Does Stress Lead to a Loss of Team Perspective?

Abstract

One of the more well-established findings in the research literature is that stress leads to a restriction or narrowing of attentional focus. In the present study, we extend this research to the group context. We propose that, in a team environment, the narrowing of attention induced by stress may result in a shift in perspective from a broad team perspective to a more narrow or individualistic self-focus, and this loss of team perspective may result in degraded team performance. The results of an empirical study found that stress resulted in a narrowing of team perspective and that team perspective was a significant predictor of team performance. Moreover, when the effects of team perspective were controlled, the effects of stress on team performance were substantially weakened. These results suggest that one way in which stress impacts team performance is by narrowing or weakening the team-level perspective required for effective team behavior. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Full text

Group Dynamics: Theory, Research, and Practice
1999,
Vol. 3,
No.
4,291-302Copyright 1999 by the Educational Publishing Foundation
1089-2699/99/S3.00
Does Stress Lead to a Loss of Team Perspective?
James E. Driskell
Florida Maxima CorporationEduardo Salas
University of Central Florida
Joan Johnston
Naval Air Warfare Center Training Systems Division
One of the more well-established findings in the research literature is that stress leads to
a restriction or narrowing of attentional focus. In the present study, we extend this
research to the group context. We propose that, in a team environment, the narrowing of
attention induced by stress may result in a shift in perspective from a broad team
perspective to a more narrow or individualistic self-focus, and this loss of team
perspective may result in degraded team performance. The results of an empirical study
found that stress resulted in a narrowing of team perspective and that team perspective
was a significant predictor of team performance. Moreover, when the effects of team
perspective were controlled, the effects of stress on team performance were substan-
tially weakened. These results suggest that one way in which stress impacts team
performance is by narrowing or weakening the team-level perspective required for
effective team behavior.
Research indicates that individuals respond to
stress by a restriction or narrowing of attentional
focus (Cohen, 1980; Combs & Taylor, 1952;
Easterbrook, 1959). For example, Salovey
(1992) showed that unexpected events may
result in a shift of individual attention from
external cues to a more internal self-focus. If we
assume that this restriction of attention extends
to social cues as well, then we may expect a
narrowing of social perspective to occur at the
group level. More specifically, we propose that a
narrowing of attention at the group level is
accompanied by a shift from a broader, team
perspective to a more narrow, individualistic
focus.
Furthermore, in a team task environment,
we expect this narrowing of perspective to lead
to a degradation of performance. However,
although considerable research has examined
James E. Driskell, Florida Maxima Corporation, Winter
Park, Florida; Eduardo Salas, Department of Psychology,
University of Central Florida; Joan Johnston, Naval Air
Warfare Center Training Systems Division, Orlando, Florida.
The views expressed herein are those of the authors and
do not reflect the opinion, policy, or views of the Department
of Defense.
Correspondence concerning this article should be ad-
dressed to James E. Driskell, Florida Maxima Corporation,
507 North New York Avenue, R-l, Winter Park, Florida
32789.
Electronic mail may be sent to james.driskell @
rollins.edu.
individual reactions to stress, comparatively
little work has examined the effects of stress on
group performance (see Driskell & Salas, 1991).
The purpose of this study is to examine the
relationship between stress, team perspective,
and performance in task groups.
Group Performance Under Stress
Real-world incidents provide anecdotal but
often vivid illustrations of team effectiveness
and ineffectiveness under stress. On July 19,
1989,
United Airlines Flight 232 experienced
the failure of an engine and complete loss of
hydraulic pressure, leaving the airplane with
virtually no flight controls. The airplane crashed
during an attempted landing at Sioux City, Iowa,
and 111 of the 296 passengers and crew
members were fatally injured. However, experts
noted that the crew's performance in landing the
plane under at least some measure of control
was just short of a miracle. The report of the
National Transportation Safety Board (NTSB,
1990) and subsequent evaluations of the cockpit
transcripts (Predmore, 1991) suggested that one
reason for the crew's effectiveness was that they
maintained a hiigh level of team coordination,
interaction, and communication during the
emergency.
To provide a contrasting example, United
291

292DRISKELL, SALAS, AND JOHNSTON
Airlines Flight 173 crashed near Portland,
Oregon, in December 1978, as it ran out of fuel
while the crew attempted to deal with a landing
gear malfunction. The NTSB report (1979) cited
a breakdown in teamwork as a primary cause of
this accident. The report indicated that the
captain was preoccupied with an individual task,
that "the first officer's main responsibility is to
monitor the captain" and this was not done, and
that "the flight engineer's responsibility is to
monitor the captain's and first officer's actions"
and this was not done (p. A-5). In a review of
crew performance in aviation, Foushee (1984)
noted that a majority of accidents are related to
breakdowns in crew or team coordination.
Foushee (1982) reported one commercial avia-
tion incident in which, after ignoring repeated
inquiries from a copilot, the captain responded,
"Just look out the damn window" (p. 1063).
Why does group coordination become more
problematic under stress or emergency condi-
tions? Research examining individual reactions
to stress suggests that individuals' focus of
attention shifts from a broader to a narrower
perspective when under stress, and we believe
this phenomenon may have significant implica-
tions for group interaction. One of the more
well-established findings in the stress literature
is that as stress or arousal increases, the
individual's breadth of attention narrows (Combs
& Taylor, 1952; Easterbrook, 1959). Perhaps the
earliest statement of this phenomenon was
William James's (1890) belief that the individu-
al's field of view varied from a broader
perspective under normal conditions to a more
narrow, restricted focus under stress. Pen-
nebaker, Czajka, Cropanzano, and Richards (1990)
provided an empirical test of the hypothesis that
normal thought processes and attentional focus
are restricted under stress, finding that individu-
als confronted with uncontrollable noise tended
to move from high to lower levels of thought,
from a broad to a narrow perspective, when
under stress. Other research showed that stress
may increase individual self-focus. Baumeister
and colleagues (Baumeister, 1984; Butler &
Baumeister, 1998) showed that performance
pressure engenders higher levels of self-focused
attention. This research adopts an attentional
theory approach that assumes that performance
pressure focuses an individual's attention in-
ward. This may increase self-awareness but
disrupt the performance of skilled tasks by
diverting attention from external task-relevant
information or by. inhibiting automated perfor-
mance processes (Lewis & Linder, 1997).
