THESIS

’5
‘

.r

1003

540mm

This is to certify that the

dissertation entitled

THE EFFECTS OF ACUTE STRESSORS ON TRANSACTIVE
MEMORY AND SHARED MENTAL MODELS IN
TEMPORARY PROJECT TEAMS: AN
INFORMATION PROCESSING APPROACH

presented by

Aleksander P.J. Ellis

has been accepted towards fulﬁllment
of the requirements for

PH . D . degree in Philosophy

amaL
/ v

 

 

‘ \
V6181“ r essor

Date January 10, 2003

 

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0- 12771

 

 

 

_UBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DaioDue.p65-p.15

THE EFFECTS OF ACUTE STRESSORS ON TRANSACTIVE MEMORY AND
SHARED MENTAL MODELS IN TEMPORARY PROJECT TEAMS: AN
INFORMATION PROCESSING APPROACH
By

Aleksander R]. Ellis

A DISSERTATION

Submitted to
Michigan State University
In partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

2003

ABSTRACT
THE EFFECTS OF ACUTE STRESSORS ON TRANSACTIVE MEMORY AND
SHARED MENTAL MODELS IN TEMPORARY PROJECT TEAMS: AN
INFORMATION PROCESSING APPROACH
By
Aleksander R]. Ellis

The purpose of this study was to develop and test a model of stress in team-based work
structures. Based on information processing theory, stress was proposed to negatively
impact team processes leading to decrements in team performance. Results indicated that
stress was negatively related to the development of both transactive memory and shared
mental models. Transactive memory and shared mental models were then shown to
mediate the relationship between stress and team performance. A number of personal and
situational characteristics were also proposed to affect the relationship between stress and
team processes. Results indicated that cognitive ability and extraversion had little impact
on team processes and failed to ameliorate the effects of stress on transactive memory
and shared mental models. Results regarding the two situational characteristics, feedback
and prior shared information, were mixed. The level of prior shared information was
positively related to the level of retrieval coordination within the team, as well as shared
mental model similarity and accuracy. Prior shared information also moderated the
negative effects of stress on the level of retrieval coordination within teams. Providing
negative feedback to the team as a whole rather than to an. individual team member did
not directly affect team processes, but it helped to moderate the negative effects of stress

on the level of information allocation within teams. The theoretical and practical

implications of these results are discussed.

Copyright by
Aleksander R]. Ellis
2003

TABLE OF CONTENTS

INTRODUCTION ...................................................................................................... 1
REVIEW OF THE LITERATURE ............................................................................. 9
The Response-Based Perspective ......................................................................... 10
General Adaptation Syndrome ....................................................................... 10
Problems with the Response-Based Perspective ............................................. 12
The Stimulus-Based Perspective .......................................................................... 12
General Adaptation Syndrome ....................................................................... 13
Problems with the Response-Based Perspective ............................................. 13
The Interactive Perspective .................................................................................. 14
The Job Demands-Control Model .................................................................. 14
The Cybernetic Framework ........................................................................... 17
The P-E Fit Model ......................................................................................... 18
Problems with the Interactive Perspective ...................................................... 20
The Transactional Perspective .............................................................................. 20
Problems with Transactional Perspective ....................................................... 21
Summary ............................................................................................................. 23
Teams as Information Processors ......................................................................... 26
Summary ....................................................................................................... 30
The Effects of Acute Stressors on Information Processing within Teams .............. 31
Summary ....................................................................................................... 34
Team Processes .................................................................................................... 35
Shared Mental Models ................................................................................... 36
Transactive Memory ...................................................................................... 40
Summary ............................................................................................................. 43
Personal and Situational Characteristics ............................................................... 45
Cognitive Ability ........................................................................................... 45
Extraversion .................................................................................................. 49
Prior Shared Information ............................................................................... 55
Feedback ....................................................................................................... 62
Summary ............................................................................................................. 66
The Effects of Shared Mental Models and Transactive Memory on
Team Performance ............................................................................................... 67
Shared Mental Models ................................................................................... 67
Transactive Memory ...................................................................................... 69
Summary ..................................................................................................... z'. ...... 72
METHODS ............................................................................................................... 74
Participants .......................................................................................................... 74
Design .................................................................................................................. 74
Task ..................................................................................................................... 75
The MSU-DDD Grid ..................................................................................... 75
Bases and Vehicles ........................................................................................ 76
Tracks ........................................................................................................... 78
Actions Taken Toward Tracks ....................................................................... 79

iv

Team Member Specialties ............................................................................. 79

Measures .............................................................................................................. 80
Cognitive Ability ........................................................................................... 80
Extraversion .................................................................................................. 81
Transactive Memory ...................................................................................... 81
Shared Mental Models ................................................................................... 88
Team Performance ........................................................................................ 91
Manipulations ...................................................................................................... 91
Acute Stressors .............................................................................................. 91
Prior Shared Information ............................................................................... 95
Feedback ....................................................................................................... 97
Procedure ............................................................................................................. 98
RESULTS ............................................................................................................... 101
Manipulation Checks ......................................................................................... 101
Acute Stressors .................................................................................................. 101
Feedback ............................................................................................................ 103
Prior Shared Information .................................................................................... 104
Tests of Hypotheses ........................................................................................... 105
Hypothesis 1 ................................................................................................ 107
Hypothesis 2 ................................................................................................ 107
Hypothesis 3 ................................................................................................ 107
Hypothesis 4 ................................................................................................ 108
Hypothesis 5 ................................................................................................ 108
Hypothesis 6 ................................................................................................ 109
Hypothesis 7 ................................................................................................ 109
Hypothesis 8 ................................................................................................ 109
Hypothesis 9 ................................................................................................ 110
Hypothesis IO .............................................................................................. 110
Hypothesis 1 l .............................................................................................. l l 1
Hypothesis 12 .............................................................................................. 1 ll
Hypothesis 13 .............................................................................................. l 12
Hypothesis 14 .............................................................................................. 112
Hypothesis 15 .............................................................................................. l 12
Hypothesis 16 .............................................................................................. 1 13
Hypothesis 17 .............................................................................................. 113
Hypothesis l8 .............................................................................................. l 14
Supplemental Analyses ...................................................................................... 1 15
DISCUSSION ......................................................................................................... 120
Theoretical Implications ..................................................................................... 120
Practical Implications ......................................................................................... 130
Limitations and Directions for Future Research .................................................. 135
Conclusion ......................................................................................................... 140
REFERENCES ....................................................................................................... 141
APPENDIX A: TABLES ........................................................................................ 187
Table 1: Summary of Assets and Tracks ............................................................. 188

Table 2: Team Member Specialties .................................................................... 189

Table 3: Summary of Hypotheses ....................................................................... 190
Table 4: Means, Standard Deviations, and Intercorrelations Among Acute

Stressor Manipulation Checks ............................................................................ 192
Table 5: Means, Standard Deviations, and Intercorrelations Among Variables
Included in Hypothesis Tests .............................................................................. 193
Table 6: Effects of Stress, Situational, and Personal Characteristics on Shared
Mental Model Similarity and Shared Mental Model Accuracy ........................... 194

Table 7: Effects of Stress, Situational, and Personal Characteristics on Directory
Updating, Information Allocation, Retrieval Coordination and Transactive

Memory ............................................................................................................. 195
Table 8: Effects of Directory Updating, Information Allocation, and Retrieval
Coordination on Shared Mental Model Similarity and Shared Mental Model
Accuracy ............................................................................................................ 196
Table 9: Effects of Directory Updating, Information Allocation, Retrieval
Coordination, Transactive Memory, Shared Mental Model Similarity, and Shared
Mental Model Accuracy on Team Performance .................................................. 197
Table 10: Regression Analysis of Information Allocation, Retrieval Coordination,
and Shared Mental Model Accuracy as Mediators of the Relationship Between
Stress and Team Performance ............................................................................ 198
Table 11: Supplemental Analyses of the Effects of Directory Updating, Information
Allocation, and Retrieval Coordination on Team Performance During the Initial
15-Minute Task .................................................................................................. 199
Table 12: Supplemental Analyses of the Effects of Directory Updating, Information
Allocation, and Retrieval Coordination on Shared Mental Model Similarity and

Shared Mental Model Accuracy During the Initial lS-Minute Task .................... 200
APPENDIX B: FIGURES ....................................................................................... 201
Figure 1: Theoretical Model of Stress in Team—Based Work Structures .............. 202
Figure 2: Generic Information-Processing Model ............................................... 203
Figure 3: The DDD Grid, Including Bases, Vehicles, and Tracks ....................... 204
Figure 4: The Effect of the Interaction Between Stress and Shared Information
on Retrieval Coordination .................................................................................. 205
Figure 5: The Effect of the Interaction Between Stress and Feedback on
Information Allocation ....................................................................................... 206
Figure 6: The Effect of the Interaction Between Shared Mental Model Similarity
and Shared Mental Model Accuracy on Team Performance ................................ 207
APPENDIX C: DISSERTATION PROTOCOL ...................................................... 208
APPENDIX D: COMMUNICATION CODING ..................................................... 231
APPENDD( E: SHARED MENTAL MODEL MEASURE ..................................... 232
APPENDIX F: AREAS OF EXPERTISE ............................................................... 234
APPENDIX G: SHARED INFORMATION MAN IPULATION ............................. 235
APPENDIX H: ACUTE STRESSOR MANIPULATION CHECK ......................... 236

vi

INTRODUCTION

Organizations have witnessed a number of drastic changes in the nature of the
workplace over the last several decades (Cooper, 1998). In the 19808, organizations were
faced with an “enterprise culture,” where privatization, mergers, joint ventures, and
process reengineering became the norm. By the end of the 19805 and into the early
19908, structures became ﬂatter as organizations scrambled to eliminate unnecessary
levels of management by “downsizing.” This created a workforce composed of
employees who had to do more work, while feeling more insecure about their jobs. The
expansion of information technology only added to the problems by overloading
employees with information and increasing the pace of work (Cooper, Dewe, &
O’Driscoll, 2001).

As the changes began to take hold, employees became familiar with words such
as “stress” and “strain.” Between 1985 and 1990, the number of employees reporting
“feeling highly stressed” more than doubled (Northwestern National Life, 1991).
Employees under stress exhibited a number of somatic complaints, including migraines,
tensions headaches, nausea, muscular discomfort, pain, burnout, anxiety, and emotional
exhaustion (e.g., Burke, 1993; Hoiberg, 1982; Lee & Ashforth, 1993; Zander & Quinn,
1962). Stress was recognized as a potential problem not only for the employee, but also
for the organization. In 1980, it was estimated that stress cost organizations
approximately $75 to $90 billion annually nationally (Ivancevich & Matteson, 1980). By
the end of the 19803, the cost surpassed $100 billion (Niehouse, 1987).

To keep up with the changes within the workplace, research on stress has

accumulated in an effort to better understand its causes and consequences. Researchers

have identiﬁed numerous factors that contribute to stress levels among employees,
including workload (e. g., Cooper & Roden, 1985; Kushmir & Melamed, 1991; Westman
& Eden, 1992), role ambiguity (e. g., O’Driscoll & Beehr, 1994; Schaubroek, Cotton, &
Jennings, 1989), role conﬂict (e.g., Jackson & Schuler, 1985; O’Driscoll & Beehr, 1994),
and interpersonal conﬂict (e. g., Keenan & Newton, 1985). These factors tend to strain
employees mentally and physically, leading to lower levels of productivity, increased
absenteeism and turnover, and decreased health and well-being (e. g., Cooper & Payne,
1988; Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964; Karasek & Theorell, 1990; Keita
& Sauter, 1992; Levi, 1981; Matteson & Ivancevich, 1982; Perrewe, 1991; Quick,
Murphy, & Hurrell, 1992). Researchers have also identiﬁed a number of moderators that
may inﬂuence the stress process, such as the Type A behavior pattern (e. g., Davidson &
Cooper, 1980; Lee, Ashford, & Bobko, 1990), negative affectivity (e.g., Parkes, 1990;
Watson & Clark, 1984), and hardiness (e.g., Allred & Smith, 1989; Kobasa, 1982). To
summarize the literature, several books and chapters have been published (e. g., Beehr,
1995; Beehr & Bhagat, 1985; COOper, Dewe, & O’Driscoll, 2001; Ivancevich &
Matteson, 1980; Jex, 1998; Jex & Beehr, 1991; Kahn & Byosiere, 1992).

Although much progress has been made toward understanding organizational
stress, the nature of work continues to change as we move into the 21“ century. A number
of internal and external forces are impinging upon organizations, forcing them to shift to
alternative work arrangements. New technologies such as computer-based
communication systems are being developed and implemented at an exponential rate
(Hesketh & Neal, 1999). Combined with the globalization of trade, the increase in

technological capability has led to reductions in the size of many organizations. These

changes demand that organizations remain ﬂexible and adaptive, ready to expand or
contract at a moment’s notice while continually adopting innovation (Cooper, Dewe, &
O’Driscoll, 2001). In such a fast-paced, ever-changing environment, employees are often
faced with tasks that exceed their capabilities. In order to remain competitive,
organizations have begun to assign tasks to groups or teams of employees (e.g., Cannon-
Bowers, Oser, & Flanagan, 1992; Hackman & Morris, 1975; Salas, Dickenson, Converse,
& Tannenbaum, 1992).

A “team” refers to a distinguishable set of two or more people who interact
interdependently toward a common and valued goal/objective/mission, and who have
each been assigned speciﬁc roles or functions to perform (e.g., Dyer, 1984; Ilgen, Major,
Hollenbeck, & Sego, 1995; Salas et al., 1992). Teams have become prevalent in most
organizations, including hospitals, the military, nuclear power plants, the airline industry,
and the automobile industry. Although organizations often credit their success to the
addition of team-based work structures (e. g., Hammer & Champy, 1993), employees
working in team contexts are still susceptible to stress (e. g., Morgan & Bowers, 1995).

Unfortunately, stress researchers have been unable to keep up with the change
from individual to team—based work structures. Little has been done to further our
understanding of the effects of stress within teams (Boff & Lincoln, 1988; Driskell &
Salas, 1991; Morgan & Bowers, 1995). This lack of research is surprising, considering
that a number of scholars have suggested that studies of stress need to move beyond the
individual to incorporate groups and teams (e.g., Bliese & Jex, 1999; Cox, 1997;
Grifﬁths, 1994). Examining the stress process in team-based work environments is

important for both practical and theoretical reasons.

Practically, stress within teams has been at least partly implicated in a number of
tragic incidents. For example, in 1972, an Eastern Airlines ﬂight crashed into the Florida
Everglades, killing 99 passengers and crew members. During the ﬂight, one of the
landing gear lights burned out. While the entire crew focused their attention on ﬁxing the
problem, the autopilot became disengaged and the plane slowly began to descend into the
ground (National Transportation Safety Board, 1972). A similar accident occurred in
1978, when a United Airlines ﬂight ran out of fuel because the crew was trying to ﬁx a
problem with the landing gear. The ﬁrst ofﬁcer and ﬂight engineer warned the captain
about the problem on several occasions, yet the captain failed to listen (National
Transportation Safety Board, 1979). In the military, anti-air warfare systems have
become increasingly complex, forcing soldiers to process a large amount of information
in a short period of time. The consequence of placing soldiers in such an intense team
environment was exempliﬁed in 1988, when a US. Navy Aegis cruiser mistakenly shot
down an Airbus 300 Iranian commercial airliner (Salas, Driskell, & Hughes, 1996).

Theoretically, teams require employees to attend to additional responsibilities that
may alter conceptualizations of the stress process. When individuals are placed in an
interdependent team setting, tasks are often completed through the interaction of the team
members. This social aspect of teams has recently received attention from scholars
interested in socially shared cognition and information processing systems (e.g.,
Gruenfeld, Martorano, & Fan, 2000; Hinsz, Tindale, & Vollrath, 1997). Employees
encode, store, and retrieve information as individuals, but teams process information
between as well as within the minds of the team members (Ickes & Gonzales, 1994). In

order for teams to be effective, there must be a coordinated exchange of information

among the team members (Oser, McCallum, Salas, & Morgan, 1989). Social interaction
has been identiﬁed as a critical factor in the stress process within teams (Hackman &
Morris, 1975; Morgan & Bowers, 1995). Researchers have suggested that stressful
environmental conditions may disrupt social interaction by focusing the attention of the
team members away from the others and toward themselves (Driskell, Salas, & Johnston,
1999).

Despite the practical and theoretical importance surrounding the study of stress
within tearm, few researchers have attempted to empirically or conceptually tackle the
issue. The purpose of this dissertation is to develop and test a theoretical model of stress
in team-based work structures. The basic model is shown in Figure 1. The hypotheses
derived from the mode] are summarized in Table 3. The model draws upon a number of
different literatures. To begin the literature review, I discuss how researchers have
conceptualized the stress process at the individual level. Then I introduce the literature on
groups as information processors in order to show how team members may respond
differently than individuals in stressful situations. In particular, when team members are
faced with stressful environmental conditions, two social interaction processes may
suffer: shared mental models and transactive memory. Next, a number of personal and
situational characteristics are introduced as possible moderators of the negative
relationship between the presence of stressful environmental conditions and team
processes. More speciﬁcally, regarding personal characteristics, I examine the level of
cognitive ability and the level of extraversion among the team members. Regarding
situational characteristics, I examine the amount of prior shared information given to

each team member and whether negative feedback is given to one team member or the

team as a whole. Finally, team processes are linked to team performance. After
introducing the model, a laboratory study is designed in order to test my hypotheses.
Before beginning the literature review, it is necessary to note several
characteristics which limit the scope of this study. First, this study is primarily interested
in examining acute stress, which Salas, Driskell, and Hughes (1996) deﬁne as “sudden,
novel, intense, and of relatively short duration, disrupts goal oriented behavior, and
requires a proximate response. Acute stress is illustrated by the prototypical ‘emergency’
situation, in which the scenario unfolds rapidly, the task must be dealt with in a short time
period, and the consequences of poor performance are immediate” (p. 6). This study will
not deal with chronic stress, which exacts its effects over a much longer period of time.
Second, this study focuses on the negatixge consequences of stressful situations
that are acute in nature. Within the organizational stress literature, most researchers have
concentrated on examining the effects of stressful situations on job performance. This has
stimulated a longstanding debate regarding whether performance decreases or increases
in stressful situations. Some have suggested that the relationship is negative (e.g., Jackson
& Schuler, 1985), others have shown that it is positive (e. g., McGrath, 1976), and still
others feel that it follows the Yerkes-Dodson law, which results in an inverted, U-shaped
curve (e.g., Jamal, 1984). Unfortunately, most empirical research has found weak and
inconsistent relationships between the presence of job stressors and performance (see Jex,
1998). As a result, Fried and Tiegs (1995) suggested that “the literature on stress and
performance would beneﬁt greatly if researchers concentrated on developing and
evaluating theoretically derived linkages between speciﬁc job stressors and the speciﬁc

duties and responsibilities of employees in a particular organizational position” (p. 282).

This study does not examine the direct relationship between acute stressors and job
performance. Instead, I theoretically link the presence of acute stressors to the social
behavior of project team members. Although I agree that acute stressors can beneﬁt team
members, I believe that the presence of acute stressors will negatively affect team
members’ social reactions. These reactions then are proposed to affect the team’s level of
performance, supporting researchers who suggest that performance is indirectly, not
directly, affected by job stressors (e. g., Beehr & Bhagat, 1985).

Third, research on stress has undergone such a drastic expansion over the last ﬁfty
years that there are now at least four broad ﬁelds specializing in the study of stress,
including medicine, clinical psychology, engineering psychology, and organizational
psychology (Beehr & Franz, 1987). Although this study incorporates aspects of each
discipline, the primary focus is on the organizational psychology literature. Attempting to
summarize all four ﬁelds would be impossrble and impractical, since many of the
concepts do not transfer well across domains.

Fourth, while there are a number of different types of teams that do work in
organizations, this study concentrates on project teams. As noted earlier, organizations
need to remain adaptive and ﬂexible in order to remain competitive. One way to
accomplish this goal is to make sure that the various subsystems within the organization
are ﬂexible and adaptive. Employees are now being asked to integrate into team
environments as leaders or members, knowing full well that the team will disband in the
near future (Allred, Snow, & Miles, 1996). These fast acting, temporary project teams
have become the norm within many organizations (e.g., Cohen & Bailey, 1997; Herriot,

1993; Sundstrom, McIntyre, Halfhill, & Richards, 2000; Zemke, 1978). A recent national

survey found that 30% of all teams are temporary project teams (Gordon, 1992).
Members of project teams are expected to be self-managing, be able to handle novel tasks
without prior training, and be willing to invest in a continuous learning process (Allred,
Snow, & Miles, 1996). Project team members’ day-to-day activities differ from those of
other employees, requiring them to move from one task to another on a ﬁequent basis.
For example, an employee may be a member of an engineering team for six months and
then a member of a production team for the next two months.

In sum, for this study, I am interested in examining team members’ social
responses to stressful environmental conditions that are time-limited, sudden, and often
unexpected. In these types of situations that can occur in a variety of team settings, quick
responses are critical to success and the consequences of failure are immediate. However,
in this study I am focusing on teams that come together for short periods of time in order
to perform a certain set of tasks together. Although this limits the generalizability of my
results, acute stress in project teams is relevant to a number of applied settings such as

aviation, military operations, the automotive industry, and other industrial occupations.

REVIEW OF THE LITERATURE

Although our knowledge of the stress process in individual employees has grown
exponentially over the past ﬁfty years, there is still much we do not know due to the
complexity of the phenomenon under study (e. g., Beehr, 1995; Jex & Beehr, 1991;
Spector, 1992). The problem has been compounded by researchers’ inability to settle on
one conceptualization, deﬁnition, and operationalization of organizational stress (e. g.,
Cooper, Dewe, & O’Driscoll, 2001; Mason, 1975; Seyle, 1975). Without a clear
conceptualization of the concept, it becomes difficult to determine the nature and
direction of research and limits the explanations that can be offered regarding any
research ﬁndings (Newton, 1995). The difﬁculties in determining what is meant by
organizational stress can be seen in the plethora of deﬁnitions offered by researchers. For
example, Selye (1956) deﬁned stress as a “nonspeciﬁc response to any demand.” Caplan
and his colleagues (1975) deﬁned it as “any characteristic of the job environment which
poses a threat to the individual.” Cannon (1929) deﬁned stress as “a condition at work
interacting with worker characteristics to disrupt psychological or physiological
homeostasis.” Others deﬁne stress as “a transaction between the person and the
environment [that] is evaluated by the person as a harm, threat, or challenge to that
person’s well-being” (Lazarus, 1991).

The variance in the deﬁnitions offered by researchers reﬂects the four different
perspectives of the stress that have been identiﬁed in the literature: response-based,
stimulus-based, interactional, and transactional. Researchers have often treated these
perspectives as distinct, although they share a number of features with one another. In

following few pages, I discuss the main thrust of each approach to organizational stress,

include examples of the theories that they generated, and note any conceptual or
operational problems that have been identiﬁed.
The Resmnse-Based Perspective

A response-based approach views organizational stress as a dependent variable, or
the employee’s response to threatening stimuli. The lay person may ﬁnd it easy to
identify with this perspective, as most people have felt “stressed” at some point in their
working lives. Responses usually consist of physical, psychological, and behavioral
components, which combine to form a level of “strain” on an employee (Cooper, Dewe,
& O’Driscoll, 2001). The origins of this approach to stress research can be linked to
medicine. Wolf and Wolf (1943) were some of the ﬁrst researchers to study stress as a
response. Their patient, Tom, exhibited changes in stomach activity when faced with
certain environmental conditions. Soon the response-based perspective became a
powerful force in stress research, especially with the introduction of Hans Selye’s general
adaptation syndrome (Selye, 1956).

General Adaptation Syndrome (GAS 1. Selye (1956) suggested that stress was a
nonspeciﬁc bodily response to any demand made upon it. An individual’s response was
invariant to the nature of the stressor, which meant that responses generally followed a
universal three-stage pattern: alarm, resistance, and collapse. When faced with a stressor
in the environment, the individual exhibits an initial alarm reaction, which is the initial
psychO-physical response. At that point, the individual’s resistance to the stressor is
lowered. After the initial shock phase, the individual enters the counter-shock phase and
resistance levels begin to increase. The reaction can be formed as either a “ﬁght or ﬂight

response” (Cannon, 1935). The individual’s body is ready to take action, as sympathetic

lO

activity increases with the release of catecholamines, the metabolism of fat and glucose,
and the delivery of oxygen to muscle ﬁbers. The initial alarm reaction is replaced with
either the adaptation response or a return to equilibrium. If resistance continues for a long
period of time, the individual could enter the last stage of collapse. Energy levels needed
for adaptation have been depleted, which could result in exhaustion or even death (Selye,
1983).

The general adaptation syndrome was an inﬂuential theory at the time for a
number of reasons. For one thing, Selye emphasized that stress reactions are not
automatically detrimental. This explanation ﬁts well with an evolutionary perspective,
which views stress as a necessary component of developing societies. In the past,
individuals had to either stand and conﬁont an enemy or run away from a potentially
dangerous situation. This is analogous to the “survival of the ﬁttest” promoted by
Darwinism. Everyone is faced with stress and those who react appropriately will carry
on, while others will be eliminated. However, despite intuitive appeal, there were
problems with applying GAS. In contemporary society, employees’ choices when faced
when responding to stress are much more limited. In the workplace, there is little
opportunity to physically ﬁght to combat stress or run away from the situation.
Employees, for the most part, lead sedentary lives and have no outlet for the psycho-
physical response associated with stressful environmental conditions (Sutherland &
Cooper, 2000). In addition to problems Speciﬁc to the GAS, there were also a number of
characteristics exhibited by response-based theories in general that reduced their

applicability to the organizational stress process.

11

Problems with the Resp_onse-B ased Perspective. The major difﬁculty in viewing
stress from a response-based perspective lies in the assumption that the stimulus
dimension of the stress experience can be ignored. Research has indicated that responses
to different stimuli do not always follow the same pattern. Responses are, in fact, often
stimulus-speciﬁc and depend on certain types of hormonal secretions. For example, the
release of adrenaline seems to occur more frequently in anxiety-provoking situations,
while noradrenaline is released in response to aggression—provoking events (Sutherland &
Cooper, 2000). Response—based theories also tend to focus on the psycho-physical
responses individuals exhibit in a stressful situation and ignore psychological responses,
which reduces their applicability in situations that stimulate psychosocial stress (Christian
& Lolas, 1985). Another problem with response-based theories is that they tend to
disregard individual differences, which are considered important aspects of the stress
process (e.g., Cox, 1990; Sutherland & Cooper, 1990). As a result, response-based
theories are seen as representing only one component of the stress process (Cooper et al.,
2001). In order to offer a more comprehensive conceptualization of the concept,
researchers began to utilize different approaches to examining stress within
organizational settings. This moved the study of stress away from response-based models
to stimulus—based models.

The Stimulus-Based Perspective

Researchers examining stress from a stimulus-based perspective assume that there
are certain forces that affect an individual in a disruptive fashion and are primarily
interested in identifying such potential “stressors” for employees (Goodell, Wolf, &

Rogers, 1986). This perspective is rooted in physics and engineering, where stress is

12

viewed as a force exerted on a certain object. Objects possess certain tolerance levels
protecting them against such external forces. When those tolerance levels are exceeded,
the object is distorted in some fashion. Depending on the strength of the force, temporary
or permanent damage could result (Sutherland & Cooper, 2000). Within organizations,
employees conﬁont external forces that arise primarily from certain aspects of their job.
In an effort to better delineate some of the speciﬁc core characteristics of jobs that act as
stressors, researchers developed the job characteristics model of stress (e.g., Beehr,
1985).

The Job Characteristics Model. The job characteristics model has identiﬁed a
number of job characteristics that are considered to be stressors, including role ambiguity,
role overload, role conﬂict, underutilization of skills, job insecurity or actual job loss, and
lack of participation in organizational decision making (e.g., Beehr, 1976; Beehr, Walsh,
& Taber, 1976; Caplan, Cobb, French, Van Harrison, & Pinneau, 1975; French &
Kaplan, 1973). These job characteristics result in feelings of uncertainty among
employees, and the longer the uncertainty continues, the more severe the levels of stress
(Beehr, 1985; Ivancevich & Matteson, 1980). Although a number of studies support the
idea that altering certain job characteristics can produce a more optimal work
environment (see Hackman & Oldham, 1980), stimulus-based theories encounter the
same problems as their response-based counterparts.

Problems with the Stimulus-BaSed Persﬁtive. Like the job characteristics
model, most stimulus-based theories of job stress subvert the importance of individual
differences in the stress process. Despite the inclusion of a small subset of individual

difference variables, such as growth need strength and experience in the job

13

characteristics model, these theories generally assume that the environment plays the
primary role. However, most researchers now agree that the situation and the person
often play equal roles in the stress process. For example, variability in tolerance levels
and expectations could explain why two individuals react differently when exposed to the
same situation (Cooper et al., 2001). The stimulus-based perspective, like the response-
based perspective, focuses too much on one end of the stimulus—response paradigm. Due
to problems with both the stimulus and response-based approaches to organizational
stress, researchers began to move toward those processes that link the individual with the
environment. The result was an effort to examine stress as an interaction between person
and Situation.
The Interactive Persp_ective

The interactive model of stress embodies aspects of both the response-based and
stimulus-based models of stress. In this combined model, the presence of certain working
conditions is thought to be associated with a number of stress responses. Various
organizational characteristics, situational factors, and individual differences then
moderate this stimulus—response paradigm. Researchers have identiﬁed a number of
interactive models of the stress process within the literature, including the job demands-
control model, the cybernetic model, and the P—E ﬁt approach (e.g., Cooper et al., 2001;
Cummings & Cooper, 1979; Edwards & Cooper, 1988; Eulberg, Weekly, & Bhagat,
1988; Kahn & Byosiere, 1992). .

The Job Demands-Control Model. The job demands—control model of stress was
initially developed by Karasek (1979) who proposed that workers become stressed as a

result of the interaction between job dermmds and job control. “Job demands are deﬁned

14

as psychological stressors, such as requirements for working fast and hard, having a great
deal to do, not having enough time, and having conﬂicting demands . . . Job decision
latitude [or control] comprises two components: the worker’s authority to make decisions
on the job (decision authority) and the variety of skills used by the worker on the job
(skill discretion). Operationally these two components are combined into one measure of
decision latitude, or control” (Ganster & Schaubroeck, 1991, pp. 241—242). Karasek felt
that stress occurs primarily in “high-strain” jobs, which are high in demands and low in
control. In “active jobs,” where demands and control are both high, employee
satisfaction, motivation, and healthful regeneration can result. “Job demands put the
employee into an aroused or motivated state. If this aroused or motivated state is
accompanied by low decision latitude or control over the job, this arousal will not be
released in the normal execution of the job. It is this non-release, according to Karasek,
which leads to negative psychological and physical consequences” (Jex & Beehr, 1991,
p.322).

The job demands-control model of stress has garnered quite a bit of interest from
researchers since its introduction and a number of its components have been empirically
supported. Researchers have found that indices of strain and ill-health are positively
related to the demands of the job and negatively related to the employee’s level of control
(e.g., Fletcher & Jones, 1993; Krasek, 1979; Landsbergis, 1988; Parkes, Mendham, &
von Rabenau, 1994). The results regarding the hypothesized interactive effects of job
demands and control, however, have been much more ambiguous. Although initial efforts
supported the interaction effect (e.g., Karasek, 1979), the statistical techniques that were

used were inadequate and possibly inﬂated the effect sizes (Edwards & Cooper, 1990;

15

Ganster & Fusilier, 1989). Subsequent efforts failed to conﬁrm the predicted interaction
effect (e.g., Carayon, 1993; Fletcher & Jones, 1993; Hurrel & McLaney, 1989;
Landisbergis, 1988; Payne & Fletcher, 1983; Spector, 1987). Despite the presence of a
few studies supporting the interaction (e.g., Dwyer & Ganster, 1991), researchers
concluded that the model’s “empirical validity has yet to be established” (Ganster &
Fusilier, 1989, p. 254).

Due to the inconsistent results, researchers have suggested that the job demands-
control model may be deﬁcient in some respect. Researchers feel that Karasek’s model
neglects to measure certain variables that could inﬂuence the interactive effects of job
demands and employee control (e. g., Schaubroeck & Merritt, 1997). Some have focused
their attention on the level of co-worker social support within the workplace and have
found results that are generally supportive of a three-way interaction between job
demands, control, and social support (e. g., Johnson & Hall, 1988; Parkes, Mendham, &
Von Rabenau, 1994). The job demands-control interaction holds only when there is little
social support within the workplace. Others have concentrated on the effects of domain-
speciﬁc individual differences. Although Karasek’s (1979) model is interactive, both
variables are environmental in nature. The model, by neglecting the interaction between
person and environment, assumes that the main propositions will remain applicable
across all employees (Xie, 1996). Schaubroeck and Merritt (1997) found that the
demands-control interaction holds for people who are high in self-efﬁcacy because they
feel conﬁdent in their ability to do their job. Although these results support the addition

of individual difference variables to Karasek’s model, other interactive stress theories,

l6

such as the cybernetic approach and P-E ﬁt theory, do a better job of combining the
person and the situation.

The Cybernetic Frmwork. Cybernetics involves the functioning of self-
regulating systems and was developed by researchers in order to explain how individuals
use information and feedback to control purposeful behavior (Weiner, 1948; Ashby,
1956). At the heart of the theory is the negative feedback loop, which operates through a
number of distinct steps. First, an individual constructs a perception of the environment,
which gets sent via an input function to the comparator. Then the comparator evaluates
the perceived environment against a relevant reference criterion and, if there is a
discrepency, an output ﬁrnction gets sent back in order to change the environment,
thereby reducing or eliminating the discrepancy. The basic premise assumes that, when
an individual realizes that he or she deviates ﬁom a certain goal state, they will be
motivated to engage in certain behaviors to rectify the discrepancy (Edwards, 1992).

Initially, researchers in the biological and physical sciences used cybemetics to
explain how systems adjust to disturbances (see Cummings & Cooper, 1979). However,
the theory can easily be translated to ﬁt into the organizational stress literature. Basically,
“the theory deﬁnes stress a discrepancy between an employee’s perceived state and
desired state, provided that the presence of this discrepancy is considered important by
the employee” (Edwards, 1992, p. 245). According to cybernetic theory, stress does not
result from individual or environmental factors alone, but rather from an ongoing
relationship between the two. Variables in the physical and social environment, the
employee’s personal characteristics, social information, and the employee’s own

construction of reality inﬂuence an employee’s perceptions. These perceptions are then

17

compared to the employee’s own desires, which refer to any state or condition that the
employee consciously wants. When the employee’s perceptions and desires do not match,
stress is often the result. This can affect the employee’s psychological and physical
health, as well as any efforts to cope with the problem (Edwards, 1992).

Although cybemetics offers a comprehensive, interactive view of the stress
process, little has been done to empirically test the model. One reason may be that it is
impossible to operationalize several of the more important concepts, such as the
employee’s perceived and desired states. Work stressors “frequently present employees
with a myriad of subtle and contradictory signals, making it difﬁcult to interpret what is
real or imaginary. Direct knowledge of the states of these situations is also rarely
available” (Cummings & Cooper, 1979, p. 404). There are both environmental and
individual variables that make the assessment of actual work conditions difﬁcult. Even if
an employee is able to evaluate their perceived and desired states, it is not clear that they
are comparable (Cummings & Cooper, 1979). Researchers have noted that there are a
number of substantive and methodological problems with measuring the discrepancy
between two variables (e. g., Cronbach & Furby, 1970; Johns, 1981). Difﬁculties in
testing the cybernetic model of stress have reduced its practical and theoretical
contribution to the literature. P-E ﬁt theory has encountered similar problems, although
there have been some initial attempts to test its validity.

The P-E Fi_t Model. The P-E ﬁt model of organizational stress has been one of the
most popular perspectives within the literature (e. g., Edwards, 1991, 1996; Edwards &
Cooper, 1988; Edwards & Van Harrison, 1993). The theory rests on the assumption that

employees feel a certain level of strain when they are out of equilibrium with the

18

environment, due to the presence of unmet needs and demands (Cooper et al., 2001). Two
major versions of P-E ﬁt have been identiﬁed: supplies-values (S-V) ﬁt and demand-
abilities (D-A) ﬁt. S-V ﬁt compares the employee’s values with the aspects of the
environment that have the potential to fulﬁll those values. Values can include a number
of different internal characteristics, such as interests, motives, and goals (e. g., Cummings
& Cooper, 1979; Edwards, 1992). When the environmental supplies cannot satisfy an
employee’s values, the employee will begin to feel more and more strained at work (e. g.,
Cummings & Cooper, 1979; Edwards, 1996; French, Caplan, & van Harrison, 1982). D-
A ﬁt, on the other hand, compares the knowledges, skills, and abilities of the employee
with the demands of the job. Demands can be quantitative, qualitative, objective, or
socially constructed. When the demands exceed the employee’s ability level, strain can
result (e. g., Edwards, 1996).

The P-E ﬁt model of stress is intuitively simple. However, empirically testing the
model has proven to be difﬁcult, due to problems measuring relevant constructs (i.e.,
environmental supplies) and specifying the nature of person-environment misﬁt
(Edwards & Cooper, 1988). Theoretically and methodologically, there are three forms of
ﬁt that can be examined. Researchers can calculate the discrepancy between P and E, the
interaction between P and E, or the proportion of P that is fulﬁlled by E. Often these three
interpretations are treated as interchangeable, despite recommendations to the contrary.
Once researchers choose one form of ﬁt, they still need to worry about the use of
difference scores, the number of ﬁt dimensions to include, and the measurement of the P
and E components (Edwards & Cooper, 1990). Furthermore, although researchers have

recently found some support for the P-E model of stress (Edwards & Van Harrison,

19

1993), there are a number of characteristics common to all interactive theories that the P-
E model cannot avoid.

Problems with the Interactive Perspective. The interactive approach to stress
research attempted to build upon previous theories, which focused solely on the stimulus
or response side of the equation. However, the interactive approach is simply based upon
the statistical interaction between the stimulus and the response. Although some theories
(e.g., cybemetics) try to break out of this rather static mold, interactions generally ignore
the process itself (Cooper et al., 2001). When restricted to simple cause and effect
relationships, the complex stress process can only be further understood by adding
additional moderator variables into the equation (e. g., Lazarus & Launier, 1978).
Researchers have noted that “it is important now to move beyond the simple
identiﬁcation of potential moderator variables to more comprehensive theories that
attempt to explain the mechanisms by which all relevant factors interact” (Cooper et al.,
2001, p.11). Dissatisfaction with incomplete versions of the stress process led researchers
to propose that the relation between environmental demands and the individual’s
response should be emphasized (McGrath, 1976). This more relational perspective led to
the development of a transactional approach to the stress process.

The Transactionalﬂrspective

The transactional approach represents a fundamental shift in how stress is
conceptualized and research is conducted. Previous theories following response-based,
stimulus-based, and interactive perspectives of the stress process viewed the causes and
consequences of stress as conceptually distinct. However, the transactional perspective,

as the name suggests, considers these constructs as “deﬁned relationally and ultimately

20

become inseparable from the context within which the stressﬁrl encounter takes place”
(Cooper et al., 2001, p. 13). Researchers espousing this viewpoint believe that the
cognitive appraisal and coping that underlie the stress process are critical for advancing
our understanding of the concept of stress within the workplace.

According to the theory, employees engage in two forms of cognitive appraisal:
primary appraisal and secondary appraisal during a stressful encounter. First, the
employee determines, through primary appraisal, whether the situation they are
encountering represents harm, threat, or challenge. If the situation is appraised as threat to
the employee’s well-being, the secondary appraisal process begins to search for the
appropriate coping resources (e. g., Lazarus, 1966; Lazarus, 1991; Lazarus & Folkman,
1984, 1987). Lazarus (1995) deﬁnes coping as “the cognitive and behavioral efforts a
person makes to manage demands that tax or exceed his or her personal resources” (p. 6).
Like other theories (e. g., P-E ﬁt theory), a stressful situation from a transactional
perspective disrupts the employee’s level of homeostasis, which must be rectiﬁed.
However, the transactional approach sets itself apart from other theories by concentrating
on the adaptive process of meaning, adjustment, and coping, instead of structural
relationships between person and environment (Dewe, Cox, & Ferguson, 1993).

Problems with the Transactiong Perspective. Although a number of researchers
feel that the transactional approach represents a major advance in our thinking regarding
the organizational stress process (e.g., Cooper et al., 2001; Harris, 1991; Salas, Driskell,
& Hughes, 1996), it is not immune from criticism. A number of problems stem from the
empirical approach researchers are forced to take when attempting to verify and test any

transactional theory of stress. In essence, the transactional approach views stress as a

21

dynamic, ongoing relationship between person and environment. As a result, typical
nomothetic research designs do not represent an adequate test of the model and must be
dropped in favor of more ideographic designs. This may help researchers understand the
stress process within one individual, but provide little insight into the cognitive processes
or emotions of others. In fact, “the limited generalizability of ideographic ﬁndings
present a chronic problem in the veriﬁcation and testing of the transaction process
approach” (Harris, 1995, p. 26).

