1995) as an appropriate method
for estimating path models with latent variable interactions.
Overall Model Fit. Path estimates are provided in Figure 1. In accordance
with procedures for interaction testing (Aiken & West, 1991), an initial model was
examined which had an added path linking task processes directly to performance.
This path was statistically nonsignificant and was removed after which the hypoth­
esized model was tested. Overall fit indexes indicated an adequate model fit. The
probability of the chi-square statistic was .05, which increased to .07 when the
model was tested again after removing the nonsignificant path from commitment

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

370 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

to performance. The goodness-of-fit index for the hypothesized model was .95, the
adjusted goodness-of-fit index was .85, and the comparative-fit index was .99,
which, in combination, indicate an overall acceptable fit.

Table 1
Means and Correlations for Group-Level Variables

Variables M SD 1 2 3 4 5

1. Commitment 6.2 .53 1.00

2. Trust 6.0 .62 .79* 1.00
3. Conflict 2.8 .78 -.29 -.46* 1.00
4. Task Processes 6.1 .46 .54* .63* -.42* 1.00
5. Performance 81.4 6.21 .19 .22 -.35* .24 1.00

Note: Data collected from surveys administered to 464 students in 80 groups in the busi­
ness school at a large midwestern university. Individual responses (not performance) were
averaged within groups to compute group scores.
*p < .01.

Evaluation of Hypotheses
The hypotheses concerning the linkages between the team interaction char­
acteristics (H1, H2, & H5) were all significant in the hypothesized directions.
Hypothesis 1 predicted that the group's task commitment would have a positive
influence on the group's trust. Results indicate a highly significant relationship
between these two variables, and .69 standard deviations is the expected increase
in group trust for each one standard deviation increase in task commitment.
Hypothesis 2 predicted that the group's trust would have a positive influence on its
task processes. Results support this hypothesis, and .63 standard deviations is the
expected increase in group task processes for each one standard deviation increase
in trust. Finally, Hypothesis 5 predicted that trust would decrease the level of con­
flict within the group. Again, results support this prediction, and .46 is the expected
decrease in group conflict for each one standard deviation increase in trust.
The second set of hypotheses focus on the effect of team interaction charac­
teristics on performance. Results indicate a non-significant direct relationship
between task commitment and performance failing to support H6. Although the
coefficient is in the direction hypothesized, the coefficient is small and insignifi­
cant. To interpret the coefficients that link commitment, conflict, and the conflict-
task process interaction to performance, we examine the unstandardized coeffi­
cients because our measure of performance is readily interpretable in its unstan­
dardized metric. Our estimated model for performance using unstandardized coef­
ficients is:
Performance = 70.59 + .72 (Commitment) - 5.8 (Conflict) + .6 (Conflict ×
Task Processes)

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

 T. W. PORTER AND B. S. LILLY 371

Note: Each arrow Corresponds to a hypothesis and is labeled with the corresponding
hypothesis number, the expected direction of effect (either positive or negative), the
estimated regression coefficient, and an indication of significance.
*p < .01.

When substituting means for commitment, conflict, and task processes in the
above equation, the resulting performance value is 81.37 which is the mean group
performance. What then is the impact of conflict on performance? A one-unit
increase in conflict for our sample corresponds to a 5.8 decrease in performance
plus an increase in performance equal to .6 multiplied by the group task process
level due to the interaction term. For a group with an average level of task pro­
cesses which is 6.2, a one unit increase in conflict corresponds to a change in per­
formance of -5.8 + .6 (6.2) which is a 2.08 decrease in performance. Further, given
that the standard deviation of conflict is .78, we would expect that an increase in
conflict by one standard deviation would correspond to an expected 1.62 decrease
in performance for groups with average levels of task processes. This result sup­
ports both H3 and H4. Conflict, in our sample, had a negative impact on perfor­
mance, but this impact was reduced by task processes.