Does this phenomenon occur at the group
level? The analogue to this process at the group
level is that group members under stress may
become more self-focused and less group
focused. Indeed, some evidence suggests this
may occur. Several studies showed that stress
leads to a decrease in prosocial behaviors, such
as helping. For example, Mathews and Canon
(1975) found that individuals were less likely to
help or assist others when exposed to loud
ambient noise. Rotton, Olszewski, Charleton,
and Soler (1978) found that loud noise reduced
participants' ability to discriminate among
people occupying different roles. Cohen (1978,
1980) concluded that the narrowing of attention
that occurs under stress may include social
phenomena as well and that stress may lead to a
neglect of social or interpersonal cues and
decreased sensitivity to others.
Some preliminary evidence on attentional
focus in groups lends peripheral support to this
analogy. For example, Mullen (1991) showed
how group composition has consequences for
attentional focus in groups. Yet despite the
explosion of research on cognition, attention,
and decision making in the past several decades,
relatively little work has addressed stress and
team performance. Certainly, the debilitating
effect of stress on team performance has been
recognized in the work of Cannon-Bowers and
Salas (1998), Karau and Kelly (1992) and
others, and interventions have been developed
to enhance individual and team decision making
under stress (Hartel & Hartel, 1997). However,
as Littlepage and Karau (1997) noted, despite
these efforts, a number of fundamental questions
remain unanswered. Most lacking is an under-
standing of the processes by which stress affects
team performance.
In summary, a large body of research
indicates that the individual's breadth of atten-
tion narrows, and individuals tend to become
more self-focused when under stress. In the
present study, we extend this research to
examine group performance by arguing that
stress decreases group focus. We argue that,
under stress, group members adopt a narrower,
more individual perspective of task activity.
With this narrowing of perspective, team
members' cognitions shift from a broader, team

TEAM PERSPECTIVE293
perspective to a more narrow, individualistic
focus.
Team Perspective
From the early days of our discipline,
theorists noted the importance of team perspec-
tive to group interaction. Mead (1934) argued
that only to the extent that individuals develop a
group concept does cooperative activity become
possible. Asch (1952) claimed that group action
was possible "only when each participant has a
representation that includes the actions of others
and their relations" (p. 251). In more recent
work describing the knowledge, skills, and
attitudes required for effective team perfor-
mance, Cannon-Bowers, Tannenbaum, Salas,
and Volpe (1995) noted the importance of group
members adopting a team concept, or collective
orientation (Driskell & Salas, 1992), that places
team considerations above individual concerns.
Taking an information-processing perspective,
Hinsz, Tindale, and Vollrath (1997) described
how the attentional resources of group members
may be directed internally (self-focused atten-
tion) or externally. They further argued that the
manner in which attention is allocated can
determine the nature and outcome of group
interaction, noting that greater self-focus can
divert attention from the task.
We define team perspective as a perception of
the interrelations of actors and actions in a group
system. Team perspective consists of two
primary components. The first component is a
collective representation of the group, or group
identity. Some argued that a critical element that
defines a functioning group is the existence of a
common group identity. In Lewin's (1951)
terms,
a key component of group membership is
that individuals perceive the dynamic interdepen-
dence of group members; this has also been
termed a mutual awareness (McGrath, 1984), or
a common social identification (Reicher, 1982).
Thus,
one component of team perspective is a
"we-ness," a sense of being part of a team
versus a more individualistic self-focus.
A second component of team perspective is a
collective representation of the task or a team
mental model of task activity. To coordinate
activity to pursue group goals, a group must
construct a common cognitive environment.
That is, team members must develop a team-
level perspective of the task and of the inter-
dependent roles that comprise the task. Cannon-
Bowers, Salas, and Converse (1993) argued that
a team mental model includes a model of task
interaction—how team members must interact
with one another to perform the task. Other
research suggested that teams are more effective
when group members share a common perspec-
tive on group resources, goals, and performance
strategies (Hackman, 1987). Group members'
mental models of the task may range from a
team orientation (i.e., viewing the task as an
interdependent team effort) to a more individual-
istic focus (i.e., viewing the task as an individual
activity).
Present Study
Our overall goal in this study was to examine
the extent to which stress degrades team
performance and, more specifically, to examine
whether the effect of stress on team performance
is due to the more fundamental effect of stress
on team perspective. We designed a study to
examine the relationship among stress, interde-
pendence, team perspective, and performance.
We first wished to examine the effects of stress
and team interdependence on team perspective.
To document the team perspective phenomenon,
we wanted to observe whether greater interde-
pendence among team members would lead to a
stronger team perspective. Further, we wished to
examine the effects of stress on narrowing of
team perspective. Second, we wished to exam-
ine the extent to which stress and team
interdependence affected team performance.
Finally, we wished to examine the role of team
perspective in mediating the effects of stress and
interdependence on performance.
We adopted the mediation model of Baron
and Kenny (1986) as an analytic strategy to test
these relationships. According to this model,
first, variation in the independent variable must
account for variation in the hypothesized
mediator. Second, variation in the independent
variable must account for variation in the
dependent variable. Third, the hypothesized
mediator must affect the dependent variable.