By focusing on the individual, the transactional approach tends to neglect the
environment. Although knowledge regarding speciﬁc stressful environmental conditions
is helpful, transactional researchers believe that it has received adequate attention within
the literature. If the study of stress is to move forward, researchers need to “generate
knowledge about the kinds of persons who are more or less vulnerable to divergent
sources of stress” (Lazarus, 1995, p. 10). However, by ignoring the situational part of the
equation, the transactional approach leaves a number of questions unanswered. For
example, what is it about the workplace that activates the stress and coping processes?
Applied psychologists are particularly interested in the answer, as one of their duties is to
inform organizations regarding potentially harmful working conditions. Brief and George
(1995) note that “the challenge to organizational researchers is the development of theory
to guide one to identify those conditions of employment likely to affect adversely the
psychological well—being of most persons exposed to them” (p. 16). Transactional
researchers, supporting an ideographic perspective, limit themselves to better
understanding the cognitive processes of stress for single subjects while ignoring the

environment (Harris, 1995).

22

Several problems also lie in the conceptualization of the coping process. Lazarus
and Folkman (1987) suggest that individuals select different coping strategies for
different situations based on their level of individual control. The coping strategies they
select then act as mediating variables, altering the relationship between the person, the
environment, and the emotional response. However, this perspective fails to recognize
that individuals may possess a particular coping style that they tend to rely on across
situations (e.g., Newton, 1989). If coping style is recognized as a variable, then the
relationship becomes a moderated one. This would change the relationships among the
variables included in the transaction and would modify the analytical strategies that could
be used to test the model (Harris, 1995). Another conceptual difﬁculty involves the
recursive nature of the theoretical arguments behind the coping process. Folkman and
Lazarus (1988) note that “the relationship between emotion and coping in stressful
encounters is bidirectional, with each affecting the other” (p. 466). However, this makes
it difﬁcult to determine whether stress affects coping or coping affects stress (Harris,
1995).
ﬁlmy.

Clearly there has been quite a bit of debate within the literature regarding the
conceptualization of stress within the workplace. Some have viewed it as a response,
others as a stimulus, others as an interaction, and still others as a transaction. Conceptual
and operational difficulties abound, no matter which perspective is adopted by
researchers. My intention in reviewing the four major perspectives of the stress process

identiﬁed in the literature was not to denigrate their contribution to our understanding of

23

occupational stress. Each has the potential to further our understanding of organizational
stress, depending on the situation (Cooper et al., 2001).

In this study, the focus is on the effects of acute stressors within temporary project
teams. Going back to the earlier deﬁnition of teams, team members must interact with
one another in order to achieve their common goals and objectives because each team
member depends on others to get the job done. The social interaction between team
members has been shown to be an integral part of team decision making and problem
solving (e.g., Gouran & Hirokawa, 1983; Hackman & Morris, 1975; Hirokawa &
Scheerhom, 1986), allowing team members to pool information and resources (e. g.,
Bamlund, 1959; Marquart, 1955; Zaleska, 1978), catch errors and reject inaccurate
statements (e.g., Shaw, 1932; Taylor & Faust, 1952), and inﬂuence the decisions of
others (e. g., Riecken, 1958). As group members interact, a certain chemistry is created
through the synthesis of ideas and viewpoints that controls much of what the team does
within the organization (Poole & Hirokawa, 1986). Although the chemistry can be faulty
(e.g., Janis, 1972), social interaction is an essential component of any team. In fact,
researchers consider interaction to be the key to understanding team behavior (Hackman
& Morris, 1975).

However, little is known about how the social chemistry between team members
reacts in stressful situations. In order to further our understanding of the stress process
within teams, the interactive nature of teams needs to be taken into account (Cannon-
Bowers & Salas, 1998; Morgan & Bowers, 1995). As a result, this dissertation focuses
primarily on the social responses of team members to the presence of acute stressors.

Given the situation, the conceptualization of the stress process that is most appropriate is

24

the resmnse-based perspective. Therefore, in Figure 1, the emphasis will be placed on
team processes and how they change when acute stressors are introduced. Aspects of

other perspectives of the stress process will be included, but will only play a secondary
role.

In order to develop my response-based model of stress, I base the remainder of
the literature review on the emerging view of groups as information processors.
Researchers have suggested that this perspective is particularly useful model for
understanding the social processes that occur within groups and holds promise as a meta-
theoretical foundation for explaining many group phenomena (Hinsz et al., 1997). In the
next section, I describe how groups and teams have been conceptualized as information
processing systems within the literature. Then I explain how the collective information
processing capabilities of team members can be disrupted by the introduction of acute
stressors.

By utilizing information processing theory to examine the implications of team
members’ behavioral responses in stressful situations, I hope to better understand stress
within teams, n_o_t_ team stress. I am not interested in developing an emergent theory of
stress at the team level. Emergence occurs when individuals interact and create a
collective construct that originates in their individual cognition, affect, behaviors, or other
characteristics (e.g., Kozlowski & Klein, 2000). I do not suggest that stress exhibits
emergent properties that alter the conceptualization of the construct at the team level.
Team members feel stress just like other employees and respond similarly when faced

with stressful environmental conditions. However, the implications of such responses

25

 

may be much different in team situations, where employees are required to work
interdependently with one another.
Teams as Information Processors.

Work groups and teams are facing tasks that are increasingly intellectual and
cognitive in nature (Galegher, Kraut, & Egido, 1990; Salas, Dickenson, Converse, &
Tannenbaum, 1992; Walsh & Ungson, 1991; Weick & Roberts, 1993). As a result,
researchers have begun to extend methodological and theoretical developments in
cognitive psychology that have traditionally been targeted toward individuals in order to
better understand how groups and teams process relevant and available information in
order to accomplish their goals (e.g., Bazerman, Mannix, & Thompson, 1988; Hastie,
1986; Hinsz et al., 1997; Hinsz, Vollrath, Nagao, & Davis, 1988; Ickes & Gonzales,
1994; McGrath & Hollingshead, 1994; Tindale, 1989).

For instance, among individuals, researchers deﬁne information processing as a
sequence of operations within the human mind that takes in information, transforrm it,
and produces some sort of output. This sequence can take on a number of different forms,
although researchers have suggested that there are certain elements that remain fairly
consistent. These components are shown in Figure 2. An individual, when placed in a
certain context, initially possesses a processing objective that directs his or her search of
the environment. Individuals generally look for information that fulﬁlls a certain
objective, mission, goal, etc. While searching, individuals attempt to actually perceive the
information in the attention phase. This information is then evaluated, interpreted and
transformed through an encoding process that prepares it for storage within the mind. To

get the information out again, the individual must retrieve it. The encoding, storage, and

26

retrieval process follows certain rules that are contained in the individual’s processing
workspace. Based on the information that comes out of the processing workspace, the
individual will likely make some sort of response that could involve making a decision or
solving a problem.

The core elements that comprise an individual’s information processing system
have recently been transferred to groups and teams (Hinsz et al., 1997). However, that
does not mean that the two systems are identical. In groups and teams, information
processing occurs between as well as within the minds of the group members (Ickes &
Gonzales, 1994). Information, ideas, and cognitive processes can all be potentially shared
within the group or team through verbal or nonverbal interaction. Hinsz and his
colleagues (1997) deﬁne information processing in groups as “the degree to which
information, ideas, or cognitive processes are shared, and are being shared, among the
group members and how this sharing of information affects both individual- and group-
level outcomes” (p. 53). The social nature of teams changes the way that each component
functions within the larger network as interactions and the interdependence between team
members enter each phase of the process.

Processing objectives within teams still represent a targeted search of the
environment. However, the team members need to have a common or shared frame of
reference for the processing objective. If the search pattern is fragmented within the team,
each team member could treat important information differently. For example, if a men’s
basketball team is playing defense and one team member thinks they are playing man-to-
man while the rest of the team thinks they are playing a zone, that team member will be

looking for different pieces of information. Instead of picking his spot on the ﬂoor, he

27

will look for his man. By ignoring or misinterpreting certain pieces of information due to
separate processing objectives, the system may break down and the other team may score
a basket.

Like individuals, teams need to pay attention to information to process it.
However, it is not necessary for all the team members to pay attention to all the
information needed by the team. Through interaction, individuals can be assigned to
perceive a certain set of information so that the team members are not overloaded. In
addition, if one team member perceives an important piece of information, he or she can
bring it to the team’s collective attention. Only a small subset of team members need to
perceive the information in order for the team to perceive it as a whole (e.g., Laughlin,
1980; Stasser & Titus, 1985, 1987). For instance, the basketball team, while playing
defense, may encounter a point guard who is too quick for their players. When the point
guard drives to the hoop, he always scores. So the team members playing defense on the
perimeter yell “drive” whenever he drives to the hoop. At that point, their teammates
converge under the basket, knowing that an opposing player is coming, although they did
not perceive the initial move themselves.

The collective nature of the information processing system within teams continues
in the encoding stage. At that point, team members’ individual representations need to be
combined into a representation for the entire team (Wilson & Canter, 1993). Sometimes
team members nmy attach different meanings to the same information, so the team would
not share a representation of the information. If encoding is allowed to occur with
variance in team members’ mental representations, problems could arise at a later point in

time (Bettenhausen & Murnighan, 1991; Kim, 1993). To avoid such problems, and to

28

facilitate the encoding process, any differences should be brought up within the team as
soon as possible through open and direct communication. Going back to the basketball
team, one team member could call a time-out if he feels that things are not going as well
as they could be. When he yells “drive,” one of his teammates does not move to the
basket and instead sets up on the perimeter. By interacting, the team members can all
converge on one interpretation of the word “drive” during the game.

Once the team has a shared representation of the information attended to by the
separate team members, they need to store it in memory for later use. Researchers have
noted that groups are often better than individuals at a number of memory tasks (e.g.,
Clark & Stephenson, 1989; Hartwick, Sheppard, & Davis, 1982; Stewart & Stasser, 1995;
Yarmey, 1992) because their storage capacity is much larger (Hinsz, 1990). When faced
with a complex task, team members do not have to remember all of the information that
the team needs because they are able to interact with one another. Each team member can
be given a set of information to store in memory. If the team needs the information, its
members can ask for it and an individual team member can retrieve it from memory. In
the basketball game, the team members do not all need to remember who goes to the
basket and who goes to the perimeter when the point guard drives. When someone yells
“drive,” each team member has their own set of information in memory that tells them
where to go on the ﬂoor.

By splitting up the storage duties, teams are better able to retrieve the information
(Stasser, Stewart, & Wittenbaum, 1995). For one thing, teams draw on multiple
memories, which helps them catch any retrieval errors made by one team member. In

addition, if one team member remembers one set of information, it can stimulate other

29

team members’ retrieval processes (e. g., Martel] & Borg, 1993; Meudell, Hitch, & Kirby,
1992). If one of the basketball players goes to the basket when he hears “drive,” it may
stimulate another player’s memory so they are both there when the opposing player
reaches the hoop. '

The culmination of the processes described above is the collective response of the
team members. When one team member yells “drive,” two team members move to the
basket, two team members converge on the ball, while the rest spread out around the
perimeter to defend against any outside shot. The team will succeed in defending its
basket as long as they have concentrated on sharing and disseminating information during
each phase of the information processing system.

Mum. Utilizing an information processing ﬁ'amework to describe how groups
and teams deal with cognitive and intellectual tasks has helped researchers understand the
importance of social processes (Fiske & Goodwin, 1994; Hastie & Pennington, 1991;
Ickes & Gonzales, 1994; Levine, Resnick, & Higgins, 1993; Neisser, 1982; Resnick,
Levine, & Teasley, 1991; Wegner, 1987). The information processing system maps out
the speciﬁc function of group interaction and communication at each phase of the
process, beginning with the team’s processing objectives and ending with the team's
response. As noted earlier, researchers have always known that team members need to
communicate with one another in order to accomplish their goals (e.g., Hackman &
Morris, 1975). However, an information processing framework provides a window into
the exact nature of those processes. It helps researchers understand when, why, and how
information is shared and combined within the team. Disrupting the ﬂow of

cormnunication could be potentially disastrous for the team. For example, going back to

30

the basketball team, it is critical that someone yells “drive” when the point guard on the
opposing team drives to the hoop. If one team member perceives an initial move to the
hoop by the point guard, but is distracted by the sight of a celebrity in the crowd, he may
delay his cry of “drive” for a couple of seconds. Those few seconds may allow the point
guard to get through the defense for an easy lay-up. If the game is close, success or
failure could depend on the timely communication of information within the team.
Clearly the ﬂow of information within the team needs to be unfettered if the team wishes
to reach its goal. By better understanding how interaction operates within teams,
researchers can more easily pinpoint possible obstacles to free and open communication.
This study examines one potential obstacle, the presence of acute stressors, which may
impair the team members’ ability to interact with one another while attempting to reach
their objective.

The Effects of Acute Strgsors on Information Processing within Teams.

Researchers have long suggested that stress may be related to interaction
processes within teams (Hackman & Morris, 1975). Although there has been some direct
support for this proposition in the organizational literature, cognitive researchers
interested in the scope of human attention have provided quite a bit of indirect support.
When individuals are immersed in stressful situations, their breadth of attention narrows
(Bacon, 1974; Baddeley, 1972; Bahrick, Fitts, & Rankin, 1952; Combs & Taylor, 1952;
Easterbrook, 1959; Pennebaker, Czajka, Cropanzano, & Richards, 1990; Wachtel, 1968).
For example, Bursill (1958) asked subjects to perform a primary pursuit tracking task
while at the same time detecting intermittent visual signals provided by a number of

lamps surrounding the tracking task. As he raised the temperature in the room, subjects’

31

experienced began to perform more poorly on the secondary detection task, indicating
that their attention span began to selectively deteriorate. Using a similar task
environment, researchers have been able to successfully replicate Bursill’s ﬁndings (e. g.,
Hockey, 1970a; Hamilton & Copeman, 1970).

Additional support for attentional narrowing in the presence of acute stressors has
been provided by a number of other researchers. Hockey and Hamilton (1970) presented
subjects with a series of words on eight separate slides in a noisy or a quiet environment.
After a short break, the subjects were asked to recall what the words were. They were
also required to indicate which corner of the slide the words appeared in: a task they were
not aware of from the start. Subjects in the noisy environment ignored the secondary task.
Hockey, Dornie, and Hamilton (1975) asked subjects to read through one of two
interleaved passages. Subjects’ recognition memory was then tested for words appearing
in both passages. Under noisy conditions, subjects focused their attention on, and were
better able to recall words from, the primary passage.

Although the results of Bursill and Hockey and his colleagues are intriguing, the
performance decrement on the secondary task could have been partially due to the
distracting nature of the environmental stressors that were manipulated. In an effort to
remedy this potential confound, Weltman, Smith, and Egstrom (1971) compared two
groups of divers on central and peripheral signal detection tasks. One group was told that
they would experience conditions similar to a 60-foot dive, even though the pressure was
kept constant across both groups. Divers who thought they were under extreme pressure
concentrated much less on the peripheral signal detection task, supporting previous

results.

32

The rather consistent ﬁndings suggest that, despite the fact that a few studies have
failed to ﬁnd similar effects (e.g., Hockey, 1970b), an individual’s scope of attention
narrows in stressful situations as they prefer to pay attention to sources of information
that are considered to be a priority (Hockey, 1979). Although individuals may
occasionally beneﬁt by ignoring secondary task information (Edland & Svenson, 1993),
it may be extremely detrimental in a team setting.

In teams, attention can be focused on the situation, the task, the self, or the group
(Hinsz et al., 1997). Researchers have suggested that, when a team member’s scope of
attention narrows in the presence of acute stressors, their level of interaction with other
team members may be signiﬁcantly reduced and they may become more self-focused and
less team-focused (Driskell & Johnston, 1998; Driskell, Salas, & Johnston, 1999). This
idea has been supported by Cohen (1978, 1980), who felt that individuals possess a
certain level of attentional capacity. When acute stressors are introduced, an individual is
forced to monitor potentially threatening stimuli, while at the same time inhibiting a
natural response to the stimuli. Therefore, an individual must set priorities for his or her
attention. The most common strategy is to focus attention on one’s own goals while
neglecting less important social cues. The threat rigidity thesis takes a similar
perspective, suggesting that individuals, groups, and organizations tend to behave rigidly
in threatening situations. In groups and teams, rigid behavior consists of narrowing ﬁelds
of attention, simplifying information codes, and reducing the number of information
channels used by each individual. This results in a system that is not sufﬁciently
diversiﬁed or ﬂexible, which can be maladaptive during radical environmental shifts

(Staw, Sandelands, & Dutton, 1981).

33

This tendency for team members to become more self-focused and less team-
focused when narrowing their attentional ﬁelds in stressful situations has been supported
by a number of researchers. Gladstein and Reilly (1985) found that group members
experiencing an external threat used fewer communication channels and exhrbited less
interaction than group members in a control group. Others have found that individuals
reduce their level of prosocial behaviors. For instance, individuals are less likely to
engage in altruistic behavior (Aderman, 197 2; Cunningham, Steinberg, & Grev, 1980;
Isen, Clark, & Schwartz, 1976; Isen & Levin, 1972; Rosenhan, Salovey, & Hargis, 1981)
and are less likely to help others (Mathews & Canon, 1975) when acute stressors are
present. These results may reﬂect on the research by Rotten, Olszewski, Charleton, and
Soler (1978), who found that individuals in stressful situations have difﬁculty
differentiating between people in different roles. As a result, “linkages between members
may become confused 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” (T orrence, 1954, p. 754).

Sm. The research described above indicates that acute stressors narrow the
scope of attention within teams, forcing team members to focus on tasks that are
considered to be a priority. Team members, as I have deﬁned teams, are interdependent
and need to divide their attention between completing their own tasks and coordinating
with other team members through interaction and communication. When team members’
attention narrows, they focus on their own tasks and neglect to interact with their
teammates. I propose that this decreases the ability of teams to process information. In a

collective information processing system, information needs to be processed within as

34

well as among the minds of the team members (Ickes & Gonzales, 1994). When acute
stressors are introduced, each team member is likely to focus on processing information
in his or her own mind, while neglecting to process the information between the minds of
his or her teammates. This weakens the interconnections between the team members and
negatively affects the ability of the team as a whole to develop processing objectives,
attend to important pieces of information, encode, store, and retrieve the information, and
come up with an appropriate response. As a result, the deﬁnitions of stress offered earlier
need to be extended to incorporate the collective information processing requirements of
teams. I suggest that, from a response-based perspective, stress occurs when, in the
presence of acute stressors, team members narrow their scope of attention and reduce
their level of interaction, thereby disrupting the team’s collective information processing
capabilities. The implications of this disruption can be quite widespread. In particular,
team processes, which are an inherent part of the team’s collective information
processing system, may suffer.
Team Processes

Team processes represent the mechanisms that determine whether team members
are able to successfully combine their capabilities and behavior into some sort of
functional output. If the team wishes to be effective, the processes working within the
team must be running smoothly (McGrath, 1964). A variety of team processes have been
identiﬁed within the organizational literature, including coordination, cooperation,
cohesion, climate, collective efﬁcacy, and team learning (see Kozlowski & Bell, 2001).
For the most part, team processes require that team members communicate and interact

with one another. However, this study focuses on two interaction-dependent team

35

processes that ﬁt particularly well into an information processing framework: shared
mental models and transactive memory.

Shared Mental Models. When individuals interact with their environment, a
psychological representation is created to explain the behavior of the world around them,
recognize and remember how things are related to one another, and predict future events
(Mathieu, Heffner, Goodwin, Salas, & Cannon-Bowers, 2000; Rouse & Morris, 1986).
These types of organized knowledge structures have been termed mental models in the
organizational literature (e.g., Mathieu et al., 2000). As a general deﬁnition, a mental
model is a “mechanism whereby humans generate descriptions of system purpose and
form, explanations of system functioning, and observed system states, and predictions of
future system states” (Rouse & Morris, 1986, p. 360). Mental models can help
individuals process information in a rapid and ﬂexible manner by providing them with a
heuristic that connects and classiﬁes different aspects of their environment (Cannon-
Bowers, Salas, Converse, 1993; Rumelhart & Ortany, 1977).

Team members also utilize mental models in order to remain adaptive in
environments that are constantly changing (Cannon-Bowers et al., 1993). However, in
addition to understanding their own environment, team members must also understand
how the characteristics, duties, and needs of their teammates ﬁt in (Prince, Chidester,
Bowers, & Cannon-Bowers, 1992). Researchers have suggested that, because they
interact within a common environment, team members develop shared mental models
(Cannon-Bowers, Salas, & Converse, 1990; Kleinman & Serfaty, 1989). According to
Oransanu and Salas (1993), shared mental models represent organized knowledge

structures that are mutually held by the team members. Team members do not have to

36

possess identical knowledge structures, but they should be compatible and they should
lead to common expectations. When confronted with a certain situation, most if not all of
the team members should think about and interpret it the same way (Cannon—Bowers et
al., 1993). Common expectations can help teams adapt to environmental shifts by
coordinating the behavior of each team member (e. g., Cannon-Bowers & Salas, 1990;
Cream, Eggemeier, & Klein, 1978; Gabarro, 1990).

Team mental models begin to develop during training, or at the “forming” stage
(Tuckman, 1965), where team members begin to learn declarative and procedural
knowledge regarding the task. As their knowledge base grows, team members may begin
to share different ideas about how to work together. Occasionally, team members’
perspectives may clash with one another (i.e., “storming”), but ultimately the team will
reach some level of understanding of the nature of the team, the task, and the rules
governing behavior (i.e., “norming”). However, the development process does not end
there. Although shared mental models become more detailed through experience
(McClure, 1990), they are constantly changing in reaction to the environment.

If team members’ shared mental models do not adapt accordingly, team decision
making and effectiveness will likely suffer (Cannon-Bowers & Salas, 1990; Cannon-
Bowers et al., 1993; Oransanu & Salas, 1993). This is likely a result of the contribution of
shared mental models to the collective information processing capabilities of teams.
Teams need some sort of strategy, heuristic, procedure, or integration technique to
organize and process information (Hinsz, 1995). In the team’s information processing
system, this relates to the encoding stage, where team members’ individual

representations need to be combined into a representation for the entire team. Team

37

members need to share their mental representations of the information they are presented
with if they want to avoid problems at a later point in time (Hinsz et al., 1997).

If teams utilize shared mental models to facilitate the collective encoding process,
any disruption to the inforrmtion processing system could have dire consequences. One
such disruption, the presence of acute stressors, creates a shift in emphasis away from the
team to the individual. Team members focus less on interacting with each other so that
they can complete their own task-related duties. Consequently, the interconnections
between the team members become weakened. This could negatively affect the encoding
process by slowing the development of shared mental models. Researchers have
suggested that shared mental models develop only if team members interact and
communicate with one another (Donnellon, Gray, & Bougon, 1986;1nnami, 1992). Once
team members enter the “norming” stage, communication and interaction may not be as
essential because team members’ mental representations have reached maturity.
Unfortunately, project teams, which form for short periods of time, still need
communication and interaction because the majority of their time is spent in the
“forming” and “storming” stages. When unusual task circumstances present themselves
(i.e., acute stressors), teams often rely on the shared mental models that are already in
place (Cannon-Bowers & Salas, 1997). This could be especially detrimental for project
teams, who still need to develop their mental representations and common behavioral
expectations. Therefore, I propose that, by reducing the amount of team-related
interaction, acute stressors prevent the team ﬁom fully developing shared mental models.

However, certain shared mental models may be affected more by stress than

others. A number of researchers have suggested that team members possess different

38

types of shared mental models (e.g., Cannon-Bowers et al., 1993; Mathieu et al., 2000;
Rentsch & Hall, 1994). Kilmoski and Muhammed (1994) note that “there can be (and
probably would be) multiple mental models co—existing among team members at a given
point in time. These would include models of task/technology, of response routines, of
team work, etc.” (p. 432). In an effort to bound the literature, Mathieu and his colleagues
(2000) developed a typology of mental models that included four broad categories. First,
team members can possess a mental model of the technology or equipment that they are
using. Second, they can hold shared job or task models that deal with procedures,
strategies, etc. that can be used to complete a task. Third, team members can collectively
understand the ﬂow of communication within the team. According to Mathieu and his
colleagues, “these models describe the roles and responsibilities of team members,
interaction patterns, information ﬂow and communication channels, role
interdependencies, and information sources” (p. 275). Fourth, team members can hold
shared team member mental models. These types of models include “information that is
speciﬁc to member’s teammates- their knowledge, skills, attitudes, preferences, strengths,
weaknesses, tendencies, and so forth” (Mathieu et al., 2000, p. 275). The last two mental
models, the interaction and team member models, are considered to focus on team-related
aspects of the situation, while the ﬁrst two models focus on task-related aspects of the
situation. The models that focus on team-related aspects of the situation develop
primarily through communication between the team members. I expect the development
of the latter two models to be particularly susceptible to the presence of acute stressors,
which are likely to focus attention away from the team and toward the task. Therefore, I

hypothesize the following:

39

Hypothesis 1. The presence of acute stressors will be negatively related to shared
interaction and shared team member mental models in temporary project teams.

Bangctive Memory. Another team process that may be affected by the presence
of acute stressors is transactive memory. Transactive memory has been utilized by
Wegner (e. g., 1987) to describe how couples in close relationships remember pieces of
information. When couples have been together for a long period of time, they begin to
learn things about each other’s memories. For instance, one partner may not be able to
ﬁnd the bath towels, but he or she knows that their signiﬁcant other has been in charge of
laundry for the last year. By communicating, the appropriate information can be
transmitted and the bath towels can be located. Each partner can therefore beneﬁt from
the collective memory of the pair by assuming responsibility for remembering certain
items and simply attending to the other person’s knowledge categories. When
information is needed, it can be retrieved from the couple’s transactive memory, which is
greater than the memory capacity of either partner.

Transactive memory has been deﬁned as a shared system for encoding, storing,
and retrieving information (Wegner, 1987; Wegner, Erber, & Raymond, 1991; Wegner,
Giuliano, & Hertel, 1985). Using the metaphor of a directory-sharing computer network,
Wegner (1995) proposed that transactive memory systems consist of three dimensions:
directory updating, information allocation, and retrieval coordination. As the directory
updating phase begins, individuals begin to learn what others know by becoming more
familiar with their domains of expertise. This can be accomplished through several
different methods. First, individuals can assign certain domains of knowledge to each

other. By agreeing to be responsible for a certain knowledge domain, an individual

40

becomes the repository for all relevant items. Second, individuals can update their
directories through perceptions of their own and their partner’s relative expertise in
different knowledge domains. This can occur through self-disclosure and shared
experience, which are part of any relationship formation process (e. g., Hollingshead,
1998). Third, directory updating can come from knowing who has access to different
pieces of information. If one partner has accessed information recently or has had to
access information for a longer period of time, it may be inferred that he or she has more
knowledge on the tOpic.

Once individuals become familiar with each other’s domains of expertise, they
can begin the second phase of information allocation, where new information is
communicated to the person who possesses the relevant area of expertise, thereby
facilitating the storage process. Normally, passing information from one person to
another can be dangerous, as information quickly degrades with each transaction (e.g.,
Bartlett, 1932). However, according to Wegner (1995), if individuals in close
relationships memorize every piece of information over a long period of time, their
memory will become scattered and their directory will expand to the point where things
become disorganized. In order to be more productive and reduce their cognitive load,
partners should pass information to each other as quickly as possrble without encoding
anything into memory.

Allocating information to the correct person can help retrieval coordination,
which is the third phase in transactive memory process. Often individuals are faced with
situations with which they are not familiar. If the individual is in a close relationship with

another person, there is more than one directory that can be accessed. Retrieval

41

coordination organizes the search process in order to maximize the speed and accuracy.
Using a “directory of directories,” an individual has a sense of who to look to ﬁrst for a
certain piece of information.

Researchers have suggested that the transactive memory system of encoding,
storing, and retrieving information can also be applied to groups and teams (e. g., Hinsz et
al., 1997; Moreland, 1999). Like couples, teams can split up different knowledge
domains and develop a collective memory that is greater than the memory capacity of the
team members by themselves. Hinsz and his colleagues (1997) note that “transactive
memory allows diﬂerent members of the group to process information, so they remember
the information that is directly related to their area of expertise” (p. 48). Through
directory updating, information allocation, and retrieval coordination, team members can
learn about each other, plan work more easily, and assign tasks to those who are most
qualiﬁed. In order for transactive memory to develop in teams, team members must
communicate with one another (Hollingshead, 1998). In fact, “communication serves
many beneﬁcial functions in the encoding and storage of new inforrmtion in transactive
memory systems” (Hollingshead, 1998b, p. 427). Team members use communication as
the medium for transferring knowledge between one another (Hinsz et al., 1997). Without
verbal interaction, team members would ﬁnd it much more difﬁcult to learn about their
teammates’ knowledge, expertise, and relevant experience (Hollingshead, 1998).
Communication also allows team members to delegate or assign responsibility for
learning and storing new information to certain individuals (Wegner, 1987).

Because communication is an essential component of transactive memory, the

presence of acute stressors could be particularly detrimental to directory updating,

42

information allocation, and retrieval coordination in teams. Acute stressors, by focusing
the attention of the team members away from the team and toward the task, could disrupt
the team’s shared system of encoding, storing, and retrieving information. If attention
narrows and interaction is reduced, team members will not be able to efﬁciently and
effectively update their directories, allocate information, or coordinate retrieval from their
collective memory. Project teams must continually repeat the initial stages of directory
updating, making them especially prone to the damaging effects of acute stressors, as
they are not given the time to become familiar with the details of everyone’s areas of
expertise. Faced with information they are not familiar with in the presence of acute
stressors, team members will not know who to allocate the information to and will not
bother to try. Instead, team members will keep the information themselves, creating
directories that are overﬂowing with useless information. By not communicating, project
team members will not be able to develop a “directory of directories,” which tells them
who has the infonnation they are looking for and retrieval coordination will suffer. Due
to the damaging effects of acute stressors on the information processing capabilities of
project teams, I hypothesize that:
Hypothesis 2. The presence of acute stressors will be negatively related to transactive
memory in temporary project teams.
Sammy

As hypothesized above, I expect the presence of acute stressors to negatively
affect shared mental models and transactive memory systems within teams. Because the
effects of acute stressors on shared mental models and transactive memory systems are

proposed to be similar, one may question whether the two constructs are sufficiently

43

different from one another to be considered distinct. Conceptually, researchers generally
agree that the two should be treated separately (e. g., Hinsz et al., 1997; Kozlowski &
Bell, 2001). Shared mental models represent an encoding system that is held by all the
team members. Transactive memory, on the other hand, represents a much larger portion
of the information processing system within teams, comprising the encoding, storage, and
retrieval stages of the process. Although shared mental models may help teams develop a
transactive memory system, there is not a one-to-one correspondence. Knowledge
regarding the interaction patterns or areas of expertise within the team will develop into a
transactive memory system only if team members are willing to engage in information
allocation and retrieval coordination. If team members do not put forth the effort, they
could share interaction and team member mental models without developing a collective
system of storage and retrieval.

Despite their conceptual differences, empirically I expect that shared mental
models will coincide with certain dimensions of transactive memory. In particular, shared
mental models may arise from communications representative of the directory updating
dimension of transactive memory systems. For example, if one team member asks
another whether they know how to access an inventory ﬁle within the organization’s
intranet, he or she is developing a mental model of his or her tearnmate’s area of
expertise and is updating his or her directory. However, once team members share mental
models regarding interaction patterns and expertise within the team, they may have
trouble properly utilizing that information. In other words, they may be reluctant to ask
their teammates for the information in their speciﬁc areas of expertise or they may

neglect to send information to the appropriate person. The presence of shared mental

models does not necessitate the presence of the information allocation or retrieval
coordination dimensions of transactive memory. Therefore, I hypothesize that:
Hypothesis 3. The directory updating dimension of transactive memory systems will be
positively related to shared interaction and shared team member mental models in
temporary project teams.

Personal and Situational Characteristics

The detrimental effects of acute stressors on shared mental models and the three
dimensions of transactive memory could be mitigated by a number of personal and
situational characteristics. As noted earlier, stress researchers have identiﬁed numerous
factors that moderate the effects of certain stressors on various outcomes, such as Type A
behavior pattern, hardiness, locus of control, autonomy, control, and social support (see
Cooper et al., 2001). Although these variables are important in the stress process, this
study focuses on personal and situational characteristics that have been identiﬁed within
the literature as potential inﬂuences on the team’s collective information processing
capabilities when faced with acute stressors. More speciﬁcally, regarding personal
characteristics, I examine the level of cognitive ability and extraversion among the team
members. Regarding situational characteristics, I am interested in the amount of shared
information and the level of feedback provided within the team.

Cognitive Ability. Cognitive ability refers to individual differences in information
processing capacity (e. g., Kanfer & Ackerman, 1989). The idea that there are limits on
attentional resources and immediate memory capacity is one of the basic tenets of
information processing theory and has been widely supported by cognitive psychologists

(e.g., Dempster, 1981; Mandler, 1975; Shiffrin, 1976). Researchers have long suggested

45

that individuals possess a certain attentional span, which determines the maximum
number of elements that they can attend to at one time (e. g., Baldwin, 1894). Working
memory, which is used to process incoming information and store the resulting products,
is led by a central executive, a ﬂexible workspace with severe restrictions on its capacity
to handle large amounts of information (e.g., Baddeley, 1986; Baddeley & Hitch, 1974).
The processing and storage of information must compete for the limited capacity in
working memory (e.g., Daneman & Carpenter, 1980; 1983).

Organizational researchers have agreed, suggesting that an employee’s attentional
resources are “an undifferentiated pool representing the limited capacity of the human
information-processing system” (Kanfer & Ackerman, 1989, p. 663). However, this
general cognitive principle is not anchored exclusively at the individual level. Consistent
with general systems theory, “the processing resources for any system are limited, and
when several processes compete for the same resources, eventually there will be a
deterioration in performance” (Norman & Bowbrow, 1975, p. 44). This could create
problems in teams, where team members must attend to the situation, the task, the self,
and the group (Hinsz et al., 1997 ). These competing task demands will dip into the team’s
collective pool of attentional resources. If team members decide to focus the majority of
their attention on the task, the pool will be too shallow to adequately deal with the rest of
their duties.

However, the depth of the team’s pool of attentional resources and memory
capacity can differ depending on the level of cognitive resources within the team In the
industrial and organizational psychology literatures, cognitive resources have been

described in terns of general cognitive ability or g (e. g., Kanfer, 1987; Kanfer &

46

Ackerman, 1989). Support for the existence of g comes from research indicating that a
single factor underlies performance on almost all tests that measure cognitive abilities

(e. g., Hunter, 1986; Jensen, 1986; Ree & Earles, 1991). The link between g and cognitive
resources has been supported by correlating scores on tests of memory span with scores
on tests measuring g (e. g., Jensen, 1970). Additional support comes from research
showing that there is a positive relationship between performance on tasks that require
active information processing and scores on tests that measure g (e. g., Hartigan &
Wigdor, 1989; Hunter & Hunter, 1984).

Teams that consist of employees who are high in general cognitive ability or g
should have a deep pool of cognitive resources that can be distributed among the
situation, the task, the self, and the group. Research has shown that task performance
increases when team members are high in cognitive ability (LePine, Hollenbeck, Ilgen, &
Hedlund, 1997; LePine, Colquitt, & Erez, 2000). But it is likely that the beneﬁts of
cognitive ability are not restricted solely to task performance. These teams should also be
able to devote more cognitive resources to processing information both within and
between the minds of the team members. Hollenbeck and his colleagues (1996) suggest
that high cognitive ability allows team members to develop effective systems of
interaction in order to effectively share and consider different pieces of information. By
being able to pay more attention to both task-related and team-related duties, team with
members high in cognitive ability should be able to collectively encode, store, and
retrieve information more efﬁciently and effectively.

Because the level of cognitive ability within teams affects collective information

processing, it also has implications for shared mental models. It addition to expanding

47

attentional resources and immediate memory capacity, cognitive ability has also been
shown to determine learning speed (e.g. Hunter, 1986; LePine et al., 1997). This can be
extremely helpful in the initial stages of shared mental model development, where team
members begin to learn declarative and procedural knowledge about the task. The more
quickly they learn the task, the more quickly they can begin to share different ideas about
how to work together, and the more quickly they can begin to develop a shared
representation of the task and the team As a result, I suggest that:

Hypothesis 4. The level of cognitive ability will be positively related to shared interaction
and shared team member mental models in temporary project teams.

Cognitive ability also has the potential to inﬂuence transactive memory within
teams. Increased task and team resources may help in all three phases of transactive
memory. In the directory updating phase, team members learn what others know by
becoming more familiar with their domains of expertise. Team members high in
cognitive ability should be better able to update their directories, helping them allocate
information to the person with the correct area of expertise. By having a more complex
and accurate “directory of directories,” team members high in cognitive ability will be
able to effectively coordinate retrieval when information in needed. The interaction
needed at each step will not present a problem to team members high in cognitive ability,
as they have plenty of resources to split among their different sub-tasks (LePine, 1998).
Therefore, I propose that:

Hypothesis 5. The level of cognitive ability will be positively related to transactive

memory in temporary project teams.

48

The positive effects of cognitive ability on shared mental models and transactive
memory within teams could be particularly important when acute stressors are present.
Researchers have shown that, when an unexpected change in the environment creates
additional information processing demands, the level of cognitive ability within the team
can determine whether the team will succeed or fail aePine et al., 2000). As noted
earlier, acute stressors increase information processing demands and narrow the attention
span of team members. When acute stressors are present, team members tend to focus on
their task-related duties and neglect their team-related duties. This shifts resources away
from the team and toward individual team members, reducing the level of interaction
within the team However, shifting resources will not become problematic until the level
of interaction falls below the level necessary for developing shared mental models and
engaging in directory updating, information allocation, and retrieval coordination.
Unfortunately, teams that are low in cognitive ability will likely be hovering at or near
the minimum level. Therefore,,any drop in resources would be particularly damaging to
their ability to collectively process information. Teams high in cognitive ability, on the
other hand, would likely have some amount of resources above the minimum level that
could be spared. A drop in resources would not affect them as much. As a result, I
hypothesize that:

Hypothesis 6. The relationship between the presence of acute stressors and shared mental
models and transactive memory in temporary project teams will be more negative for
teams that are low in cognitive ability.

Extraversion. Another personal characteristic that has the potential to impact the

relationship between acute stressors and team processes is extraversion, which is one of

49

the Big Five personality factors. For years, research has suffered from a lack of consensus
regarding how personality should be deﬁned and measured (e.g., Driskell, Hogan, &
Salas, 1987). Researchers have developed a number of taxonomies in order to organize
the wide array of personality variables identiﬁed in the literature. Some have suggested
that personality consists of 16 primary and 8 second-order factors (Cattell, 1946), while
others have proposed three primary factors (Eysenck, 1991), and still others have
examined six primary factors (Hogan, 1986). However, the ﬁve factor model has
received the most comprehensive and expansive empirical support (see Goldberg, 1993).
Researchers have replicated the ﬁve-factor structure across cultures and rating scales and
have found that the ﬁve factors are heritable and remain stable over time (Costa &
McCrae, 1992). Although the argument is far from over (e.g., Block, 1995), “the
development and validation of the ﬁve factor approach to personality offers a broad-
based, empirically manageable, and demonstrably relevant avenue for examining
personality in work organizations” (Barry & Stewart, 1997, p. 63). The ﬁve factors, or
the Big Five, consist of agreeableness, conscientiousness, neuroticism, openness to
experience, and extraversion. Agreeable individuals are ﬂexible, trusting, good-natured,
and cooperative. Conscientiousness is associated with responsibility, organization,
meticulousness, and dependability. Neurotic individuals are anxious, depressed, angry,
worried, and insecure. Openness to experience relates to being imaginative, curious, and
open-minded. Extraversion is associated with being sociable, active, assertive, talkative,
and gregarious (Barrick & Mount, 1991).

Out of the Big Five, conscientiousness has been found to be the most consistent

predictor of performance across a wide variety of jobs (Barrick & Mount, 1991; Mount &

50

Barrick, 1995). However, Barrick and Mount (1991) do not suggest that the other four
factors are irrelevant. Each of the ﬁve factors has the potential to explain variance in
employee behavior in certain situations. Extraverts, because they are enthusiastic,
outgoing, and ﬁiendly, should perform well when interpersonal skills and social
interactions are required (McCrae & Costa, 1989). Research has supported this
proposition, ﬁnding that employees high in extraversion tend to perform better in
occupations where interactions with others are a signiﬁcant part of the job (Barrick &
Mount, 1991; Mount, Barrick, & Stewart, 1998). For example, Barrick and his colleagues
(2001) examined the results of 15 prior meta—analytic studies correlating the Big Five and
job performance and concluded that extraversion is highly related to sales performance,
where there is a high degree of interaction between employees and customers.