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

372 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

Discussion
Primary Conclusions
In general, this study's results confirm the usefulness of considering team
interaction characteristics in understanding the effectiveness of task-performing
project teams. The results indicate that trust and task commitment are important
variables to consider, not because of their direct influence on performance (they
appear to have little impact when the effects of conflict and task processes are
taken into account), but because they influence other important interaction charac­
teristics (conflict, task processes), that do influence performance. Our findings also
demonstrate the critical role that conflict plays in influencing team performance. In
the sample studied, in which the team's task requires only moderate creativity, con­
flict tends to reduce team performance. Thus, for complex, optimizing tasks
requiring limited creativity, the benefits of conflict do not overcome the cost of
distraction from the focal task.
On the other hand, the adverse effects of conflict are partially mitigated when
the team's task processes are effectively managed. The significant interaction
between conflict and task processes may be related to the phenomenon that others
describe as functional conflict. However, whereas Deutsch (1969) points to some
conflict that is good (functional) and other conflict that is not, our results suggest
that it is the quality of the task processes that influences the effects of task related
conflict on team performance. Furthermore, in our study, the impact of conflict on
performance was always negative. For conflict to have positively influenced per­
formance in our sample, task processes would have to have been 9.7, while our
scale for task processes ranged only from one to seven.
This study contributes to the literature in several important ways. First, we
demonstrated the usefulness of including interaction characteristics in a model of
team effectiveness. Trust and task commitment do influence performance, just not
directly. Since these beliefs are likely influenced by the environment in which the
project takes place, our results seem to support a view that trust and commitment
mediate environmental effects on the team. Although our findings suggest that
conflict as a process tends to impede performance (in task-performing groups),
when the team is able to effectively manage its internal task processes, these
effects are reduced.
Our study also contributes to understanding the effects of conflict on perfor­
mance in task-performing project teams. Our study demonstrates that task-related
conflict seems to have a very different impart on project team performance as
compared to previous findings which focused on strategic decision making teams.
The implications from decision making research are that management should
attempt to increase cognitive conflict and consensus, while minimizing negative
affective outcomes. In the type of project groups studied here, our findings suggest
that management would do well to structure the team so that task conflict is mini­
mized.

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

 T. W.PORTER AND B.S. LILLY 373

Limitations
As is often the case, several potential limitations of this study should be rec­
ognized and taken into account when interpreting its findings. First, it is necessary
to recognize that this study uses cross-sectional data to examine a phenomenon that
is inherently dynamic. However, since the teams received no feedback on their per­
formance prior to completing the questionnaire, we can be somewhat confident
about direction of causality inferences when applied to performance, which is our
ultimate dependent variable.
A second potential limitation is that although we focused on task-related con­
flict, it is possible that team members reported conflict which was over equity
issues (how much each member contributes) rather than content issues. Another
interpretation of our findings would be that task processes indicates that different
kinds or degrees of conflict are going on within the groups. Future research may
benefit by seeking to identify the source of the conflict (content issues, process
issues, or equity issues).
A third potential limitation concerns the timing of the conflict. Project team
activity can be viewed as a form of problem-solving. Problem-solving is often
viewed as a multistage process that includes problem identification, alternative
generation, alternative evaluation and choice, decision implementation, and deci­
sion control (Bass, 1983). A failure by the team may occur at any time in the
problem-solving process. Also, these stages differ in the degree to which they rely
on divergent or convergent thinking (Aldag & Fuller, 1993). For example, the
alternative generation stage requires primarily divergent thinking, while evaluation
and choice requires more convergence. Thus, the type of critical analysis inherent
in cognitive conflict may be most highly related to effectiveness in the early or
middle stages of the problem-solving process. Our measures do not detect differ­
ences between instances of conflict according to their timing in relation to the
problem-solving process. Here, we treat all conflict equally. Adding conflict timing
would be an interesting area for future inquiry.
A fourth limitation is that the measure of performance used here is but one
aspect of group effectiveness. Hackman (1987) defines effectiveness in terms of
three criteria: productive output, personal need satisfaction, and capacity for future
cooperation. Trust and task commitment may have a direct relationship with these
other aspects of group effectiveness.
A fifth limitation concerns the external validity of these findings. The teams
in this study were made up of students working on a new product case in an aca­
demic setting. Whether these results generalize to project teams doing real work in
a corporate environment is an empirical question. However, the duration and com­
plexity of the project, the intensity on the part of the team members required to
accomplish the task, the relative importance of the task, the nature of the project
(being a new product commercialization case), and the way the team members do
the work all contribute to make this project an attractive forum for studying project
teams.

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

374 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

While recognizing the above limitations, the potential contribution of this

research to better understanding the effectiveness of project teams is high. It
should be possible to examine environmental and individual factors as antecedents
to the formation of trust, commitment, and conflict among team members. In the
present study, trust both reduces cognitive conflict, and improves task processes
which in concert make trust a highly important variable. One interesting question
for future research would be to see if the negative relationship between trust and
conflict would occur in mature groups. Additionally, it may be that trust improves
the affective outcomes associated with conflict. For example, under conditions of
high trust (versus low trust), team members may be less inclined to react nega­
tively to cognitive conflict. This was not tested and represents an opportunity for
future research.