Finally, when the mediator is controlled, the
relationship between the independent and de-
pendent variable is substantially decreased.
This strategy is formalized in the following
hypotheses.

294DRISKELL, SALAS, AND JOHNSTON
Hypotheses
We reasoned that the greater the degree of
interdependence among group members, the
more likely group members are to develop a
strong team perspective. That is, one factor that
distinguishes an interdependent team from an
aggregate of individuals is that the interdepen-
dent team members develop a broad team
perspective. To examine this proposition, we
designed an experimental task in which partici-
pants performed either independently in a
coacting team or as members of an interdepen-
dent team. Those who comprised the coacting
teams were introduced as members of a team
(thus,
they had a "nominal" team identity) and
were told that their team's performance would
be evaluated against that of other teams (thus,
team members possessed a common fate).
However, coacting team members performed
the task independently with no interaction with
other group members (i.e., the coacting group
members were simultaneously engaged in a
similar activity).1 By contrast, in the interdepen-
dent teams, group members were mutually
dependent and worked together to perform the
task. We predict that greater interdependence
among group members should lead to a broader
team perspective. Thus, we proposed the
following:
Hypothesis la: Greater interdependence among group
members will result in a broader team perspective.
As we noted earlier, research indicates that
individuals respond to stress with a narrowing of
attention and increased self-focus. Extending
this proposition to the group level, we predicted
that increased stress would result in a restriction
of social cues and a narrowing of team
perspective. To examine this proposition, we
had half of the teams perform under high-stress
conditions (increased auditory distraction, task
load, and time pressure) and half of the teams
perform under low-stress conditions without
these distractions. We predicted that, under
stress,
group members would shift from a
broader team perspective to a more narrow,
individualistic perspective. Thus, we proposed
the following:
Hypothesis lb: Increased stress will result in a
narrowing of team perspective.
We predicted that stress and team interdepen-
dence would affect team performance. Stress
has been shown to degrade performance in a
number of studies (Driskell & Salas, 1991;
Johnston, Driskell, & Salas, 1997), and we
predicted that high-stress conditions would lead
to poorer performance on the present task than
lower stress conditions. Furthermore, given that
the task used in this study is a team-level task,
we expected that greater interdependence among
group members would lead to more effective
performance. Thus, we proposed the following:
Hypothesis 2a: Greater interdependence among group
members will lead to more effective performance.
Hypothesis 2b: Increased stress will lead to degraded
task performance.
However, the more interesting issue is the
role of team perspective in mediating the effects
of interdependence and stress on performance,
that is, the extent to which the effects of stress
and team interdependence on performance are
due to the more fundamental effects of stress
and team interdependence on team perspective.
We propose that one way stress degrades team
performance is by narrowing team perspective.
If this is the case, then the effects of stress on
performance should weaken when variability as
a result of team perspective is partialed out.
Similarly, if we argue that one way in which
interdependence enhances team performance is
by strengthening team perspective, then the
effects of interdependence on performance
should weaken when the variability as a result of
team perspective is partialed out. Thus, we
proposed the following:
Hypothesis 3: When team perspective is controlled, the
effects of stress and interdependence on performance
will be substantially reduced.
Method
Participants
Participants in this study were 95 U.S. Navy
technical school personnel who volunteered to
take part in a study of decision making. The
study was a 2 (type of group: coacting vs.
1 Many hold that a coacting group is not a task group at
all.
However, others fsuch as Shaw (1981) noted that group
tasks differ in the degree of cooperation required, from
instances in which group members must coordinate their
actions to instances in which group members work
independently.

TEAM PERSPECTIVE295
interdependent) X 2 (stress: high vs. low) de-
sign. Participants were randomly assigned to
one of four experimental conditions.
Procedure
The participants arrived at the experimental
laboratory to take part in a study of team
performance and were assigned to a three-
person group. Participants were seated at
individual work stations in an open room and
interacted over a computer network. For the first
half hour of the session, participants were
briefed on the nature of the experimental task,
first listening to a video training tape and then
receiving individualized instruction and prac-
tice.
The task was a computer simulation of a
naval decision-making task. The task required
that participants monitor a radar screen that
contained their own ship at the center and
numerous unidentified contacts positioned at
concentric rings away from the ship. The
objective was to identify and label each contact
according to three classifications: type of craft
(aircraft, surface craft, or subsurface), its status
(civilian or military), and intentions of the craft
(hostile or peaceful). To make each classifica-
tion, the participant would access one of three
information fields or menus: A, B, and C,
corresponding to these headings. For example,
Menu A contained five items of information
used to help identify the type of craft. Within
Menu
A,
the participant could access the altitude
item to ascertain whether the contact was above
the surface, on the surface, or below the surface
of
the
water. After the participant determined the
type of craft, he or she then proceeded to gather
identifying information from Menu B and Menu
C. Once the contact had been labeled as to the
type of craft, its status, and its intentions, the
contact would then be either cleared from the
screen if it was determined not to be a threat or
engaged (targeted) if it was determined to be
hostile. Participants were told to work as quickly
and as accurately as possible to identify each
contact before it reached their ship.
After receiving detailed task instructions and
practice, participants performed the task. After
15 min, action was stopped and participants
completed a task questionnaire. At this point,
they were asked to complete several items:
measures of their perception of the group, task
focus,
and collective orientation (these measures
are described later). We chose to administer
these items during task performance rather than
at completion of the task to enhance the
immediacy of their responses. Participants then
continued the task for an additional
15
min, after
which play was stopped and participants re-
ceived a postexperimental questionnaire. After
the questionnaire was completed, participants
were interviewed individually, received a full
explanation of the study, and were thanked for
their participation.