Extraversion may also exert an inﬂuence in team settings. Team members high in
extraversion are much more likely to participate in group discussions (Littlepage,
Schmidt, Whisler, & Frost, 1995). They also feel conﬁdent that they can perform well in
a team environment and prefer working in teams over working alone (Thoms, Moore, &
Scott, 1996). Barry and Stewart (1997) note that extraverted team members “offer verbal
contributions from the outset and are presumably apt to communicate easily and ﬁeely
without fear of intimidation by their peers” (p. 66). This can create an atmosphere of
open communication within the team that can persist through task completion.

The beneﬁts of extraversion within teams have only been empirically examined in
a small number of studies. Barry and Stewart (1997), utilizing 63 work groups consisting
of graduate MBA students, found that group members high in extraversion were

perceived as having a greater impact on group performance though their increased

51

socioemotional and task inputs. According to Barry and Stewart, socioemotional inputs
refer to the group members’ ability to facilitate constructive intragroup relations.
Interestingly, the proportion of extraverted group members was curvilinearly related to
task focus and group performance. AS the number of group members who were high on
extraversion increased, task focus decreased and then plateaued. Group performance, on
the other hand, increased and then leveled off.

However, the groups examined by Barry and Stewart completed a number of
analytical exercises involving business case problems and team composition effects. To
perform well, the group members simply had to turn in a written solution. There was no
indication of the extent to which social interaction was required within the groups. In
fact, each group member could be independently assigned one of the exercises without
any input from the other group members. It is likely that teams, when performing an
interdependent task requiring a high degree of interaction, would beneﬁt from high levels
of extraversion. That is exactly what Barrick and his colleagues (1998) found when they
examined the behavior of 51 teams that assembled small appliances and electronic
equipment. As the number of extraverted team members increased, so did team viability
and team performnce. These results seem to suggest that, when interaction is necessary,
teams beneﬁt from high levels of extraversion.

Because collective information processing within teams is highly dependent on
team member interaction, the effects of extraversion on shared mental models and
transactive memory should be even more powerful. Shared mental models develop when
team members discuss different ways of working together and combine their individual

representations into a combined representation for the entire team By interacting with

52

one another, team members can better understand the ﬂow of communication within the
team and they can discover each other’s speciﬁc knowledges, skills, etc. Shared
interaction and shared team member mental models enable team members to collectively
encode incoming information more efﬁciently and effectively than if they were working
separately. However, these shared mental models will be slow to emerge if team
members are shy and prefer not to talk to each other. Any hesitation could lead to
problems for project teams, where team members are not given a lot of time to develop
shared mental models. In these situations, high levels of extraversion within the team
could help to speed up the process. Therefore, I hypothesize the following:

Hypothesis 7. The level of extraversion will be positively related to shared interaction and
shared team member mental models in temporary project teams.

Extraversion could also help with all three phases of transactive memory systems
in teams. When updating their directories, team members have to learn who knows what.
If team members feel reluctant to ask their teammates, or they neglect to divulge any
important information, they will not be able to get a collective picture of the distribution
of expertise within the team During information allocation, information is verbally sent
to the appropriate team member. If team members hesitate to communicate with each
other and begin to concentrate on other tasks, they may forget important pieces of
information and the transactive memory system will break down. Shy or non-talkative
team members will also be reluctant to ask their teammates for any information that falls
under their area of expertise, thereby handicapping retrieval coordination. Teams high in
extraversion should be better able to handle the collective encoding, storage, and retrieval

requirements of transactive memory systems. Consequently, I hypothesize that:

53

Hypothesis 8. The level of extraversion will be positively related to transactive memory
in temporary project teams.

Extraversion could also mitigate the effects of acute stressors on shared mental
models and transactive memory. Researchers have suggested that extraverted team
members prefer team-related duties over task-related duties, seeking out opportunities to
interact socially with their teammates (Barry & Stewart, 1997). As noted earlier, Barry
and Stewart found that teams composed of highly extraverted team members focused
signiﬁcantly less on task-related duties. This could be particularly important when teams
encounter stressful situations. When acute stressors are present, information processing
demands increase and team members tend to reduce their level of interaction, focusing
' more on task-related duties and less on team-related duties. However, extraverted team
members may be more resistant to any shift in resources away from the team. Because
they enjoy interacting with their teammates, team members high in extraversion will ﬁnd
it difﬁcult to eliminate that aspect of their job. Team members low on extraversion, on
the other hand, may see the presence of acute stressors as the excuse they need to reduce
their level of interaction with their teammates. Therefore, when choosing where to
allocate attention in stressful situations, high levels of extraversion may enable teams to
evidence less of a shift in resources away from team-related duties and toward task-
related duties. As a result, I hypothesize that:

Hypothesis 9. The relationship between the presence of acute stressors and shared mental
models and transactive memory in temporary project teams will be more negative for

teams that are low in extraversion.

54

Prior Shared Information. Aside from team member personality and cognitive
ability, there are a number of situational characteristics that can intrude upon the direct
relationship between acute stressors and team processes by affecting the team’s ability to
collectively process information. One such characteristic is the level of prior shared
information within the team Prior shared information refers to information that is
provided by the organization to more than one team member before the team begins its
task. Unique or unshared information, on the other hand, is given to only one team
member. When team members begin working on their task, they are unaware of which
pieces of information are contained in each of their teammates’ knowledge domains.
Stasser and his colleagues (e. g., Stasser & Stewart, 1992) have used a similar approach in
a number of their studies.

Although it has been suggested that unique information offers a number of
beneﬁts in team situations (Shiﬂett, 1979; Steiner, 1972), researchers have shown that
decision-making groups are signiﬁcantly more likely to discuss information that is shared
rather than unshared (Larson, Christensen, Abbott, & Franz, 1998; Larson, Foster-
Fishman, & Keys, 1994; Stasser, 1999; Stasser & Stewart, 1992; Stewart & Stasser,
1995; Wittenbaum 1998). For example, Stasser and his colleagues (1989), using three
and six person groups, distributed information regarding three candidates for student
body president. Several pieces of information were given to all the group members and
several pieces were only given to one group member. The three person groups mentioned
37% of the shared information and only 18% of the unshared information. The effects
were even more pronounced in the six person groups, where 58% of the shared

information got mentioned compared to only 18% of the unshared information. These

55

effects seem to occur in most group situations, even when group members are aware that
they each have speciﬁc domains of knowledge that contain a lot of unshared information
(Stewart & Stasser, 1995).

Even if unique information is presented, it is less likely that group members will
repeat it. For example, Larson and his colleagues (1994) had three—person groups study
written proﬁles of three hypothetical faculty candidates. Each proﬁle contained 18 items
of information, 33% of which were shared, and 67% of which were unshared, among all
the group members. When the group members brought up items of information, those
that were shared were signiﬁcantly more likely to be brought up again in the group’s
discussion. Other researchers have found similar results, supporting the idea that shared
information holds a repetition advantage over unshared information (Stasser, Taylor, &
Hanna, 1989).

Furthermore, when group members try to remember information, they have an
easier time recalling prior shared information mentioned during group discussion.
Stewart and Stasser (1995) found that, contrary to their hypotheses, assigning expertise to
group members failed to eliminate the sampling advantage for prior shared information,
especially during collective recall. The three person groups in the assigned expertise
conditions were able to remember 56% of the shared information and only 42% of the
unshared information.

Because prior Shared information is discussed more frequently, repeated more
often, and recalled more easily, the amount of prior shared information within groups has
been associated with their ability to collectively process information. Hinsz and his

colleagues (1997) note that “at the group level, information processing involves the

56

degree to which information, ideas, or cognitive processes are shared, and are being
shared, among the group members” (p. 43). They continue by suggesting that “the shared
and sharing aspects of group information processing are interdependent of each other” (p.
44). Therefore, in order for a transmission of information to occur, there must be at least
some level of prior shared information within the group. In project teams, the level of
prior shared information needed may be even higher, because when individuals have little
experience working together, they tend to focus even less on unshared items of
information (Wittenbaum, 1998).

If the collective processing capabilities of project teams rely on a certain amount
of prior shared information, excessive amounts of unshared information could disrupt the
team’s ability to construct shared mental models and a transactive memory system As
described earlier, shared mental models develop when team members hold mental
representations that provide them with common expectations about the task and their
behavior. Shared interaction models represent the network of communications within the
team including who needs to communicate with who and when. Shared team member
models represent the various knowledges, skills, and abilities of each team member. If
project teanm wish to construct representations regarding each team member and the
communication patterns that link them all the team members must frequently talk to one
another. However, talking is not enough. Team members also need to remember the
information that gets presented if their collective representations are to fully develop.
This process could be much more efﬁcient and effective if team members share items of

information with their teammates. Therefore, I hypothesize that:

57

Hypothesis 10. The level of prior shared information in project teams will be positively
related to shared interaction and shared team member mental models.

Transactive memory could also be positively affected by the presence of a high
degree of prior shared information. Directory updating requires team members to learn
each person’s area of expertise, even if it is unfamiliar. If team members hold information
that is unshared, the team will be less likely to get a complete picture of each team
member’s speciﬁc skills. Even if the unshared information is distrrbuted, team members
still need to remember it and with whom it was associated. Recall becomes much more
difﬁcult when team members are faced with unshared information. This could also create
problems for information allocation and retrieval coordination, both of which require
knowledge of who needs what and who knows what. As a result, I hypothesize the
following:

Hypothesis 11. The level of prior shared information will be positively related to
transactive memory in temporary project teams.

Although prior shared information may directly affect team processes, things
become more complicated in the presence of acute stressors. When team members
perceive that they are under pressure from acute stressors, they initially restrict their
interaction and choose to focus on themselves instead of the team as a whole, breaking
the collective information processing cycle and negatively affecting their ability to
develop shared mental models and transactive memory systems. If the remaining
interaction within the team is restricted to prior shared information, the problems could
become even worse, especially if the team members possess a lot of information that is

unshared. Although there are few studies examining the effects of prior shared

58

information on team in stressful situations, there is some evidence suggesting that
preferences for shared information become even more powerful when acute stressors are
present.

When groups initially begin communicating, they tend to rapidly deplete the pool
of shared information and neglect the unshared information. As the discussion
progresses, the pool of unshared information becomes larger and larger in comparison to
the pool of shared information. Eventually, there will be more opportunities to discuss
unshared rather than shared information (e. g., Larson, Foster-Fishman, & Franz, 1998;
Larson et al., 1994). However, when additional attentional demands such as time pressure
are placed on group members, they are more likely to continue to focus on shared
information (Stasser, 1999). This suggests that preferences for shared information are
magniﬁed in times of stress, which can affect how team members choose to allocate their
attention. More speciﬁcally, because they prefer to discuss shared information in the
presence of acute stressors, project team members could further shift their attention span
away from their team—related duties when provided with a large amount of unshared
information.

The potential buffering effect of prior shared information on the relationship
between acute stressors and team processes is also indirectly supported by research
examining social support. In the organizational stress literature, social support has been
inconsistently and vaguely deﬁned (Beehr, 1995). As noted by Vaux (1988), “people
assist each other in an astonishing variety of ways” (p. 17). One of the most popular
conceptualizations of social support has been offered by House (1981), who differentiates

between four kinds of support. Emotional support involves showing interest in, or

59

sympathy for, an individual’s diﬁculties. Appraisal support gives feedback about an
individual’s functioning that may help to bolster his or her self-esteem Informational
support supplies the individual with information that may help him or her deal with
problems. Instrumental support refers to providing help that is direct and of a practical
nature.

Social support can exert inﬂuence on the stress process in a number of ways.
First, there may be a main effect of social support, whereby strains are reduced
irrespective of the stressors that are encountered. Second, social support may mediate the
effects of stressors on psychological and physiological strains. Third, in what is referred
to as the stress-buffering model, social support may moderate the relationship between
stressors and strains. That is, when confronted with aversive events, individuals who
receive social support will be shielded or protected from any potentially harmful
consequences. Although support for each model varies, in this situation I am particularly
interested in research examining the stress-buffering hypothesis.

Because organizational psychologists believe that other people can exert an
inﬂuence on the stress process, there has been a “quick, widespread acceptance of the
buffering hypothesis in regard to social support and work-related stress” (Beehr, 1995, p.
182). However, the results ﬁ'om a number of studies conducted over the years have been
much more inconclusive (Cooper et al., 2001). Cohen and Wills (1985), in their review of
the literature, concluded that there was considerable support for the buffering effects of
social support in the occupational literature. Despite their positive conclusions, they only
managed to review three studies. In a more comprehensive review, Beehr (1985) found

that results are mixed, with some studies ﬁnding a buffering effect, others ﬁnding a

reverse-buffering effect, and some ﬁnding no buffering effects whatsoever. Since Beehr’s
review, there have been a number of additional studies examining the stress-buffering
effects of social support (see Cooper et al., 2001; Beehr, 1995). Although several studies
have found support for the buffering hypothesis (e. g., Iwata & Suzuki, 1997; Moyle &
Parkes, 1999), only a small set of the interactions they examined reached signiﬁcance.
For example, across six studies (Beehr, King, & King, 1990; Etzion, 1984; Ganster,
Fusilier, & Mays, 1986; Kaufmann & Beehr, 1986, 1989; Seers, McGee, Serey, & Graen,
1983), researchers found 26 signiﬁcant interactions across 153 tests (Beehr, 1995).

Although the buffering effects of social support have been fairly weak within the
organizational literature, the support provided by shared information within project teams
could be much more powerful. Researchers have noted that research on social support
suffers from a lack of speciﬁcity. In order for social support to be effective, the source of
support should be consistent with the nature of the outcome and the source of the stressor
(Beehr, 1985, 1995; Cooper et al., 2001; Ganster et al., 1986; Russell, Altmaier, & Van
Velzen, 1987). In addition, social support represents a meta-construct and should be
broken down into its components if researchers hope to ﬁnd any signiﬁcant interactions.
Shared information represents a speciﬁc form of social support that matches up with the
relationship between acute stressors and team processes examined in this study.
Speciﬁcally, shared information gives informational and instrumental support that can
help to maintain the collective information processing capabilities of project teams faced
with a combination of acute stressors. Team members who share information understand
what each other are experiencing, which could offer solace in stressful situations,

providing an easy way to maintain their team-related duties. Because the degree of prior

61

shared information within project teams may have a greater inﬂuence on team processes
when acute stressors are present, I hypothesize that:

Hypothesis 12. The relationship between the presence of acute stressors and shared
mental models and transactive memory in temporary project teams will be more negative
for teams that possess high levels of prior unshared information.

Feedback. Another situational characteristic that has the potential to affect the
information processing capabilities of teams in the presence of acute stressors is the
provision of performance feedback. Feedback can be generally deﬁned as the process of
communicating information regarding the results and outcomes of actions or behaviors
(Taylor, Fisher, & Ilgen, 1984). Because of its informational and motivational
components (Ilgen, Fisher, & Taylor, 1979), organizations have recognized feedback as
an integral tool that can be used to effectively manage human resources (Barley,
Northcraft, Lee, & Lituchy, 1990). Due to its popularity, a number of studies have been
conducted within the organizational literature examining the effects of feedback on the
performance of individual employees (e. g., Atwater, Waldman, Atwater, & Cartier, 2000;
Earley, Northcraft, Lee, & Lituchy, 1990). Although there has been some debate
regarding the effectiveness of feedback interventions (Kluger & DeNisi, 1996),
researchers have generally agreed that feedback represents an important component of an
individual’s information processing system (e.g., Carver & Scheier, 1981; Kanfer &
Ackerman, 1989; Naylor, Pritchard, & Ilgen, 1980).

Feedback has also been associated with the collective information processing
capabilities of groups and teams (Hinsz et al., 1997; Tinsdale, 1989). However, providing

feedback to teams of employees differs from providing feedback to individuals. Because

62

team members operate in an interactive, social setting, organizations have two feedback
options.’ Feedback can be given to one team member regarding his or her behavior or it
can be given to the entire team regarding their collective behavior (e. g., Conlon & Barr,
1989). Although empirical evidence is lacking, researchers have suggested that feedback
provided to one team member could be particularly damaging to the team’s information
processing system (Hinsz et al., 1997). When feedback is given to one team member, the
team may begin to focus less on team-related duties in favor of individual task-related
duties (Smither, 1998). This is supported by theories of vicarious reinforcement
(Bandura, 1977, 1986), where observers are aware that someone is getting punished or
rewarded for a speciﬁc set of behaviors. Positive or negative feedback can signal to the
observers what behaviors should be matched or avoided to produce particular outcomes.
If the correct behavior is unclear, the observer will develop hypotheses about the types of
responses that are required to obtain rewards and avoid punishment.

The effects of providing feedback to individual team members or the team as a
whole will be especially powerful if the feedback is negative. Taylor and his colleagues
(1984) propose that negative feedback forces individuals to look for more effective
alternative behaviors, while positive feedback is less likely to lead to behavioral change.
Kluger and DeNisi (1996) agree, citing evidence indicating that individuals reduce or
maintain their effort when receiving positive feedback and increase their effort when
receiving negative feedback.

Despite the paucity of empirical evidence, the literature suggests that providing
negative feedback to one team member could focus the attention of all the team members

on their own task-related duties, damaging the team’s collective information processing

63

capabilities. This could have repercussions for shared mental models and transactive
memory among the team members. Shared team-member and interaction models require
that team members focus not only on what they are doing, but also on what the their
teammates are doing, in order to work together efﬁciently and effectively. If team
members are successful in building these types of shared representations, it could help
them with the encoding stage of information processing (Hinsz et al., 1997). However, if
team members are focused on their own behavior, they will have problems encoding the
communication networks within the team and the skill domains of their teammates.
Giving one team member feedback about his or her behavior could be one factor that
contributes to the self-focused attention of all the team members. As a result, I
hypothesize that:

Hypothesis 13. Providing negative feedback to one team member, rather than the team as
a whole, will be negatively related to shared interaction and shared team member mental
models in temporary project teams.

In addition, it is likely that the attention team members pay to their own task
duties relates to transactive memory. If team members wish to update their directories,
they need to ﬁnd out what their team members know. When negative feedback is given to
one team member, the most important tasks become those that have the potential to avoid
punishment (i.e., each team member’s individual responsrbilities). This may make it more
difﬁcult to learn each team member’s area of expertise. Negative feedback given to one
team member could also have an inﬂuence on information allocation. Team members
will not know who to send important pieces of information to, and distributing

information will not become a priority. Retrieval coordination could also suffer due to

team members’ inability to remember who has what piece of information in their
knowledge domain. Because negative feedback has the potential to affect all three phases
of transactive memory, I hypothesize the following:

Hypothesis 14. Providing negative feedback to one team member, rather than the team as
a whole, will be negatively related to transactive memory in temporary project teams.

The beneﬁts of providing negative feedback to the team as a whole could also
help to insulate teams against the debilitating consequences of acute stressors by
stimulating proactive coping behaviors. Proactive coping “involves the accumulation of
resources and the acquisition of skills that are not designed to address any particular
stressor but to prepare in general, given the recognition that stressors do occur and that to
be forearmed is to be well prepar ” (Aspinwall & Taylor, 1997). By accumulating
resources, which act as moderators, team members can minimize the negative effects of a
stressful encounter (Hobfoll, 1988, 1989). When acute stressors are present, teams with
the correct resources will continue to interact and collectively process information,
thereby experiencing less damage to their shared mental models and their ability to
engage in directory updating, information allocation, and retrieval coordination.

One resource that the team could use when faced with acute stressors is feedback.
Perhaps the most important characteristic of feedback is its informational value, which
refers to the extent to which feedback provides meaningful information about the
“correctness, accuracy, or adequacy of the response” (Ilgen et al., 1979). Barley (1988)
suggests that speciﬁc feedback activates planning by directing an individual’s attention to
Speciﬁc domains. After receiving feedback, an individual focuses his or her attention on

job relevant behaviors and away ﬁ'om irrelevant or inappropriate ones (Earley,

65

N orthcraft, Lee, & Lituchy, 1990). Individuals tend to match their attention and behavior
to the dimensions of performance feedback that are provided (Ilgen & Moore, 1987). If
team members match their behavior to the feedback that is provided, then negative
feedback that is given to the team as a collective unit should serve to concentrate their
efforts on their team-related duties, which they should see as more job-relevant. When
acute stressors are introduced, the team should have the correct set of resources that they
need to remain resilient in the face of difficult circumstances. On the other hand, if
individual team members receive negative feedback, the team as whole could one on the
wrong set of behaviors. When the team enters a stressful situation, they will not have the
resources they need to proactively cope with the situation. This could further shift the
concentration of the team members away from their team-related duties and toward their
task-related duties. Because the negative effects of acute stressors on team process may
be exacerbated when feedback is given to only one team member, I hypothesize that:
Hypothesis 15. The relationship between the presence of acute stressors and shared
mental models and transactive memory in temporary project teams will be more negative
when negative feedback is provided to one team member rather than the team as a whole.
ﬁlm-

A number of characteristics have been examined as moderators of the stress
process at the individual level, such as Type A behavior pattern and hardiness. Often
these characteristics are included without much theoretical rationale. As a result, ﬁrm
conclusions regarding the exact nature of their effects have been lacking (Cooper et al.,
2001). In this study, I introduce several personal and situational characteristics that can

be directly linked to the collective information processing capabilities of temporary

66

project teams, including cognitive ability, extraversion, shared information, and feedback.
Theoretically, these characteristics have the potential to inﬂuence the deterioration of
shared mental models and transactive memory in stressful situations. The purpose of this
study is to empirically examine the issue. These results have important implications
because, as discussed in the next section, shared mental models and transactive memory
can be linked to team performance.

The Effects oﬁLared Mml Modelsgand Trmctive Memory on Team Performing

McGrath (1964), in his input-process-outcome model, suggested that team
processes exhibit a direct link with team outcomes. Although shared mental models and
transactive memory could have implications for a number of important team outcomes, I
am interested in investigating their effects on team performance.

Shared Ment_al Mode_ls. Empirically, researchers have, for the most part, only
been able to establish an indirect link between team mental models and performance
(e.g., Hammond, 1965; Kleinman & Serfaty, 1989). For example, Oser and his colleagues
(1990) found that team members who offered information to teammates before they
requested it performed better. Foushee and his colleagues (1986) discovered that teams
performed better under difﬁcult circumstances when team members possessed a number
of shared experiences. Researchers also found that team members worked together better
when they shared internal frames of reference (Mitchell, 1986) and schema
representations (Rentsch, Heffner, & Duffy, 1993).

To more directly test the relationship between team mental models and
performance, Mathieu and his colleagues (2000) ran 56 two-person teams through a ﬂight

simulator task. They found that team mental models had a positive inﬂuence on team

67

performance by increasing the communication between the team members. These results
were recently supported by Marks and her colleagues (2000), who found that leader
brieﬁngs and team interaction training led to shared mental models, which subsequently
led to increased team performance.

Despite a few detractors (e. g., Adelrnan, Zirk, Lehner, Moffett, & Hall, 1986;
Brehmer, 1972), the majority of evidence seems to indicate that team mental models
improve team performance. That does not mean that team mental models are always
beneﬁcial. For one thing, sharing knowledge structures can hurt the team, as exempliﬁed
by groupthink, whereby team members blindly follow the wrong path and end up in
disastrous circumstances (e. g., Janis, 1972). This is a direct result of underutilizing the
resources of the team However, groupthink is much more likely to occur when team
members’ mental models completely overlap with one another (Wellens, 1993). In this
study, teams work on a highly interdependent task (see the Methods section), which
should reduce extensive overlap between the team members’ mental representations
while still covering the task domain.

Team mental models can also be inaccurate representations of the team
performance environment, which could be extremely detrimental to team success (Hall,
Volpe, & Cannon-Bowers, 1992; Marks et al., 2000; Stout, 1994). If team members send
or request information from the wrong person, but believe they are correct, the team will
not be able to beneﬁt from shared interaction and team member mental models. The
problem will be exacerbated if team members continue to follow the wrong system for an

extended period of time. Therefore, I hypothesize the following:

68

Hypothesis 16. Accurate shared interaction and team member mental models will be
positively related to team performance in temporary project teams.

Transactive Memory. The level of directory updating, information allocation, and
retrieval coordination within project teams also has the potential to affect team
performance. Moreland (1999) notes that “the potential beneﬁts of transactive memory
for a work group’s performance are clear. When group members know more about each
other, they can plan their work more sensibly, assigning tasks to the people who will
perform them best” (p. 5). By anticipating each other’s behavior, team members may be
better able to coordinate the collective behavior of the team (e.g., Murnighan & Conlon,
1991). Problem solving may also improve, as team members assign problems to those
individuals who have the expertise to solve them (e. g., Moreland & Levine, 1992).

So does transactive memory actually translate into improved team performance?
Unfortunately much of the empirical evidence focuses on couples (e. g., Wegner et al.,
1991), which is not easily translated to the organizational environment. There is,
however, some indirect evidence in the group decision-making literature. Researchers
have shown that recognizing expertise, which is an essential component of transactive
memory (Moreland, 1999), can lead to better group decisions (e. g., Henry, 1993; Henry,
Strickland, Yorges, & Ladd, 1996; Littlepage, Schmidt, Whisler, & Frost, 1995;
Littlepage & Silbiger, 1992). Henry and her colleagues, in their research, provide subjects
with a number of trivia questions that they must solve as a group. When group members
are able to determine each person’s actual domain of expertise, shared representations
can develop within the group and more accurate solutions can result. Littlepage and his

colleagues, who are also interested in group problem solving, ask their subjects to ﬁgure

69

out how to negotiate several hypothetical situations, such as surviving in the desert
without enough supplies. Like Henry, Littlepage has found that group members’ beliefs
about relative expertise can have a signiﬁcant impact on their ability to successfully solve
the problem at hand.

Another line of research that may relate to the relationship between transactive
memory and team performance examines the effects of group member familiarity. It is
likely that directory updating, information allocation, and retrieval coordination begin to
be utilized as employees work together on the same set of tasks for an extended period of
time (Moreland, 1999). No one has studied this process directly, but there is a wealth of
research showing that groups tend to perform better when their members are more
familiar with one another (e.g., Argote, 1993; Goodman & Shah, 1992; Jehn & Shah,
1997; Murnighan & Conlon, 1991). Although there are a number of potential underlying
causes, familiarity may impact performance because it gives group members the chance
to recognize one another’s areas of expertise (Moreland, 1999).

As described above, research on group problem solving and group member
familiarity are only tangentially relevant. More direct evidence for the positive effects of
transactive memory on group performance has recently begun to appear in the
organizational literature. For example, Liang, Moreland, and Argote (1995) had small
groups of students assemble the AM portion of an AM-FM radio. Group members were
trained individually or as a group, and then were given a chance to practice assembling
the radio. After practicing for about 30 minutes, the testing session began, at which point
each group was asked to recall how to assemble the radio. Groups who were trained

together exhibited more memory differentiation, task coordination, and task credibility,

7O

which were considered to be three factors of transactive memory. Memory differentiation
referred to the tendency of group members to specialize in one aspect of the process.
Task coordination was assessed by determining how well the group members worked
together. Task credibility was deﬁned as the level of trust the group members placed in
each person’s knowledge domain. These three factors were then positively related to the
number of assembly errors made by the groups. These results have since been replicated
in a number of experiments by Moreland and his colleagues (Moreland, 1999; Moreland
& Myaskovsky, 2000), indicating that directory updating, information allocation, and
retrieval coordination play a role in group performance. Therefore, I hypothesize that:
Hypothesis 1 7. Transactive memory will be positively related to team performance in
temporary project teams.

The positive effects of shared mental models and transactive memory on team
performance seem to be supported both theoretically and empirically within the literature.
As noted in the introduction, however, the presence of acute stressors is no_t linked to
team performance in this study. That relationship has been the subject of much debate.
Some see it as negative, some see it as positive, and some see it as curvilinear. I do not
argue that, when individuals are placed in stressful situations, there can be a number of
positive consequences. In fact, in this study, I argue that team members will focus their
attention on their individual task-related duties within the team in the presence of acute
stressors. However, this study focuses on the negative effects by examining team
members’ social reactions to stressful situations. It is these reactions that have the
potential to affect the team’s level of performance. In other words, I believe the presence

of acute stressors indirectly inﬂuences team performance by disrupting accurate shared

71

mental models and transactive memory. Because the effects of acute stressors on team
performance are thought to be indirect, not direct, I hypothesize the following:
Hypothesis 18. Accurate shared mental models and transactive memory mediate the
relationship between the presence of acute stressors and team performance in temporary
project teams.

Simm-

The literature review that has encompassed the last seventy pages or so was
designed to develop a model of the effects of acute stress within temporary project teams.
The model includes components of a number of different perspectives of the stress
process. However, the focus of the model is on the responses of project team members to
the introduction of acute stressors. In particular, I proposed that, in the presence of acute
stressors, team members’ shift their attention away from the interactive duties necessary
for successful task completion. Based on information processing theory, this will result in
a deterioration of the team’s ability to fully develop shared mental models and utilize the
directory updating, information allocation, and retrieval coordination dimensions of
transactive memory systems. The negative effects of acute stressors on team processes
can then be compounded or attenuated by a number of personal and situational
characteristics that relate to the information processing capabilities of teams. In
particular, regarding personal characteristics, I introduced the level of cognitive ability
and extraversion within the team Regarding situational characteristics, I discussed the
effects of negative feedback given to the team versus an individual team member as well
as the amount of prior shared information provided to each team member. Finally, I

suggested that accurate shared mental models and transactive memory will affect team

72

performance. The next section of this dissertation describes a lab study designed to test

the hypotheses generated throughout the literature review.

73

METHODS
Participants

Participants included 396 students from an introductory management course at a
large Midwestern University who were arrayed into 99 four-person teams. Out of the 396
students, 223 (56.3%) were male and 342 (86.4%) were white with an average age of 21.
In exchange for their participation, each earned class credit and all were eligible for cash
prizes ($400 per team) based upon the team’s performance.

m

The basic design was a 2 X 2 X 2 completely crossed factorial design conducted
in a laboratory setting. Acute stressors (present versus absent), shared information (high
degree versus low degree), and negative feedback (individual versus team) were
manipulated. Cognitive ability and extraversion were not manipulated in this study and
were operationalized as continuous variables.

A laboratory setting was chosen for a number of reasons. It allows for the
observation of numerous teams performing the same task under the same type of
experimental conditions. It also allows for the manipulation acute stressors, shared
information, and feedback while being able to directly observe the interaction patterns
within the team, which would be extremely difﬁcult to do in a ﬁeld setting. I am
concerned more with the developing and testing a model of stress within teams based on
information processing theory than with task itself. Because there is no reason to think
that information processing theory would not be applicable in this context, this context
serves as a meaningful venue for testing our hypotheses. I am simply asking the “can it

happen” question which, according to Ilgen (1986), is exactly the type of question that

74

bears investigation in this type of a laboratory setting. Furthermore, researchers have
estimated that the correlation between the effect sizes obtained in the ﬁeld and those
obtained in the lab generally exceed .70 (Anderson, Lindsey, & Bushman, 1999). For this
and other reasons, as Cook and Campbell (1979) note, “a strong case can be made that
external validity is enhanced more by many heterogeneous small experiments than by one
large experiment employing random selection of subjects, tasks, and times” (p. 80).

Ia_s_lg

Participants engaged in a modiﬁed version of the Distributed Dynamic Decision-
making (DDD) Simulation (see Miller, Young, Kleinman, & Serfaty, 1998). The DDD is
a dynamic command and control simulation requiring team members to monitor activity
in a geographic region and defend it against invasion from unﬁiendly air or ground tracks
or targets that enter the region.

The speciﬁc variant of this task used in this research, MSU-DDD, was developed
for use in contexts where teams are comprised of anywhere from 2 to 5 members with
little or no military experience. In this version of the simulation, each participant has a
networked PC at his or her workstation, and uses a computer mouse to control various
military sub-platforms such as tanks, helicopters, jets and AWACS reconnaissance
planes. These sub-platforms are used in an effort to monitor and control a speciﬁc
geographic area represented in a 20 by 20 grid.

The MSU-DDD Grid. Figure 3 is a display of the geographic region, which is
partitioned into four quadrants of equal size. Each team member in a four-person team is
assigned responsibility for one of the four quadrants and operates from a workstation

labeled DM-l through DM-4 in the ﬁgure. Team members are referred to as “Decision

75

Makers,” thus the DMi notation, with DMl the southeast (SE) quadrant, DM2 in the
northwest (NW), DM3 in the southwest (SW), and DM4 in the northeast (NE) quadrant.
In the center of the screen is a 4 by4 square designated as the “higth restricted zone”
which is nested within a larger 12 by 12 square called the “restricted zone.” Outside the
restricted zone is a neutral space. Each team member in the conﬁguration illustrated in
Figure 3 is responsible for an equal portion of highly restricted, restricted and neutral
space.

The objective of the simulation is to identify any tracks that enter the space,
determine whether they are ﬁiendly or unfriendly, and, if unfriendly, keep them out of
the restricted zones. Each team starts with a set number of points, and loses points for each
unit of time (seconds) that an unﬁiendly vehicle resides in a restricted or highly restricted
zone. Teams also lose points whenever they disable a friendly track in any area or an
unfriendly track in neutral territory. Teane with the most points were awarded cash prizes.

gases and Vehicles. In tenns of monitoring the geographic space, each team
member's base (see the small black rectangles labeled DMl, DMZ, etc. in Figure 3) has the
sane radar capacity as every other team member. Speciﬁcally, each base has a _d_etept'§p
ing of roughly six grid units (denmrcated by tle circle surrounding each DM’s base, as
shown in Figure 3). The team member can detect tle presence or absence of any track
within the radius of his or fer base’s detection ring. To detect tracks outside of their bases’
detection rings, team nembers have two options. They can rely on their teammates to
monitor regions of the space outside their own quadrant or they can rely on the mobile

assets, or vehicles, located at their base to view different sections of the screen.

76

Each DM has control ofvarious types ofassets that can be launched from his or her
base, and then moved to different areas of the screen. These assets are semi-intelligent
agents that can automatically perform certain functions (e. g., return to base to reﬁreL etc.).
Each DM manages these semi-intelligent agents. Most of the MSU-DDD simulation is
played via tle assets or vehicles, and hence understanding the unique characteristics of each
asset is critical to appreciating the complex nature of this task.

There are four different types of assets used in MSU-DDD; (a) AWACS planes, (b)
tanks, (c) helicopters, and ((1) jets. Each of these assets varies in capability across ﬁve major
dimensions; (a) range of vision, (b) speed of movenent, (c) duration of operability, ((1)
weapons capacity, and (e) identiﬁcation capacity.

Capabilities are distributed among the assets so that each has both strengths and
weaknesses, as shown in Table 1. Range of vision refers to the width of the detection ring
around each vehicle (see Figure 3). The AWACS has the largest range of vision (radius of
4 grid units), followed by the jet, the helicopter and ﬁnally the tank (radius of 2 grid units).
In tenns of speed of movement, the jet moves the fastest (1 grid unit per second), followed
by the AWACS, the helicopter, and ﬁnally the tank (.1 grid units per second). While the
tank is limited in terns of speed and vision, it is the best asset in terms of duration of
operation. It can be away from tie base for 8 minutes without having to refuel. The
AWACS can operate away from the base for 6 minutes, followed by the helicopter at 4
minutes, and the jet at 2 minutes. Three of the vehicles, the tank, the helicopter, and the jet,
have weapons capacity, which is represented by the inner ring surrounding each vehicle.
The tank has a power of 5, followed by the helicopter with a power of 3, and ﬁnally the jet

with a power of 1. Tie AWACS has a power of zero, which neans that it cannot disable any

77

tracks on the screen. However, the AWACS plane is the only vehicle that has the ability to
identify tracks on the screen. The inner ring around tle AWACS plane represents its
identiﬁcation ring.

Tracks. Tracks enter the screen from the sides of the grid with a line (i.e., a

 

vector) attached to them indicating the direction they are proceeding through the space.
Initially, when tracks enter someone’s detection ring, they show up as unidentiﬁed, which
is represented by a small diamond with a question mark in the middle. For example, in
Figure 3, track number 201 in the northwest quadrant has not been identiﬁed. Once the
track enters an identiﬁcation ring, it can be identiﬁed. When tracks are identiﬁed, the
diamond turns into a box with a letter inside of it, as shown by track number 200 in
DM2’s restricted zone in Figure 3. Number 200 has been identiﬁed as a B target. In this
study, teams were faced with four different types of targets: A, B, C, and D. Each target
represented a ground target of power 1, 5, 3, and 0, respectively.

Each team faced a total of 40 targets in the ﬁrst 15-minute game and 100 targets
in the second 30-minute game. The 140 targets consisted of 35 of each of the four types
(A, B, C, and D), and each team member saw 35 targets enter their quadrant across both
games. During the second 30-minute game, a subset of tracks (four meg consisting of 8
tracks per wave = 32 tracks) originated near the comers of tie Northwest, Northeast,
Southwest, and Southeast quadrants and proceeded in a straight line diagonally toward the
opposite comer of the task screen. Each of the DM’s experienced one wave of traCks during
the second 30-minute gane. These waves consisted of two of each type of track (A, B, C,
and D). Of these eight tracks, ﬁve stopped once they reached the highly restricted zone of

the quadrant ﬁom which they entered and stayed tlere for ten minutes, or until they were

78

successfully engaged. In this way, each DM had primary responsrbility for locating,
identifying, and engaging if necessary, a disproportionately greater amount of tracks at some
point during each task.

Amns Taken Towards Tracks. Once a track moves inside the identiﬁcation ring
of an AWACS plane, it can be identiﬁed as either A, B, C, or D. If it is an A, B, or C
track, a team nember can engage the track by moving an asset near enough so that the
track is within the attack ring. If the asset has the correct level of power, the track can be
disabled (see Table 1). The A target can only be disabled with the tank, the B target can
only be disabled with the jet, and the C target can only be disabled with the helicopter.
When team members are able to quickly disable unﬁiendly tracks inside the restricted
zones, the team will avoid losing large amounts of points. However, to maximize their
score, team members also have to make sure that they are not disabling a ﬁiendly track or
disabling an unfriendly outside the restricted zones.

Team Member Specialties. During the two games, knowledge regarding the
targets and possession of the four different vehicles was split up in order to maximize the
interdependence within the team (see Table 2). At the beginning of the ﬁrst lS-minute
game, each team member knew the power level of one target. DM4 knew that target A
was a Ground target with a power of 1, DM3 knew that target C was a G3, DM2 knew
that target B was a GS, and DM1 knew that target D was a G0. Each team nember was _
also be responsible for one type of vehicle. DM2 had the tanks, DM3 had the helicopters,-

DM4 had the jets, and DM1 had the AWACS planes.

79

Measures

Cognitive Ability. Cognitive ability was measured by the Wonderlic Personnel
Test (Form IV), which has been shown to be reliable in the educational and psychological
literatures (e.g., Dodrill, 1983; McKelvie, 1989; Weeless & Serpento, 1982). Internal
consistency reliabilities generally range between .88 and .94 across all forms of the
Wonderlic (Wonderlic & Associates, 1983). Researchers have also found that the
Wonderlic neasures the verbal, quantitative, and spatial abilities that comprise g (Hunter,
1986; Jensen, 1977).

Although the Wonderlic is designed to measure g at the individual level, this
study intends to examine cognitive ability at the team level. To aggregate the four team
member cognitive ability scores, the attribute in question and the nature of the task need
to be taken into consideration (LePine, Hollenbeck, Ilgen, & Hedlund, 1997). Steiner
(1972) offers an aggregation guide based on the resource requirenents of disjunctive,
conjunctive, and additive tasks. Disjunctive tasks are structured such that there is
signiﬁcant overlap between member activities. This creates a situation where one team
member can perform the team’s task efﬁciently on his or her own. Conjunctive tasks, on
the hand, evidence little overlap. So if one team member is not performing up to par, the
rest of the team members cannot compensate. Additive tasks are an internediary between
conjunctive and disjunctive tasks. Out of Steiner’s three categories, the additive model
best represents the team task used in this study (see pp.75-79). The activities of the team
members only evidence partial overlap with one another and each team nember

contributes to task performance in proportion to his or her resource level. As a result, the

80

sum of the team nembers’ Wonderlic scores was used to represent cognitive ability at the
team level.

Extraversion. Extraversion was measured at the individual level using the NEO
Five-Factor Inventory (NEO FFI). On this questionnaire, 12 items are used to measure
each of the ﬁve dimensions of personality. Answers are given on a ﬁve-point Likert-type
scale depending on how much they agree or disagree with the statement. The reliability
and validity of this test has been demonstrated by a variety of researchers and internal
consistency values for extraversion have generally been shown to be well above .80
(Costa & McCrae, 1992). In this study, coefﬁcient alpha reached .82 for the extraversion
scale. Consistent with the rationale outlined above, extraversion was operationalized at
the team level using the sum of the four team members’ scores.