References

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interac-
tions. Newbury Park, CA: Sage.
Aldag, R. J., & Fuller, S. R. (1993). Beyond fiasco: A reappraisal of the groupthink phe­
nomenon and a new model of group decision processes. Psychological Bulletin, 113,
533-552.
Amason, A. C. (1996). Distinguishing the effects of functional and dysfunctional conflict on
strategic decision making: Resolving a paradox for top management teams. Academy of
Management Journal, 39, 123-148.
Amason, A. C, & Schweiger, D. M. (1994). Resolving the paradox of conflict, strategic
decision making, and organizational performance. International Journal of Conflict
Management, 5, 239-253.
Ancona, D. G., & Caldwell, D. F. (1992). Demography and design: Predictors of new prod­
uct team performance. Organization Science, 3,321-343.
Anderson, J. C, & Narus, J. A. (1990). A model of distributor firm and manufacturer firm
working partnerships. Journal ofMarketing, 54, 42-58.
Bales, R. F., Cohen, S. P., & Williamson, S. A. (1979). SYMLOG. New York: Holt, Rhein-
hart & Winston.
Bass, B. M. (1983). Organizational decision making. Homewood, IL: Irwin.
Bennis, W. G., & Nanus, B. (1985). Leaders: Strategies for taking charge. New York:
Harper and Row.
Bollen, K. A. (1995). Structural equation models that are nonlinear in latent variables: A
least-squares estimation. In P. V. Marsden (Ed.), Sociological methodology (pp.
223-252). Washington, DC: American Sociological Society.
Bryk, A. S., & Raudenbush S. W. (1992). Hierarchical linear models. Newbury Park, CA:
Sage.
Churchill, G. A. (1979). A paradigm for developing better measures of marketing constructs.
Journal of Marketing Research, 16, 64-73.
Deutsch, M. (1969). Conflicts: Productive and destructive, Journal of Social Issues, 25,
7-41.
Festinger, L. (1950). Informal social communication. Psychological Review, 57, 271-282.
Goodman, P. S., Ravlin, E., & Schminke, M. (1987). Understanding groups in organizations.
Research in Organizational Behavior, 9, 121-173.
The International Journal of Conflict Management, Vol. 7, No. 4, October 1996
 T. W. PORTER AND B. S. LILLY 375

Guzzo, R. A., Wagner, D. B., Maguire, E., Herr, B., & Hawley, C. (1986). Implicit theories
and the evaluation of group process and performance. Organizational Behavior and
Human Decision Processes, 37, 279-295.
Hackman, R. J. (1987). The design of work teams. In J. W. Lorsch (Ed.), Handbook of
organizational behavior (pp. 315-342). Englewood Cliffs, NJ: Prentice-Hall.
Hackman, R. J., & Morris, C. G. (1975). Group tasks, group interaction process, and group
performance effectiveness: A review and proposed integration. In L. Berkowitz (Ed.),
Advances in experimental psychology (8th ed., pp. 45-99). New York: Academic Press.
James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group interrater relia­
bility with and without response bias, Journal of Applied Psychology, 69, 85-98.
Janis, I. L. (1982). Groupthink (2nd ed.). Boston: Houghton Mifflin.
Jehn, K. (1994). Enhancing effectiveness: An investigation of advantages and disadvantages
of value-based intragroup conflict. International Journal of Conflict Management, 5,
223-238.
Kramer, R. M. (1990). The effects of resource scarcity on group conflict and cooperation.
Advances in Group Processes, 7, 151-177.
Katzenbach J. R., & Smith, D. K. (1993). The wisdom of teams. Boston, MA: Harvard Busi­
ness School.
Leana, C. R. (1985). A partial test of Janis' groupthink model: Effects of group cohesiveness
and leader behavior of defective decision making. Journal of Management, 11, 5-17.
McGrath, J. E. (1964). Social psychology: A brief introduction. New York: Holt.
McGrath, J. E. (1991). Time, interaction, and performance (TIP): A theory of groups. Small
Group Research, 22, 147-174.
Moorman, C , Deshpande, R., & Zaltman, G. (1993). Factors affecting trust in market
research relationships. Journal of Marketing, 57, 81-101.
Moorman, C , Zaltman, G., & Deshpande, R. (1992). Relationships between providers and
users of market research: The dynamics of trust within and between organizations. Jour-
nal of Marketing Research, 29, 314-329.
Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship market­
ing. Journal of Marketing, 58, 20-38.
Mullen, B., & Copper, C. (1994). The relation between group cohesiveness and perfor­
mance: An integration. Psychological Bulletin, 115, 210-227.
Nunnally, J. C. (1978). Psychometric theory. New York: McGraw-Hill.
Page, A. L. (1993). Assessing new product development practices and performance: Estab­
lishing crucial norms. Journal of Product Innovation Management, 10, 273-290.
Philip, H., & Dunphy, D. (1959). Developmental trends in small groups. Sociometry, 22,
162-174.
Ping, R. A., Jr. (1995). A parsimonious estimating technique for interaction and quadratic
latent variables. Journal of Marketing Research, 32, 336-347.
Saavedra, R. P., Earley, P. C , & Van Dyne, L. (1993). Complex interdependence in task-
performing groups. Journal of Applied Psychology, 78, 61-72.
Salancik, G. R. (1977). Commitment and the control of organizational behavior and belief.
In B. M. Staw & G. R. Salancik (Eds.), New directions in organizational behavior (pp.
1-54). Chicago: St. Clair.
Sessa, V. I. (1996). Using perspective taking to manage conflict and affect in teams. Journal
of Applied Behavioral Research, 32, 101-115.
Steiner, I. D. (1972). Group process and productivity. New York: Academic Press.
Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological Bulletin,
63, 384-399.