Experimental Manipulations
Type
of Group
The task was configured so that participants
performed as part of a three-person coacting
team or an interdependent team. In the coacting
group condition, participants were introduced to
one another as team members Alpha, Bravo, and
Charlie. They were told that they would be
working as a team on a computer-based task and
their team's performance would be compared
with that of other teams. However, the task was
structured so that each group member performed
the task independently. That is, Alpha's task was
to select an unidentified contact, gather informa-
tion from Menu A, Menu B, and Menu C, and
then clear or engage the contact. Bravo and
Charlie did the same. Thus, in the coacting
group condition, group members worked on the
task independently, with no interdependence
required.
In the interdependent team condition, the task
was structured so that
Alpha,
Bravo, and Charlie
worked interdependently. Menu A was only
available to Alpha, Menu B was only available
to Bravo, and Menu C was only available to
Charlie. Furthermore, the information items
within each menu were shared across partici-
pants.
For example, Alpha was responsible for
ascertaining the five items of information that
would allow him or her to determine the type of
craft in Menu
A;
however, two of these pieces of
information were held by the other team
members. That is, Alpha would have to get one
item of information from Bravo and one item
from Charlie to combine with the information in
Menu A to determine the type of craft. Thus, to
process a contact, each team member had to
request needed information from others and
provide required information to others. There-

296DRISKELL, SALAS, AND JOHNSTON
fore,
in the interdependent team condition, the
task was configured so that participants had to
work interdependently to perform the task.
Stress
Participants performed the task in either a
normal-stress or a high-stress environment. To
induce high stress, we manipulated three fac-
tors:
auditory distraction, task load, and time
pressure. Auditory distraction was implemented
by playing a multitrack audio recording of
task-related chatter over the participants' head-
phones during the task. Task load was imple-
mented by increasing the rate at which contacts
were presented on the screen. Thus, in the
high-stress conditions, participants were pre-
sented with a greater number of potentially
threatening contacts. Finally, time pressure was
induced by the experimenters telling the partici-
pants to "hurry up" and "work harder" at 5-min
intervals during the task. We implemented
multiple stressors rather than a single stressor to
provide a more robust manipulation of task
stress.
Measures
We defined team perspective as consisting of
two primary components: (a) a collective
representation of the group, or group identity,
and (b) a collective representation of the task.
Collective Representation of the Group
This component of team perspective repre-
sents the extent to which group members
perceive themselves as belonging to a distinct
social unit or team. We assessed this component
of team perspective using two measures: (a)
group members' perception of the group and (b)
the extent to which group members evaluated
the outcome of performance in egocentric
versus group terms.
Perception of the group. We adapted an
item previously used by Gaertner, Mann,
Murrell, and Dovidio (1989) to assess the extent
to which participants perceived themselves as a
group. This item was designed to assess par-
ticipants' conceptual representations of the
group by asking them to rate the extent to which
they perceived themselves as a team or as three
separate individuals. This measure has been
successfully used in previous research to dif-
ferentiate groups_of varying salience (Gaertner,
Mann, Dovidio, Murrell, & Pomare, 1990;
Gaertner et al., 1989). This item ("To what
extent do you feel like a team, or do you feel
more like three individuals?") was presented on
a scale ranging from 1 (team) to 7 (individuals).
We administered this item immediately when
play was paused after the first 15 min of task
performance.
Egocentricity. This measure of group iden-
tity was designed to assess the extent to which
participants evaluated task performance in
individual versus team
terms.
Beck (1993) noted
that one aspect of egocentricity is the tendency
to interpret the meaning of events in terms of
what they mean to the individual rather than the
meaning of the event to the group as a whole.
Therefore, we devised an item to measure the
extent to which participants interpreted task
performance in terms of how they performed
versus how well the team performed. At the end
of the task, each participant was given a printout
showing purported individual and team scores.
The experimenter described the chart, which
indicated that the individual's score was in the
higher range (80—90th percentile) compared
with how others had performed on the task.
However, the team score was shown to be in the
lower performance range (20-30th percentile).
After participants were left for several minutes
to review their score sheets, they were given a
postexperimental questionnaire, which asked
them to rate how well the task was performed on
a scale ranging from 0 (extremely poor) to 100
(extremely well). Note that this question was
intentionally broad in that it asked participants
how well "the task" was performed, thus
leaving the participants to interpret this question
in terms of their individual performance or their
team's performance. Group members providing
a higher evaluation of performance were per-
ceived as interpreting the task according to how
well they themselves did. Group members
providing a lower evaluation of performance
were perceived as interpreting the task accord-
ing to how well the group did.
Collective Representation of the Task
The second component of team perspective is
a collective representation of the task, or a
team-level model of task activity. To assess this

TEAM PERSPECTIVE297
aspect of team perspective, we used measures of
(a) task focus, (b) collective orientation, and (c)
the elaboration of a team task model.
Task focus. The extent to which individuals
were focused on individual versus team activi-
ties was assessed by three items. The first item
asked participants to rate where their attention
had been focused on a scale ranging from 1
{interacting with the team) to 7 (doing indi-
vidual tasks). The remaining two items asked
participants whether they felt more responsible
for their own efforts rather than how others were
doing and whether they felt they concentrated
more on their own tasks rather than interacting
with others. This three-item scale rendered an
alpha of
.93.