Transactive Memory. Transactive memory has been a difficult construct to
neasure, as evidenced by the lack of empirical work within the organizational literature.
Moreland and his colleagues (Liang et al., 1995; Moreland, 1999; Moreland, Argote, &
Krishnan, 1996; 1998; Moreland & Myakovsky, 2000) are among the ﬁrst researchers to
attempt to measure transactive memory in a group setting. Using a research paradigm
where participants attempt to assemble transistor radios as a group, Moreland and his
colleagues examine behaviors that are thought to reﬂect memory differentiation, task
coordination, and task credibility. Memory differentiation refers to the level of
specialization among the group members, task coordination refers to ability of the group
nembers to work together efﬁciently, and task credibility refers to the level of trust in

other group members’ radio building knowledge. Moreland and his colleagues watch

81

each group perform the task and then rate the levels of memory differentiation, task
coordination, and task credibility within the group on a 7-point scale.

The neasurement technique of Moreland and his colleagues was _r_1_o_t used in this
study. It is not clear how the memory differentiation, task coordination, and task
credibility relate to directory updating, information allocation, and retrieval coordination,
which are the three dimensions of transactive memory that have been identiﬁed by
Wegner (e.g., 1995). For example, trusting in your teammates’ knowledge regarding the
task does not mean that speciﬁc pieces of information will be retrieved from the person
with the requisite expertise. Part of the problem stems from the task used by Moreland
and his colleagues. Team members were not responsible for speciﬁc areas of the
assembly process, which is a necessary component of transactive memory (e.g., Wegner,
Erber, & Raymond, 1991). Therefore, it is unlikely that a system of encoding, storing,
and retrieving information needs to develop among the group members in order to
complete the task. Another problem lies in the fact that, although they videotaped each
group session, Moreland and his colleagues did not examine speciﬁc verbal
communications between the team members. Researchers have suggested that
communicative behaviors are the best way to assess the three dinensions of transactive
memory (e.g., Hollingshead, 1998).

In an attempt to remedy problems with previous measures of transactive memory,
this study focused on the verbal communications between highly specialized team
nembers that speciﬁcally related to directory updating, information allocation, and
retrieval coordination (see Appendix B). Researchers have shown that coding verbal

communication coding represents a valid measure of transactive memory (e. g.,

82

Hollinshead, 1998). Through directory updating, individuals learn what others know by
becoming more familiar with their domains of expertise (e.g., Wegner, 1995). This can
occur through self—disclosure and shared experience, whereby team members share their
own area of expertise or ask about their teammates areas of expertise. Hollingshead
(1998), when coding the verbal behavior of couples, considered both assertions about
one’s own expertise and questions about the other’s expertise. Therefore, in this study,
directory updating was measured by coding verbal communications where team members
shared expertise (SE) or requested expertise (RE). Sharing expertise was deﬁned as the
distribution of knowledge by one team member regarding his or her target or vehicle
specialty. For example, “I’m DM2 and I have the tanks.” Requesting expertise was
deﬁned as the solicitation of information regarding another team member’s target or
vehicle specialty. For instance, “Who knows what the D target is?” Sharing expertise and
requesting expertise were correlated .54 (p<.01) with one another. As a result, there was
both conceptual and empirical evidence supporting the creation of an index of directory
updating by combining sharing expertise and requesting expertise.

Information allocation, which is the second dimension of transactive memory,
was deﬁned as the distribution of information to the individual with the relevant area of
expertise (e. g., Wegner, 1995). In this study, information allocation (IA) was measured
by coding verbal communications where team members sent information to the person
with the correct target or vehicle specialty. For example, “DM3, I have several C targets
in my restricted zone.”

Retrieval coordination, the third dimension of transactive memory, was deﬁned as

a “directory of directories,” whereby individuals know exactly where to look when a

83

speciﬁc piece of information is needed (Wegner, 1995). In this study, retrieval
coordination (RC) was measured by coding verbal communications where team members
requested information that is known to be part of someone’s target or vehicle specialty.
For instance, “DM3, what is the C target again?” Hollingshead (1998) coded similar
communications demonstrating a cooperative effort to retrieve needed knowledge, which
she termed transactive information search.

Two experimenters were in charge of coding transactive nemory related
communication within the 99 teams. In order to ensure that the coding was accurate and
consistent, both experimenters participated in a two-hour training session. The session
included a review of the construct deﬁnitions for each dimension of transactive nemory
described above as well as the coding of two practice teams. After coding each of the
teams, the two experimenters received detailed performance feedback and discussed any
coding discrepancies in order to ensure that they understood how to correctly identify
transactive memory related communication. After the two-hour training session, both
experinenters coded 17 (17%) of the teams to assess the level of interrater agreement.
Although a number of indexes of interrater agreement have been utilized within the
literature (Zwick, 1988), Cohen’s (1960) K has been supported as a good index of
agreement when presence/absence coding schemes are used (e. g., Ellis, West, Deshon, &
Ryan, in press; Stevens & Kristof, 1995). In this study, K = .77 across all three
dinensions of transactive nemory (K = .76 for directory updating, K = .76 for
information allocation, K = .72 for retrieval coordination), which indicated that the two

experinenters evidenced acceptable levels of interrater agreement. As a result, the

84

remaining 82 teams were divided between the two experinenters (62 teams were coded
by the ﬁrst experinenter while 20 were coded by the second experinenter).

The frequency of communication related to directory updating, information
allocation, and retrieval coordination was calculated for each of the four team members.
The four team nembers’ scores were then pooled to give an index of the team’s level of
directory updating, information allocation, and retrieval coordination. Aggregation was
supported by calculating intraclass correlation statistics, which measure the consistency
of scores within the team by comparing the between and within-team variance (see
Bliese, 2000). Directory updating evidenced an ICC(l) of .41 and an ICC(2) of .74.
Information allocation evidenced an ICC(l) of .35 and an ICC(2) of .68. Retrieval
coordination evidenced an ICC(l) of .38 and an ICC(2) of .71. Researchers have
suggested that consistencies above .30 are high enough to justify aggregation to the team
level using the sum of the four team members’ scores (see Klein, Bliese, Kozlowski,
Dansereau, Gavin, Grifﬁn, Hofrnann, James, Yammarino, & Bligh, 2000).

Schwab (1980) deﬁned construct validity as “the correspondence between a
construct (conceptual deﬁnition of a variable) and the operational procedure to measure
or manipulate that construct” (p.6). By specifying what the measure should or should not
reﬂect, the research must convey information about the dimensionality of the construct.
Information must also be provided regarding the way the dimensions relate to the overall
construct in order to comnence with any sort of model testing (Law & Wong, 1999).
Misspecifying this relationship can result in errors when tests of signiﬁcance are being
interpreted. Wegner (1987; 1995) proposed that the construct of transactive memory

consists of three dimensions: directory updating, information allocation, and retrieval

85

coordination. Law, Wong, and Mobley (1998) describe several ways by which the three
dinensions could relate to the overall construct.

Iftransactive memory conforms to a “latent” model, transactive memory would
be the latent variable that partially causes directory updating, information allocation, and
retrieval coordination. Measures of the three dimensions would therefore contain some
variance reﬂecting transactive memory, some systematic variance not related to
transactive memory, and error variance (Lepine, Erez, & Johnson, 2002). Law et a1.
(1998) descrrbe cognitive ability as an example of a latent construct because an
underlying general intelligence, or g, is thought to inﬂuence measures of more speciﬁc
abilities.

Iftransactive memory conforne to an “aggregate” model, directory updating,
information allocation, and retrieval coordination would “cause” the transactive memory
construct. Transactive memory could be constructed by capturing and adding or
multiplying systematic variance from each dirrension (Lepine et al., 2002). Law et a1.
(1998) view the construct of job satisfaction as an aggregate construct because it can be
formed by summing scores on a variety of dinensions such as satisfaction with pay,
satisfaction with coworkers, etc.

This study considers transactive memory an aggregate construct for several
reasons. First, if transactive nemory were a latent construct, it could be deﬁned as the
commonality among its dimensions. Conﬁrmatory factor analysis could then be used to
estimate transactive memory from the variance—covariance matrix among directory
updating, information allocation, and retrieval coordination (Law et al., 1998). In this

study, a CPA was run using AMOS (version 4.01), with transactive memory as the latent

86

variable and directory updating, information allocation, and retrieval coordination as its
three indicators. The one-factor model evidenced a very poor ﬁt to the data (GFI = .83,
AGFI = .67, RMSEA = .31). Second, the three dinensions of transactive nemory seem
to be conceptually distinct. Engaging in behavior that relates to one dinension does not
necessarily mean that the team will engage in behavior that relates to the other two
dimensions. For example, a team could continually update their directories regarding
separate areas of expertise yet neglect to allocate information to the correct individual. Or
teane may allocate information to the correct individual but forget to update their
directories when team members gain expertise in a certain area. Empirically this is
supported by the insigniﬁcant correlation between directory updating and information
allocation in this study (r = .10, ns). Lepine and his colleagues (2002) suggest that this
provides evidence that the construct is aggregate in nature. Aggregate constructs can be
formed from dimensions that are completely unrelated to one another. Third, in order for
a construct conform to the latent model, it must exist “at a deeper and more embedded
level than its dimensions” (Law et al., 1998, p. 742). However, transactive nemory could
not exist within teams without directory updating, information allocation, and retrieval
coordination. For example, if team members do not becone more familiar with each
other’s domains of expertise, encoding, storing, and retrieving information would be an
individual rather than shared process. This suggests that transactive nemory represents a
multidimensional construct that exists at the same level as its dimensions, unlike
constructs such as general mental ability and personality. Conceptually, the structural
arrows seem point from directory updating, information allocation, and retrieval

coordination to transactive memory (this does not imply causation, but rather that the

87

dimensions are part of the deﬁnition of transactive nemory), which provides evidence
that the multidimensional construct is aggregate in nature (Law et al., 1998).

Aggregate constructs can be formed by algebraically combining the dimensions.
For example, overall job satisfaction has been calculated by simply summing its
dimensions (Locke, 1969). Other constructs, such as overall job characteristics, have
been calculated by utilizing a multiplicative nonlinear function (e. g., Hackman &
Oldham 1976). Law and his colleagues (1998) suggest that the dimensions of an
aggregate construct should be combined algebraically based on theory. Going back to
Wegner’s (1987; 1995) initial theory, directory updating, information allocation, and
retrieval coordination are all treated as equal contributors to transactive nemory. As a
result, studies investigating transactive nemory have weighted all three dimensions
equally (e.g., Liang et al., 1995; Moreland & Myaskovsky, 2000). Because there is no
theoretical evidence supporting the domination of one or more dimensions of transactive
memory, directory updating, information allocation, and retrieval coordination were
standardized and combined equally to form the aggregate transactive memory construct.
Furthennore, because the dimensions of transactive nemory are considered distinct,
analyses will focus on directory updating, information allocation, and retrieval
coordination along with the aggregate transactive memory variable.

Shared Ment_al Models. Although several techniques have been suggested,
researchers have focused on the use of concept mapping as a method of neasuring shared
mental models in teams (e. g., Marks et al., 2000). Concept maps provide participants

with a variety of prelabeled concepts, which are then placed in a prespeciﬁed hierarchical

88

structure. The overlap between team nembers’ concept maps indicates the level of shared
knowledge within the team

Using this technique offers a number of advantages. Concept maps graphically
link concepts together (Marks et al., 2000), they can be administered to participants quite
easily (Novak, 1990), and they have been used successfully within the knowledge
assessment literature (Ruiz-Primo & Shavelson, 1995). However, Marks and her
colleagues (2000) note that the typical use of concept maps within the literature presents
a number of disadvantages. First, team members’ concept maps are usually compared to
one expert model and, if there are discrepancies, the model is automatically discounted as
incorrect. Using concept maps in this way precludes the possibility that there are
numerous ways to successfully approach a task. Second, concept maps normally depict a
linear ordering of events and do not allow for the ordering of events across various
branches of the map.

To try to maximize the advantages and minimize the disadvantages of using
traditional concept maps, Marks and her colleagues (2000) introduced team-interaction
concept maps, a variant of which was utilized in this study. Team-interaction concept
maps allow for more than one correct response, computing a similarity score separately
from an accuracy score. Team members are also asked about what all their teammates are
doing at the sane time to complete the task. More speciﬁcally, team members are given a
map of a possible task scenario along with a number of concepts that represent different
aspects of the task domain. Each team member completes a map by placing concepts that

best represent the actions of each team nember on the map (see Appendix C). The four

89

rows represent the sequence of actions for each team nember. The columns represent
what the team members are doing at about the sane time.

To calculate the similarity between the four concept maps, I used the scoring
algorithm utilized by Marks and her colleagues (2000). One point was given when two
team nembers shared a linked set of concepts (A-B), three points were given when three
team members shared a linked set of concepts, and nine points were given when all four
team nembers shared a linked set of concepts. Therefore, similarity scores can range
from 0 (no similarity among any of the four team nember concept maps) to 36 (four
identical concept maps) for ﬁrst concept map pictured in Appendix C. For the second
concept map, two points were given when two team members shared three linked
concepts (A-B-C), six points were given when three team members shared three linked
concepts, and eighteen points were given when four team nembers shared three linked
concepts. Thus, similarity scores could have ranged from 0 to 72. The similarity scores
for each of the two concept maps were added together to form a total shared nental
model similarity score.

Accuracy calculations also followed the work of Marks and her colleagues
(2000). The accuracy of team nember mental models was assessed by two judges who
were SMEs in the DDD command and control simulation. Each team member’s concept
map was rated on a Likert-type scale of structural accuracy ranging from 1 (inaccurate) to
7 (highly accurate). Judges paid particular attention to (1) the critical DDD functions, (2)
the appropriate role assignments for each team member, and (3) a reasonable sequence of
actions for successful completion of the task. The two judges’ evaluations were highly

correlated for both the ﬁrst (I = .95) and the second (r = .94) mental models. As a result,

90

team accuracy scores were formed by averaging the two judges’ accuracy ratings for
individual mental models and then taking the sum of the four team members’ accuracy
scores.

Team Performance. The two main objectives of every DDD team are to (1)
engage enemy tracks as quickly as possible once they enter one of the forbidden zones
and (2) allow friendly tracks to roam ﬁ'eely across the screen. These two objectives are
reﬂected in the team’s offensive and defensive scores during the game. The team’s
offensive score starts at 1,000 points and goes up 5 points every time an enemy target is
cleared from one of the forbidden zones. If a team member clears an enemy target outside
the forbidden zones or clears a ﬁiendly target anywhere on the screen, the team’s
offensive score drops by 25 points. The team’s defensive score starts at 50,000 and
decreases 1 point for every second an enemy resides within the restricted zone. If an
enemy enters the highly restricted zone, the team’s defensive score begins to drop by 2
points per second. Team performance was measured by standardizing and combining
each team’s offensive and defensive scores.

Manipulations

Acute Stressors. Researchers have identiﬁed numerous aspects of the environment
that can act as stressors, including job scope (e.g., Xie & Johns, 1995), role conﬂict and
role ambiguity (e.g., Jackson & Schuler, 1985; O’Driscoll & Beehr, 1994; Schaubroeck,
Cotton, & Jennings, 1989), lack of control over one’s job (e.g., Averill, 1973; Friedland,
Keinan, & Regev, 1992; Spector, 1986), responsibility (e. g., Cooper & Kelly, 1984;
Martin & Wall, 1989; Sutherland & Cooper, 1986), temperature (e.g., Jewell, 1998;

Surry, 1968), workload (e. g., Cooper & Roden, 1985; Kushmir & Melamed, 1991;

91

Westman & Eden, 1992), noise (e.g., Poulton, 1978; Teichner, Arees, & Reilly, 1963),
and interpersonal conﬂict (e. g., Keenan & Newton, 1985).

The problem with many of the stressors identiﬁed in the literature is that they are
impractical for laboratory investigation. For example, manipulating the workload level
within the team would likely create stress and team nembers would likely interact less
with one another. However, it would be difﬁcult to determine whether any behavioral
effects were due to the stress or the increased task requirements. The level of interaction
within the team could simply be a function of more individual task-related duties.

Therefore, I chose to examine the combined effects of time pressure and threat, which

 

can be manipulated without altering the task. Researchers have shown that a combination
of stressors results in an exponentially greater response rather than an additively greater
one (Kahn & Byosiere, 1992).

Tine pressure, which is present, at least to sone degree, in much of the working
world (Kelly & McGrath, 1985), has been deﬁned as a “restriction in the time required to
perform a task” (Salas, Driskell, & Hughes, 1996, p. 32) or as “a ratio of time to perform
required tasks divided by the time available” (Orasanu & Backer, 1996, p. 100). It is
merely the parception that there is not enough time to complete a given amount of work
(Cooper et al., 2001). These perceptions have been shown to increase arousal and
psychological stress levels among employees (e.g., Driskell, Salas, & Johnston, 1999;
Gladstein & Reilly, 1985; Holsti, 1971; Keinan, Friedland, & Ben-Porath, 1987;
Oransanu & Backer, 1996; Salas, Driskell, & Hughes, 1996; Svenson & Maule, 1993).

There is also sone evidence that perceptions of tine pressure lead to a breakdown

in information processing. Individuals under tine pressure must deal with the additional

92

cognitive demands resulting from the requirement to process a large amount of
information in a limited amount of time (Wright, 1974), forcing them to cone up with a
strategy to reduce the information overload. Researchers have suggested that this can lead
to a more selective information processing system (Hinsz et al., 1997; Karau & Kelly,
1992; Payne, Bettman, & Johnson, 1988). A number of studies have shown that
employees under tine pressure focus on speciﬁc pieces of information and pay attention
to a smaller set of cues when working alone (e.g., Janis, 1982; Janis & Mann, 1977;
Rothstein, 1986; Svenson, Edland, & Karlsson, 1985; Wright, 1974) and when working
in a team setting (Driskell et al., 1999; Gladstein & Reilly, 1985). These results have
been linked to the ﬁnding that individuals narrow their attention under stress (e. g.,
Driskell et al., 1999; Gladstein & Reilly, 1985).

To manipulate time pressure, I extended the work of Driskell and his colleagues
(1999) who told subjects to “hurry up” and “work harder” at ﬁve—minute intervals during
the task. In this study, team nembers were informed that, because most teams fail to
attend to the clock, the experinenter will make sure to keep track of the time and will let
them know how much tine they have left at various intervals (see Appendix A). At the
10 minute-mark, the ﬁrst warning was given. Team members were told that “You now
have only 20 minutes left to work on the task, which is not a lot of time. In order to
perform well, you need to hurry up and work harder at keeping the forbidden zones free
from enemy targets.” Similar warnings were given at the 15, 20, 22, 24, 26, and 28-
minute marks, letting the team nembers know that they had 15, 10, 8, 6, 4, and 2 minutes

left respectively.

93

Threatening the group can only enhance the effects of the time pressure
manipulation. Threat has been broadly deﬁned “as an environmental event that has
impending negative or harmful consequences for the entity" (Staw, Sandelands, &
Dutton, 1981, p. 502). Researchers have shown that threats represent a signiﬁcant cause
of psychological and physiological stress and strain among employees (e. g., Cobb &
Kasl, 1977; Jick & Murray, 1982; McGrath, 1976). Threatening employees can also
affect the way that they process information. Researchers have shown that individuals
restrict the amount of information that they have to deal with by focusing on internal
hypotheses, prior expectations, and dominant cues and tend to behave according to well-
learned or dominant response patterns (Beehr, 1995; Smart & Vertinsky, 1977; Staw et
al., 1981).

One reason why such threats are experienced as stressful events is because they
create a high level of uncertainty among the employees. “If any one variable were to be
singled out as the predominant underlying source of occupational stress, it would be
uncertainty” (Sharit & Salvendy, 1982, p. 150). The signiﬁcance of uncertainty in
feelings of stress and strain has been derived from expectancy theory, which examines
the degree to which two organizational events are related to one another (Beehr &
Bhagat, 1985).

To manipulate threat in this study, teams in the stress conditions were told the
following at the 15-minute mark:

Each semester, we notice that there are a few teams that seem to lag far behind the others.
We feel that this is due to a lack of motivation among certain students. Everyone gets
credit for participating, so some students feel that they can slack off during their 3-hour
session in the lab. When one person doesn’t pull their weight in the team, the other team

nembers need to pick up the slack for them If one person isn’t taking care of his or her
duties, the team’s performance will suffer. Therefore, in order to make sure you give us

94

your full attention and eﬂ'ort for the next 30 minutes, we are going to put a little pressure
on you. Right now, there are almost 600 students enrolled in your MGT 315 course. We
assune that you would prefer not to be displayed in front of all your friends if you aren’t
pulling your weight in the lab today. Therefore, we are going to videotape your team’s
performance. If your team is one of the three lowest performers, Dr. Morgeson will show
the tape to the entire MGT 315 class the last week of the senester as an example of
ineffective team behavior. I don’t think any of you want everything you say picked apart
by Dr. Morgenson during class. So just remember that the canera will be on you for the
next 30 minutes.

The speciﬁc threat utilized in this study has received little attention within the
literature. However, there have been numerous studies docunenting the stressful effects
of performance monitoring (e.g., Aiello & Kolb, 1995; Aiello & Shao, 1993; Amick &
Smith, 1992). Stress results when individual workers feel that their behavior is under
observation, even if they are aware that their work is aggregated with others before being
monitored (Aiello et al., 1991). Researchers have suggested that performance monitoring
creates a stressful work environment by reducing the amount of control workers have
over their jobs (Smith et al., 1992) and by restricting the amount of social support
available within the workgroup (Amick & Smith, 1992). In this study, participants
believed that their performance was being monitored via videotape. As in the workplace,
participants were aware of the negative consequences that could result from non-
productive behavior. Instead of the possibility of losing their jobs, participants were
aware of the possibility of public humiliation.

Prior Shared Information. Stasser and his colleagues (e. g., Stasser, Taylor, &
Hanna, 1989; Stasser & Titus, 1985; 1987; Stewart & Stasser, 1995) have conducted a
number of studies focusing on the discussion of unshared and shared information within

decision-making groups. Generally, they follow the same basic research paradigm For

example, Stasser and Titus (1987) distributed 24 pieces of information regarding several

95

hypothetical candidates for a political position. In the 33% shared condition, each
candidate received the 8 pieces of shared information and 4 pieces of unshared
information. In the 66% shared condition, each candidate received 16 pieces of shared
information and 2 pieces of unshared information.

The manipulation of shared information in this study differed slightly from the
work of Stasser and his colleagues. Four pieces of inforrmtion regarding the wave attacks
were distributed among the four team members (see Appendix E). Across all conditions,
each team nember received 2 pieces of information. Collectively, each team possessed
all the informtion about the wave targets, but team nembers individually received only
part of the information.

The distribution of information within the team was based on “truth supported
wins” models, which suggest that it is enough for two team nembers to share access to
the sane set or information in order for the group to attend to and acquire the information
collectively (Hinsz et al., 1997). Researchers have shown that groups tend to be more
successful when group members have a partner to corroborate their decisions or behavior
(e.g., Laughlin, 1980; Laughlin & Ellis, 1986). Pairing up two teammates together has
been shown to beneﬁt a number of different outcomes, including team learning (Ellis,
Hollenbeck, Ilgen, Porter, West, & Moon, under review-a). Given these ﬁndings, creating
“role partners” (two individuals who share expertise and information processing
responsibilities) may allow for the optimal level of shared information within the team

To generate a high level of prior shared information, where 100% of the
information was shared, two sets of role partners were created within the team The ﬁrst

set of role partners (DM1 and DM2) received information regarding the ﬁrst two wave

96

attacks, while the second set of role partners (DM3 and DM4) received information
regarding the last two wave attacks. To generate a low level of prior shared information,
where 50% of the information was shared, only one set of role partners was created. DM1
and DM2 received the same information as in the high prior shared information
conditions regarding the ﬁrst two wave attacks. However, DM3 received information
regarding the ﬁrst and last wave attacks, while DM4 received information regarding the
second and third wave attacks. As a result, no role partner was available to corroborate
information regarding the third and fourth wave attacks in the low prior shared
information conditions.

Feedback. The feedback manipulation used in this study was modeled after Porter
(2001). Across all eight conditions, teams received three speciﬁc goals during training
(see Appendix A). First, the team nembers were informed that they should never allow
enemy tracks to remain in the restricted zone for more than 3 minutes. Second, the
experimenter advised team nembers to try and engage enemy tracks residing in the
highly forbidden zone within 1 nrinute after entry. Third, team nembers were told that
they should try not to shoot down more than 1 friendly track during the 45-minute task.

At the 15-minute mark, teams in the appropriate conditions received the
following feedback manipulation, where the experinenter provided negative individual
level or team level performance feedback regarding the extent to which the team member
met the goals discussed prior to the ﬁrst task (see Appendix A):

DM2, after watching your performance during the ﬁrst 15 minutes, it appears that you
have been having sonre difﬁculty with the goals we set earlier. In particular, you allowed
a number of tracks to remain in your restricted zone for more than 3 nrinutes and several

tracks remained in your highly restricted zone for well over 1 minute, which is much
longer than we discussed. In addition, you have already shot down more than 1 friendly

97

track in your quadrant. By not doing these things, your performance has probably hurt the
team’s score.

Although the individual level feedback could have potentially been given to any
of the team members, DM2 was randomly selected as the recipient for the duration of the
study. Negative feedback was always be given in the presence of all four team members.
Procedure.

Immediately after entering the laboratory and ﬁlling in a consent form,
participants were randomly assigned to a four-person DDD team The team was then
escorted into a room in which they worked on the team task. Once in the room, each
participant was randomly assigned to one of four computer stations (e.g., DMl, DM2,
DM3, or DM4). After being seated at their respective stations, participants completed the
Wonderlic Personnel Test and the NBC FFI.

For the next hour, the participants were trained on the declarative and procedural
knowledge necessary for successful task completion. Participants were ﬁrst trained on
various aspects of the task, including the location of the bases, the scoring, the functions
of the various rings, the different vehicles, and the tracks (see Appendix A). Participants
then played a 45-minute practice game, where they learned how to launch and move
vehicles, identify tracks, and attack targets. During the practice gane, team nembers
owned one tank, one AWACS plane, one helicopter, and one jet. They were also told that
the A, B, C, and D targets corresponded to ground targets with power 0, 1, 3, and 5,
respectively.

Imnediately following the training, the goals, which were described earlier, were
set for the team All participants, regardless of experimental condition, received the same

goals. Therefore, goals were constant across all teams and all participants and were not

98

confounded with the experinental manipulations. The goal setting intervention was
nerely a component of the feedback intervention, which provided information regarding
the extent to which team members met the performance standards set forth during the
goal setting intervention.

Team members were also told that they would each have their own specialties
during the game. Speciﬁcally, each team nember possessed only one vehicle during the
game and only knew the power of one target. Each team member was given a sheet that
illustrated their own speciﬁc role, which they were able to keep during both games (see
Appendix D). They were inforned that a few pieces of information may overlap with the
information given to other team nembers (depending on the condition they are in), but
most would only be known by one team member.

At the end of the ﬁrst lS-minute task, the trainer recorded the team’s scores. The
trainer then provided negative performance feedback to either DM2 or the team as a
whole, depending on condition. If the team was in one of the acute stressor conditions,
they were then told that their performance would be videotaped and they might be in
jeopardy of having the tape shown in front of the entire class as an example of ineffective
teamwork. They were also told that the experinenter would keep track of the time during
the second game and would warn the team at various intervals when time was beginning
to run short.

During both games, the experimenter remained in the room in order to record any
transactive-related communication within the team After the second 30—minute game,
team members ﬁrst completed the manipulation check for the acute stressor

manipulations, then the shared nental models measure, and then the manipulation check

99

for feedback manipulation. Once all of the team nembers completed the post-task
measures, they were debriefed and dismissed. The debrieﬁng ﬁrst informed participants
that their team perforned well enough that there was no chance that their performance
would be shown and critiqued in front of the whole class. They were also told that, if they
received feedback regarding their performance, it may or may not have been related to

their actual behavior.

lOO

RESULTS

Manipulation Checks.

Acute Stgsors. To examine the effectiveness of the acute stressor manipulation

 

introduced in this study, several scales were utilized. First, participants completed the
Pressure/Tension scale from the Intrinsic Motivation Inventory (see Appendix F), which
has been developed and utilized by Deci and Ryan and their colleagues (e. g., Ryan, 1982;
Ryan, Mims, & Koestner, 1983; Ryan, Koestner, & Deci, 1991; Deci, Eghrari, Patrick, &
Leone, 1994). The scale has been shown to be both reliable and valid (e. g., McAuley,
Duncan, & Tammen, 1987). Coefﬁcient alpha for this study was .85. Answers ranged
from 1 (not at all true) to 7 (very true). As expected, individuals in the acute stress
conditions felt more tense (M = 4.62, SD = 1.31) than individuals in the non-stress
conditions (M = 4.29, S_D = 1.28), t(394) = 2.53, p < .05. Second, participants completed
the Affective (Negative) Thought scale, as shown in Appendix F, which has evidenced
acceptable psychonetric properties across a number of empirical studies (e.g., Kanfer &
Ackerman, 1989; Kanfer, Ackernran, Murtha, Dugdale, & Nelson, 1994). In this study,
coefﬁcient alpha reached .77. Responses to the Affective (Negative) Thought scale were
on a Likert-type 5-point scale from strongly disagree to strongly agree. Individuals in the
acute stress conditions experienced more negative thoughts (M = 2.90, SD = .82) than
individuals in the non-stress conditions (M = 2.69, E = .81), t(394) = 2.50, p < .05.
Third, participants completed the 204tem STAI State Anxiety scale (Spielberger, 1983).
The scale has been shown to evidence acceptable levels of reliability and validity (see
Spielberger & Reheiser, 1995) and has been frequently utilized within the stress literature

(e.g., Spector, Dwyer, & Jex, 1988). Coefﬁcient alpha was .92 in this study. Higher levels

101

of state anxiety were reported in the acute stress conditions (M = 1.99, SD = .56) than in
the non-stress conditions (M = 1.79, M)_ = .49), t(394) = 3.78, p < .01. Fourth,
participants completed four questions designed to assess reactions to each of the speciﬁc
acute stressor manipulations. Regarding time pressure, participants were asked the extent
to which they “felt a lot of tine pressure during this task” and “felt stressed because there
was not enough time to complete the task.” Coefﬁcient alpha reached .87 for these two
items. Individuals in the acute stress conditions felt more time pressure (M = 4.38, S_D =
1.64) than individuals in the non-stress conditions (M = 3.58, S_D = 1.67), 1(394) = 4.80, p
< .01. Regarding threat, participants were asked the extent to which “the idea that other
students may be aware of my performance on this task was very stressful” and the extent
to which they “felt a lot of pressure to perform well on this task because there was a
chance that others could observe my behavior.” Coefficient alpha reached .88 for these
two items. Individuals in the threat conditions felt more stress (M = 3.60, S12 = 1.76) than
individuals in the non-threat conditions (M = 2.52, S_D = 1.38), t(394) = 6.75, p < .01.
Finally, participants indicated, on a scale from 1 (not at all stressed) to 7 (very stressed),
how stressed they felt while playing the DDD command and control simulation.
Participants in the stress conditions reported feeling more overall stress (M = 4.18, E =
1.49) than participants in the non-stress conditions (M = 3.54, S]; = 1.35), t(394) = 4.49,
p < .01. The manipulation checks all point to the sane conclusion: the acute stressor
manipulation was highly effective. Table 4 provides the neans, standard deviations, and
intercorrelations among the acute stressor manipulation checks. A number of team

members in the stress conditions also verbally supported the effectiveness of the acute

102

stressor manipulation with comments such as “I’m starting to sweat. This is stressing me
out!” “I want to punch the screen right now!” and (my favorite) “Dude, back off.”

Feedback. The following three items assessed the effectiveness of the feedback
manipulation used in this study: (1) “Over the course of the two games you just played,
did the experimenter provide any information regarding progress toward the goals that
were set during training?" (2) “If the experinenter did provide such information, was it
generally positive or negative in nature?” (3) “If the experimenter did provide such
information, who was it given to (i.e., DM2, DMl, DM3, DM4, or the team as a whole)?”
Regarding the ﬁrst question, 99.0% of the participants agreed that they received some
sort of feedback concerning the three goals that were set during the training session. The
majority of participants (93.7%) also agreed that the feedback they received was
generally negative in nature. In the individual negative feedback conditions, 87.5% of the
students thought that DM2 received the information. In the team negative feedback
conditions, 99.5% of the students recognized that the team as whole received the
information. Participants in the team negative feedback conditions indicated that the
feedback was directed at the team as a whole more ﬁ‘equently (M = 5.00, E = .00) than
participants in the individual negative feedback conditions (M = 2.32, SQ = .95), 1(394) =
39.30, p < .01. In sum, 88.4% of the participants responded correctly to all three
questions, suggesting that the feedback manipulation was successful.

One concern was that the individual negative feedback manipulation could have
put undue stress on DM2. As a result, DM2’s scores on the acute stressor manipulation
checks were compared to DMl, DM3, and DM4 across the individual negative feedback

conditions. DM2 did not report feeling more tense (M = 4.43, S_D = 1.32) than the rest of

103

the team (M = 4.34, SQ = 1.26), t(198) = .40, ns. DM2 also failed to report higher levels
of state anxiety (M = 1.97, $2 = .57) than his or her teammates (M = 1.81, SQ = .51),
t(198) = 1.92, ns. Furthermore, DM2 did not report feeling more time pressure (M = 3.88,
S1; = 1.76) than the rest of the team (M = 3.94, S_D = 1.65), t(198) = -.22, ns, or more
threat (M = 3.38, S_D = 1.75) than his or her teammates (M = 2.91, S_D = 1.58), t(198) = -
1.79, ns. DM2 differed only when his or her scores on the Affective (Negative) Thought
scale (M = 3.18, S_D = .88) were compared with the rest of the team (M = 2.64, SD = .78),
t(198) = 4.06, p < .01. These results suggest that, in general, DM2’s level of stress was
not inﬂuenced by the individual negative feedback manipulation.

Another concern was that participants’ perceptions of procedural justice would be
affected by the time pressure and threat manipulations, even though they were told about
the possibility that their performance would be recorded and they were ﬂee to leave at
any time. Therefore, for the ﬁrst ten teams, participants completed a four-item procedural
justice measure adapted from Greenberg (1993; 1994). Answers were given on a ﬁve-
point scale from strongly disagree to strongly agree. An example item was “the
procedures used in the lab today were fair and aboveboard.” Participants in the acute
stressor conditions did not report lower levels of procedural justice (M = 2.38, SD = .78)
than participants in the non-stress conditions (M = 2.41, E = .78), t(40) = .10, ns.

Prior Shared Information. According to the formulas developed by Stasser. and
his colleagues (e.g., Stasser, Taylor, & Hanna, 1989; Stasser & Titus, 1985; 1987)
regarding the discussion of shared vs. unshared information, information regarding the
four wave attacks should be discussed more frequently in the high prior shared

information conditions vs. the low prior shared information conditions. Teams with high

104

levels of prior shared information should also discuss the third and fourth wave attacks
more ﬁequently than teams with low levels of shared information. Results indicated that
the ﬁrst two waves were discussed among team members with high levels of prior shared

information (M = 5.73, SD = 3.39) about as frequently as team members with low levels

 

of prior shared information (M = 4.64, S_D = 2.43), 1(97) = 1.83, ns. The third and fourth
wave attacks, on the other hand, were discussed among team members with high levels of
prior shared information (M = 3.72, S_D = 2.61) signiﬁcantly more frequently than team
members with low levels of prior shared information (M = 2.16, S_D = 1.69), t(97) = 3.52,
p<.01. Overall, the wave attacks were discussed among team nembers with high levels of
prior Shared information (M = 9.45, S_D = 5.38) signiﬁcantly more frequently than team
nembers with low levels of prior shared information (M = 6.81, S_D = 3.70), t(97) = 2.84,
p<.01. This suggests that the prior shared information manipulation inﬂuenced the
discussion of shared vs. unshared information as expected.
Test; of Hypotheses

Means, standard deviations, and intercorrelations among all the variables included
in the hypothesis tests are included in Table 5. Two teams were eliminated from the
analyses due to changes in the task that occurred after their participation. The timing of
the targets was altered, making the task much more difﬁcult for subsequent teams. As
shown in Table 5, shared nental model similarity and shared nental model accuracy
were highly related to one another. Marks and her colleagues (2000) reported a similar
relationship between Shared mental model similarity and accuracy. Yet the two variables
were kept distinct. Their results indicated that the two variables exhibited an interactive

effect on team communication processes above and beyond the direct effects. In order to

105

investigate the interactive effects of nental model similarity and accuracy on team
performance in this study, the two variables were not combined together.

The hierarchical regression analyses designed to test the hypotheses proceeded in
several stages. First, the variables utilized in testing the hypotheses were centered.
Second, a series of regressions were run to examine the direct and interactive effects of
the ﬁve independent variables on shared mental model similarity, shared nental model
accuracy, directory updating, information allocation, retrieval coordination, and
transactive nemory. Each regression consisted of three steps. Stress was entered in the
ﬁrst step, followed by the situational or personal characteristic of interest in the second
step, followed by their interaction in the third step. Because the independent variables
were not signiﬁcantly related to one another, order of entry did not affect signiﬁcance
levels. It also allowed for an examination of the amount of variance in each dependent
variable that could be attributed to the acute stressor manipulation and each situational
and personal characteristic. Separate regressions were run for each situational and
personal characteristic because, with only 97 teams, degrees of freedom becones an issue
when all ﬁve independent variables and their interactions are in one equation. Third,
regressions were run to determine whether directory updating, information allocation, and
retrieval coordination affected shared nental model similarity and shared nental model
accuracy. Fourth, regressions were run to examine the effects of shared nental model
similarity, shared nental model accuracy, directory updating, information allocation,
retrieval coordination, and transactive nemory on team performance. Fifth, regressions
were run to test for shared mental model similarity, shared mental model accuracy,

directory updating, information allocation, retrieval coordination, and transactive

106

memory as mediators of the relationship between stress and team performance based on
the recommendations of Baron and Kenny (1986) and James and Brett (1984).

Hyppthesis 1. Hypothesis 1 proposed that the presence of acute stressors would be
negatively related to shared mental model development. As shown in Table 6, the
presence of acute stressors, entered as the ﬁrst hierarchical step, accounted for a
signiﬁcant 22% of the between—team variance in shared mental model similarity and a
signiﬁcant 28% of the between-team variance in shared mental model accuracy. The
direction of the relationship indicated that acute stressors negatively impacted similar (B
= -.47, p<.01) and accurate (B = -.53, p<.01) shared nental models, supporting
Hypothesis 1.

Hypothesis 2. Hypothesis 2 predicted that the presence of acute stressors would
be negatively related to transactive nemory. Table 7 shows that teams faced with acute
stressors evidenced signiﬁcantly lower levels of directory updating (B = -.32, p<.01),
information allocation (B = -.32, p<.01), and retrieval coordination (B = -.49, p<.01).
Entering the presence of acute stressors as the ﬁrst hierarchical step in the regression also
negatively affected the aggregate transactive nemory variable (B = -.52, p<.01). The
presence of acute stressors explained 10% of the between-team variance in directory
updating, 10% of the between-team variance in information allocation, 24% of the
variance in retrieval coordination, and 27% of the variance in transactive nemory. These
results offer strong support for Hypothesis 2.

Hymthesis 3. Table 8 addresses Hypothesis 3 by providing the results of
regressing shared nental model similarity and accuracy on directory updating,

information allocation, retrieval coordination, and transactive memory. Directory

107

' updating, information allocation, and retrieval coordination were entered in one
hierarchical step to assess their unique contrrbutions to shared mental models. Retrieval
coordination explained a signiﬁcant percentage of unique variance in shared mental
model similarity (B = .39, p<.01), while information allocation (B = .20, p<.05) and
retrieval coordination (B = .32, p<.01) explained a signiﬁcant portion of variance in
shared mental model accuracy. The aggregate transactive memory variable was highly
related to both shared mental model similarity (B = .48, p<.01) and shared mental model
accuracy (B = .46, p<.01), explaining 23% and 21% of the between-teams variance
respectively. These results do not support Hypothesis 3, which predicted that directory
updating would be the primary determinant of shared mental model developnent.

Hypothesis 4. The results for the effects of cognitive ability on shared mental
model similarity and shared nental model accuracy did not support Hypothesis 4, which
predicted that teams high in cognitive ability would be in a better position to develop
shared nental models. As shown in Table 6, entered as a second step, cognitive ability
failed to predict shared mental model similarity (B = .10, ns) and shared mental model
accuracy (B = -.01, ns). Cognitive ability explained almost no variance in shared mental
model similarity (AR2=.01) and shared nental model accuracy (AR2=.00) beyond the
effects of acute stressors.