The International Journal of Conflict Management, Vol. 7, No. 4, October 1996

376 GROUP CONFLICT, TRUST, AND TASK COMMITMENT

Wall, V. D., & Nolan, L. (1986). Perceptions of inequity, satisfaction, and conflict in task-
oriented groups. Human Relations, 39, 1033-1052.

APPENDIX
Latent Variable Construct Definitions and Scale Items
Task commitment, trust, and conflict items were measured with Likert ratings
anchored by 1 = Strongly Disagree . . . 7 = Strongly Agree. Task process items were mea­
sured with Likert scales with 1 = Very Poor. . . 7 = Very Good. The fourth trust item and
the second conflict item were reverse coded which is indicated by the (R) after these items.
Chronbach a s (a) are provided for data at the individual level with N = 464 and at the
aggregated group level with N = 80.
Task Commitment—the individual's perception of his or her group's level of com­
mitment to task performance that is superior to acceptable standards of performance. 1.
Superior performance is very important to our group. 2. This group has a "commitment to
excellence." (a = .84 at individual level, a = .91 at group level)
Trust—confidence in an exchange partner's reliability and expertise. 1. If I were
absent from a group meeting, I would be confident in the other group members' ability to
make decisions without my involvement. 2. If I were unable to monitor the other group
members' work, I would be willing to trust them to get the job done right. 3.1 trust the team
to do things I can't do myself. 4.1 generally do not trust my group. (R) ( a = .84 at individual
level, a = .93 at group level)
Conflict—the group member's perception of the amount of opposing work-related
views articulated within the group. 1. The members of the group have frequent disagree­
ments in the process of doing these tasks. 2. Our group generally sees "eye-to-eye" on all
issues. (R) 3. We often have quite heated debates in the process of doing these projects. 4.
Differences of opinion within the group are quite common. (a = .82 at individual level, a =
.92 at group level)
Task Processes—behaviors aimed at organizing members to get work done. Please
rate your group's ability to: 1. Identify task objectives. 2. Prioritize work. 3. Develop work­
able plans. ( a = .85 at individual level, a = .88 at group level)

Biographical Note
Thomas Porter
Marketing Department, BU 328
Indiana University
Bloomington, IN 47405
Phone/Fax: 812-855-8878/6440
E-Mail: thporte@ndiana.edu
Thorn Porter is a doctoral candidate in Marketing, Indiana University. His current research
focuses on decision making in new product development teams, cross-functional interaction,
and marketing education.
Bryan Lilly is a doctoral candidate in Marketing, Indiana University. His current research
focuses on new product introductions, and on research methodology. He has published in
Journal of Product Innovation Management.

Received: April 11, 1996

Accepted after two revisions: October 10, 1996
♥

TheInternational Journal of Conflict Management, Vol. 7, No. 4, October 1996