Collective orientation. Collective orienta-
tion refers to the mutual interdependence of
team members (Driskell & Salas, 1992). Group
members who are collectively oriented perceive
themselves as working together as a group to
solve the task. Group members who are less
collectively oriented view their actions in a
more autonomous manner. We assessed the
extent of collective orientation in the following
manner. After participants had performed the
task for 15 min, we paused the action and
administered a task questionnaire. We asked
participants to describe the events that had
occurred during the task. Participants were
asked, "If you were asked to explain to someone
what has been going on the last 10 to 15
minutes, what would you say?" Participants
were given 5 min to write down their response
on a blank form.
The measure of collective orientation used
was the proportion of first person plural pronoun
usage (e.g., we, us, our, ours, ourselves) in the
written transcripts. This approach is an adapta-
tion of
a
technique used by Wegner and Giuliano
(1980) and Mullen, Chapman, and Peaugh
(1989) to assess the self-focus of individuals.
We expected that participants with a more
collective or group focus would respond with a
greater proportional usage of first person plural
pronouns (i.e., We identify targets close to our
ship").
For each subject, we calculated the total
number of first person plural pronouns used in
describing task activities divided by the total
word count.
Team task model. We assessed the extent to
which group members elaborated a team-level
model of the task by having three independent
judges who were unaware of the conditions rate
each written transcript according to the extent to
which it illustrated a team perspective of the
task versus an individual perspective. A team
task perspective was defined as emphasizing the
interdependency of group members (i.e., "Al-
pha, Bravo, and Charlie must exchange informa-
tion to identify targets"). An individual task
perspective was defined as emphasizing indi-
vidual actions or perceptions (i.e., "I have to
determine whether a target is peaceful or
hostile"). Transcripts were rated on a
5-point
scale, ranging from 1 (emphasis on individual
perceptions, feelings, and actions) to 5 (empha-
sis on group perceptions and activities of other
group members or the group as a whole). The
three judges' ratings were reliable, yielding a
mean interrater correlation of .86 and Spearman-
Brown effective reliability of .95.
Performance. During the 30-min perfor-
mance period, group members were required to
determine three classifications for each contact:
type of craft, its status, and its intentions. Group
member performance was assessed by the
proportion of classifications that were correctly
identified.
Results
Manipulation Check
As a check on the success of the stress
manipulation, participants were asked to rate the
extent to which they felt distracted and pres-
sured during die task. Participants performing
under high-stress conditions reported feeling
more distracted (M = 2.29, SD = 1.06) than
those in the no-stress conditions (M = 1.50,
SD = .86), t(93) = 4.00, p < .01, and reported
feeling more pressured (M = 2.11, SD = 1.06)
than those in the no-stress conditions (M = 1.86,
SD =
1.00),
r(92) = 4.02, p <
.01.
There were
no significant effects of team interdependence
on the distraction or pressure measures (ps > .1).
Team
Perspective
To gauge th^ effects of stress and team
interdependence on team perspective, we exam-
ined the various indicators of team perspective
and found that they were highly intercorrelated

298DRISKELL, SALAS, AND JOHNSTON
(mean r = .70).2 Therefore, these various in-
dicators were standardized into Z scores and
aggregated into a composite measure of team
perspective (for which a higher score indicates
greater team perspective). Then this composite
indicator of team perspective was subjected to a
2 (type of
group:
coacting vs. interdependent) X
2 (stress: low stress vs. high stress) analysis of
variance (ANOVA). Means and standard devia-
tions for team perspective scores are presented
in Table 1. The expected main effect of stress
was significant, F(l, 91) = 11.99, p < .001.
Group members performing under high-stress
conditions exhibited lower team perspective
(M = -.010) than those operating under low
stress (M = .086). Similarly, the expected main
effect of type of group was significant, F(l,
91) = 396.84,/? < .001. Participants operating
in interdependent teams exhibited higher levels
of team perspective (M = .766) than those
operating in mere coaction (M = —.783).
The interaction between stress and type of
group was also significant, F(l, 91) =
8.91,/?
=
.004.
In the interdependent teams, there was a
significant loss of team perspective from low
stress (M =
1.016)
to high stress (M = .515),
r(91) = 4.60, p < .001. In the coacting teams,
there was no loss of team perspective from low
stress (M = -.766) to high stress (M = -.803),
t(9l) =
0.34,
p = .
369.
Performance
To gauge the effects of stress and team
interdependence on performance, a 2 (type of
group: coacting vs. interdependent) X 2 (stress:
low stress vs. high stress) ANOVA was con-
ducted on performance scores. Means and
standard deviations for performance scores are
Table 1
Means and Standard Deviations for Team
Perspective and Performance as a Function
of Stress and
Type
of Group
Low stressHigh stress
MeasureMSD MSD
Team perspective
Interdependent 1.016 0.393 24 0.515 0.492 24
Coacting -0.766 0.319 26 -0.803 0.260 21
Performance
Interdependent 0.950 0.112 24 0.888 0.156 24
Coacting 0.892 0.145 23 0.859 0.160 20
also presented in Table 1. The expected main
effect of stress was marginally significant, F(l,
87) = 2.47, p = .059. Participants operating
under conditions of low stress performed better
(M = .922) than those operating under condi-
tions of high stress (M = .875). Similarly, the
expected main effect of type of group was
marginally significant, F(l, 87) = 2.09, p =
.076.
Participants operating in interdependent
teams performed better (M = .919) than those
operating in mere coaction (M = .877). The
interaction between stress and type of group was
not significant, F(l, 87) = 0.22, p = .639.
Mediating Role of
Team
Perspective
The foregoing analyses reveal that stress and
team interdependence exert predictable effects
on team perspective and performance. However,
the critical issue that drives this research effort is
the extent to which the effects of stress and
interdependence on performance are due to the
more fundamental effects of stress and interde-
pendence on team perspective. The most direct
way to test the proposed relationship is to test
for mediation.