Hymthesis 5. Hypothesis 5 predicted that cognitive ability would positively affect
transactive memory. Table 7 shows that cognitive ability, entered in the second step of
the regression, was unrelated to directory updating (B = -.07, ns), information allocation
(B = .15, ns), and retrieval coordination (B = .10, ns). Cognitive ability was also unrelated

to transactive nemory (B = .08, ns). Cognitive ability explained little variance in

108

directory updating (AR2=.01), information allocation (AR2=.02), retrieval coordination
(AR2=.01), and transactive memory (AR2=.01) beyond the effects of acute stressors.
Thus, Hypothesis 5 was not supported.

Hymthesis 6. Hypothesis 6 proposed that cognitive ability would moderate the
relationship between the presence of acute stressors and team processes. As shown in
Table 6, the interaction between stress and cognitive ability entered in the third
hierarchical step failed to relate to shared nental model similarity (B = .04, ns) and shared
nental model accuracy (B = .11, ns). Table 7 shows that the interaction between stress
and cognitive ability was not signiﬁcantly related to directory updating (B = .09, ns),
information allocation (B = -.05, ns), retrieval coordination (B = -.02, ns), or transactive
memory (B = .01 , ns). As the interaction failed to explain a signiﬁcant amount of variance
in any of the team processes above and beyond the direct effects, Hypothesis 6 was not
supported.

Hyp_othesis 7. Hypothesis 7 proposed that high levels of extraversion would
strengthen shared nental models in temporary project teams. The nean level of
extraversion within the teams, entered as the second hierarchical step as shown in Table
6, was unrelated to shared mental model similarity (B = .11, ns) and shared nental model
accuracy (B = .02, ns). Because extraversion explained an insigniﬁcant 1% of the
between-teams variance in shared nental model similarity and 0% of the between-teams
variance in shared nental model accuracy, Hypothesis 7 was not supported.

Hyp_othesis 8. Table 7 shows that extraversion, entered as the second step, failed
to predict directory updating (B = .08, ns), information allocation (B = .04, ns), and

retrieval coordination (B = .01, ns). Extraversion was also unrelated to the aggregate

109

transactive memory variable (B = .06, ns). In total, extraversion explained 1% of the
between-teams variance in directory updating, 1% of the between-teams variance in
information allocation, 0% of the between-teams variance in retrieval coordination, and
0% of the between-team variance in transactive nemory. Therefore, Hypothesis 8,
which predicted that extraverted teams would exfubit nrore directory updating,
information allocation, and retrieval coordination, was not supported.

Hymthesis 9. Hypothesis 9 proposed that extraversion would moderate the effects
of acute stressors on team processes. However, as shown in Table 6, the interaction
between extraversion and stress, entered as the third step in the regression, failed to
predict shared mental model similarity (B = -.03, ns) and shared mental model accuracy
(B = -.07, ns). The results were similar for directory updating (B = -.07, ns), information
allocation (B = -.03, ns), and retrieval coordination (B = .04, ns). The interaction term also
failed to predict transactive nemory (B = -.03, ns). Because the interaction between
extraversion and acute stressors did nOt explain a signiﬁcant percentage of variance in
any of the team process variables, Hypothesis 9 was not supported.

Hyppthesis 10. Hypothesis 10 proposed that the level of prior shared information
would positively relate to shared mental model similarity and shared nental model
accuracy. Table 6 shows that prior shared information, entered as the second hierarchical
step, was signiﬁcantly and positively related to both shared nental model similarity (B =
.21, p<.05) and shared nental accuracy (B = .21, p<.05). Despite the strong effects of
stress, the level of prior shared information explained a signiﬁcant 4% of the between-
tearns variance in shared nental model similarity and 5% of the between-teams variance

in shared nental model accuracy. Thus, Hypothesis 10 was supported.

llO

Hypqthesis 11. Hypothesis 11 examined the relationship between prior shared
information and transactive memory. As shown in Table 7, the amount of shared
information, entered as the second step, failed to predict directory updating (B = .13, ns),
information allocation (B = -.07, ns), and the aggregate transactive memory variable (B =
.10, ns). However, the level of prior shared information was positively related to the level
of retrieval coordination within the team (B = .15, p<. 10). Prior shared information
explained a signiﬁcant 2% of the variance in retrieval coordination over and above the
effects of stress, partially supporting Hypothesis 11.

Hyp_othesis 12. Hypothesis 12 predicted that the level of prior shared information
would help to buffer the effects of acute stressors on team processes. Prior shared
information had no inﬂuence on the relationship between acute stressors and shared
mental models, as shown in Table 6. The interaction between stress and shared
information, entered as the third hierarchical step, was unrelated to shared mental model
similarity (B = .08, ns) and shared mental model accuracy (B = .16, ns), explaining 0%
and 1% of the variance respectively. The results regarding directory updating,
information allocation, and retrieval coordination were slightly more promising. Table 7
shows that, although the interaction between stress and shared information fails to predict
directory updating (B = -.17, ns), information allocation (B = .08, ns), and transactive
nemory (B = -.20, ns), the interaction relates to the level of retrieval cordination within
the team (B = -.34, p<.05), explaining a signiﬁcant 4% of the variance. As Shown in
Figure 4, teams experiencing acute stressors exhibited less of a reduction in retrieval
coordination when there was a high level of prior shared information. As a result,

Hypothesis 12 was partially supported.

111

Hyp_othesis 13. Hypothesis 13 proposed that providing negative feedback to one
individual rather than the team as a whole would damage shared nental models. Table 6
indicates that the type of negative feedback provided to the team, entered as the second
hierarchical step, had little impact of shared nental model similarity (B = .07, ns) and
shared mental model accuracy (B = .03, ns). Feedback explained an insigniﬁcant 1% of
the between-teams variance in shared nental model similarity and 0% of the between-
teams variance in shared nental model accuracy. So Hypothesis 13 was not supported.

Hymthesis 14. Table 7 shows that feedback, entered in the second step, had little
impact on directory updating (B = .01, ns), information allocation (B = .02, ns), retrieval
coordination (B = .10, ns), and transactive nemory (B = .06, ns) beyond the effects of
acute stressors. Feedback explained an insigniﬁcant 1% of the variance in retrieval
coordination and 0% of the variance in directory updating, information allocation, and
transactive nemory. Therefore, Hypothesis 14, which proposed that providing negative
feedback to one team member would negatively impact transactive nemory, was not
supported.

Hypgthesis 15. Hypothesis 15 proposed that providing negative feedback to the
team as a whole would ameliorate the effects of acute stressors on team processes. As
shown in Table 6, the interaction between feedback and stress, entered as the third step in
the regression, was unrelated to shared mental model similarity (B = -.14, ns) and shared
nental model accuracy (B = -. 12, us). The interaction term explained an insigniﬁcant 1%
of the variance in shared nental model similarity and 0% of the variance in shared mental
model accuracy. Table 7 shows that the interaction between feedback and stress was also

unrelated to directory updating (B = .04, ns), retrieval coordination (B = -.24, ns), and

112

transactive nemory (B = -.23, ns). However, the interaction was related to the level of
information allocation within the team (B = -.29, p<.10), explaining a signiﬁcant 3% of
the variance. As shown in Figure 5, the negative impact of acute stressors on information
allocation was exacerbated by the provision of negative individual feedback. Therefore,
Hypothesis 15 was partially supported.

Hymthesis 16. Hypothesis 16 proposed that shared nental model accuracy would
explain a signiﬁcant portion of variance in team performance. As shown in Table 9,
shared mental model accuracy and shared nental model similarity, entered together in the
ﬁrst step of the regression, explained a signiﬁcant 23% of the variance in team
performance. However, shared nental model accuracy was signiﬁcantly related to team
performance (B = .40, p<.05), while shared nental model similarity was unrelated to team
performance (B = .08, ns). Because Shared mental model accuracy explained a signiﬁcant
portion of unique variance in team performance, Hypothesis 16 was supported.

Hymthesis 17. Hypothesis 17 predicted that transactive nemory would relate to
team performance. Table 9 indicates that, entered as the ﬁrst block in the regression,
directory updating, information allocation, and retrieval coordination explained a
signiﬁcant 16% of the variance in team performance. However, information allocation (B
= .18, p<. 10) and retrieval coordination (B = .26, p<.05) were positively related to team
performance, while directory updating was unrelated to team performance (B = . 10, ns).
Only information allocation and retrieval coordination explained a signiﬁcant portion of
unique variance in team performance. As an aggregate variable, transactive memory was
positively related to team performance (B = .39, p<.01), explaining 15% of the variance.

Consequently, Hypothesis 17 was partially supported.

113

Hymthesis 18. Hypothesis 18 proposed that shared mental models and transactive
memory would mediate the relationship between stress and team performance. In order to
test for mediation, it is necessary to demonstrate that (a) both the independent (stress) and
the mediating (shared mental models, directory updating, information allocation, retrieval
coordination) variables relate to the dependent variable (team performance), (b) the
independent variable relates to the mediating variables, and (c) the relationship between
the independent variable and the dependent variable becomes negligible or is reduced
signiﬁcantly when controlling for the mediating variables (Baron & Kenny, 1986; James
& Brett, 1984).

Regarding shared nental models, Table 10 shows that stress is signiﬁcantly
related to team performance (B = -.43, p<.01). Table 9 shows that shared mental model
similarity is not related to team performance (B = .08, ns), while shared mental model
accuracy is signiﬁcantly related to team performance (B = .40, p<.05). Therefore, stress
and shared mental model accuracy pass the ﬁrst nediation requirement. Stress is
signiﬁcantly related to shared mental model accuracy (B = -.53, p<.01), as shown in
Table 6, which passes the second nediation requirement. As shown in Table 10, the
amount of variance in team performance explained by stress drops signiﬁcantly ﬁ'om
18% to 4% when controlling for shared mental model accuracy, indicating that shared
mental model accuracy mediates the relationship between stress and team performance.

Regarding directory updating, information allocation, and retrieval coordination,
Table 9 shows that information allocation (B = . 18, p<.10) and retrieval coordination (B =
.26, p<.05) relate to team performance, while directory updating does not (B = . 10, ns).

As noted above, stress is signiﬁcantly related to team performance. Therefore, stress,

114

information allocation, and retrieval coordination pass the ﬁrst mediation requirenent.
Table 7 indicates that stress is related to both information allocation (B = -.32, p<.01) and
retrieval coordination (B = -.49, p<.01), which passes the second mediation requirement.
As shown in Table 10, the amount of variance in team performance explained by stress
drops signiﬁcantly from 18% to 7% when controlling for information allocation and
retrieval coordination, indicating that information allocation and retrieval coordination
mediate the relationship between stress and team performance.

The aggregate transactive nemory variable was also examined as a potential
nediator. Table 9 shows that transactive nemory signiﬁcantly relates to team
performance (B = .39, p<.01). Because stress also predicts team performance, the ﬁrst
mediation requirement is fulﬁlled. Stress also signiﬁcantly relates to transactive nemory
(B = -.52, p<.01), passing the second nediation requirenent (see Table 7). As shown in
Table 10, the amount of variance in team performance explained by stress drops
signiﬁcantly from 18% to 7% when controlling for transactive memory, indicating that
transactive memory nediates the relationship between stress and team performance. In
sum, these results offer partial support for Hypothesis 18.

Supplemental Analﬂs

Table 3 provides a summary of which hypotheses were supported and which were
. not. However, in this section, I go beyond speciﬁc hypotheses and examine some issues
that bear further investigation. In particular, I investigate an alternative operationalization
of cognitive ability and extraversion, shared nental models as mediators of the
relationship between shared information and team performance, the interactive effects of

shared mental nrodel similarity and accuracy on team performance, and the relationship

115

between team processes and team performance in gane 1. Finally, I double-check the
regression results regarding the three manipulations utilizing MANOVA.

Hypotheses 4—9, which examined the effects of cognitive ability and extraversion
on the relationship between stress and team processes, were all unsupported. Because the
task was considered to be additive in nature, the mean was used to represent the level of
cognitive ability and extraversion in the team However, it could be argued that the task is
more conjunctive than additive in nature. For example, only one team member was able
to identify targets on the screen. Therefore, Hypotheses 4-9 were re-exarnined using the
minimum score on the Wonderlic and the NBC within the team Only the results for
Hypothesis 4 changed. Cognitive ability, entered in the second step after stress, was
signiﬁcantly related to shared mental model similarity (B = .15, p<. 10), partially
supporting Hypothesis 4. However, the marginality of these results suggest that the
individual differences examined in this study had little inﬂuence on the relationship
between stress and team processes.

The results regarding Hypothesis 10 indicated that the level of prior shared
information was positively related to shared mental model similarity and shared nental
model accuracy. Although it was not speciﬁcally hypothesized, shared mental models
could mediate the relationship between the level of prior shared information and team
performance. The ﬁrst requirenent for nediation is that both the independent (prior
shared information) and the nediating (shared mental models) variables relate to the
dependent variable (team performance). Regressing team performance on the level of

prior shared information, it was found that prior shared information explained an

116

insigniﬁcant 1% of the variance in team performance. Therefore, prior shared
information failed to pass the requirements for nediation.

Testing for Hypothesis 16, it was deternrined that shared nental model accuracy,
not shared mental model similarity, explained a signiﬁcant amount of variance in team
performance. Although the hypothesis was conﬁrmed, another possible relationship could
emerge. Marks et a1. (2000) found that shared mental model accuracy and shared mental
model similarity interacted in their prediction of team outcomes. In order to test this
relationship, team performance was regressed on shared nental model similarity and
shared mental model accuracy in the ﬁrst step and their interaction in the second step.
The results indicated that the interaction explained a signiﬁcant 4% of the variance in
team performance (B = .25, p<.05). As shown in Figure 6, shared mental model similarity
reduced team perforrmnce when shared nental model accuracy was low. Shared nental
model similarity had little effect on team performance when shared mental model
accuracy was high.

According to Hypothesis 17, directory updating, information allocation, and
retrieval coordination were expected to relate to team performance. Only information
allocation and retrieval coordination explained a signiﬁcant amount of unique variance in
team performance. However, Wegner (1987; 1995) suggested that, when groups initially
fornr, directory updating is most important. The effects should only be enhanced for tasks
similar to the one used in this study, where team members’ areas of expertise remain
static. Team members did not have to concentrate on constantly updating their
directories. Once their directories were initially updated, team members could

concentrate on information allocation and retrieval coordination. As shown in Table 11,

117

during the initial 15-minute task, directory updating was highly related to team
performance (B = .32, p<.01), while information allocation (B = -.08, ns) and retrieval
coordination (B = .02, ns) were not. Table 12 shows that the level of directory updating
during the ﬁrst 15 minutes also contributed to shared nental model similarity (B = .18,
p<.10) and shared nental model accuracy (B = .20, p<.05). This provides further
evidence that, while directory updating, information allocation, and retrieval coordination
all contribute to transactive nemory, they should be viewed as distinct dinensions and
transactive nemory should be viewed as an aggregate construct.

Because this study examined the direct and interactive effects of continuous
individual difference variables on team processes, multiple regression was the
appropriate analytic strategy. However, the categorical nature of the manipulations would
suggest the use of multivariate analysis of variance (MANOVA). Therefore, to double-
check the results, a MANOVA was run examining the effects of the three manipulations
on the team process variables. The results revealed a signiﬁcant effect of the acute
stressor manipulation on shared mental model similarity (F(5, 85) = 26.99, p<.01), shared
mental model accuracy (F(5, 85) = 37.71, p<.01), directory updating (F(5, 85) = 10.64,
p<.01), information allocation (F(5, 85) = 10.40, p<.01), retrieval coordination (F(5, 85)
= 32.65, p<.01), and transactive nemory (F(5, 85) = 36.05, p<.01). The prior shared
information manipulation evidenced direct effects on shared nental model similarity
(F(5, 85) = 5.03, p<.05), shared nental model accuracy (F(5, 85) = 6.15, p<.05), and
retrieval coordination (F(5, 85) = 2.68, p<.10). Regarding interactions, there was a
signiﬁcant interaction between the feedback and acute stressor manipulations on

information allocation (F(5, 85) = 3.12, p<. 10). There was also a signiﬁcant interaction

118

between the prior shared information manipulation and the acute stressor manipulation on
retrieval coordination (F(5, 85) = 4.81, p<.05). The results of the MANOVA analyses
mirror the results of the regression analyses, offering additional support for the

hypotheses tests descrrbed in the previous sections.

119

DISCUSSION

The purpose of this study was to develop and test a model of stress in team-based
work structures. Based on information processing theory, stress was proposed to
negatively impact team processes leading to decrements in team performance. The results
supported this hypothesis, indicating that stress was negatively related to both transactive
nemory and shared mental models. Transactive memory and shared nental models were
then shown to nediate the relationship between stress and team performance. A number
of personal and situational characteristics were also proposed to affect the relationship
between stress and team processes. Results indicated that cognitive ability and
extraversion had little impact on team processes and failed to ameliorate the effects of
stress on transactive nemory and shared mental models. Results regarding the two
situational characteristics, feedback and prior shared information, were mixed. The level
of prior shared information was positively related to the level of retrieval coordination
within the team, as well as shared mental model similarity and accuracy. Prior shared
information also buffered the negative effects of stress on the level of retrieval
coordination within teams. Providing negative feedback to the team as a whole rather
than to an individual team nember did not directly affect team processes, but it helped to
moderate the negative effects of stress on the level of information allocation within
teams. These results have a number of theoretical and practical implications.
Theoretical Implications

Cognitive psychologists generally accept the proposition that an individual’s
attention narrows under stress, focusing on sources of information that are considered a

priority (e.g., Bacon, 1974; Easterbrook, 1959). In teams, attention can be focused on the

120

task, the self, or the group (Hinsz et al., 1997). Researchers have suggested that, under
stress, team nembers tend to becone more self-focused and less team-focused (e. g.,
Driskell et a1, 1999). A number of studies have supported this assertion by showing that
interaction is reduced in stressful situations (e. g., Gladstein & Reilly, 1985). In this study,
results indicated that team nembers focused less on secondary task duties and engaged in
less interaction under stress, supporting past research.

However, this study went a step further by also proposing that, by reducing
interaction and becoming more team—focused, stressed teams have trouble collectively
processing information. Information processing systems within teams require team
members to process information between as well as within the minds of the team
members (Ickes & Gonzales, 1994). Team members need to develop a common frane of
reference, attend to the correct set of information, combine pieces on information into a
shared representation for the entire team, store a section of the shared representation in
nemory, and retrieve it from nemory when needed (Hinsz et al., 1997). Shared nental
models and transactive memory each represent portions of the team’s collective
information processing system The results of this study showed that shared nental
models and transactive nemory were both negatively affected under stress. Shared
nental models became less similar and less accurate, and team nembers engaged in less
directory updating, information allocation, and retrieval coordination. Stressed team
members had trouble developing a shared representation of the roles and responsrbilities
of their teammates and failed to utilize a transactive nemory system to distribute the
information within the team Team members found it difﬁcult to split up the team’s

information processing duties. When stressful environnental conditions presented

121

themselves, teams clearly failed to process information between the minds of the team

 

members, supporting the idea that stress undermines the team’s collective information
processing system

It is possible that the negative effects of stress on transactive memory and shared
nental models could have been due to decreased willingness to communicate within the
team That is, under stress, teams sirrrply evidence less of a willingness to speak to one
another. This could have been of particular concern in this study, where participants
agreed to be videotaped during the task. However, the data indicate that teams under
stress exhibited similar amounts of overall communication (M = 128.95, S1; = 47.18)
than teams in the non-stress conditions (M = 140.80, $2 = 53.13), t(95) = 1.16, ns. These
results indicate that team members still talked under stress, but they talked about the
wrong things. They failed to utilize speciﬁc conrrnunications that would beneﬁt the
development of transactive nemory systems and shared mental models. This also
provides additional evidence for the validity of the transactive nemory measure utilized
in this study.

Although the results of this study supported the hypothesized disruption in teams’
information processing systems, a number of personal and situational characteristics were
proposed to inﬂuence the relationship between stress and team processes. The results
regarding cognitive ability and extraversion, the two personal characteristics examined in
. this study, were disappointing. Neither individual difference variable had any effect on
transactive nemory or shared nental models. Individual differences also failed to
moderate the negative effects of stress on the two team processes. This does not support

the idea that teams high in cognitive ability are able to develop effective systems of

122

interaction in order to effectively share and consider different pieces of information
(Hollenbeck et al., 1996). Results suggested that teams high in cognitive ability were not
able to pay nrore attention to both task-related and team-related duties, thereby improving
the efficiency and effectiveness of their encoding, storage, and retrieval systems in
stressful situations. Results also failed to support the notion that teams high in
extraversion tend to develop more effective communication networks when the task
requires a high level of interaction (Banick et al., 1998). Under stress, teams high in
extraversion shifted the same amount of attention away from the interactive duties of the
task.

A possible explanation for the lackluster ﬁndings regarding cognitive ability and
extraversion deals with the complexity of the task experienced by participants. Mischel
(1977) proposed that situations can be characterized by the degree to which there are
well-recognized and accepted rules of conduct that constrain and direct behavior (Weiss
& Adler, 1984). According to Mischel (1977), when (1) everyone is pressured to view the
situation in the same way, (2) there are uniform expectancies regarding appropriate
behavior, (3) adequate incentives are offered for the correct response pattern, and (4)
everyone has the skills the task requires, the task rmy not be complex enough to be
susceptible to the effects of individual difference variables. Although the laboratory is an
accepted local for testing theory-derived predictions under controlled conditions,
researchers often underestimate the effects of task complexity. In fact, the goal of most
lab experiments is to create situations in order to maximize variance attributed to the
study’s manipulations. Adding the use of homogenous samples with restricted variance

on individual difference variables creates a Situation where individual differences have a

123

I very small chance of exhibiting direct or moderated effects (Weiss & Adler, 1984). For
example, Ganster (1980) examined the effects of several individual difference variables
on the relationship between task characteristics, task perceptions, and satisfaction. The
task manipulations explained between 45% and 74% of the variance in task perceptions
and 58% of the variance in satisfaction. The individual differences accounted for little
additional variance in the dependent variables. Weiss and Adler (1984) attributed the
weak and inconsistent effects of the individual differences to aspects of the task. In this
study, stress accounted for between 10% and 28% of the variance in team processes. In
addition, the DDD command and control simulation tends to provide team members with
uniform expectancies regarding appropriate response patterns and those types of
behaviors are rewarded (e. g., shooting enemy targets inside versus outside the restricted
zones). Teams received feedback regarding their performance, which “provides clear
information about where teams need to focus their effort in order to be effective”
(LePine, Hollenbeck, Ilgen, Colquitt, & Ellis, 2002). Although this may have contributed
to the reduction or elimination of individual difference effects, past research using the
DDD has found effects for individual differences (e.g., Hollenbeck et al., 2002).
However, this is the ﬁrst DDD study to combine a functional structure with information
about the workload distribution during the task, speciﬁc performance feedback regarding
progress toward their three goals, and speciﬁc areas of expertise for each team nember.
Participants’ behaviOr was constrained and restricted to a greater degree in this study than
in the past, which may have reduced task complexity, thereby limiting the potential

explanatory power of cognitive ability and extraversion within teams.

124

Although the effects of cognitive ability and extraversion were nonexistent, the
results regarding prior shared information were more positive. Teams that shared 100%
of the information regarding future workload distribution problems during the task
developed more similar and more accurate shared nental models and engaged in more
retrieval coordination. Prior shared information also helped to moderate the negative
effects of stress on retrieval coordination. Results indicated that teams distributed
information within the team more frequently when it was shared. This supports the work
of other researchers interested in the discussion of shared and unshared information
within groups (e.g., Stasser & Titus, 1985). However, in this study, team nembers in the
high prior shared information conditions were paired up. Studies have shown that pairing
team members together may positively affect the level of information processing within
teams (Ellis, Hollenbeck, Ilgen, Porter, West, & Moon, under review-a; Hinsz et al.,
1997). Bion (1961) proposed that unconscious forces pressure team nembers to pair up
with other individuals. Team members want to feel uniﬁed with, not isolated from,
others. A number of beneﬁts, including improved team processes, may result. The results
of this study support this assertion by showing that giving two team members access to
the sane information improves shared nental models and retrieval coordination. The
results also partially support the notion that pairing team members together creates a
social support network that helps to lessen the negative effects of stress on transactive
memory. .

Feedback is another situational characteristic that was proposed to inﬂuence the
relationship between stress and team processes. Results indicated that providing negative

feedback to one team nember rather than the team as a whole did not directly affect

125

shared nental models or transactive memory. This did not support the proposed
relationship between feedback type and information processing efﬁciency and
effectiveness. However, results partially supported the idea that team level feedback
provides information that can help team nembers effectively cope with stressful
situations. Feedback moderated the relationship between stress and the information
allocation dimension of transactive nemory. When DM2 received the negative feedback,
teams evidenced lower levels of information allocation under stress than when the team
as a whole received the feedback. Team nembers receiving team level feedback appeared
to shift less of their attentional resources away from their interactive team duties in
stressful situations.

The weakness of the feedback nmnipulation may have been partially due to the
task environment. Each team was given negative feedback concentrating on three areas:
(1) clearing the restricted area of enemy tracks within 3 minutes, (2) clearing the highly
restricted zone of enemy tracks within 1 minutes, and (3) avoiding engaging ﬁiendly
tracks. It was proposed that giving this feedback to one team member would pressure
team nembers to concentrate more on their own duties and ignore the collective duties of
the team However, by giving each team member a very speciﬁc area of expertise, I may
have made it very difﬁcult to accomplish their goals without their teamrmtes’
participation. For example, if DM4 encountered a track with a power of 5 in his or her
restricted area, he or she had to ask DM2 for help because he or she only had 4 jets (i.e.,
power of 1). In other words, both types of feedback may have motivated team nembers

to try to work together as a collective unit because it was very difﬁcult for team nembers

126

to concentrate solely on their own duties. This may have eliminated any differential
effects between the individual and team negative feedback conditions.

Clearly the strongest results revolve around the negative effects of stress on
transactive memory and shared nental models. However, the disruption in information
processing due to the introduction of acute stressors would mean little if team
performance were not affected. Results showed that shared nental models and transactive
memory mediated the relationship between stress and team performance. This supports
the theory that the effects of stress on team performance are indirect, not direct,
functioning according to the input-process-outcome model propounded by researchers
(e.g., Hackman & Oldham, 1975). These results also support the importance of
information processing in the relationship between stress and team performance. Without
the ability to collectively process information, teams may be at a serious disadvantage
and may have trouble completing the tasks they are given, as suggested by researchers
(e.g., Hinsz et al., 1997).

The speciﬁc effects of shared mental models on team performance support the
ﬁndings of other researchers (Marks et al., 2000). While shared nental model accuracy
and shared mental model similarity were highly related to one another, they evidenced
interactive effects on team performance. Shared nental model similarity reduced team
performance when shared mental model accuracy was low but had little effect on team
performance when shared mental model accuracy was high. The results of this study
suggest that, when operationalizing and analyzmg shared mental models, accuracy and

similarity should be kept distinct ﬁ'om one another.

127

The speciﬁc effects of transactive nemory on team performance are more
intriguing and have a number of theoretical implications. Although Moreland and his
colleagues (e. g., Moreland, 1999; Moreland & Myaskovsky, 2000) have found that task
credibility, task coordination, and memory differentiation relate to team performance, this
is the ﬁrst study to show that directory updating, information allocation, and retrieval
coordination all have the potential to affect team performance. Results showed that, in
this study, directory updating was the primary inﬂuence early on in the process, while
information allocation and retrieval coordination becanre more important as time
progressed. However, the task in this study remained fairly static and team nembers’
areas of expertise did not expand or contract. In these situations, continually updating
one’s directory may become useless and perhaps even counterproductive. When new
responsibilities arise or team nembers gain additional knowledge or skill, which would
likely happen in most organizations, directory updating could become more important.

The results regarding transactive nemory also provide additional evidence that
the construct may be aggregate in nature. The dimensions exhibited differential
relationships with team performance depending on the duration of the task. Initially,
directory updating determined whether teams were effective or not. However, after team
nembers gained knowledge regarding each other’s areas of expertise, they began to
utilize information allocation and retrieval coordination to become more eﬁ'ective. With
latent constructs, the dinensions are nerely manifestations of some deeper, more
embedded construct (Law et al., 1998). Measures of each dinension should therefore be
highly related to one another because, if transactive memory is operating as a latent

construct, team members should utilize similar amounts of directory updating,

128

information allocation, and retrieval coordination. However, the three dimensions
evidenced low correlations with one another because team nembers may not need to use
all three to enjoy a fully functioning transactive memory system This supports the idea
that the arrows point to the transactive nemory construct and not away from it (Lepine et
al., 2002). Lepine and his colleagues suggest that, if the construct is aggregate, future
research should examine the directory updating, information allocation, and directory
updating as separate dimensions. Although the results of this study found few differential
effects, it is likely that certain predictors of transactive memory exhibit different
relationships with directory updating, information allocation, and retrieval coordination.

If transactive memory is an aggregate construct, it can be formed by algebraically
combining directory updating, information allocation, and retrieval coordination (Lepine
et al., 2002). In this study, there was a lack of previous conceptual and empirical
evidence supporting the domination of one or more dimensions of transactive nemory, so
they were combined equally to form the aggregate transactive memory construct.
However, results indicate that this may be inappropriate in some situations. For example,
if the team has been together for an extended period of tine and team nembers’ areas of
expertise have remained static, it may be useful to weight information allocation and
retrieval coordination more heavily. Researchers need to determine which dinensions are
most important, given the situation. This is the ﬁrst study to suggest that it. may make
conceptual and empirical sense to combine directory updating, information allocation,
and retrieval coordination unequally.

Finally, this is the ﬁrst study to examine the relationship between transactive

memory and shared mental models. The results show that both were related to one

129

another, supporting the proposition that both comprise portions of the team’s collective
information processing system However, the three dimensions of transactive nemory
were not equally strong predictors of shared mental model similarity and shared nental
model accuracy. When the team members initially began the task, and were unaware of
each other’s areas of expertise, directory updating played the major role in determining
the degree to which their shared mental models were similar to one another and accurate.
As the task progressed, directory updating began to become less inﬂuential and
information allocation and retrieval coordination began to contribute to the development
of shared mental models. Conceptually, this indicates that, once team members are aware
of each other’s areas of expertise, they may need to utilize information allocation and
retrieval coordination to keep the roles and responsibilities of their teammates fresh in
their minds. If team members fail to use retrieval coordination and information allocation,
their shared interaction and team member mental models may becone less similar and
less accurate.

Practical ImMications.

Practically, the results of this study may help to explain why certain problems
arise in real-world stressful situations. As noted in the introduction, the USS Vincennes
mistakenly shot down an Iranian passenger airline in 1988. At the tine of the accident,
the Vincennes was engaged in a gun battle with Iranian gunboats. The situation became
increasingly hectic as US. military personnel scrambled to avoid further loss of life.
Suddenly the Iranian airliner appeared on US. radar screens about 20 miles away
heading directly for the Vincennes. The airliner was one of hundreds of blips on the radar

screen. Without speaking to the ofﬁcer responsible for coordinating anti-air warfare, two

130

console operators reported that the airliner was descending. The airliner was actually
ascending, which the ofﬁcer would have known because it was his area of expertise. If
the console operators would have remembered who was responsible for what, the
accident may have been avoided. It appeared that shared interaction and team member
nental models deteriorated under the stress of the attacks (Collyer & Malecki, 1998). The
results of this study provide empirical evidence to support the negative effects of stress
on shared mental models in team contexts.

Similar problems arose during the terrorist attacks of 9/11. News of the attacks
dominated television stations and newpapers, ingraining the pictures of suffering and
grief in the mind of people in and outside of the United States. Several days after the
attacks, a national survey was conducted to determine how far reaching the effects were.
Results showed that 90% of those polled reported at least one or more symptoms of stress
(Schuster et al., 2001). The stress caused by 9/11 may have hit those closest to the attacks
the hardest. Government agencies such as the FBI and the CIA were responsible
investigating the incident and ensuring that all US. citizens remained safe and secure in
the aftermath of the attack on national security. Weeks later, the behavior of agencies
such as the CIA and the FBI immediately before and after the incident underwent close
scrutiny. It was found that the various agencies, who were supposed to be working
together as a collective unit, had trouble “connecting the dots.” More speciﬁcally, they
failed to share vital pieces of information with each other and local police departments
(e.g., Gates, 2002; Tumulty, 2002). The various agencies did not know who had what or
who needed what information. It was also unclear who could be given access to certain

pieces of information. Each agency possessed a speciﬁc area of expertise that others were

131

unaware of and they failed to ask each other for the information. As a result, information
could not be given to the people who needed it. To connect the dots, the agencies needed
to update their directories, allocate information, and coordinate the retrieval of
information. However, the results of this study show that, during stressful encounters,
transactive nenrory breaks down, which could have created some of the interagency
problems during 9/11.

Fortunately, there may be a practical solution to the problems associated with
stress in teams: team training. After the USS Vincennes mistakenly shot down the Iranian
passenger airline, a group of researchers began a research and development program
called Tactical Decision Making Under Stress (T ADMUS). The program had a number
of goals, including the development of training and simulation principles for stressful
situations. Collyer and Malecki (1998) noted that “the focus of this task is on developing
and demonstrating a variety of individual and team training strategies and techniques to
minimize the adverse effects of stress” (p. 11). The strategies that were developed by
TADMU S focused on shared situational models of the task environnent and the task
itself, as well as shared models of interaction requirements and responsibilities.

TADMU S researchers suggested that shared mental model training could help individuals
and teams perform under stress (Serfaty, Entin, & Johnston, 1998). However, Serfaty and
his colleagues merely assuned that shared mental models suffer during stressful
situations. The TADMU S project never examined the effects of stress on any team
process variable. In order to develop a training program, a needs assessment should

always be performed to determine where trainees lack the requisite knowledge, skill, or

132

ability (e.g., Ostroﬁ & Ford, 1989). If stress does not negatively affect shared nental
models, training would be more beneﬁcial in other areas.

This study provides empirical justiﬁcation for the development and use of training
programs designed to improve not only shared mental models, but also transactive
nemory, in teams under stress. Two training programs in particular have the potential to
provide the knowledge, skill, and ability to weather demanding environmental conditions
and strengthen the interconnections within teams: cross-training and transportable
teamwork skills training. Cross-training has been deﬁned as “an instructional strategy in
which each team member is trained in the duties of his or her teammates” (Volpe,
Cannon-Bowers, Salas, & Spector, 1996, p. 87). Researchers have suggested that cross-
training encourages team members to understand the behavior of their teammates, which
positively affects team processes such as communication and coordination
(Blickensdorfer, Cannon-Bowers, & Salas, 1998). Volpe and her colleagues (1996) and
Cannon-Bowers and her colleagues (1998) rotated team members so that each could
experience the roles and responsibilities of their teammates. Both studies found that the
positional rotation form of cross-training improved information distribution and team
performance. Marks and her colleagues (2002) examined two less in-depth forms of
cross-training: positional clariﬁcation and positional modeling. Positional clariﬁcation
provides team nembers with verbal information regarding their teammates’ jobs.
Positional modeling involves verbally discussing and observing team members’ roles.
Marks and her colleagues found that both forms of cross-training signiﬁcantly affected

the level of shared nental models within teams.

133

While cross-training is task speciﬁc, transportable teamwork skills training can be
given to employees even if they switch teams and tasks on a frequent basis. Transportable
teamwork skills training has been deﬁned as an instructional strategy in which team
members are trained in team- and task-generic knowledges, skills, and attitudes (Ellis,
Bell, & Ployhart, under review-b). Smith-Jentsch and her colleagues (1996), in one of the
ﬁrst studies examining the effects of transportable teamwork skills training, trained 60
undergraduates in team performance-related assertiveness, which can be deﬁned as the
ability of team members to share their opinions with their teammates in a manner that is
persuasive to others. Participants were trained as individuals using one of three training
nethods: behavioral role modeling, lecture with demonstration, and lecture based
training. All three training methods were team- and task—generic. Participants were then
paired up to complete a PC-based ﬂight simulation task as a two-person team Results
indicated that the three transportable teamwork skills training programs produced nrore
positive attitudes regarding team performance-related assertiveness. Behavioral role
modeling also had a signiﬁcant positive effect on performance-related assertive behavior.
Cannon-Bowers and her colleagues (1995) suggested that transportable teamwork skills
training positively inﬂuences team effectiveness by strengthening the interconnections
between team nembers. Ellis and his colleagues (under review-b) went a step further by
proposing that transportable teamwork skills training improves the information
processing systems operating within teams. In their study, participants were trained in a
variety of transportable skills before being placed in a team environnent. Teams then
completed a command and control simulation, where they were allowed to interact freely

with one another. Results indicated that training positively affects both team learning and

134

backup behavior by increasing the amount of knowledge sharing and backup requests in
temporary project teams.

Cross-training and transportable teamwork skills training have been conceptually
and empirically linked to the information processing capabilities of teams. Researchers
have shown that these types of training programs have the potential to impact shared
mental models (Marks et al., 2002). Similar types of training have also been linked to
transactive memory (Liang, Moreland, & Argote, 1995). Team training likely affects
shared nental models and transactive memory because processing information between
group nembers became easier, allowing them “to better coordinate and differentiate
nember skills and knowledge of the task” (Hinsz et al., 1997, p. 53). The results of this
study suggest that these types of training programs may be the perfect solution in times of
stress.

Limitations and Directions for Future Research.

Despite the theoretical and practical implications of the results, there are a number
of limitations that bear further investigation. For one thing, this study took a response-
based perspective to the study of stress in temporary project teams. Although aspects of
stimulus-based, interactive, and transactional theories were included, the main focus was
on team processes. The stress literature would beneﬁt ﬁom a further investigation of
stress in teams utilizing one of the other three perspectives.

From a stimulus-based perspective, research is needed to investigate the effects of
different environnental characteristics on the level of stress within teams. The
combination of tine pressure and threat were used to maximize the level of stress within

teams. Results could not differentiate the unique effects of each stressor. Because the two

135

manipulations were combined together, this presents a confound that could inﬂuence the
effects of stress. Tine pressure varied by time period but the stress of being videotaped
was consistent across time. The impact of this limitation is minimal, however, because
time did not signiﬁcantly affect the relationship between stress and transactive memory
and shared nental models.

There are also a number of potential stressors that are unique to team contexts,
such as team member conﬂict, breakdowns in communication, and layoffs within the
team, that were not investigated in this study. Because team members are interdependent,
stressors that affect one team member rmy still inﬂuence the team’s collective
information processing capabilities. For instance, threatening to layoff one team member
may affect the level of transactive memory and shared nental models within the team In
addition, this study concentrated on the effects of acute stress. Chronic stress exerts an
inﬂuence on employees over a much longer period of time, which could inﬂuence
transactive memory and shared mental models differently than acute stress. For example,
with acute stressors, the effect on team processes may be greatest immediately after the
incident. With chronic stressors, the effects may take years to manifest themselves within
teams.

Future research should also continue to take an interactive perspective by
investigating other personal and situational characteristics as possible moderators of the
relationship between stress and team processes. This study concentrated on cognitive
ability, extraversion, feedback, and shared information, which can be linked to the team’s
collective information processing capabilities. However, stress researchers have

identiﬁed a number of other personal variables, including Type A behavior pattern,

136

negative affectivity, and hardiness, that have been shown to ameliorate the negative
effects of stress at the individual level (see Cooper et al., 2001). Little has been done to
examine such variables at the team level. Aggregating these types of personal
characteristics in teane may deternrine whether transactive memory and shared nental
models remain intact during stressful situations.

There are also a number of situational characteristics that bear further
investigation. For instance, the type of team encountering stress could inﬂuence the
degree to which transactive memory and shared mental models deteriorate. This study
focused on project teams, where team nembers cone together for a short period of time
to handle a set number of tasks (Allred et al., 1996). Because project team members have
little experience with one another, they spend most of their tine in the “forming” and
“storming” stages, where communication and interaction are essential. They often fail to
reach the “norming” stage, where team nembers’ nental representations reach maturity
and communication plays less of a role (Cannon-Bowers & Salas, 1997). However,
organizations utilize a variety of teams, including top management teams, virtual teams,
sales teams, service teams, assembly teams, etc. Sundstrom and Altman (1989)
categorized work groups according to four characteristics: work team differentiation
(redundancy of member roles), external integration (synchronization with constituents
external to the group), work cycles (the length and uniqueness of a performance event),
and typical outputs (physical versus non-physical outcones). Teams with little team
member diﬁemntiation or teams with long work cycles may be more impervious to stress.
In these types of teams, communication and interaction play less of a role or have becone

routine. Future research needs to examine the stress process in these types of teams.