To test the mediating role of team perspective,
we conducted an analysis of covariance
(ANCOVA), wherein a 2 (type of group: co-
acting vs. interdependent) X 2 (stress: low stress
vs.
high stress) analysis was conducted on per-
formance scores, with team perspective scores
serving as a covariate. If
the
effects of stress and
team interdependence on performance are pri-
marily a result of their effects on team per-
spective, then the effects of stress and team
interdependence on performance should be
reduced when variability resulting from team
perspective is partialed out. Indeed, this is
precisely what happened: In this ANCOVA, the
expected main effect of stress was now reduced
to negligible magnitude, F(l, 86) = 0.93, p =
.169.
Similarly, the expected main effect of type
of group was now reduced to negligible
magnitude, F(l, 86) = 0.62, p = .216. Once
again, the interaction between stress and type of
group was not significant, F(l, 86) = 0.0002,
2 The egocentricity measure did not show the overall
pattern of results that was consistently shown by the other
measures of team perspective and failed to correlate highly
with the other measures of team perspective (average
r = .31). Therefore, this measure was not included in the
composite measure of team perspective.

TEAM PERSPECTIVE299
p = .987. The results of this ANCOVA are
illustrated in Figure 1. It can be seen that the
main effect and interactive effects of stress and
interdependence on performance are reduced
when the effects of stress and interdependence
on team perspective are first partialed out.
Discussion
Teams have become a critical means to
accomplish complex tasks from the company
boardroom to the commercial airliner. Teams are
often called on to respond effectively under
crisis or emergency conditions. Sometimes they
are able to coordinate their resources and
activities to land an immobilized airliner, and
sometimes team performance deteriorates to a
point at which team members barely acknowl-
edge one another.
The idea that stress narrows the scope of
attentional focus is well documented in the
research literature. There has been some sugges-
tion that stress may affect how group members
attend to social information as well. In an early
examination of small group behavior under
stress,
Torrance (1954) wrote, "Under stress ...
linkages between members may become con-
fused and thus_people do not have a clear
perception of what they can expect from one
another, with whom they can relate, [and] how
they can relate to one another" (p. 754).
Although the effects of stress on disrupting
team performance have been well documented
in the research literature, reviewers often
bemoan the fact that we lack a clear understand-
ing of the processes by which this occurs (cf.
Littlepage & Karau, 1997). The results of this
study provide one explanation for the tendency
for groups to function less effectively under
stress:
Stress leads to a loss of team perspective.
The results of this study demonstrate that
interdependence leads to
a
broader team perspec-
tive and that stress results in a narrowing of team
perspective. The results further indicate that the
narrowing of perspective that occurred under
stress led to impaired team performance. This is
not altogether surprising; however, what is
striking is the central role played by team
perspective. After partialing out the effects of
team perspective, both group interdependence
and stress became insignificant predictors of
performance. These results suggest that team
b = -.165(p = .O59)
Stress
Stress X b = -.
Interdependence
Interdependence
b = -.019(p = .860)
Figure 1. Path analysis illustrating the interrelations of stress, interdependence, team
perspective, and performance. Values above lines represent direct effects; values below lines
represent indirect effects (after the mediating role of team perspective is partialed out).

300DRISKELL, SALAS, AND JOHNSTON
perspective is a critical factor in group perfor-
mance and that one way in which stress and
group interdependence impact performance is
through their influence on team perspective.
We now examine the theoretical rationale for
why stress narrows team perspective. The
classic arousal perspective argues that stress
results in heightened arousal and that arousal
leads to a narrowing of attention (see Broadbent,
1971;
Easterbrook, 1959). As attention narrows,
peripheral (less relevant) task cues are first
ignored, followed by further restriction of
central or task-relevant cues. To the extent that
task-relevant cues are neglected, performance
suffers. Accordingly, tasks that demand atten-
tion to a wide range of
cues
are more susceptible
to degradation under stress. Team tasks require
attention to both direct task-related activities
and interpersonal or teamwork activities such a
coordination and communication. To the extent
that these social or team cues are marginalized
as attention is narrowed under stress, team
perspective is weakened and performance may
suffer.
Cohen (1980) presented a related argument,
noting that stress leads to increased demand as
the individual must attend to novel and distract-
ing stimuli. This information overload results in
a narrowing of attentional capacity. To reduce
this informational overload, attention is re-
stricted to those cues most relevant to the task.
Cohen proposed that this restriction may affect
both social and nonsocial cues. As important
social cues (such as attention to others' requests
or actions) are neglected, individuals become
less socially cognizant. Accordingly, a team-
level perspective is likely to become weakened
in task groups under stress as these social or
interpersonal cues are disregarded. In summary,
both the arousal and the information overload
positions maintain that stress leads to a narrow-
ing of perspective as attention is focused on the
most central or salient task cues. The results of
the present research provide evidence that this
theoretical approach can be extended to the
group level as well and that the narrowing of
attentional focus under stress may include a
weakening of team perspective.
It is further worthwhile to consider the
practical implications of this research. The
results indicate that teams are indeed vulnerable
to stress, and one mechanism through which
stress affects team performance is by narrowing
team perspective._Both the arousal and informa-
tion overload theories contend that stress results
in a narrowing of
attention.
As external demands
increase, teamwork behaviors, such as attending
to others, may be neglected to the benefit of
more central task cues, such as those involved in
performing one's own immediate task. If we
assume that the narrowing of attention is an
adaptive response to the stress environment, an
attempt to simplify an increasing complex and
demanding task environment, then one approach
to maintaining effective performance is to
simplify the task environment. For those set-
tings in which effective teamwork is critical, it
may be necessary to structure the task to make it
less demanding (i.e., by delegating subtasks), so
that attention can be maintained on essential
task and teamwork cues. In
brief,
if stress
restricts attentional resources, the task environ-
ment may need to be restructured so that group
members are not forced to sacrifice attention to
teamwork matters to maintain performance.
Unfortunately, for many real-world tasks, reduc-
ing the complexity of the task environment is a
difficult undertaking.
A second approach to counter the effect of
stress on narrowing team perspective is the
attempt to enhance or strengthen team perspec-
tive.
Wickens (1996) noted that the attentional
narrowing resulting from stress is, at least to
some degree, determined by subjective impor-
tance. That is, attention is maintained on
high-priority items to the exclusion of informa-
tion that is perceived to be of lower importance.
It is possible that, for many team tasks, the
importance of teamwork behaviors such as
coordination and communication may be per-
ceived as secondary to other basic, individual
task demands. Anecdotally, highly interdepen-
dent sports such as soccer seem to illustrate this
phenomenon: When teams get behind, team
members often ignore team play and each
person tries to win the game on his or her own,
with predictable results. (Interestingly, this
seems to occur more on inexperienced vs. more
experienced teams.) Interventions that attempt
to enhance team perspective, such as team
building (see Satas, Rozell, Driskell, & Mullen,
1999),
or the use of preparatory information to
reinforce the interrelations of actors and the
interdependent nature of the task (Inzana,

TEAM PERSPECTIVE301
Driskell, Salas, & Johnston, 1996) may be
effective approaches to counter stress effects.
We should also note several limitations of the
present study. First, although we conducted this
research in an applied setting with real-world
participants, the teams that were formed were
short-lived, and the nature of the experimental
task did not allow the observation of other
factors that may be important in more natural
settings. Whether these results hold for groups
that have been together for longer periods of
time or that juggle multiple tasks is unknown.
Second, the primary objective of this study
was to examine the extent to which team
perspective mediates the relationship between
stress and team performance. A variable such as
team perspective functions as a mediator if it
meets several conditions: (a) Variations in levels
of the independent variable (stress) account for
variations in the level of the mediator, (b)
variations in the mediator account for variations
in the dependent variable (team performance),
and (c) when the mediator is controlled, the
relationship between the independent variable
and the dependent variable is substantially
weakened (Baron & Kenny, 1986). These
conditions were all upheld. Although the media-
tion analysis was the most direct way to test the
hypothesized relations, mediation is best done
when there is a strong relation between the
independent variable and the dependent vari-
able,
and in the present case this effect was
modest. For example, the main effect of stress
on team performance translates into a small
effect size of .17 (p = .059). Future researchers
should be advised to test these relationships with
a larger sample size, rendering greater statistical
power.
Finally, the present research has revealed a
novel effect of stress on team perspective that
has not been previously demonstrated. We have
identified one mechanism through which stress
impacts team performance: by narrowing team
perspective. If we adopt Hinsz et al.'s (1997)
model of groups as information processors, it is
likely that, in addition to affecting attentional
processes, stress may disrupt other aspects of
group functioning, such as memory, retrieval,
processing, and response. Further research
should be directed toward gaining a more
complete understanding of the processes by
which stress impacts team performance.
References
Asch, S. E. (1952). Social psychology. Englewood
Cliffs,
NJ: Prentice Hall.
Baron, R. M., & Kenny, D. A. (1986). The
moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality
and Social
Psychology,
51, 1173-1182.
Baumeister, R. F. (1984). Choking under pressure:
Self-consciousness and paradoxical effects of
incentives on skilled performance. Journal of
Personality and Social
Psychology,
46, 610-620.
Beck, A. T. (1993). Cognitive approaches to stress. In
P.
M. Lehrer & R. L. Woolfolk (Eds.), Principles
and practice of stress management (pp. 333-372).
New York: Guilford Press.
Broadbent, D. E. (1971). Decision and stress. New
York: Academic Press.
Butler, J. L., & Baumeister, R. F. (1998). The trouble
with friendly faces: Skilled performance with a
supportive audience. Journal of Personality and
Social
Psychology,
75, 1213-1230.
Cannon-Bowers, J. A., & Salas, E. (1998). Team
performance and training in complex environ-
ments: Recent findings from applied research.
Current Directions in Psychological Science, 7,
83-87.
Cannon-Bowers, J. A., Salas, E., & Converse, S.
(1993).
Shared mental models in expert team
decision making. In J. Castellan (Ed.), Current
issues in individual and group decision making (pp.
221-246). Hillsdale, NJ: Erlbaum.
Cannon-Bowers, J. A., Tannenbaum, S. I., Salas, E.,
& Volpe, C. E. (1995). Denning competencies and
establishing team training requirements. In R. A.
Guzzo & E. Salas (Eds.), Team effectiveness and
decision making in organizations (pp. 333—380).
San Francisco: Jossey-Bass.
Cohen, S. (1978). Environmental load and the
allocation of attention. In A. Baum, J. E. Singer,
& S. Valins (Eds.), Advances in environmental
psychology (Vol. 1, pp.
1-29).
Hillsdale, NJ:
Erlbaum.
Cohen, S. (1980). Aftereffects of stress on human
performance and social behavior: A review of
research and theory. Psychological Bulletin, 88,
82-108.
Combs, A. W., & Taylor, C. (1952). The effect of the
perception of mild degrees of threat on perfor-
mance. Journal of Abnormal and Social Psychol-
ogy, 47, 420-424.
Driskell, J. E., & Salas, E. (1991). Group decision
making under stress. Journal of Applied Psychol-
ogy, 76, 473^78.
Driskell, J. E., & Salas, E. (1992). Collective
behavior and team performance. Human Factors,
34, 277-288.