137

Matching the environnent to the team’s structure may also represent an important
characteristic of the Situation. In this study, teams were structured functionally and team
nembers were grouped based on the similarity of the work that they performed.
Functional structures create narrow and specialized roles with high interdependence
requirements. According to Structural Contingency Theory, functionally structured teams
tend to perform best in relatively predictable and stable environments (Burns & Stalker,
1961). However, the environnent faced by teams in this study was unstable and
unpredictable, which “create[d] changing and complex contingencies that overwhelm the
simple and specialized subunits” (Hollenbeck, Moon, Ellis, West, Ilgen, Porter, &
Wagner, 2002). Hollenbeck and his colleagues have shown that teams in situations of
external misﬁt, where the environnent and the structure do not match, tend to be much
less effective. Future research is needed to determine whether external ﬁt has
implications for the relationship between stress and team processes. In functional
structures, behavior may be much more routine when the environment is relatively stable
and predictable. This could offer an advantage to teams by moderating the negative
effects of stress on transactive memory and shared nental models.

The transactional perspective may also offer an avenue for further stress research
in teams. Folkman and his colleagues (e. g., 1984) suggest that the way individuals
cognitively appraise and cope with the situation could also have beneﬁcial effects in the
stress process. When individuals encounter stress, they determine whether the situation
represents harm, threat, or challenge through the primary appraisal process. The
secondary appraisal process then searches for the appropriate coping resources. However,

little has been done to investigate stress in teane from a transactional perspective. Do

138

team nembers appraise the sane situation differently or is there sone sort of shared
primary appraisal? Can team nembers utilize their teammates’ coping resources? Are the
coping strategies used by teams the same as those used by individuals? Future research is
needed to examine the effects of cognitive appraisal on the stress process in teams.

While viewing stress in teams from stimulus-based, interactive, and transactional
perspectives may be beneﬁcial, there is much more work that needs to be done regarding
team nembers’ responses to stress. In this study, we focused on transactive nemory and
shared nental models. However, there are a number of other outcomes that need to be
examined. For example, Ellis and his colleagues (under review-a) used an information
processing framework to investigate the degree to which certain personal and situational
characteristics inﬂuenced team learning. In line with information processing theory, they
found that teams learned more when composed of individuals who were high in cognitive
ability, when the workload was distributed evenly, and when the team utilized a paired
structure. Conversely, team learning was negatively affected when teams were conrposed
of individuals who were high in agreeableness. The results of Ellis and his colleagues
support the notion that the team’s ability to collectively process information is critical for
the team learning process. Because team learning appears to rely on the collective
information processing capabilities of the team members, it may be particularly
susceptible to the effects of stress. Other researchers have suggested that backup
behavior, which represents one of the critical aspects of teamwork (Dickenson &
McIntyre, 1997; McIntyre & Salas, 1995), depends on the team’s infornmtion processing
system (Ellis et al., under review-b). Marks and her colleagues (2001) deﬁne backup

behavior as “assisting team nembers to perform their tasks, which may occur by ( 1)

139

providing a teammate verbal feedback or coaching, (2) assisting a teammate behaviorally
in carrying out actions, or (3) assuming and completing a task for a teammate” (p. 367).
Ellis and his colleagues (under review-b) found that training teams in transportable
teamwork skills strengthened the interconnections between the team members and
resulted in higher levels of backup behavior. If stress weakens the interconnections
between team members, backup behavior may be reduced or eliminated within teams
under stress.

Conclusion.

The results of this dissertation suggest that stress exerts a strong inﬂuence on the
interaction processes within teams. In this study, the effects were particularly detrimental
to transactive nenrory and shared mental models. Consequently, teams under stress were
much less effective than teams in the control condition. It is suggested that this reﬂects a
disruption in the teams’ collective information processing systems. Although these results
further our understanding of the stress process in teams, few prescriptions are offered to
alleviate the negative effects of stress on team processes. The personal and situational
characteristics examined generally failed to moderate the relationship between stress and
transactive nemory and shared mental models. However, it is hoped that this dissertation
serves as an impetus for further research examining not only the stress process within
teams but also methods to bolster the information processing capabilities of teams under

SITCSS.

140

References
Adelrnan, L., Zirk, D.A., Lehner, P.E., Moffett, R.J., & Hall, R. (1986). Distributed
tactical decision making: Conceptual framework and empirical results. IEEE

Transactions on Systems, Man, and Cametics, 16, 794-805.
Aderrnan, D. (1972). Elation, depression, and helping behavior. Journal of Personality

and Social Psychology, 24, 91-101.

Aiello, J .R., & Kolb, KJ. (1995). Electronic performance monitoring and social context:
Impact on productivity and stress. Journal of Applied Psychology, 80, 339-353.

Aiello, J.R., & Shao, Y. (1993). Electronic performance monitoring and stress: The role
of feedback and goal setting. In M.J. Smith & G. Salvendy (Eds), Human-computer
interaction: Applications and case studie_s (pp. 1011-1016). Amsterdam: Elsevier
Science.

Aiello, J.R., DeNisi, A.S., Kirkhoff, K., Shao, Y., Lund, M.A., & Chomiak, AA. (1991,
June). The Qp' act of feedback and individual/gmup computer monitoring on work
affgptp Paper presented at the meeting of the American Psychological Society,
Washington, DC.

Allred, B.B., Snow, C.C., & Miles, RE. (1996). Characteristics of managerial careers in
the 21St century. Academy of Management Executive, 10, 17-27.

Allred, K., & Smith, T. (1989). The hardy personality. Cognitive and physiological

. responses to evaluative threat. Journal of Personality and Social Psychology, 56, 257-
266.
Amick, B.C., & Smith, M.J. (1992). Stress, conrputer-based work monitoring and

measurement systems: A conceptual overview. Applied Ergonomics, 23, 6- 16.

141

Anderson, C., Lindsay, J .L., & Bushman, J. (1999). Research in the psychological
laboratory: Truth or triviality? Curren; Directiona in Psychological Science, 8, 3-9.

Argote, L. (1993). Group and organizational learning curves: Individual, system, and
environmental components. Britisir Journal of Social Psychologyﬂ, 31-51.

Ashby, W.R. (1966). An introduction to cmeetics. New York: Wiley.

Aspinwall, L.G., & Taylor, SE. (1997). A stitch in time: Self-regulation and proactive
coping. Psychological Bulletin. 121. 417-436.

Atwater, L.E., Waldman, D.A., Atwater, D., & Cartier, P. (2000). An upward feedback
ﬁeld experiment: Supervisors’ cynicism reactions, and commitnent to subordinates.
Personnel Psyaholagy,5_3, 275-297.

Averill, JR. (1973). Personal control over aversive stimuli and its relationship to stress.

Psychological Bulletin, 80, 286-303.

Bacon, 8.]. (1974). Arousal and the range of cue utilization. Journal of Experimenpal
Psychology, 102, 81-87.

Baddeley, AD. (1972). Selective attention and performance in dangerous environnents.
Britis_h Journal of Psychology, 63, 537-546.

Baddeley, AD. (1986). Working memog. Oxford, England: Oxford University Press.

Baddeley, A.D., & Hitch, G. (1974). Working nemory. In G. Bower (Ed.), Advances in
learning and motivation (Vol. 8). New York: Academic Press, 1974.

Bahrick, H.P., Fitts, P.M., & Rankin, RE. (1952). Effects of incentives upon reactions to
peripheral stimuli. Journal of Exp_e_rimental Psychology, 44, 400-406.

Baldwin, J.M. (1894). Mental development in the child and the rag. New York:

Macmillan.

142

Bandura, A. (1977). Social learning theogy. Englewoods Cliffs, NJ: Prentice-Hall, Inc.

Bandura, A. (1986). Social foundation of thought and action: A social cognitive theogy.
Englewood Cliffs, NJ: Prentice-Hall.

Banrlund, DC. (1959). A comparative study of individual, majority, and group judgnent.
Journal of Abnormal and Social Psychology, 58, 55-60.

Baron, R.M., & Kenny, D.A. (1986). The mediator-moderator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations,
Journal of Persoaaﬂyand Socarl Psychology, 51, 1173-1182.

Barrick, M.R., & Mount, M.K (1991). The Big Five personality dimensions and job
performance: A meta-analysis. Personnel Psychology, 44, 1-26.

Barrick, M.R., Mount, M.K, & Judge, TA. (2001). The FFM personality dimensions and
Job Performance: Meta-Analysis of Meta-analyses. Invited submission to a spacial

“selection” issue of International Journal of Selection and Assessnent 9, 9-30.

 

Barrick, M.R., Stewart, G.L., Neubert, M.J., & Mount, MK. (1998). Relating nember
ability and personality to work-team processes and team effectiveness. Journal of
Applied Psychology, 83, 377-391.

Barry, B., & Stewart, G. (1997). Composition, process, and performance in self-managed
groups: The role of personality. Journal of Applied Psychology, 82, 62-78.

Bartlett, EC. (1932). Remembering: A study in exparimental and social psychology.
Cambridge, England: Cambridge University Press.

Bazerman, M.H., Mannix, E..A., & Thonrpson, LL. (1988). Groups as mixed-motive

negotiations. Advances in Group Processes, 5, 195-216.

143

Beehr, T.A. (1976). Perceived situational moderators of the relationship between
subjective role ambiguity and role strain. Journal of Applied Psychology, 61 , 35—40.

Beehr, T.A. (1985). The role of social support in coping with organizational stress. In
T.A. Beehr & R.S. Bhagat, (Eds.), Human stress and cognition in organizations (pp.
375-398). New York: Wiley.

Beehr, T.A. (1995). Psychological stress in the worm. New York: Routledge.

Beehr, T.A., & Bhagat, RS. (1985). Introduction to human stress and cognition in
organizations. In T.A. Beehr & R.S. Bhagat (Eds.), Human stress and cogm'tion in
organizations (pp. 3-19). New York: John Wiley.

Beehr, T.A., & Franz, TM. (1987). The current debate about the meaning of job stress. In
J.M. Ivancevich & D.C. Ganster (Eds.), Job stress: From theo to su estion
M). New York: Hawthorne Press.

Beehr, T.A., King, L.A., & King, D.W. (1990). Social support and occupational stress:

Talking to supervisors. Journal of Vocational Behavior 36, 61-81.

 

Beehr, T.A., Walsh, J .T., & Taber, TD. (1976). Relationships of stress to individually
and organizationally valued states: Higher order needs as a moderator. Journal of
Applied Psychology, 61, 41-47.

Bettenhausen, KL, & Murnighan, J .K. (1991). The development of an intragroup norm
and the effects of interpersonaland structural challenges. Administrative Science
Qparterly, 36, 20-35. .

Bion, W.R. (1961). Experiences in gpoups. New York: Basic Books.

Blickensderfer, E., Cannon-Bowers, J.A., & Salas, E. (1998). Cross-training and team

performance. In J.A. Cannon-Bowers & E. Salas (Eds.), Making decisions under

144

stress: Implicatioas ftﬂndividual a_n_d team trai_n_i_ng (pp. 299-311). Washington, DC:
American Psychological Association.

Bliese, PD. (2000). Within-group agreement, non-independence, and reliability:
Implications for data aggregation and analysis. In KJ. Klein & S.W.J. Kozlowski
(Eds.), Multilevel theog, research, and methods in organizations. San Fransisco, CA:
J ossey Bass.

Bliese, P.D., & Jex, SM. (1999). Incorporating multiple levels of analysis into

occupational stress research. Work and Stress 13, 1-6.

 

Block, J. (1995). A contrarian view of the ﬁve-factor approach to personality description.
Psychological Bulletin, 117, 187-215.

Boff, KR., & Lincoln, J.E. (1988). Engineering data commndium: Human mrception and
mrformance. Wright-Patterson Air Force Base, OH: Harry G. Armstrong Medical
Research Laboratory.

Brehner, B. (1972). Cue utilization and cue conﬁguration in multiple-cue probability
learning. Organizational Behavior and Human Perfmnce. 8. 1-36.

Brief, A.P., & George, J.M. (1995). Psychological stress and the workplace: A brief
comment on Lazarus’ outlook. . In R. Crandell & P.L. Perrewe (Eds.), Occupational

stress: A handbook (pp. 15-19). Washington, DC: Taylor & Francis.

Burke, R]. (1993). Toward an understanding of psychological burnout among police

ofﬁcers. Journal of Social Behavior and Personality, 8, 425-438.

Burns, T., & Stalker, GM. (1961). The management of innovation. London: Tavistock.

Bursill, AB. (1958). The restriction of peripheral vision during exposure to hot and

humid conditions. Qaarterly Journal of Experimental Psychology, 10, 113-119.

145

Cannon, W. (1935). Stresses and strain of homeostasis. American Journal of Medical

 

Science 189,1-14.

 

Cannon, W.B. (1929). Organization for physiological homeostasis. Physiological

Review 9, 339—430.

 

Cannon-Bowers, J.A., & Salas, E. (1990). Cognitive psychology and team training:
Shared mental models in complex systems. Pap_er presented at the annual neeting of
the Society for Industrial and Organizational Psychology, Miami.

Cannon-Bowers, J.A., & Salas, E. (1997). A framework for developing team performance
measures in training. In M.T. Brannick, E. Salas, & C. Prince (Eds.), 3am
mrformance assessment and neasurenrent (pp. 45-62). Mahwah, NJ: Erlbaum

Cannon-Bowers, J.A., & Salas, E. (1998). Making decisions un_der stress: Implications
for individual and team training. Washington, DC: American Psychological
Association.

Cannon-Bowers, J .A., Oser, R., & Flanagan, D.L. (1992). Work teams in industry: A
selected review and proposed ﬁamework. In R.W. Swezey & E. Salas (Eds.), Bram;
Their training and p_erformance (pp. 355-377). Norwood, NJ: Ablex.

Cannon-Bowers, J .A., Salas, E., & Converse, SA. (1990). Cognitive psychology and
team training: Shared nental models in complex systems. Human Factors Society

Bulletin 33, 1-4.

 

Cannon-Bowers, J.A., Salas, E., & Converse, SA. (1993). Shared mental models in

expert decision nmking. In N .J . Castellan, Jr. (Ed.), Current issues in individual and

group decision making (pp. 221-246). Hillsdale, NJ: Erlbaum

146

Cannon-Bowers, J .A., Salas, E., Blickensderfer, E., & Bowers, CA. (1998). Impact of
cross training and workload on team functioning: A replication and extension of
initial ﬁndings. Human Factors. 40. 92-101.

Cannon-Bowers, J.A., Tannenbaum S.I., Salas, E., & Volpe, CE. (1995). Deﬁning
competencies and establishing team training requirements. In R.A. Guzzo & E. Salas
(Eds.), Team effectiveness and decision making in organizations (pp.262-290). San
Francisco, CA: Jossey-Bass.

Caplan, R.D., Cobb, 8., French, J.R., Jr., Harrison, R.V., & Pirnneau, SR. (1975). 191;
demanLds and wcﬂcer health: Main effects and occmgational differences. Washington,
DC: Government Printing Ofﬁce.

Carayon, P. (1993). A longitudinal test of Karasek’s job strain model among ofﬁce
workers. Work and Stress, 7. 299-314.

Carver, C.S., & Scheier, M.F. (1981). Attention and self-regulation: A control theog to
human behavior. New York: Springer-Verlag.

Cattell, RB. (1946). The description and measurement of parsonality. Yonkers, NY:
World Book.

Christian, P., & Lolas, F. (1985). The stress concept as a problem for a theoretical

 

pathology. Social Science and Medicine 2, 363-365.

Clark, N .K., & Stephenson, GM. (1989). Group remembering. In P. Paulus (Ed.),

Psychology of group inﬂuence (pp. 357-391). Hillsdale, NJ: Erlbaum
Cobb, S., & Kasl, S.V. (1977). Termination: The consequences of 19' b loss. US.

Department of Health, Education, and Welfare. Washington, DC: US. Govemment

Printing Office.

147

Cohen, J. (1960). A coefﬁcient of agreement for nominal scales. Educational and
Psychological Measurenerapgg, 37-46.

Cohen, J. (1988). Statistical p_ower analysis for the behavioral sciences (2"d ed.).
Hillsdale, NJ: Erlbaum

Cohen, J ., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the
mvioml sciances (2"d ed.). Hillsdale, NJ: Erlbaum

Cohen, S. (1978). Environmental load and the allocation of attention. In A. Baum J.E.
Singer, & S. Valins (Eds.), Advances in experimental psychology. 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.

Cohen, S., & Wills, T. (1985). Stress, social support, and the buffering hypothesis.
Psychological Bulletin. 92. 257-310.

Cohen, S.G., & Bailey, DE. (1997). What makes teams work: Group effectiveness
research from the shop ﬂoor to the executive suite. Journal of Management, 23, 239-
290.

Collyer, S.C., & Malecki, GS (1998). Tactical decision making under stress. In J.A.
Cannon-Bowers & E. Salas (Eds.), Making decisions under StTCSS:Mn_S_f(M
individual_ and team trainLng (pp. 3-15). Washington, DC: American Psychological
Association.

Combs, A.W., & Taylor, C. (1952). The effect of the perception of mild degrees of threat

on performance. Journal of Abnormal and Social Psychology, 47, 420-424.

148

Conlon, E.J., & Barr, SH. (1989). A framework for understanding group feedback.
Advances in Group Processes, 6, 27-48.

Cook, T.D., & Campbell, JP. (1979). Quasi-experimentation: Desigagalysis issues for
ﬁeld settings. Boston, MA: Houghton Mifﬂin Company.

Cooper, CL. (1998). Theories—of organizational stress. New York: Oxford University
Press.

Cooper, C.L., & Kelly, M. (1984). Stress among crane operators. Journal of Occupaaarﬂ
Medicine. 26. 575-578.

Cooper, C.L., & Payne, R. (1988). Causes, coping, and cmraequences of stresﬁwork.
Chichester, England: Wiley.

Cooper, C.L., & Roden, J. (1985). Mental health and satisfaction among tax officers.
SociMScieﬁnce aM Meiicine. 21. 747-751.

Cooper, C.L., Dewe, P.J., & O’Driscoll, M.P. (2001). Organizational stress: A revie_waad
We of theory. research, and applications. Thousand Oaks, CA: Sage Publications,
Inc.

Costa, P.T., & McCrae, RR. (1992). Revised NEO p_ersonality inventogt. Odessa, FL:
Psychological Assessment Resources.

Cox, T. (1990). The recognition and measurement of stress: Conceptual and
nethodological issues. In N. Corlett & J. Wilson (Eds.), W._
London: Taylor & Francis. .

Cox, T. (1997). Workplace health promotion. Wrm and Stress. 11. 1-5.

Cream B.W., Eggemeier, P.T., & Klein, GA. (1978). A strategy for development of

training devices. HUMactors. 20. 145-158.

149

Cronbach, L.J., & Furby, L. (1970). How should we neasure “change”- or should we?
Psychological Bulletin, 74, 68-80.

Cummings, T.G., & Cooper, CL. (1979). A cybernetic framework for studying
occupational stress. Hman Relations. 32. 395-418.

Cunningham M.R., Steinberg, J., & Grev, R. (1980). Wanting to and having to help:
Separate motivations for positive mood and guilt induced helping. Journal of
Personality and Social Psychology. 38. 181-192.

Daneman, M., & Carpenter, RA. (1980). Individual differences in working nemory and
reading. Journal of VerbaMearn_i_ngand Verbal Behaiior. 19. 450-466.

Daneman, M., & Carpenter, RA. (1983). Individual differences in integrating
information between and within sentences. Jouraai of Experimental Psychology:
Learning, memogt, and cognition. 9. 561-583.

Davidson, M., & Cooper, C. (1980). Type A coronary-prone behavior in the work

 

environment. Journal of Occupational Medicine. 22. 375-383.

Deci, E.L., Eghrari, H., Patrick, B.C., & Leone, D. (1994). Facilitating internalization:
The self-determination theory perspective. Journal of Personm, 119-142.

Dempster, RN. (1981). Memory span: Sources of individual and developnental
differences. Psychological Bulletin. 89. 63-100.

Dewe, P., Cox, T., & Ferguson, E. (1993). Individual strategies for coping with stress and
work: A review. Work!and Stress. 7. 5-15.

Dickenson, T.L., & McIntyre, RM. (1997). A conceptual franework for teamwork

measurement. In M.T. Brannick, E. Salas, & C. Prince (Eds.), Team performance and

150

neasurement: Theory. methods. and applications (pp. 19-43). Mahwah, NJ: Lawrence
Erlbaum

Dodrill, CB. (1983). Long term reliability of the Wonderlic personnel test. Journal of
Consulting and Clinical Psychology, 51, 316-317.

Donnellon, A., Gray, B., & Bougon, MG. (1986). Communication, neaning, and
organized action. Administrative Science Quarterly. 31. 43-55.

Driskell, J .E., & Johnston, J.H. (1998). Stress exposure training. In J.A. Cannon-Bowers
& E. Salas (Eds.), Making decisions under stress: Implications for individualMaad
team training. Washington, DC: Anerican Psychological Association.

Driskell, J .E., & Salas, E. (1991). Group decision making under stress. Journal of
Applied Psychology, 76, 473—478.

Driskell, J .E., Hogan, R., & Salas, E. (1987). Personality and group performance. In C.
Hendrick (Ed.), Group processes and intergroup relations (pp. 91-112). Newbury
Park, CA: Sage.

Driskell, J .E., Salas, E., & Johnston, J. (1999). Does stress lead to a loss of team

perspective? Group Mamics: Theory. Research. and Practice 3, 291-302.

 

Dwyer, D.J., & Ganster, DC. (1991). The effects of job demands and control on

employee attendance and satisfaction. Journal of Organizational Behavior. 12. 595-
608.

Dyer, J. (1984). Team research and training: A state-of-the-art review. In F.A. Muckler
(Ed.), Human factors review: 1984 (pp. 285-323). Santa Monica, CA: Human Factors

Society.

151

Earley, RC. (1988). Computer-generated performance feedback in the magazine
subscription industry. Organizational Behavior and Human Decision Processes. 41.
50-64.

Earley, P.C., Northcraft, G.B., Lee, C., & Lituchy, TR. (1990). Impact of process and
outcone feedback on the relation of goal setting to task performance. Academy of
Magenta; Journal. 33. 87-105.

Easterbrook, J.A. (1959). The effect of emotion on cue utilization and the organization of
behavior. Psychological Review. 66. 183-201.

Edland, A., & Svenson, O. (1993). Judgment and decision making under tine pressure:
Studies and ﬁndings. In 0. Svenson & J. Maule (Eds.), Time pressure a_nd;stress in
human jr_r' dgment_ and decision making (pp. 27-40). New York: Plenum

Edwards, J .R. (1991). Person-job ﬁt: A conceptual integration, literature review, and
methodological critique. In C.L. Cooper & I.T. Robertson (Eds.), International review
of industrial and organizational psychology (pp. 283-357). New York: Wiley.

Edwards, J .R. (1992). A cybernetic theory of stress, coping, and well-being in
organizations. Academy of Management Journal. 17. 238-274.

Edwards, JR. (1996). An examination of competing versions of the person-environment
approach to stress. Academy of Management Journal. 39. 292-339.

Edwards, J.R., & Cooper, CL. (1988). Research in stress, coping, and health: Theoretical
and nethodological issues. Psychological Medicine. 18. 15-20.

Edwards, J.R., & Cooper, CL. (1990). The person-environnent ﬁt approach to stress:
Recurring problems and some suggested solutions. Journal of Organizational

Behavior 10, 293-307.

 

152

Edwards, J.R., & van Harrison, R. (1993). Job demands and worker health: Three
dimensional reexamination of the relationship between person-environnent ﬁt and
strain. Journal of Applied Psychology, 78, 628-648.

Ellis, A.P.J, West, B.J., Ryan, A.M., & Deshon, R.P. (in press). The use of impression
management tactics in structured interviews: A function of question type? Journal of
Applied Psychology.

Ellis, A.P.J., Bell, B.S., & Ployhart, R.E. (under review-b). Connecting the dots: The
effect of transportable teamwork skills training on project team effectiveness.
Administrative Science Quarterly.

Ellis, A.P.J., Hollenbeck, J.R., Ilgen, D.R., Porter, C.O.L.H., West, B.J., & Moon, H.K
(under review-a). Project team learning: The effects of personal and situational

characteristics. Journal of Applied Psychology.

Etzion, D. (1984). Moderating effect of social support on the stress-bumout relationship.

Journal of Applied Psychology, 69, 615-622.
Eulberg, J .R., Weekley, J .A., & Bhagat, RS (1988). Models of stress in organizational

research: A metatheoretical perspective. Human Relations 41, 331-350.

 

Eysenck, HI. (1991). Dimensions of personality: 16, 5, or 3? Criteria for a taxonomic
paradigm. Personality and Individual Differences, 12, 773-790.

Fiske, S.T., & Goodwin, SA. (1994). Social cognition research and small group research,
a West Side Story or . . . . S_mall Group Research. 25. 147-171.

Fletcher, B.C., & Jones, F. (1993). A refutation of Karasek’s demand-discretion model of
occupational stress with a range of dependent measures. Journal of Organizational

Behavior 14, 319-330.

 

153

Folkman, S., & Lazarus, RS. (1988). Coping as a mediator of emotion. Journal of

Applied Psychology, 52, 466-475.
Foushee, H.C., Lauber, J.K., Baetge, M.M., & Acomb, DB. (1986). Crew factors la

ﬂight omrations: III. The oparational signiﬁcance of exppsure to short-haul air
transpgrt op_erations (NASA Technical Memorandum 88322). Moffett Field, CA:
National Aeronautics and Space Administration.

French, J., Caplan, R., & van Harrison, R. (1982). The mechanisms of job stress and

strain. New York: Wiley.

 

French, J.R.P., Jr., & Caplan, RD. (1973). Organizational stress and individual strain. In
A.J. Marrow (Ed.), The failure of success. New York: AMACOM.

Fried, Y., & Tiegs, RB. (1995). Supervisors’ role conﬂict and role ambiguity differential
relations with performance ratings of subordinates and the moderating effect of
screening ability. Journal of Applied Psychology, 80, 282-291.

Friedland, N ., Keinan, G., & Regev, Y. (1992). Controlling the uncontrollable: Effects of
stress on illusory perceptions of controllability. Journal of Personality and Social

Psychology, 63, 923-931.
Gabarro, J.J. (1990). The development of working relationships. In J. Galegher, R.E.

Kraut, & C. Egido (Eds.), Intellectual teamwork (pp. 79-110). Hillsdale, NJ:

Lawrence Erlbaum Associates.
Galegher, J., Kraut, R.E., & Egido, C. (1990). Intellectual teamwork: Social and
technological foundations of cooperative wglg. Hillsdale, NJ: Erlbaum

Ganster, DC. (1980). Individual differences and task design: A laboratory experiment.

Organizational Behavior and Human Performance, 26, 131-148.

154

Ganster, D.C., & Fusilier, MR. (1989). Control in the workplace. In C.L. Cooper & I.T.
Robertson (Eds.), International Review of Industrial and Organizational Psychology
(pp. 235-280). Chichester: Wiley.

Ganster, D.C., & Schaubroeck, J. (1991). Work stress and employee health. Journal pf
Management, 17, 235-271.

Ganster, D.C., Fusilier, M.R., & Mayes, B.T. (1986). Role of social support in the
experience of stress at work. Journal of Applied Psychology, 71, 102-110.

Gates, RM. (2002, May). A former CIA chief on “connecting the dots.” Time, 152 (21 ),
30.

Gladstein, D.L., & Reilly, NP. (1985). Group decision making under threat: The tycoon
game. Academy of Management Journal. 28. 613-627.

Goldberg, LR. (1993). The structure of phenotypic personality traits. American

Psychologist, 48, 26-34.
Goodell, H., Wolf, S., & Rogers, RB. (1986). Historical perspective. In S. Wolfe & A.J.

Finestone (Eds.), Occupational stress, health, and parfonnance at work. Littleton,
MA: PSG, Inc.

Goodman, P.S., & Shah, S. (1992). Familiarity and work group outcones. In S. Worchel,
W. Wood, & J.A. Simpson (Eds.), Group process and productivity (pp, 226-228).
Newbury Park, CA: Sage.

Gordon, J. (1992). Work teams: How far have they come? Training, 29, 59-62.

Gouran, D.S., & Hirokawa, R.Y. (1983). The role of communication in decision-making

groups: A functional perspective. In M.S. Mandler (Ed.), Communication in

transition (pp. 168-185). New York: Praeger.

155

Greenberg, J. (1993). Stealing in the nane of justice: Informational and interpersonal
moderators of theft reactions to underpayment inequity. Organizational Behavior and

Human Decision Processes 54, 81-103.

 

Greenberg, J. (1994). Using socially fair treatment to promote acceptance of a work site
smoking ban. Journal of Applied Psychology, 79, 288-297.

Grifﬁths, A. (1994). Editorial: Musculoskeletal upper-limb disorders and white-collar
work in the UK. Work__and Stress. 11. 1-5.

Gruenfeld, D.H., Martorana, P.V., & Fan, B.T. (2000). What do groups learn from their
worldliest member? Direct and indirect inﬂuence in dynamic teams. Organizational

Behavior and Human Decision Processes 82, 45-59.

 

Hackman, J.R., & Monis, CG. (1975). Group tasks, group interaction process, and group
performance effectiveness: A review and proposed integration. Advances in
Experimental Social Psycholggy,_8, 45-99.

Hackman, J .R., & Oldham GR. (1976). Motivation through the design of work: Test of a
theory. Organizational Behavior and Human Performance. 16, 250-279.

Hackman, J .R., & Oldham, GR. (1980). Work redesign. Reading, MA: Addison-Wesley.

Hall, J .K., Volpe, C.E., & Cannon-Bowers, J .A. (1992). Mitigating the effects of stress:

A look at potential team training strategies. Papar presented at the meeting of the
American Psychological Association, Washington, DC.

Hamilton, P., & Copeman, A. (1970). The effect of alcohol and noise on components of a

tracking and monitoring task. British Journal of Psychology, 61, 144-156.

Harnner, M., & Champy, J. (1993). Reengineering the cogp_oration: A manisfesto for

business revolution. New York: Harper Business.

156

Hammond, KR. (1965). New directions in research on conﬂict resolution. Journal of

Social Issues 21, 44-6.

 

Harris, J .R. (1991). The utility of the transactional approach for occupational stress
research. Joumal of Social_BtaraaIior and Personaﬂy,_§, 21-29.

Harris, J.R. (1995). An examination of the transactional approach in occupational stress
research. In R. Crandell & P.L. Perrewe (Eds.), Occupational stress: A handbooLLppa
M. Washington, DC: Taylor & Francis.

Hartigan, J .A., & Wigdor, AK. (1989). Fairness in employn_lent testing: Validity
gmalization. minority issues. and the general aptitude test battery. Washington, DC:
National Academy Press.

Hartwick, J ., Sheppard, B.H., & Davis, J.H. (1982). Group remembering: Research and
implications. In R.A. Guzzo (Ed.), Improving group decision making in organizations
(pp. 41-72). New York: Academic Press.

Hastie, R. (1986). Experimental evidence on group accuracy. In B. Grofrnan & G. Owen
(Eds.), Decision research (Vol. 2. pp. 129-157). Greenwich, CT: JAI Press.

Hastie, R., & Pennington, N. (1991). Cognitive and social processes in decision making.
In L.B. Resnick, J .M. Levine, & S.D. Teasley (Eds.), Persmtives on socially shared
cognition (pp. 308-327). Washington, DC: Anerican Psychological Association.

Henry, RA. (1993). Group judgment accuracy : Reliability and validity of post-
discussion conﬁdence judgments. Organizational Behavior and Human Decision

Processes. 56. 11-27.

157

Henry, R.A., Strickland, O.J., Yorges, S.L., & Ladd, D. (1996). Helping groups
determine their most accurate member: The role of outcone feedback. Journal of
Applied Social Psychology, 26, 1153-1170.

Herriot, P. (1993). Commentary: A paradigm bursting at the seams. Journal of
Organizational Behayior. 14. 371-376.

Hesketh, B., & Neal, A. (1999). Technology and performance. In D. Ilgen & E. Pulakos
(Eds.), The changing nature of performance: Implications for staff'mg, motivation,
aad development (pp. 21-55 ). San Francisco: Jossey-Bass.

Hinsz, VB. (1990). Cognitive and consensus processes in group recognition nemory
performance. Jourﬂl of Personality and Social Psychology, 59, 705-718.

Hinsz, VB. (1995). Mental models of groups as social systems: Considerations of
speciﬁcation and assessment. S_mall Group ResearchJ6. 200-233.

Hinsz, V.B., Tindale, R.S., & Vollrath, D.A. (1997). The emerging conceptualization of
groups as information processors. Psychological Bulletin, 121, 43-64.

Hinsz, V.B., Vollrath, D.A., Nagao, D.H., & Davis, J.H. (1988). Comparing the structure
of individual and small group perceptions. International Journal of Small Group

Research 4, 159-168.

 

Hirokawa, R.Y., & Scheerhom, DR. (1986). Communication in faulty group decision-
making. In R.Y. Hirokawa & M.S. Poole (Eds.), Comunication and gpoup decision
m. Beverly Hills, CA: Sage.

Hobfall, SE. (1988). The ecology of stress. Washington, DC: Hemisphere.

Hobfall, SE. (1989). Conservation of resources: A new attempt at conceptualizing stress.

Amerign Psychologist. 44. 513-524.

158

Hockey, G.R.J. (1970a). Effect of loud noise on attentional selectivity. Quarterly Journal
of Exparimental Psychology, 22, 26-28.

Hockey, G.R.J. (1970b). Signal probability and spatial location as possible bases for
increased selectivity in noise. Quarterly Journal of Experimental Psychologyng, 37-
42.

Hockey, G.R.J., & Hamilton, P. (1970). Arousal and information selection in short-term
memory. Nature. 226. 866-867.

Hockey, G.R.J., Dornie, S., & Hamilton, P. (1975). Selective attention during reading:
The effect of noise. University of Stoakholm Institute of Applied Psycholagy
Rep_orts, 66.

Hockey, R. (1979). Stress and the cognitive components of skilled performance. In V.
Hamilton & D.M. Warburton (Eds.), Human stress and cogﬁnition: An informagion
processing approach (pp. 141-178). New York: Wiley.

Hogan, R. (1986). _M_an_pal for the Hgan Personality Inventopy. Minneapolis, MN:
National Computer Systems.

Hoiberg, A. (1982). Occupational stress and illness incidence. Journal of Occupa;i_o_na_l

Medicine 24, 445-451.

 

Hollenbeck, J .R., LePine, J .A., & Ilgen, DR. (1996). Adapting to roles in decision-
making teams. In KR. Murphy (Ed.), Individual diﬁeremeaand behavior in
organizations (pp. 300-333). San Francisco, CA: Jossey-Bass.

Hollenbeck, J.R., Moon, H., Ellis, A.P.J., West, B.J., Ilgen, D.R., Sheppard, L., Porter,

C.O.L.H., & Wagner, J .A. (2002). Structural contingency theory and individual

159

differences: Examination of external and internal person-team ﬁt. Journal of Applied
Psychology, 87, 599-606.
Hollingshead, AB. (1998). Communication, learning, and retrieval in transactive

nemory systems. Journal of Expprinental Social Psychology, 34, 423-442.

Hollingshead, AB. (1998). Retrieval processes in transactive memory systems. Journal
of Personality and Social Psychology, 74, 659-671.
Holsti, OR. (1971). Crisis, stress, and decision making. lntemational Social Science

Journal 23, 53-67.

 

House, J. (1981). Work stress and social supp_ort. Reading, MA: Addison-Wesley.

Hunter, J .E. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job
performance. Journal of Vocatimaal Behavior. 29. 340-362.

Hunter, J .E., & Hunter, RF. (1984). Validity and utility of alternate predictors of
performance. Psychological Bulletin. 96. 72-98.

Hurrell, J .J., & McLaney, M. (1989). Control, job demands and job satisfaction. In S.L.
Sauter, J.J Hurrel], & C.L. Cooper (Eds.), Job control and worker health (pp. 97-103 ).
Chichester, UK: Wiley.

Ickes, W., & Gonzalez, R. (1994). “Social” cognition and social cognition: From the
subjective to the intersubjective. Small Group Research. 25. 294-315.

Ilgen, DR. (1986). Laboratory research: A question of when, not if. In E.A. Locke (Ed.),
Generalizing from lab to ﬁeld settings (pp. 257-267). Lexington, MA: Heath.

Ilgen, D.R., & Moore, CF. (1987). Types and choices of performance feedback. Journal

of Applied Psychology, 72, 401-406.

160

Ilgen, D.R., Fisher, & Taylor, MS (1979). Consequences of individual feedback on
behavior in organizations. Journal of Applied Psychology, 64, 349-371.

Ilgen, D.R., Major, D.A., Hollenbeck, J.R., & Sego, DJ. (1993). Team research in the
19908. In M.M. Cheners & R. Ayman (Eds.), Leadership theogy and research (pp.
245-270). New York: Academic Press.

Innami, I. (1992). Determinants of the quality of group decisions and the effect of the
consensual conﬂict resolution. Academy of Management Best Pap_ers Proceedings,
217-221.

Isen, A.M., & Levin, AF. (1972). Effect of feeling good on helping: Cookies and
kindness. Journal of Personality and Social Psychology, 21, 384-388.

Isen, A.M., Clark, M., & Schwartz, M.F. (1976). Duration of the effect of mood on
helping: “Footprints on the sands of time.” Journal of Personality and Social
Psychology, 34, 385-393.

Ivancevich, J .M., & Matteson, M.T. (1980). Stressa and work. Glenview, IL: Scott,
Foresman.

Iwata, N., & Suzuki, K. (1997). Role stress-mental health relations in Japanese bank

 

workers: A moderating effect of social support. Applied Psychology: An International

Review 40, 207-218.

 

Jackson, S., & Schuler, R., (1985). A meta-analysis and conceptual critique of research
on role ambiguity and role conﬂict in work settings. Organizational Behavior and
Human Deci_sion Processes. 36. 16-78.

Jamal, M. (1984). Job stress and job performance controversy: An empirical assessment.

Organizational Behavior and Human Performarnce. 33. 1-21.

161

Janis, I. (1982). Groupthink (2" ed.). Boston: Houghton-Mifflin.
Janis, LL. (1972). Victims of ggoupthink. Boston: Houghton Mifflin.
Janis, I.L., & Mann, L. (1977). Decision making: A psychological anaysis of conﬂict.

choice and commitment. New York: The Free Press.

 

Jehn, KA., & Shah, RP. (1997). Interpersonal relationships and task performance: An
examination of mediating processes in friendship and acquaintance groups. Journal of
Personality and Social Psychology, 72, 775-790.

Jensen, AR. (1970). A theory of primary and secondary familial mental retardation. In
N.R. Ellis (Ed.), International Reviaw of Research in MenLaLRetaLdation (Vol. 4).
New York: Academic Press.

Jensen, AR. (1977). An examination of culture bias in the Wonderlic Personnel Test.

Intelligence,1, 51-64.

Jensen, AR. (1986). g: artifact or reality? Journal of Vocational Behaviom29. 301-331.

 

Jewell, L. (1998). Contempprmy industrial/organizational psychology (3rd ed. ). Paciﬁc
Grove, CA: Brooks/Cole.

Jex, SM. (1998). Stress and job mrformanceihaory. research. and implications for
managerial practice. Thousand Oaks, CA: Sage.

Jex, S.M., & Beehr, T.A. (1991). Energing theoretical and methodological issues in the
study of work-related stress. In G.R. Ferris & KM. Rowland (Eds.), Research in
personnel and haman resources management (pp. 311-364). Greenwich, CT: JAI.

Jick, T.D., & Murray, V.V. (1982). The management of hard times: Budget cutbacks in

the public sector. Organizational Studies, 3, 141-169.

162

Johns, G. (1981). Difference score measures of organizational behavior variables: A
critique. Organizational Behavior and Human Performance, 27, 443-463.

Johnson, J .V., & Hall, E. (1988). Job strain, workplace social support, and cardiovascular
disease: A cross-sectional study of a random sample of the Swedish working
population. American Jomf Public HeMthJ6. 1336-1342.

Kahn, R.L., & Byosiere, P. (1992). Stress in organizations. In M.D. Dunnette & I.M.

Hough (Eds.), Handbook of industrial and organizational psychology (2nd ed., Vol. 2,
pp. 571-650). Palo Alto, CA: Consulting Psychologists Press.

 

Kahn, R.L., Wolfe, D.M., Quinn, R.P., Snoek, J.D., & Rosenthal, RA. (1964).

Organizational stress: Studies in role conﬂict and ambigpity. New York: John Wiley.

Kanfer, R. (1987). Task-speciﬁc motivation: An integrative approach to issues of

neasurenent, nechanisms, processes, and determinants. Journal of Social and

 

Clinical Psychology, 5, 237-264.

Kanfer, R., & Ackerman, PL (1989). Motivation and cognitive abilities: All integrative /
aptitude-treatment interaction approach to Skill acquisition. Journal of Applied
Psychology, 74, 657-690.