302DRISKELL, SALAS, AND JOHNSTON
Easterbrook, J. A. (1959). The effect of emotion on
cue utilization and the organization of behavior.
Psychological
Review,
66,
183-201.
Foushee, H. C. (1982). The role of communications,
socio-psychological, and personality factors in the
maintenance of crew coordination.
Aviation,
Space,
and Environmental Medicine, 53, 1062-1066.
Foushee, H. C. (1984). Dyads and triads at 35,000
feet: Factors affecting group process and aircrew
performance. American Psychologist, 39, 885-
893.
Gaertner, S. L., Mann, J. A., Dovidio, J. F., Murrell,
A. J., & Pomare, M. (1990). How does cooperation
reduce intergroup bias? Journal of Personality and
Social
Psychology,
59, 692-704.
Gaertner, S. L., Mann, J., Murrell, A., & Dovidio, J. F.
(1989).
Reducing intergroup bias: The benefits of
recategorization. Journal of Personality and Social
Psychology, 57, 239-249.
Hackman, J. R. (1987). The design of work teams. In
J. Lorsch (Ed.), Handbook of organizational
behavior (pp. 315-342). Englewood Cliffs, NJ:
Prentice Hall.
Hartel, C. E. J., & Hartel, G. F. (1997). SHAPE-
assisted decision making and problem solving:
Information-processing-based training for condi-
tions of cognitive business. Group Dynamics, 1,
187-199.
Hinsz, V. B., Tindale, R. S., & Vollrath, D. A. (1997).
The emerging conceptualization of groups as
information processors. Psychological Bulletin,
121,
43-64.
Inzana, C. M., Driskell, J. E., Salas, E., & Johnston, J.
(1996).
Effects of preparatory information on
enhancing performance under stress. Journal of
Applied
Psychology,
81, 429-435.
James, W. (1890). The principles of psychology (Vol.
1).
New York: Holt.
Johnston, J., Driskell, J. E., & Salas, E. (1997).
Vigilant and hypervigilant decision making.
Jour-
nal of Applied
Psychology,
82, 614-622.
Karau, S. J., & Kelly, J. R. (1992). The effects of time
scarcity and time abundance on group performance
quality and interaction process. Journal of Experi-
mental Social
Psychology,
28,
542-571.
Lewin, K. (1951). Field theory
in
social
science.
New
York: Harper.
Lewis, B. P., & Under, D. E. (1997). Thinking about
choking: Attentional processes and paradoxical
performance. Personality and Social Psychology
Bulletin, 23, 937-944.
Littlepage, G. E., & Karau, S. J. (1997). Utility and
limitations of the SHAPE-assisted intuitive deci-
sion-making procedure. Group Dynamics, 1, 200-
207.
Mathews, K. E., & Canon, L. K. (1975). Environmen-
tal noise level as a determinant of helping behavior.
Journal of Personality and Social Psychology, 32,
571-577.
McGrath, J. E. (1984). Groups: Interaction and
performance. Englewood Cliffs, NJ: Prentice Hall.
Mead, G. H. (1934).
Mind,
self,
and
society.
Chicago:
University of Chicago Press.
Mullen, B. (1991). Group composition, salience, and
cognitive representations: The phenomenology of
being in a group. Journal of Experimental Social
Psychology, 27, 297-323.
Mullen, B., Chapman, J. G., & Peaugh, S. (1989).
Focus of attention in groups: A self-attention
perspective. Journal of Social Psychology, 129,
807-817.
National Transportation Safety Board. (1979). Air-
craft accident
report:
United Airlines Flight 173,
DC-8-61,
N8082U (NTSB/AAR-79/12). Washing-
ton,
DC:
Author.
National Transportation Safety Board. (1990). Air-
craft accident
report:
United Airlines DC-10-10
engine explosion and landing at Sioux City Iowa
(NTSB/AAR-90/06). Washington, DC: Author.
Pennebaker, J. W., Czajka, J. A., Cropanzano, R., &
Richards, B. C. (1990). Levels of thinking.
Personality and Social Psychology Bulletin, 16,
743-757.
Predmore, S. C. (1991). Microcoding of communica-
tions in accident analyses: Crew coordination in
United 811 and United 232. In R. S. Jensen (Ed.),
Proceedings of the sixth annual international
symposium on aviation psychology (pp. 353-358).
Columbus: Ohio State University.
Reicher, S. (1982). The determination of collective
behavior. In H. Tajfel (Ed.), Social identity and
intergroup relations (pp. 41-83). Cambridge,
England: Cambridge University Press.
Rotton, J., Olszewski, D., Charleton, M., & Soler, E.
(1978).
Loud speech, conglomerate noise, and
behavioral aftereffects. Journal of Applied
Psychol-
ogy, 63, 360-365.
Salas,
E., Rozell, D., Driskell, J. E., & Mullen, B.
(1999).
The effect of team building on perfor-
mance: An integration. Small Group Research, 30,
309-329.
Salovey, P. (1992). Mood-induced self-focused atten-
tion. Journal of Personality and Social Psychology,
62,
699-707.
Shaw, M. E. (1981).
Group
dynamics: The psychology
of small group
behavior.
New York: McGraw-Hill.
Torrance, E. P. (1954). The behavior of small groups
under the stress conditions of survival. American
Sociological
Review,
19, 751-755.
Wegner, D. M., & Giuliano, T. (1980). Arousal-
induced attention to
self.
Journal of Personality
and Social
Psychology,
38, 719-726.
Wickens, C. D. (1996). Designing for stress. In J.
Driskell & E. Salas (Eds.), Stress and human
performance
(pp.
279-295). Hillsdale,
NJ:
Erlbaum.