Kanfer, R., Ackerman, P.L., Murtha, T.C., Dugdale, B., & Nelson, L. (1994). Goal
setting, conditions of practice, and task performance: A resource allocation

perspective. Journal of Applied Psychology, 79, 826-835.
Karasek, RA (1979). Job demands, job decision latitude, and mental strain: Implications

for job redesign. Administrative Science (Marterly, 24, 285-308.
Karasek, R.A., Jr., & Theorell, T. (1990). Healthy wok: Stress, productivity, and the

reconstruction of working life. New York: Basic Books.

163

Karau, S.J., & Kelly, J .R. (1992). The effects of tine scarcity and tine abundance on
group performance quality and interaction process. Journal of Exparimental Social
Psychology, 28, 542-571.

Kaufrnann, G.M., & Beehr, T.A. (1986). Interactions between job stressors and social
support: Sone counterintuitive results. Journal of Applied Psychology, 71 , 522-526.

Kaufmann, G.M., & Beehr, T.A. (1989). Occupational stressors, individual strains, and

social supports among police ofﬁcers. Human Relations 42, 185-197.

 

Keenen, A., & Newton, TJ. (1985). Stressful events, stressors, and psychological strains
in young professional engineers. Joumal of Occupaﬂal Behavior. 6. 151-156.

Keinan, G., Friedland, N ., & Ben-Porath, Y. (1987). Decision making under stress:
Scanning of alternatives under physical threat. Acta Psychologica, 64, 219-228.

Keita, G.P., & Sauter, S.L. (1992). Work and well being: An agenda for the 19908.
Washington, DC: American Psychological Association.

Kelly, J.R., & McGrath, J.E. (1985). Effects of tine limits and task types on task
performance and interaction of four-person groups. Journal of Personality and Social

Psychology, 49, 395-407.

 

Kim DH. (1993). The link between individual and organizational learning. Sloan
Management Review, 35, 37-50.

Klein, KJ., Bliese, P.D., Kozlowski, S.W.J., Dnasereau, F., Gavin, M.B., Griffin, M.A.,
Hofrnann, D.A., James, L.R., Yammarino, F.J., & Bligh, M.C. (2000). Multi-level
analytical techniques: Commonalities, differences, and continuing questions. In KJ.
Klein & S.W.J. Kozlowski (Eds.), Multilevel theog), research, and methods in

organizations (pp. 512-556). San Fransisco, CA: Jossey Bass.

164

Kleinman, D.L., & Serfaty, D. (1989). Team performance assessment in distrlbuted
decision making. In R. Gilson, J.P. Kincaid, & B. Goldiez (Eds.), Proceedinms for
Interactive Networked Simarlation for Training Conference (pp. 22-27). Orlando:
Institute for Simulation and Training.

Klimoski, R., & Mohammed, S. (1994). Team mental model: Construct or metaphor?
Jorml of Management, 20, 403-437.

Kluger, A.N., & DeNisi, AS (1996). Effects of feedback interventions on performance:
A historical review, a neta-analysis, and a preliminary feedback intervention theory.
Psychological Bulletin. 119. 254-284.

Kobasa, S. (1982). The hardy personality: Toward a social psychology of stress and
health. In G. Sanders & J. Suls (Ed.), Social psychology of health and illness (pp. 3-
32). Hillsdale, NJ: Lawrence-Erlbaum

Kozlowski, S.J.W., & Bell, BS. (in press). Work groups and teams in organizations. In
W.C. Borman, D.R. Ilgen, and R. Klimoski (Eds.), Comprehensive handboo_k_o_f
psychology ( 201. 12): Industrial and organizational psychology. New York: Wiley.

Kozlowski, S.W.J., & Klein, KJ. (2000). A levels approach to theory and research in
organizations: Contextual, temporal, and emergent processes. In KJ. Klein & S.J.W.
Kozlowski (Eds.), Multilevel theom, research, and nethods in prganizations. San
Francisco, CA: Jossey-Bass.

Kushmir, T., & Melamed, S. (1991). Workload, perceived control, and psychological
distress in Type A/B industrial workers. WEI—2, 155-

168.

165

Landsbergis, RA. (1988). Occupational stress among health care workers: A test of the
job demands-control model. Journal of Organizational BeMvior. 9. 217-240.

Larson, J.R., Foster-Fishman, P.G., & Franz, T.M. (1998). Leadership style and the
discussion of shared and unshared information in decision-making groups.
Personality and Somﬂ Psychology Bulletin, 24, 482-495.

Larson, J.R., Jr., Christensen, C., Abbott, A.S., & Franz, T.M. (1998). Diagnosing
groups: Charting the ﬂow of information in medical decision making teams. lgp_rn_al
of Personality and Social Psychology, 71, 315-330.

Larson, J.R., Jr., Foster-Fishman, P.G., & Keys, CB. (1994). The discussion of shared
and unshared information in decision-making groups. Joumal of Personality and
Social Psychology, 67, 446-461.

Laughlin, RR. (1980). Social combination processes in cooperative problem-solving
groups on verbal intellective tasks. In M. Fishbein (Ed.), Progr_'ess in social
psychology (pp. 127-155). Hillsdale, NJ: Erlbaum

Laughlin, P.R., & Ellis, AL. (1986). Demonstrability and social combination processes
on mathematical intellective tasks. Journal of Exmrimental Social Psychology, 22,
177-189.

Law, KS., & Wong, C. (1999). Multidimensional constructs in structural equation
analysis: An illustration using the job perception and job satisfaction constructs.
Journal of Management, 25, 143-160.

Law, KS., Wong, C.S., & Mobley, W.H. (1998). Toward a taxonomy of

multidinensional constructs. Academy of Management Review 23 741-755.

 

166

Lazarus, R.L., & Launier, R. (1978). Stress-related transactions between person and
environnent. In LA. Pervin & M. Lewis (Eds.), Peaspectives in intemm
psychology (pp. 287-327). New York: Plenum

Lazarus, RS (1966). Psychological stress and the coping process. New York: McGraw-
Hill.

Lazarus, R.S. (1991). Psychological stress in the workplace. Journal of SociA Behavior
and Perosnality, 6, 1-13.

Lazarus, RS. (1995). Psychological stress in the workplace. In R. Crandell & P.L.
Perrewe (Eds.), Occupational stress: A handbook (pp. 3-14). Washington, DC: Taylor
& Francis.

Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.

Lazarus, R.S., & Folkman, S. (1987). Transactional theory and research on emotions and
coping. In L. Laux & G. Vossel (Eds.), Personality in biographical stress and coping
research . 141-169 .

Lee, C., & Ashford, S., & Bobko, P. (1990). Interactive effects of "Type A” behavior and
perceived control on worker performance, job satisfaction, and somatic complaints.
Academy of Management Journal, 33, 870-881.

Lee, R.T., & Ashforth, BE. (1993). A further examination of managerial burnout:
Toward an integrated model. Journal of Organizational Behavior. 14. 3-20.

LePine, J.A. (1998). An integrative model of team adaptation. Unpublished doctoral

dissertation, Michigan State University.

167

LePine, J .A., Colquitt, J.A., & Erez, A. (2000). Adaptability to changing task contexts:
Effects of general cognitive ability, conscientiousness, and openness to experience.
Personnel Psychology, 53, 563-593.

LePine, J .A., Erez, A., & Johnson, DE. (2002). The nature and dimensionality of
organizational citizenship behavior: A critical review and neta-analysis. Journal of
Applied Psychology, 87, 52-65.

LePine, J.A., Hollenbeck, J.R., Ilgen, D.R., Colquitt, J.A., & Ellis, A. (2002). Gender
composition, situational strength, and team decision-making accuracy; A criterion
decomposition approach. Organizational Behavior and Human Decision Processes,
_8_8_, 445-475.

LePine, J .A., Hollenbeck, J.R., Ilgen, D.R., & Hedlund, J. (1997). Effects of individual
differences on the performance of hierarchical decision-making teams: Much more
than g. Journal of Applied Psychology, 82, 803-811.

Levi, L. (1981). Preventing work stress. Reading, MA: Addison-Wesley.

Levine, J .M., Resnick, L.B., & Higgins, ET. (1993). Social foundations of cognition.
Annual Review of Psychology, 44, 585-612.

Liang, D.W., Moreland, R.L., & Argote, L. (1995). Group versus individual training and

group performance: The mediating role of transactive nemory. Personality and Social

Psychology Bulletin, 21, 384-393.

Liang, D.W., Moreland, R.L., & Argote, L. (1995). Group versus individual training and

group performance: The mediating role of transactive nemory. Personality and Social

Psychology Bulletin, 21, 384-393.

168

Littlepage, G.E., & Silbiger, H. (1992). Recognition of expertise in decision-making

groups: Effects of group size and participation patterns. Small Group Research 23,

 

344-355.

Littlepage, G.E., Schmidt, G.W., Whisler, B.W., & Frost, AG. (1995). An input-process-
output analysis of inﬂuence and performance in problem-solving groups. Journal of
Personality and Social Psychology, 69, 877-889.

Locke, EA. (1969). What is job satisfaction? Organizational Behmm and Humaa
Performance. 4. 309-336.

Mandler, G. (1975). Memory storage and retrieval: Sone limits on the reach of attention
and consciousness. In P.M.A. Rabbitt & S. Dornic (Eds.), Attention and parformance
y. London: Academic Press.

Marks, M.A., Mathieu, J.E., & Zaccaro, SJ. (2001). A temporally based framework and
taxonomy of team processes. The Academy of Management Review, 26, 356-376.

Marks, M.A., Zaccaro, S.J., & Mathieu, J .E. (2000). Performance implications of leader
brieﬁngs and team-interaction training for team adaptation to novel environnents.
Journal of Applied Psychology, 85, 971-986.

Marquart, DJ. (1955). Group problem solving. Journal of Social Psychology, 41, 103-

1 13.

Martel], R.F., & Borg, MR. (1993). A comparison of the behavioral rating accuracy of

groups and individuals. Journal of Applied Psychology, 78, 43-50.

Martin, R., & Wall, T. (1989). Attentional demand and cost responsrbility as stressors in

shopﬂoor jobs. Academy of Management Journal, 32, 69-86.

169

Mason, J.W. (1975). A historical view of the stress ﬁeld. Journal of Human Stress 1 6-

 

12 and 22-36.

Mathews, KE., & Canon, L.K (1975). Environnental noise level as a determinant of
helping behavior. Journal of Personality and Social Psychology, 32, 571-577.

Mathieu, J.E., Heffner, T.S., Goodwin, G.F., Salas, E., & Cannon-Bowers, J.A. (2000).
The inﬂuence of shared nental models on team process and performance. Journal pf
Applied Psychology, 85, 273-283.

Matteson, M.T., & Ivancevich, J .M. (1982). The how, what and why of stress
management training. Paramel Journal. 61. 768-774.

McAuley, E., Duncan, T., & Tammen, V.V. (1987). Psychonetric properties of the
Intrinsic Motivation Inventory in a competitive sport setting: A conﬁrmatory factor
analysis. Research Quarterly for Exercise and Spart, 60, 48-58.

McClure, BA. (1990). The group mind: Generative and regressive groups. The Journal
for SpeciaSsts in Group Work. 15. 159-170.

McCrae, R.R., & Costa, P.T., Jr. (1989). The structure of interpersonal traits: Wiggins’s
circumplex and the ﬁve-factor model. Journal of Personality and Social Psychology,
M, 586-595.

McGrath, J.E. (1964). Social psychology. A brief introduction. New York: Holt,
Rinehart, & Winston. .

McGrath, J .E. (1976). Stress and behavior in organizations. In M. Dunnette (Ed.),

Handbook of Industrial and Organizational Paychology (pp. 1351-1326). Chicago:
Rand-McNally.

170

McGrath, J .E., & Hollingshead, AB. (1994). Groups interacting with technology.
Thousand Oaks, CA: Sage.
McIntyre, R.M., & Salas, E. (1995). Measuring and managing for team performance:

Lessons from complex environments. In R.A. Guzzo & E. Salas (Eds.), Team

 

effectivenesamd decision making in organizations (pp.9-45). San Francisco: Jossey-
Bass.

McKelvie, SJ. (1989). The Wonderlic Personnel Test: Reliability and validity in an
academic setting. Psychological Reports. 65. 161-162.

Meudell, P.R., Hitch, G.J., & Kirby, P. (1992). Are two heads better that one?
Experimental investigations of the social facilitation of nemory. Applied Cognitive
Psychology, 6, 525-543.

Miller, D.L., Young, R, Kleinman, D., & Serfaty, D. (1998). Distributed dymamic
decision-making simulation phase 1: Release notes and user’s manual. Woburn, MA:
Aptirna, Inc.

Mischel, W. (1977). On the future of personality neasurement. American Psmhologist,
32, 246-254.

Mitchell, R. (1986). Team building by disclosure of internal ﬂames of reference. 1119
Journal of Applied Behavioral Science, 22, 15-28.

Moreland, R.L. (1999). Transactive nemory: Learning who knows what in work groups
and organizations. In L.L. Thompson, J.M. Levine, & D.M. Messick (Eds.), Shared

cognition in organizations: The management of knowledge (pp, 3-32). Mahwah, NJ:

Erlbaum

 

171

Moreland, R.L., & Levine, J .M. (1992). Problem identiﬁcation in groups. In S. Worchel,
W. Wood, & J.A. Simpson (Eds.), Group process and productivity (pp. 17-47).
Newbury Park, CA: Sage.

Moreland, R.L., & Myaskovsky, L. (2000). Exploring the performance beneﬁts of group
training: Transactive nemory or improved communication? Organizational Behavior
and Human DeciSion Processes. 82. 117-133.

Moreland, R.L., Argote, L., & Krishnan, R. (1996). Socially shared cognition at work:
Transactive nemory and group performance. In J.L. Nye & A.M. Brower (Eds.),
What’s social about social cognition? Research on socially shared cognition in small
groups (pp. 57-84). Thousand Oaks, CA: Sage.

Moreland, R.L., Argote, L., & Krishnan, R. (1998). Training people to work in groups. In
R.S Tindale, L. Heath, J. Edwards, E. Posavac, F.B. Bryant, Y. Suarez-Balcazar, E.
Henderson-King, & J. Myers (Eds.), Theory an_cL_research on small gpoups (pp. 36-
6_0). New York: Plenum

Morgan, B.B., Jr., & Bowers, CA. (1995). Teamwork stress: Implications for team
decision making. In R.A. Guzzo & E. Salas (Eds.), Team effectiveness and decision
making in organizations (pp.262-290). San Francisco, CA: Jossey-Bass.

Mount, M.K., & Barrick, MR. (1995). The Big Five personality dinensions:
Implications for research and practice in human resources management. Research in
Personnel and Hum Resources Management, 13, 153-200.

Mount, M.K, Barrick, M.R., & Stewart, G.L. (1998). Five-factor model of personality
and performance in jobs involving interpersonal interactions. Human Performance,

11, 145-165.

172

Moyle, P., & Parkes, K (1999). The effects of transition stress: A relocation study.
Journal of Organizational Be_h_avior. 20. 625-646.
Murnighan, J .K, & Conlon, DE. (1991). The dynamics of intense work groups: A study

of British string quartets. Administrative Science Qparterly, 36, 165-186.
National Transportation Safety Board. (1972). Aircmft accident report- Eastern Air

Lines. Inc. L-1011. N310EA. Miami, Florida, December 29, 1972. (Rep. No. NTSB-
AAR-73-l4). Washington, DC: Author.

 

National Transportation Safety Board. (1979). Aircraft accident repart: United Airlines,
Inc.. McDonnell-Douglas DC-8-61. N8082U. Portlmd. Oregon, December 2S. B78
(NTSB/AAR-79-7). Washington, DC: Author.

Naylor, J.C., Pritchard, R.D., & Ilgen, DR. (1980). A thmlrmf behavior in
organizations. New York: Academic Press.

Neisser, U. (1982). Memogy observed: Remembering in natural contextS. New York:
Freeman.

Newton, T.J. (1989). Occupational stress and coping with stress: A critique. Emma

Relations 42, 441-461.

 

Newton, T.J. (1995). Managing stress: Emotion and pgwer at work. Thousand Oaks, CA:

Sage.

 

Niehouse, O.L. (1987). Life after burnout. Business 37, 42—46.

Norman, D.A., & Bobrow, DB. (1975). On data-limited and resource-limited processes.

Cognitive Psychology, 7, 44-64.

Northwestern National Life (1991). Employee burnout: America’s newest epidemic.

Minneapolis, MN: Northwestern National Life Insurance Company.

173

Novak, JD. (1990). Concept maps and Venn Diagrams: Two netacognitive tools to

facilitate meaningful learning. Instructional Science 19, 29-52.

 

O’Driscoll, M., & Beehr, T. (1994). Supervisor behaviors, role stressors, and uncertainty
as predictors of personal outcomes for subordinates. Journal of Organizational
Behavior. 15. 141-155.

Oransanu, J ., & Salas, E. (1993). Team decision making in complex environments. In
G.A. Klein, J. Oransanu, R. Calderwood, & C.E. Zsambok (Eds.), Decision making in
action: Models and methods. Norwood, NJ: Ablex.

Oransanu, J .M., & Backer, P. (1996). Stress and military performance. In J.E. Driskell &
E. Salas (Ed.), Stress and human performance (pp. 89-126). Mahwah, NJ: Erlbaum

Oser, R., Prince, C., & Morgan, B.B., Jr. (1990). Differences in aircrew communication
content as a function of ﬂight requirement: Implications for operational aircrew
training. Posteapresented at the 34‘h Annual Meeting of the Human Factors Socie_ty,

Orlando FL.

 

Oser, R.L., McCallum G.A., Salas, E., & Morgan, B.B., Jr. (1989). Toward a deﬁnition
of teamwork: An aaalysis of cri_ti_c_a_l_taam behﬂiors (NTSC Technical Remrt No. 89-
9_1_8_). Orlando, FL: Naval Training Systems Center.

Ostroff, C., & Ford, J .K (1989). Assessing training needs: Critical levels of analysis. In
I.L. Goldstein (Ed.), Training and development in organizations (pp. 25-62). San
Francisco: Jossey-Bass. .

Parkes, K (1990). Coping, negative affectivity, and the work environnent: Additive and

interactive predictors of mental health. Journal of Applied Psychology, 75, 399-409.

174

Parkes, KR., Mendham, C.A., & von Rabenau, C. (1994). Social support and the
demands-discretion model of job stress: Tests of additive and interactive effects in

two samples. Journal of Vocational Behavior 44, 91-113.

 

Payne, J.W., Bettman, J.R., & Johnson, EL (1988). Adaptive strategy selection in
decision-nraking. Journal of Exparimental Psychology: Learning, Memog, and
Cognition, 14, 534-552.

Payne, R.L., & Fletcher, B. (1983). Job demands, support, and constraints as predictors of
psychological strain among schoolteachers. Journal of Vocational Behmtior. 22. 136—
147.

Pennebaker, J .W., Czajka, J.A., Cropanzano, R., & Richards, BC. (1990). levels of
thinking. Personality aad Social Psychology Bulletin, 16, 743-757.

Perrewe, PL. (1991). Handbook on job stress [special issue]. Journal of Social Behavior
and Personality, 6.

Poole, M.S., & Hirokawa, R.Y. (1986). Communication and group decision-making: A
critical assessment. In R.Y. Hirokawa & M.S. Poole (Eds.), Communicatiomamd
group decision-making (pp. 15-34). Beverly Hills, CA: Sage.

Porter, C.O.L.H. (2001). Singled out: The task and social implications of providing
individual level feedback in teams. Unpublished doctoral dissertation. Michigan Sta_ta
University. .

Porter, C.O.L.H., Hollenbeck, J.R., Ilgen, D.R., Ellis, A.P.J., West, B.J., & Moon, H.K.
(in press). Towards understanding backing up behaviors in work teams: The role of

personality and the legitimacy of need for backing up others. Journal of Applied

Psychology.

175

Poulton, E. (1978). Blue collar stressors. In C. Cooper & R. Payne (Eds.), Stress at work.
New York: John Wiley.

Prince, C., Chidester, T.R., Bowers, C.A., & Cannon-Bowers, J.A. (1992). Aircrew
coordination: Achieving teamwork in the cockpit. In R. Swezey & E. Salas (Eds.),
Teams: Their training and performance (pp. 329-353). Norwood, NJ: Ablex.

Quick, J.C., Murphy, L.R., & Hurrel], J.J., Jr. (1992). Stress & well beiagat work:
Assessments and interventions for occupational nental health. Washington, DC:
American Psychological Association.

Ree, M.J., & Earles, J.A. (1991). Predicting training success: Not much more than g.
Personnel Psychology. 44. 321-332.

Rentsch, J.R., & Hall, R]. (1994). Members of great teams think alike: A model of the
effectiveness and schema similarity among team nembers. Advames in
Interdisciplinary Studies of Work Teams. 1. 223-261.

Rentsch, J.R., Heffner, T.S., & Duffy, LT. (1993). Teamwork schema representations:
The role of team experience. Paper presented aghe annual meet'mgofmaSopiatyfp;
Industrial and Organizational Psychology, San Francisco, CA.

Resnick, L.B., Levine, J.M., & Teasley, SD. (1991). Persppctives on socially shared
cognition. Washington, DC: American Psychological Association.

Riecken, H.W. (1958). The effect of talkativeness on ability to inﬂuence group solutions .

 

to problems. Sociometry. 21. 309-321.
Rosenhan, D.L., Salovey, P., & Hargis, K (1981). The joys of helping: Focus of helping:
Focus of attention mediates the impact of positive affect on altruism Journal of

Persmrality and Social Psychology, 40, 899-905.

176

Rothstein, H.G. (1986). The effects of tine pressure on judgment in multiple cue

 

probability learning. Organizational Behavior apd Human Decision Processes 37 83-
92.

Rotten, J ., Olszewski, D., Charleton, M., & Soler, E. (1978). Loud speech, conglomerate
noise, and behavioral aftereffects. Journal of Applied Psychology, 63, 360—365.

Rouse, W.B., & Morris, N .M. (1986). On looking into the black box: Prospects and limits
in the search for mental models. Psychological Bulletin. 100. 349-363.

Ruiz-Prime, M.A., & Shavelson, R.J. (1995). Concept maps as potential alternative

assessments in science. Pamr presented at the meeting of AERA, San Francisco, CA.

Rumelhart, D.D., & Ortony, A. (1977). The representation of knowledge in nemory. In
R.C. Anderson & R.J. Spiro (Eds.), Schooling and the acquisition of knowledge (pp.
9_9_-_1_3S). Hillsdale, NJ: Lawrence Erlbaum Associates.

Russell, D., Altmaier, E., & van Velzen, D. (1987). Job-related stress, social support, and
burnout among classroom teachers. Journal of Applied Psychology, 72, 269-274.
Ryan, A.M. (1982). Control and information in the intrapersonal sphere: An extension of
cognitive evaluation theory. Journal of Personality and Social Psychology, 43, 450-

461.
Ryan, A.M., Koestner, R., & Deci, EL. (1991). Varied forms of persistence: When free-

choice behavior is not intrinsically motivated. Motivation and Emotion 15, 185-205.

 

Ryan, A.M., Mims, V., & Koestner, R. (1983). Relation of reward contingency and
interpersonal context to intrinsic motivation: A review and test using cognitive

evaluation theory. Journal of Personality and Social Psychology, 45, 736-750.

177

Salas, E., Dickinson, T., Converse, S.A., & Tannenbaum 8.1. (1992). Toward an
understanding of team performance and training. In R.W. Swezey & E. Salas (Eds.),
Teams: Their training and parformance (pp. 3-29). Norwood, NJ: Ablex.

Salas, E., Driskell, J .E., & Hughes, S. (1996). Introduction: The study of stress and
human performance. In J.E. Driskell & E. Salas (Eds.), Stress and hupnap
p_erformance (pp. 1-46). Mahwah, NJ: Erlbaum

Schaubroeck, J., & Merritt, DE. (1997). Divergent effects of job control on coping with
work stressors: The key role of self-efﬁcacy. Academy of Maaagement Journak40.

7 38-754.

Schaubroeck, J., Cotton, J ., & Jennings, K. (1989). Antecedents and consequences of role
stress: A covariance structure analysis. Jourmll of Orgamizatiomrl BQaVIOI‘. 10. 33-58.

Schuster, M.A., Stein, B.D., Jaycox, L.H., Collins, R.L., Marshall, G.N., Elliot, M.N.,
Zhou, A.J., Kanouse, D.E., Morrison, J.L., & Berry, SH. (2001). A national survey of
stress reactions after the September 11, 2001, terrorist attacks. New England Journal
of Medic'me. 345. 1507-1512.

Schwab, DP. (1980). Construct validity in organizational behavior. In B.M. Staw & L.L.
Cummings (Eds.), Research in organizational behavior (Vol. 2, pp. 3-43). Greenwich,
CT: JAI Press.

Seers, A., McGee, G.W., Serey, T.T., & Graen, GB. (1983). The interaction of job stress
and social support: A strong inference investigation. Academy of Management

Journal 26, 273-284.

 

Selye, H. (1956). The stress of life. New York: McGraw-Hill.

178

Selye, H. (1983). Tm stress concept: Past, present, and future. In C.L. Cooper (Ed.),
Stress research. New York: Wiley.

Serfaty, D., Entin, E.E., & Johnston, J.H. (1998). Team coordination training. In J.A.
Cannon-Bowers & E. Salas (Eds.), Making decisions under stress: Implications for
individual and team training (PP. 221-245). Washington, DC: American
Psychological Association.

Seyle, H. (1975). Confusion and controversy in the stress ﬁeld. Journal of Human Stress.
2, 36-44.

Sharit, J ., & Salvendy, G. (1982). Occupational stress: Review and reappraisal. ﬂumaa

Factors 24, 129-162.

 

Shaw, M. (1932). A comparison of individuals and small groups in the rational solution
of complex problems. American Journal of Psychology, 44, 491-504.

Shiffiin, R.M. (1976). Capacity limitations in information processing, attention, and
memory. In W.K Estes (Ed.), Handbook of leaming and cognitive processes (Vol. 4).
Hillsdale, NJ: Erlbaum

Shiﬂett, S. (1979). Toward a general model of snell group productivity. Psychological
Bulletin. 86. 67-79.

Smart, M., & Vertinsky, I. (1977). Designs for crisis decision units. Administrative
Science Quarterly. 22. 640-657..

Smith, M.J., Carayon, P., Sanders, KJ., Lirn, S.-Y., & LeGrande, D. (1992). Employee
stress and health complaints in jobs with and without electronic performance

monitoring. Applied Ergonomics. 23. 17-28.

179

Smither, J .W. (1998). Lessons learned: Research implications for performance appraisal
and management practice. In J.W. Srnither (Ed.), Performance appraisal: State of the
art in practice (pp. 537-547 ). San Francisco, CA: Jossey-Bass Publishers.

Smith-Jentsch, K, Salas, E., & Baker, DP. (1996). Training team performance-related
assertiveness. Personnel PsychologyAQ, 1 10-116.

Spector, PE. (1986). Perceived control by employees: A neta-analysis of studies

concerning autonomy and participation at work. Human Relations 39, 1005-1016.

 

Spector, PE. (1987). Interactive effects of perceived control and job stressors on

affective reactions and health outcomes for clerical workers. Work and Stress 1, 155-

 

162.

Spector, PE. (1992). A consideration of the validity and neaning of self-report measures
of job conditions. In C.L. Cooper & I.T. Robertson (Eds.), International review of
industrial and organizational psychology 1992 (pp. 123-151). Chichester, UK: Wiley.

Spector, P.E., Dwyer, D.J., & Jex, SM. (1988). Relation of job stressors to affective,
health, and performance outcomes: A comparison of multiple data sources. Journal of
Applied Psychology,7_3, 11-19.

Spielberger, CD. (1983). Manual for the State-Trait Anxiet Invent orm . Palo
Alto, CA: Consulting Psychologists Press.

Spielberger, C.D., & Reheiser, EC. (1995). Measuring occupational stress: The Job
Stress Survey. In R. Crandall & P.L. Perrewe (Eds.), Occupational stress: A

handbook (pp. 51-69). Washington, DC: Taylor & Francis.

180

Stasser, G. (1999). The uncertain role of unshared information in collective choice. In
L.L. Thompson, J .M. Levine, & D.M. Messick (Eds.), ﬂared cogn_MorLip
organizations: The management of knowledge (pp. 49-70). Mahwah, NJ: Erlbaum

Stasser, G., & Stewart, D. (1992). Discovery of hidden proﬁles by decision-making
groups: Solving a problem versus making a judgnent. Journal of Personality and
Social Psychology, 63, 426-434.

Stasser, G., & Titus, W. (1985). Pooling of unshared information in group decision
making: Biased information sampling during discussion. Journal of Personality and
Social Psychology, 48, 1467-1478.

Stasser, G., & Titus, W. (1987). Effects of information load and percentage of shared
information on the dissemination of unshared inforrmtion during group discussion.
Journal of Personality and Social Psychology, 53, 81-93.

Stasser, G., Stewart, D.D., & Wittenbaum GM. (1995). Expert roles and information
exchange during discussion: The importance of knowing who knows what. Journal of
Experinental Social Psychology, 31, 244-265.

Stasser, G., Taylor, L.A., & Hanna, C. (1989). Information sampling in structured and
unstructured discussions of three- and six-person groups. MW
Social Psychology, 57, 67-78.

Staw, B.M., Sandelands, L.E., & Dutton, J.E. (1981). Threat-rigidity effects in
organizational behavior: A multilevel analysis. AW,
501-524.

Steiner, ID. (1972). Group process and productivipz. New York: Academic Press.

181

Stevens, C.K, & Kristof, AL. (1995). Making the right impression: A ﬁeld study of
applicant impression management during job interviews. Journal of Applied
Psychology, 80(5), 587-606.

Stewart, D.D., & Stasser, G. (1995). Expert role assignment and information sanrpling
during collective recall and decision making. Journal of Personality and Social
Psychology, 69, 619-628.

Stout, R.J. (1994). Effects of planning on the shared understanding of team nember
information requirements and efﬁcient communication strategies. Unpublished
doctoral dissertation, University of South Florida.

Sundstrom E., & Altman, I. (1989). Physical environments and work-group
effectiveness. In L.L. Cummings & B. Staw (Eds.), Remarch in organizational
peaaag; (Vol. 11, pp. 175-209). Greenwich, CT : JAI Press.

Sundstrom E., McIntyre, M., Halﬂlill, T., & Richards, H. (2000). Work groups: From the
Hawthorne studies to work teams of the 19908 and beyond. Group Dynamics, 4, 44-
67.

Surry, J. (1968). Industrial accident research: A human engineering appraisal. Ontario,
Canada: Department of Labor.

Sutherland, V., & Cooper, C. (1986). Man and accidents offs_hore: W
among workers on oil and gas rigs. London: Lloyds List/Dietsmann.

Sutherland, V.J., & Cooper, C. (2000). Strategic stress management: An organizational
approach. London, Macmillan Press Ltd.

Svenson, O., & Maule, A]. (1993). Time pressure and stress in humaangMana

decision making. New York: Plenum Press.

182

Svenson, O., Edland, A., & Karlsson, G. (1985). The effect Of numerical and verbal
information and time stress on judgnents of the attractiveness of decision
alternatives. In L.B. Methlie & R. Sprague (Eds.), Knowledge representation for
decision supmrt systems. Amsterdam: North-Holland.

Taylor, D.W., & Faust, W.L. (1952). Twenty questions: Efﬁciency in problem-solving as
a function of size of group. Journal of Experimental Psychology,_4_~4, 360-368.

Taylor, M.S., Fisher, C.D., & Ilgen, DR. (1984). Individuals’ reactions to performance
feedback in organizations: A control theory perspective. In KM. Rowland & G.R.
Ferris (Eds.), Research in human resource managenent (Vol. 2, pp. 81—124).
Stanford, CT: JAI Press, Inc.

Teichner, W.H., Arees, E., & Reilly, R. (1963). Noise and human performance: A
psychological approach. Ergonomics, 6, 83-97.

Thorns, R, Moore, K.S., & Scott, KS. (1996). The relationship between self-efﬁcacy for
participating in self-managed work groups and the Big Five personality dimensions.
Journal of Organizational Belavior. 17. 349-362.

Tinsdale, RS. (1989). Group vs. individual information processing: The effects of
outcome feedback on decision making. Organizational Behavior and Human Decision
Processes. 44. 454-473.

Torrence, ER (1954). The behavior of small groups under the stress conditions of
survival. American Sociological Review. 19. 751-755. .

Tuckman, B. (1965). Developmental sequence in small groups. Psychological Bulletin,
6;, 384-399.

Tumulty, K. (2002, June). Inside Bush’s big plan. Time, 159 (24), 30-33.

183

Vaux, A. (1988). Social support theogy, research, and intervention. New York: Praeger.

Volpe, C.E., Cannon-Bowers, J.A., Salas, E., & Spector, PE. (1996). The irrrpact of
cross-training on team functioning: An empirical investigation. Human Factors. 38.
87-100.

Wachtel, PL. (1968). Anxiety, attention, and coping with threat. Joumal of Abnormal
Psychology, 73, 137- 143.

Walsh, J.P., & Ungson, GR. (1991). Organizational nemory. Academy of Management

Review 16, 57-91.

 

Watson, D., & Clark, L. (1984). Negative affectivity: The disposition to experience
negative aversive emotional states. chhological Bulletin. 96, 465-498.

Weeless, V.B., & Serpento, ST. (1982). An analysis of the Wondtﬂc Personnel Test
LForm A) as metor in hiring clerical parsonnel in institutions of higher educatiop.
Norgantown, West Virginia: West Virginia University, Ofﬁce of Institutional
Research.

Wegner, D.M. (1987). Transactive nemory: A contemporary analysis of the group mind.
In B. Mullen & G.R. Goethals (Eds.), Theories of group behavior (pp. 185-208). New
York: Springer-Verlag.

Wegner, D.M. (1995). A computer network model of human transactive nemory. Mal
Cognition, 13, 319-339.

Wegner, D.M., Erber, R., & Raymond, P. (1991). Transactive nemory in close

relationships. Journal of Personality and Social Psychology, 61, 923-929.

184

Wegner, D.M., Giuliano, T. & Hertel, RT. (1985). Cognitive interdependence in close
relationships. In W.J. Ickes (Ed.), Compatible and incommuible rtﬂationships (pp.
253-276). New York: Springer-Verlag.

Weick, KE., & Roberts, KH. (1993). Collective mind in organizations: Heedful

interacting on ﬂight decks. Administrative Science (Marterly, 38, 357-381.

Weiner, N. (1948). Cybernetics: Control and communication in the animal and the
machine. New York: Wiley.

Weiss, H.M., & Adler, S. (1984). Personality and organizational behavior. In L.L.
Cummings & B.M. Staw (Eds.), Research in organizational behavior (Vol. 6, pp. 1-
50). Greenwich, CT: JAI Press.

Wellens, AR. (1993). Group situation awareness and distributed decision making: From
military to civilian applications. In NJ. Castellan (Ed.), Individual and ggoup decision
m_akln_'_g. Hillsdale, NJ: Erlbaum

Weltman, G., Smith, J.E., & Egstrom GB. (1971). Perceptual narrowing during
simulated pressure-chamber exposure. Human Factors. 13. 99-107.

Westman, M., & Eden, D. (1992). Excessive role demand and subsequent performance.
Journal of Organizational Behavior. 13. 519-529.

Wilson, M., & Canter, D. (1993). Shared concepts in group decision making: A model for
decisions based on qualitative data. British Journal of Social Psychology, 32, 159-
172.

Wittenbaum GM. (1998). Information sampling in decision-making groups: The impact

of nembers’ task-relevant status. Small Group Research, 29, 57-84.

185

Wolf, S., & Wolff, H.G. (1943). Gastric function: An exparimental study of a man and

his stomach. New York: Oxford University Press.

Wonderlic & Associates (1983). Wonderlic mrsonnel test manual. Northﬁeld, IL: E.F.
Wonderlic and Associates.

Wright, P. (1974). The harassed decision maker: Time pressure, distractions, and the use
of evidence. Jourml of Marketing Research. 44. 429-443.

Xie, J.L. (1996). Karasek’s model in the People’s Republic of China: Effects of job
demands, control, and individual differences. Academy of Mamnt Joumal. 39.
6-59.

Xie, J.L., & Johns, G. (1995). Job scope and stress: Can job scope be too high? Academy

of Management Journal, 38, 1288-1309.

Yarney, D.A. (1992). The effects of dyadic discussion on earwitness recall. Basic and
Applied Social Psychology, 13, 251-263.

Zaleska, M. (1978). Individual and group choices among solutions of a problem when
solution veriﬁability is moderately low. Europaan Journal of Social Psychology, 8,
37-53.

Zander, A., & Quinn, R. (1962). The social environment and mental health: A review of

 

past research at the Institute for Social Research. Journal of Social Issues 18, 48-66.
Zemke, R. (1978). Team building: Helping people learn to work together. Training, 15,
23.
Zwick, R. (1988). Another look at interrater agreement. Psychological Bulletin, 103, 374-

378.

186

APPENDIX A: TABLES

187

Table 1

 

 

 

Summary of Assets and Tracks
Assets Tracks
Duration Speed Vision Power? Identify Speed Power Nature Need to
(in nrin.) rig? Disable
Assets
Tank 8:00 slow very yes (5) no
limited
Helicopter 4:00 medium limited yes (3) no
Jet 2:00 very far yes (1) no
fast
AWACS 6:00 fast very no yes
far
Tracks
D Slow none Friendly TK,HE,
JT
B Slow low(l) Enemy JT
C Slow med. Enemy HE
(3)
A Slow high Enemy TK

(5)

 

Notes: For vehicles: duration = amount of time a vehicle may stay away from the base
before needing to refuel, speed = how fast the vehicle travels across the gane screen,
vision = refers to the range of vision the vehicle has to both see and identify tracks, power
= the ability of the vehicle to engage enemy tracks. For tracks: nature = whether the track
is an enemy or friend, speed = how fast the track travels across the game screen, need to
disable = which of the vehicles can successfully engage the track.

188

Table 2

 

 

 

 

 

 

Team Member Specialties
Team Member Target Vehicle
DMl D (GO) AWACS
DM2 B (GS) Tank
DM3 C (G3) Helicopter
DM4 A (G1) Jet

 

 

 

189

 

Table 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Summary of Hypotheses
Hypothesis Supported?
Yes
H1: Acute Stressors 5 Shared Mental Models
Yes
H2: Acute Stressors F Transactive Memory
No
H3: Directory Updating ’ Shared Mental
Models
No
H4: Cognitive Ability b Shared Mental Models
No
H5: Cognitive Ability p Transactive Memory
Cognit'lvp Ability No
H6: Acute Stressors > Shared Mental Models
Transactive Memory
No
H7: Extraversion » Shared Mental Models
No
H8: Extraversion F Transactive Memory
Extraversion N 0
H9: Acute Stressors 5 Shared Mental Models
Transactive Memory
Yes
H10: Prior Shared a Shared Mental Models
Information
Partially
H11: Prior Shared p Transactive Memory
Information
Prior Sharellnformation , Partially
H12: Acute Stressors P Shared Mental Models

 

Transactive Memory

 

 

190

 

 

 

No

 

 

 

 

 

 

 

 

 

 

H13: Feedback > Shared Mental Models
No
H14: Feedback Transactive Memory
Feedback Partially
H15: Acute Stressors L Shared Mental Models
Transactive Memory
Yes
H16: Shared Mental Team Performance
Models
Partially
H17: Transactive Memory % Team
Performance
Partially

H18: Acute ___§ Shared Mental Models—p Team

Stressors

Transactive MemorL

Performance

 

 

191

 

Table 4

Means, Standard Deviations, and Intercorrelations Among Acute Stressor Manipulation

 

 

Checks
Variable Mean SD 1 2 3 4 5 6
1 . Pressure/Tension 4.46 1.31 --
2. Affective 2.80 .82 .44** --
(Negative)Thought
3. STAI State 1.89 .53 .64** .54** --
Anxiety
4. Time Pressure 3.99 1.70 .48** .41** .45** --
5. Threat 3.07 1.67 .38** .37** .49** .45** --
6. Overall Stress 3.86 1.45 .66** .42** .59** .64** .54** --

 

Note: N=396. **p < .01.

192

.8. v at 8V? 2.x... 1888.. 2682.
:88 8a _ e5 x8980.“ gamma: 33365 e8 o R68 83 Monaco-am dosage—£5 33% 3E .33 he ~ 93 comma-.85
38% 8E 32 8... o e38 83 cease-SE weane- eotm .885 not. ~ e5 82% e: he 0 cocoa 83 82% .houz ”Boz

 

 

 

cacao-tom
.- .Ik-v. :mv. 1am. 1mm. *mN. Slum. 2. mo. mor *mN. 3.9x- on; 3.- SOHN~
8584‘ 332
-- :8. $9.. 1.9.. :8. ram. 8. .8. 8.- 8. :8.- 88 3.9. 352838.:
55328on
-- 2.9.. $9.. *8. ram. 8. .8. 8. 2. :9..- 88 8.8 382 838.2
boEoE
- :8. :8. 1:. 8. :. 8. 2. :8.- ﬁ. 8. 2823.58
ﬁning-500
-- *8. 3.9.. 2. t. 8.- E. :9..- 88 and nausea
c3300:
-- 2. 8. 8.- 8. .9. :8.- _m.: on: Suwanee:
maﬁa:
-- 8. E. 8. 8.- :8.- s: 8.2 ego-.598
-- 8. 3. :. 8. on. 9.. 83886
.. GOUEOE
-- 2.- 8. 8.- on. 9.. Baamcoeae
-- 8.- 8. R. 88 come-seam .m
baa...
-- E.- 8.m 8.8 325.8.“
- on. 9.. noun .9
9 : 2 o w A o m a. m a 9 mm 5o: 235,

 

38.—- 98509»: E eons—2: 8395; mace? 2283:0885 Ea .EOUEEQ 3355 .9822

m 2an

193

Table 6

Effects of Stress, Situational, and Personal Characteristics on Shared Mental Model

Similarity and Shared Mental Model Accuracy

 

 

 

 

SMMS SMMA
Hierarchical Independent B Total R2 A R2 B Total R2 A R2
Step Variable

1 Stress -.47** .22** .22** -.53** .28** .28”

2 Cognitive .10 .23** .01 -.01 .28** .00
Ability

3 Stress x .04 .23** .00 .11 .29** .01
Cognitive
Ability

2 Extraversio .1 1 .23** .01 -.02 .28** .00
n

3 Stress x -.03 .23** .00 -.07 .29** .00
Extraversio
n

2 Shared .21* .26** .04* .21* .32** .05*
Information

3 Stress x .08 .27** .00 .16 .34** .01
Shared
Information

2 Feedback .07 .23** .01 .03 .29** .00

3 Stress x -.14 .23** .01 -.12 .29** .00
Feedback

 

Note: N=97. SMMS=Shared Mental Model Similarity; SMMA: Shared Mental Model

Accuracy. * p < .05. ** p < .01.

194

.8. v a r. .8. v a .. Eva.

 

 

 

 

.boaoz 0>UOQ§PHH2H¢ ”GOmudﬂmUuOOw-v R>Ogom "UM ”ﬂomudUO: GOUEOEHS .wﬂﬁdab bOuQOumn—HDQ .FOHZ ”DHOZ

Mount-eon

No. 3.0N. mNr No. .3.hN. «N..- .Pmo. 33.2. ON... oo- 3.o_ . 3. x mac-am m

00. 3.wN. co. #0. 3hN. 2 . 00. 3A: . No. 8. 3.2 . ﬁo. xoanvoom N
GOUM—HCOHEH
vogm

S. 3.0m. ON.- *3. from”. tum..- oo. 3; - wo. 5. 3.3. S .- x 325 m
conﬁned:

5. 3&N. S . +No. 3.0N. em - 5. 3.2. Kb.- No. 3.N~ . mm. madam N
GOmmHD>NHHKm

co. 3.xN. nor 8. 3.VN. we. co. 3.2 . mo.. 8. 3L a . no: x $0.5m m

oo- 3.wN. oo. oo. 3.?N. 3. co. 3.2 . we. 3. 3L _. mo. :23?“be N
has...
czaimoo

oo. 3.wN. 5. oo. .3.mN. No.. 00. 3.N~ . mor oo- 3._ - mo. x $0.5m m
5:5

8. 3&N. mo. ﬂo- 3.mN. 2 . No. 3.2. mm. 5. 3L _. no... 03::on N

3R-N. 3k-N. 3.Nm... 3%N. 3%N. 3.3V.- *soﬁ. 3.o_. 3an 3.2 . 3.2 . 3.Nm.- $0.5m m

mm .m .m a... 233, mam
NM 4 Rue-- m— «M < Rue-H m NM G Each m mm d BEE a Someone—OS 3020.835
2H UM .3 DO

 

 

 

 

boEoE 0383:“; 98 6385800
ESEDM £03822 cease-85 $5325 .088an no 8558820 Enema-em 98 4.83336 88% mo 88mm

N. 2an

195

Table 8

Eﬂects of Directory Updating, Information Allocation, and Retrieval Coordination on
Shared Mental Model Similarity and Shared Mental Model Accuracy

 

 

 

 

SMMS SMMA
Hierarchical Independent 8 Total A R2 B Total A R2
Step Variable R2 R2
.12
1 Directory .25** .25** .1 1 .22** .22**
Updating
Information . 14 20*
Allocation
Retrieval .39** .32**
Coordination
l Transactive .48** .23** .23** .46** .21** .21**
Memory

 

Note: N=97. SMMS=Shared Mental Model Similarity; SMMA: Shared Mental Model
Accuracy. * p < .05. ** p < .01.

196

Table 9

Effects of Directory Updating, Information Allocation, Retrieval Coordination,

Transactive Memory, Shared Mental Model Similarity, and Shared Mental Model
Accuracy on Team Performance

 

Hierarchical
Step

1

Independent
Variable

Shared
Mental Model
Similarity
Shared
Mental Model
Accuracy

Directory
Updating
Information
Allocation
Retrieval
Coordination

Transactive
Memory

Team
Performance

 

B Total R2 A R2

.08 .23" .23"

.40*

.10 .16** .16**

.18T

.26*

.39** .15** .15**

 

Note: N=97. SMMS=Shared Mental Model Similarity; SMMA: Shared Mental Model

Accuracy. 1‘p<.10 * p < .05. ** p < .01.

197

Table 10

Regression Analysis of Information Allocation, Retrieval Coordination, and Shared
Mental Model Accuracy as Mediators of the Relationship between Stress and Team

Performance

 

Variable

B B8

.ARF

 

Phase 1: Direct Effect of Stress on Team
Performance
Step 1
Stress

Phase 2a: Effect of Stress on Team Performance
After Controlling for Shared Mental Model
Accuracy
Step 1
Shared Mental Model Accuracy
Step 2
Stress

Phase 2b: Effect of Stress on Team Performance
After Controlling for Information Allocation and
Retrieval Coordination
Step 1
Information Allocation
Retrieval Coordination
Step 2
Stress

Phase 2c: Effect of Stress on Team Performance
After Controlling for Transactive Memory
Step 1
Transactive Memory

Step 2
Stress

.18**
-.43**

.22**
.34**
.26**
.24*

.15**
.ll
.17
.22**
—.31**

.15**
.23*

.22**
—.31**

.18**

.22**

.04*

.15**

.O7**

.15**

.O7**

 

Note: N=97. B is the standardized regression coefﬁcient from the full regression equation
with all the predictor variables. Increments for variables entered at the R2 signiﬁcance
levels are based on F tests for that step.* p < .05. ** p < .01.

198

Table 11

Supplemental Analyses of the Effects of Directory Updating, Information Allocation, and
Retrieval Coordination on Team Performance During the Initial 15-Minute Task

 

 

 

Team
Performance
Hierarchical Independent B Total R2 A R2
Step Variable
1 Directory .32** . 10* . 10*

Updating
Information o.08
Allocation
Retrieval .02
Coordination

 

Note: N=97. ** p < .01.

199

Table 12

Supplemental Analysis of the Effects of Directory Updating, Information Allocation, and
Retrieval Coordination on Shared Mental Model Similarity and Shared Mental Model
Accuracy During the Initial 15-Minute Task

 

SMMS SMMA

 

 

Hierarchical Independent B Total A R2 B Total R2 A R2
Step Variable R2

1 Directory .18“ .09* .09* 20* .08* .08*
Updating
Information -. 15 -. l 3
Allocation
Retrieval -.13 -.O9
Coordination

 

Note: N=97. SMMS=Shared Mental Model Similarity; SMMA: Shared Mental Model
Accuracy. *p<.10 * p < .05. ** p < .01.

200

APPENDIX B: FIGURES

201

 

 

Aoogotom
Snob
$8850
:83.

A

 

 

 

 

63m .mﬁmv

 

 

 

 

 

 

 

 

 

 

 

982:.
.2385 25.5

whammobm
38¢.

 

 

ttttttttttttttttttttttttttttttttttttttt hﬁmw .tttttttttttttttttttttttttttttttttttttttttttttttrttttttttttttttttttt
Quanwoom dosage?“ @828
83332230 Raoumamm
Amﬁm .ﬁm .2: .25
GAS J :6
3.8822 @3885; A ‘
.2252 352 Base > 3: .EV
$3805 88,—t
$3 .26

Ah: .0: .3.“ .mmv

 

 

 

Acoﬁhgabxm 5:31.. 25638
8558830 Raoﬂom

 

 

 

moSHosbm who? 3396me 5 38% mo E52 3552;

2 8:3

202

Figure 2

Generic Infomation-Processing Model

 

 

Processing
Objective

Information

 

 

J,

V Attention

 

 

 

 

 

 

 

 

 

 

Processing

Space

 

j' Encoding]

 

b------------

 

 

Work ——rStorlage ]
T

Feedback

 

 

J] Retrieval ]

 

l

 

 

 

 

 

 

Response k ..............................

203

'7

L----

 

Figure 3

The DDD Grid Includin Bases Vehicles and Tracks

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\\L_ 7A/

 

 

h\\\ +5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

204

Figure4

E fthelnt ' n e tressan Sharedlnfomlati o Rt' v
Cmm' gtion

 

 

No Stress Stress
Sins Level

 

 

 

205

FigureS
Eff t ft t

'n n andF

k 0 ormatio Alloc io

 

Information Allocation

 

 

No Stress Stress
Shes Level

-.—— lnd'lvithal Fastback
+ Team Fastback

 

206

 

Figure6

The Effect of Intera ' n Be 11 Shared Mental Model S' ' 't and S
Mental Model Accgacy on Tm Performance

 

, +High Similarity
+ Low Simila '

Team Performance

    

Low Accuracy High Accuracy
Shared Mental Model Accuracy

 

 

 

207

APPENDIX C
DISSERTATION PROTOCOL

208

Part 1. Study Overview

The present investigation seeks to explore the effects of acute stressors on
transactive memory and shared mental models in temporary project teams by taking an
information processing approach. The presence or absence of acute stressors represents
the 11min manipulation. However, feedback and the amount of shared information
possessed by each team member will also be manipulated. Therefore, a 2 (acute stressor
vs. no stressors) x 2 (negative individual feedback vs. negative team feedback) x 2 (high
level of shared information vs. low level of shared information) between team design will
be used.

Condition 1: Acute stressors x negative team feedback x high level of shared information
Condition 2: Acute stressors x negative individual (DM2) feedback x high level of shared
information

Condition 3: Acute stressors x negative team feedback x low level of shared information
Condition 4: Acute stressors x negative individual (DM2) feedback x low level of shared
information.

. Condition 5: No acute stressors x negative team feedback x high level of shared
information

Condition 6: N o acute stressors x negative individual (DM2) feedback x high level of
shared information

Condition 7: No acute stressors x negative team feedback x low level of shared
information

Condition 8: No acute stressors x negative individual (DM2) feedback x low level of
shared information.

Teams will be randomly assigned to one of the eight conditions. Within each
team, the four participants will be randomly assigned to one of four positions in the team
(i.e., DM1-4). Each team will go through a 40 minute practice task, one 15-minute
experimental task, and one 30 minute experimental task.

After the practice task, the shared information manipulation will be introduced
depending on condition. Each team member will be given their specialty sheet informing
them about their speciﬁc area of expertise during the task. For example, DM2 will know
that DM4 will be in charge of the jets and that the C target represents a G3. DMl will
also have information regarding the wave targets that will enter the screen during the
task. That information may be shared with the other team members, depending on the
condition they are in. The sheets labeled LS in the upper right corner are for the low
shared information condition and the sheets labeled HS in the upper right hand comer are
for the high shared information condition.

After the ﬁrst 15 minute experimental task, the feedback and acute stressor
manipulations will be introduced depending on the condition. Negative feedback will be
given to one individual team member or the team as a whole. If acute stressors are to be
manipulated, the clocks will be taken off each team member’s screen and the team
members will be informed that the experimenter will warn them at speciﬁc intervals
during the task when time is running short. They will also be informed that their credit is
not guaranteed and may be in jeopardy of they don’t perform up to par.

209

Part 2. Preparation for the Experiment and Participant Orientation

Also, before participants enter the lab, the experimenter should:

1. Put keyboards on top of monitors.
2. Wonderlic and short form NEO w/ small bubble sheet at each station.
3. Call up the training task using the steps below.

a

r—r tonnes?

. At the dos prompt at L0, type “con.” A blue controller window will appear. In

this blue con window:

Select “PilotOl” from the Team Name section (can actually select any name)

Select the training task (alekstrain) ﬁ'om the Expt # section J
Click on “Start New Game.”

Click “OK to overwrite log ﬁle?” if necessary

Highlight the xterm window, you should see “Start Five Locals, Please.”

Go to station DM1 and bring up the task-playing screen by entering Ll at the

command prompt.

Repeat step g at each station by entering L2-L4 at the command prompt

depending on the station.

As participants enter the laboratory, the experimenter should have the participants read
and sign the Consent Form, which will be used to award course credit.

Bring four participants into the laboratory and let them choose their own station.

 

210

Part 3. The Wonderlic and NEO

Welcome to the Teamlab. As we told you in your first section meeting for MGT 315,
you will be playing a command and control computer simulation as a team today.
However, before we get to the computer task, there are two short questionnaires
that you need to complete. Any answers you provide are conﬁdential, have no
impact on any credit you may receive for participating in this study, and have no
impact on your team’s performance.

The ﬁrst is called the Wonderlic, which is on your desk right now. Please take it out
and write your name and PID on the top somewhere. The Wonderlic is a timed, 12-
minute test of problem solving ability that is used to select employees for a variety of
jobs from civil service workers to investment bankers. If you look at section 4 on the
first page, you can see the types of questions that you will be faced with. There are
verbal questions, math questions, analytical reasoning questions, and others. There
are 50 questions total, and you need to try to complete as many as you can in the 12
minutes provided. Students rarely complete all 50 questions, so don’t worry if you
don’t make it all the way through. Please write your answers directly on the page in
the spaces that are provided. If there are no questions, please turn the page and
begin.

Using a stopwatch, make sure that they work on the Wonderlic for exactly 12 minutes.
Then pick up the Wonderlics and ﬁle them.

The second questionnaire is a short personality questionnaire called the NEO, which
is also on your desks. Please take one of the small bubble sheets and fill in your last
name and PID. The NEO consists of 60 statements that you either agree or disagree
with. Read each statement and then answer on a scale from l-strongly disagree to 5-
strongly agree. All answers go on the bubble sheet. This questionnaire is not timed,
so go through at your own pace.

211

Part 4. Binder Training

The experimenter should now instruct the participants to follow along as he/she covers
the information in the binders located at each station. The information that should be
covered is below. NOTE THAT YOU SHOULD REFER TO THE “TASK”
RATHER THAN THE “GAME”

Page 1 “Welcome to the team effectiveness lab”

Please take the binder at your station and open it up to the first page. This page
offers some background into the teamlab, so just follow along as I go through each
point. Our purpose here is to improve individual and team performance in
organizations. Who beneﬁts? Society, MSU, and, most importantly, you the
students. You get your course credit and experience in teams. Plus we will award
cash prizes to top performing teams. In fact, each team member in the highest
performing teams will receive a check for $100 at the end of the semester.

Page 2 “The Task”

The task you are going work on is called DDD. It’s on your screens right now, but
nothing’s going to happen until I start it up. Basically the task simulates a military
command and control context, where you own and operate various vehicles, such as
helicopters, jets, tanks, and radar planes. The object is to monitor restricted
airspace and prevent enemy vehicles from entering forbidden locations by detecting
them, identifying them, and attacking them if necessary. However, you must not
attack any friendly vehicles that are operating in the same locations. As a side note,
only the vehicles get destroyed, not any of the people. Everyone escapes safely.

Your time here will be broken down into three parts. The first part is what we are
doing now. I will give you a quick general overview of the task. Here I will explain
the task screen, the scoring, the vehicles you will operate and the targets you will
face. Next, I will give you an opportunity to practice working on the task for an hour
or so. This is where you will really learn the mechanics of the task and begin
interacting as a team. Finally, you will work on the actual task, which consists of a
lS-minute task followed by a 30-minute tasks. Your performance during the 15-
minute and 30-minute tasks will decide whether or not your team receives a cash
pnze.

Pages 4 The Task Screen and Scoring
This page shows you a diagram of your task screen. You can also refer to your
monitor screen if that’s more clear for you. There are two sections of the screen: the

task area which is the gridded section, and the report area which is shaded in blue.
The first thing we’re going to do is go over the task area.

212

 

The task area is basically one geographic area broken down into four separate
quadrants. Each of you has a home base inside your quadrant — that base shows up
in your binder as a black rectangle. If you look closely you will see your station’s
name inside, DMl, 2, 3, or 4. These names correspond to the ones posted above your
cubicles. Does everyone see where their base is? Also note that the writing above
your base is in a specific color. Your vehicles will also having writing in the same
color above them during the task. So DMl is red, DM2 is purple, DM3 is green, and
DM4 is yellowish-orange.

Now, within the grid system is a large green square, this is a restricted area which is
called “THE FORBIDDEN ZONE.” This is the area that you must keep enemies
from entering. Within the green Forbidden Zone is a red square. This is “THE
REALLY FORBIDDEN ZONE.” This is the area you DEFINITELY must keep free
from enemies. The way the task is scored is that, if an enemy were to enter the
Forbidden Zone - the green square -- you will lose 1 point per second from your
score. If an enemy were to enter the Really Forbidden Zone - the red square - you
will lose 2 points per second from your score.

So you’re going to want to destroy enemy targets as soon as they enter the
Forbidden Zone. In fact, if you do that, you’ll get 5 points. However, you have to be
careful not to destroy them before they get into the Forbidden Zone - if you do that
you’ll actually lose 25 points. You’ll also lose 25 points if you accidentally destroy a
friendly target, no matter where it is on the screen.

If you lose or gain points, it registers on the scoring bars, which are to the right in
the blue-shaded report area. If you notice, you have six different scoring bars-
offensive and defensive scoring bars for the individual, the group, and the team.
You don’t need to worry about the scoring bars for the group. You will only be
looking at the individual and team scoring bars during the task. Let’s say that an
enemy target enters the screen at the top left corner and makes its way into DM2’s
green zone. Since it is an enemy, DM2 will be losing 1 point per second on his/her
individual defensive score and 1 point per second on the team’s defensive score.
However, no one else’s individual defensive scoring bar will go down because it’s
only in DM2’s quadrant. As you can see, the defensive scoring bars are completely
full. That means defensive scores can ONLY go down. And they will go down
because you cannot shoot down any enemy target before it gets within the green

_ forbidden zone.

You can, however, gain offensive points. So if DM2 were to destroy that target
helshe would get 5 points on his/her individual offensive score and the team would
get 5 offensive points. However, none of the other individual offensive scoring bars
would go up because DM2 was the one who shot it down. If DM3 sent a vehicle to
DM2 to shoot the target down, DM3 and the team would get 5 points. You can go
anywhere on the screen and, as long as the target is an enemy and within the big
green box, you will get 5 points for shooting it down. However, if one of you screws

213

up and shoots down a friendly or shoots an enemy outside of the forbidden zone,
your offensive score is the one that drops by 25 points.

To reiterate the scoring, your offensive score is affected by attacking things. You
can either gain 5 points or lose 25 points. Your defensive score is affected by having
things in the green or red Forbidden Zones. You can either lose 1 point per second
or 2 points per second.

Now I’m going to quickly go over the rest of the blue-shaded report area. Above the
scoring bars is a clock, which tells you how much time has expired in the task.
Under the scoring bars is a black bar, which is called the busy signal. Basically,
when you launch something, it indicates the length of time your base requires before
it can launch again. Under the busy signal are a set of buttons. Again, when the
practice task starts, I’ll tell you what each of those buttons does. Under the buttons
is the message box. You will be able to talk as much as you want during the task, so
you will never need to use the message box. In fact, you should feel free to talk as
much or as little you like during the task. However, remember that there are four
members of your team, all of whom will be staring at a computer screen during the
entire task. So when you communicate with your teammates, be speciﬁc. Make sure
they know who you are talking to, who you are, and exactly what you want. For
example, if you are DM3 and you need to ask DM2 something, say “DM2? This is
DM3. Is that target in your red box an enemy?” Or, “I’m DM3 and I need DM2 to
send a helicopter to the upper left comer of my forbidden zone.” When everyone is
concentrating on what they are doing, they will often ignore communication within
the team unless it is speciﬁcally directed at them. And recognizing voices is
extremely difﬁcult. So tell them your station number as well. If you communicate
eﬁiciently and effectively with one another, you will have a much better chance at
getting the $100 cash prize. Finally, on the bottom of the screen are two long
windows: the report and confirmation windows. Basically, everything you do is
written in text in one of those windows. You will not use either of the windows
during the actual task, so don’t worry about them.

The last thing on your screens that we need to discuss is the black ring around your
base. Any black ring that you see on your screen is what’s called a detection ring.
Detection rings let you see targets on the screen. That means that the detection ring
around your base allows you to see targets on about 1/4 of the screen. Each team
member can see a different portion of the screen with his or her base. You will not
be able to see targets in your teammates’ detection rings. When a target enters your
detection ring, you will be able to see it, but you won’t know whether it is a friend or
foe. Targets initially enter as unidentiﬁed. In order to identify targets on the screen
or detect targets at points on the screen that aren’t covered by your base’s detection
ring, you need to use the vehicles that are stationed at your base.

Pages 6-7 The Vehicles

214

As you can see, there are four different types of vehicles. You each will have a
certain combination of these vehicles during the task. The vehicle on the left is the
jet (it says JT above it), on the top is the tank (TK), on the bottom is the helicopter
(HE), and on the right is the radar plane (AW), which is technically called an
AWACS plane. If you notice, all four vehicles have two rings around them. The
outer black ring works like the black ring around your base. So you can launch any
of your vehicles, send them into one of your teammate’s quadrants, and, as soon as
targets get within the detection ring of your vehicle, they will pop up on your screen.
But they will still pop up as unidentiﬁed. That is what the inner blue ring
surrounding your AWACS is for. Once they get within the blue ring of your
AWACS plane, you will be identify them as friendly or enemy by clicking on them a
couple of times with the mouse. If you notice, the AWACS is the only vehicle with a
blue identiﬁcation ring. You can ONLY identify targets as friendly or enemy with
the AWACS plane.

If you notice, the second ring around the tank, the jet, and the helicopter is red.
That is your attack ring. Once a target is inside the attack ring of the tank, the jet,
or the helicopter, you can try to shoot it down. You CANNOT shoot anything down
with the AWACS plane. And you can only shoot down speciﬁc targets with speciﬁc
vehicles, as shown in the power column in the chart on the facing page.

The tank has a power of 5, the helicopter has a power of 3, the jet has a power of 1,
and the AWACS has a power of 0. The tank is the most powerful vehicle and can
shoot down targets with a power of 5 or less, the helicopter can only shoot down
targets with a power of 3 or less, and the jet can only shoot down targets with a
power of 1 or less. So the power level of your vehicle must be EQUAL TO OR
GREATER THAN the power level of the target. The AWACS, as I said before, is a
radar plane so it doesn’t have any power to destroy enemies. There are also a
number of other differences between the various vehicles. You can see that the ring
diameter is different, so you can see things that are far away with the AWACS but
you have to be right next to them with the tank. Also notice the red line coming out
of each of the vehicles. That denotes the speed of the vehicle. The tank is the slowest,
while the jet is the fastest.

One thing you must remember is that you only get one shot with each vehicle, then
you are out of ammunition and you must return that vehicle to the base to reload.
After attacking, you MUST ALWAYS return your vehicle to your base in order to
reload. I’ll explain how to do that once the practice task starts.

The last thing you see regarding the vehicles is fuel time. The chart shows you how
much time you have to move each vehicle around the screen until they run out of
gas. For example, you have 8 minutes to move the tank around the screen. After the
8 minutes is up, the tank will turn into an X and it will automatically return to its
base. As soon as it gets back to the base you can launch it again for another 8
minutes. So if you see your vehicle X out during the task for no reason, you
probably just ran out of gas.

215

Pages 8-9 The Targets

If you look at the chart, you can see that there will be four different types of targets
entering the screen. The G0 is a friendly target, while the G1, G3, and G5 are all

enemy targets. However, the targets will not show up as G0-G5 on your screens.

If you look at the close-up of DM2’s quadrant, you can see what the targets are
going to look like. When targets ﬁrst enter your detection ring, which is the black
ring, they are going to show up looking like #216 in your notebook- a diamond with
a question mark in the middle. They will always look that way no matter if they are
a friend or foe. You have to manually identify them as friendly or enemy once they
enter the blue ring of an AWACS plane. As you can see, #209 has entered the blue
ring of DM2’s AWACS plane and DM2 has manually identiﬁed it. When targets are
identiﬁed, they show up as a box. Inside that box is what type of target it is. So you
can see that #209 is a B target. The B target corresponds to one of the G targets.
There are four diﬂ'erent symbols, A, B, C, and D, that represent G0, G1, G3, and
G5s.

Information about the power level of the A, B, C, and D targets, as well as the
vehicles at each team member’s base, is contained on the specialty sheets for each
base.

Give them the specialty sheets for the practice task

During the practice task, each team member’s specialty sheet will contain the exact
same information. Team members will not have speciﬁc areas of expertise. Each
team member will have one jet, one helicopter, one tank, and one AWACS plane at
his or her base and will know that the A, B, C, and D targets will represent ground
targets with power 0, 1, 3, and 5, respectively. During the lS-minute and 30-minute
task, team members will be given separate specialty sheets containing information
regarding each team member’s speciﬁc area of expertise, which will be discussed
later.

That means that, for the practice task, B, C, and D are enemy targets while the A
target is friendly, as you can see on your specialty sheets. So target #209, a B target,
is an enemy target with a power of 1. Looking at where it is on the screen, would you
want to shoot it down right now? . . . (Yes) You can see that there is a helicopter
next to target #209. Can the helicopter shoot it down? . . . . (Yes) What other vehicle
ir vehicles could you use to shoot it down? . . . . (A jet or a tank). So you can shoot
down #209 with any of your attacking vehicles. Which one would be the better
choice during the game and why? . . (The jet, because you would want to save your
helicopter and tank for targets that require more power). So if you want to be really
efﬁcient, you should try to match the power level of your vehicle with the power
level of the target. #205 has also has been identiﬁed by DM2 because it too turned

216

into a box on the screen. Would you want to shoot this target down? . . . . (No) Why
not? . . . (Because it is a friendly target) What about #212? Would you want to shoot,
that target down? . . . . (No) Why not? . . . . (Because it is outside of the forbidden
zones)

The last thing we have to talk about are the U targets, or the unknown targets.
During the 15-minute and 30-minute tasks, you will encounter four additional target
symbols, U+, UX, U-, and U#. They will appear on your screens as usual. That is, as
a diamond with a question mark in the middle. However, when you identify them,
they will not show up as A, B, C, or D inside the box. They will show up as U+, UX,
U-, and U#. Each will correspond to a G0, G1, G3, or G5. However, the power level
of the U targets will not be contained on any of the team members’ specialty sheets.
As a team, through trial and error learning, you need to ﬁgure out which is which,
especially since one of the U targets is friendly. How do you do that? Let’s say that
the UX enters DM2’s portion of the forbidden zone and DM2 notices that his or her
individual defensive score is NOT going down. DM2 would immediately know that
the UX target was what? ..... (a GO) If DM2 is losing points, and he or she sends a
TANK over to shoot it down, and is successful, what would DM2 know about the
UX? ..... (That it is a GS, G3, or G1) If DM2 sent a jet over to the UX and
successfully shot it down, what would DM2 know about the UX? ..... (That it’s a
61). Clearly you need to learn the power level of each U target in order to be
successful. This process will move much more quickly if team members
communicate with one another, especially since the power level of the U targets will
remain consistent through both tasks.

So during the tasks, you will encounter eight different target symbols: A, B, C, D,
U+, U-, U#, and UX. Each symbol corresponds to a G0, G1, G3, or G5. You will be
provided information regarding the power level of the A, B, C, and D targets, but
you will M be provided information regarding the power level of the U+, U-, U#, or
UX targets. Those you have to ﬁgure out through trial and error.

Page 10 A Successful Attack

So just to reinforce what we have been talking about so far, I want to reiterate what
constitutes a successful and an unsuccessful attack. To get ﬁve points you must
attack an ENEMY target ANYWHERE inside the big green box. Plus, it has to be
within the RED ring of your vehicle and you MUST have ENOUGH power.
Remember that, if you do shoot something, you must always return the vehicle to
your base before attacking again, because you lose all of your power after one
attack. If you look at your score and you lost 25 points, that means that you either
attacked something OUTSIDE of the forbidden zone or you attacked a FRIENDLY.
If the enemy is too far away or you don’t have enough power, you will be
unsuccessful, but your points will remain the same. You will only be losing time.

217

Part 5. Hands-On Training

You will be using the mouse to do everything in the task. The left button does pretty
much everything in the task. So when I say click on something, I mean left click. The
right click opens up speciﬁc windows on the screen, which you will see in a little bit.
Using the mouse, click on the start button on the right hand side of the screen in the
report area. It should say refresh if you clicked correctly.

(Click “start” on DMO’s computer).

Ok, so now you can see that the clock has started. The practice targets don’t come
up for a while yet, so we’re going to practice a few basic things. The ﬁrst thing we
are going to practice is zooming in- BUT DON’T DO ANYTHING YET. If there are
a lot of targets in one portion of the screen, you will ﬁnd it difficult to specify which
one you want to identify and which you want to attack. So you can zoom in on that
portion of the screen. Ok, everyone go ahead and click the “zoom in” button and
move your cursor back onto the task area- DON’T CLICK AGAIN UNTIL I TELL
YOU WHAT TO DO. So you can see that your cursors have changed into a weird
shape.

If you click on something, and your cursor changes, but you wanted to click
something else, you can go over to the cancel button next to the refresh button.
Everyone go ahead and click cancel. Your cursors are now back to normal.

Ok, go ahead and click on zoom in again and move the strange cursor to the exact
center of the grid, where the four Forbidden Zones meet. Now click on that point,
and KEEP THE BU'ITON CLICKED DOWN. Keeping the button down, drag your
mouse to the opposite corner of your Forbidden Zone. You should be creating a box
around your Forbidden Zone. Now let go.

(Make sure that each person does it correctly)

Ok, to zoom out you just have to click on the “zoom out” button in the report area.
Once you’ve done that, I want you to practice zooming in one more time. This time
zoom in on your portion of the red box in the center of the screen.

(Make sure each person does it correctly)

Remember the zoom in function during the task. Most people forget about it and it
makes the task much more difﬁcult.

Now we are going to do is launch something from your base. To do that you have to
open up a menu from your base. So put your cursor right on top of your base and
RIGHT click. You’ll see a long menu. Go all the way to the bottom and click on
“Info on Asset.” When you do that, your launch window will pop up.

218

(Make sure they have the correct window up).

Before I talk about what’s inside this box, I want to explain a couple of things about
these windows. This is UNIX, not Windows, so you CANNOT close the windows
using one of the little buttons in the upper right hand corner. If you do, it will exit
you out of the system and I will have to restart the task. The only way to close one of
these windows is to click on the “cancel” or “o ” buttons on the bottom.

In the launch window, you always want to look at the columns in the middle of the
window. The ﬁrst column says “Sub,” which lists the vehicle that are stationed at
your base. You can only launch things one at a time, so you want to pick out the one
vehicle you want to launch. Then look to the “Aboard” column. If it says yes in the
aboard column, you’re ok. If it says no, the vehicle is already on the screen and you
can’t launch something twice. After you make sure it says yes, move to the launch
column. The left pointing arrows don’t do anything. You always want to click the
right pointing arrows. In this case, I want you to launch the tank, so please click the
right pointing arrow for the tank. The arrow will highlight in black and then go
down and click the “ok” button. Your base will turn into an X for a few seconds and
your vehicle will appear next to it.

(Make sure everyone launches their tank correctly).

The detection and attack rings will appear around your tank as soon as you start
moving it, which we are going to do now. You can ﬁnd the Move option by right-
clicking on your vehicle. So everyone right click on you tank, then click on Move
Fast. Now your cursor looks like a plus sign. Take the plus sign, put it in someone
else’s quadrant, and click. The vehicle will move to the exact spot you just clicked
on. While it’s moving, right click on it again, get the phrs sign up, and put it
somewhere else on the screen to change its direction. You can also make it stop on
the screen. Right click on it again and click on stop, which is the fourth button down
in the menu.

(Make sure that they move their vehicles correctly).

During the task, you will have all four vehicles on the screen. Sometimes you will get
confused and forget which ones you used to attack with. Remember, you only get
one shot per vehicle. However, you can check how much power you have left by
right clicking on your vehicle and going down to “Info on Asset.” You will get a
square window in the middle of the screen. This window tells you two important
things. First, it tells you how much time your vehicle has left before it’s going to'
return to your base, indicated by the “time remaining for use.” Second, it tells you
how much power you have left. In this case, it says power 5 because you haven’t
attacked anything with your tank. Go ahead and click on “o ” to close this window.

Let’s pretend that you attacked a target with your vehicle, so you used your one
shot. That means that you have to return that vehicle to your base. To do that, you

219

have to right click on your vehicle again. This time go about half way down in this
box to where it says “Return.” Go ahead and click on “Return” and then click on
“ok” when the next box comes up. You’ll see that your vehicle automatically returns
to your base when you do that. You MUST use the “Return” option to return your
vehicle. You CANNOT just move your vehicle near your base.

Ok, now I want you to practice what we have been doing. I want you to launch all
four of your vehicles and then I want you to move each one to a diﬁ‘erent corner of
your Forbidden Zone. Remember, you can only launch one at a time. So launch one
and move it to one corner. Then, while the ﬁrst one is moving, go back and launch
another from your base. When you launch everything, pay attention to the diﬁ'erent
ring radiuses and the different vehicles speeds.

(Again, make sure they are doing everything correctly and wait for everyone to deploy all
their vehicles).

Ok, now go ahead and return your vehicles to your base.

(Make sure everything returns correctly).

 

 

 

10:00 Identifying a Friendly

 

Four targets will enter at the comers of the screen. Before they do, please launch
your AWACS. Once it is launched, move it to the corner of the forbidden zone
which you do not share with any of your teammates.

(Wait until they get their AWACS close to the target).

Once the target gets within the blue ring of your AWACS, you will be allowed to
identify it as either a friend or enemy. Right-click on the target, then choose the
“Identify” option. Now you identiﬁcation window will pop up in the middle of the
screen. This window will appear every time you want to identify a target.

(Make sure they have the identify window on their screens).

You always do the same thing to identify a target using this window- ﬁrst click on
O-the “Fused” button at the top of the window and then go down and click on “o ”

Now you can see that the target has an “A” in the middle of its box, so, looking at
your specialty sheet, you know that it’s a friendly target. Now look at your
individual defensive score -- it’s NOT going down. So that’s another way to
determine whether the target is a friendly or an enemy without having to manually
identify it.

Another option you have is to transfer the identity of that target to your teammates.
To do that, you right click on the target again and go down to “Transfer Info.” Then

220

click on “all linked DMs” in the next box that appears. You just transferred the
identity of that target to the rest of your team. So now when that target enters your
teammates DETECTION RING, it will appear as an “A” and not a question mark.
This does NOT make these targets automatically appear on everyone’s screens. The
target still MUST be in someone’s detection ring in order for them to see it.

Just leave your AWACS plane at the outside comer of your forbidden zones.

 

 

 

12:00 Practicing Identifying and Transferring Info

 

Another set of friendly ground targets will come into the corners of the screen.
Practice identifying and transferring that identity to your teammates. Watch your
defensive scores again.

 

 

 

13:00 Attacking Some G1 ’3

 

The next targets will come in from the corners again. Identify the targets with your
AWACS plane. You do not have to transfer the information. As you can see, they
are B targets. So launch your jets and move them close enough so that the target is
inside the jet’s attack ring.

(Make sure they are following 0k)

You attack something by right-clicking on your vehicle and choosing Attack. Once
you choose Attack your cursor will turn into a big black “X.” Put it directly over the
target and click. Your Jet and the target will then X-out and you will see a red
explosion and it will say “hit- +5 points.”

(Make sure everyone is following along. If they missed the ﬁrst target, they may be able
to catch one coming from another quadrant).

Always look for the ﬁre. If you don’t see the ﬁre and the hit sign, you did something
wrong. Now look at your offensive scores. Each individual score went up 5 and the
team offensive score went up 20. Since you attacked with your Jet you have to
ALWAYS remember to return it to your base. So go ahead and do that now. And
return your AWACS plane as well.

 

 

15:00 . Attacking Something Outside the Forbidden Zone

 

 

The next targets are going to enter at the corners again, but this time they are going
to stop just outside your forbidden zone. So launch your AWACS and move it close
to the target so you can identify it.

(Wait for everyone to identify it as an enemy).

221

Ok, do you want to attack this target right now? . . . . No you don’t because it’s
outside your Forbidden Zone, but you do want to watch it because it could come in
at anytime. However, just to show you what happens when you make a mistake, go
ahead and launch your helicopter and attack the target outside the Forbidden Zone.
Watch what happens.

(Make sure they all do it correctly).

So when you attack an enemy outside of the Forbidden Zone, or you attack a
friendly anywhere on the screen, you’ll get an X and it will say “error -25 points.”
Make sure you always look for the ﬁre when you attack something or you’ll start
losing a lot of points. Now that you used up your Helicopter’s power, you have to
return it to your base, so go ahead and do that before the next targets come onto the
screen.

 

 

 

18:00 Attacking Enemies Inside the Forbidden Zone

 

Four targets are going to come in from the comers again, but this time they are
going to sit just inside of your forbidden zone and you will be losing points off your
defensive score. So, launch you AWACS to go see what it is. Then launch an
attacking vehicle of your choice and go destroy the enemy target.

(Make sure they do everything okay).

Since you used your Jet to attack the target, what do you have to do with it now? . . .
Right, return it to your base.

 

 

 

20:00 Dealing With a Wave Attack

 

The next thing that you are going to have to deal with is a wave attack. Wave
attacks will occur at regular intervals during the task. When wave attacks occur,
one team member will get bombarded by a eight targets at once. I want everyone to
look at DM2’s screen ........ As you can see, DM2 is experiencing a wave attack
where eight targets are entering his or her forbidden zone. Wave targets will move
into the forbidden zone, where they will sit, causing you and your team to lose a lot
of points. To effectively deal with the wave attack, all the team members need to
help out. Here’s where your communication skill becomes very important. You can
see your teammates’ vehicles moving across the screen when they are helping you
out. So direct them to the comet targets, tell them where targets are, which target
numbers they should get, etc. Also keep them informed when targets are eliminated
so they don’t send their vehicles all the way over to you only to ﬁnd out that they
aren’t needed. Because you can only see what is within your rings, you need to
speciﬁcally direct other team members so that you can maximize your score. Right
now, I want each of you to practice some of these things by ﬁrst launching your
vehicles and then moving them into DM2’s zone.

222

 

5:00 Practice Time

 

 

 

For the next 20 nrinutes I’m going to let you work at the task on your own for
practice. This is a good time for you to get accustomed to doing the things that you
have learned and get used to working together as a team. Remember, when you are
communicating with one another, make sure that you tell people which DM you are,
which DM you are talking to, and speciﬁcally what you want. I will remain in the
room while you practice, so that I can answer any questions you may have or help
you if let me know you need help.

(The experimenter should now watch the team work at the task for the next 20 minutes
and answer questions and help the participants when asked. MAKE SURE that the team
members at least try to use station numbers when asking for or requesting speciﬁc pieces
of information).

(At 40 minutes, the quit box will come up on their screens indicating the end of the
practice task. Click on “terminate experiment” in the controller window).

223

Part 6. Before the lS-rninute task

Give them all a bathroom break while you set up the ﬁrst task:
a. At the dos prompt at LO, type “con.” A blue controller window will appear. In
this blue con window:
b. Select the appropriate Team Name based on list in the log book (e.g., c8t5a . .
. condition 8, team 5, ﬁrst task).
c. Select the ﬁrst task (alekslS) from the Expt # section. All teams will
complete this task ﬁrst.
(1. Click on “Start New Game.”
e. Highlight the xterm window, you should see “Start Five Locals, Please.”
f. Go to each station and enter LO-L4 at the command prompt.
When they all get back m the room, ask them if they have any questions regarding any of
the training.

We are almost ready to start the actual task. Before we do, there are several things
we need to discuss.

First, to help you be more effective at the task, I’m going to set three goals related to
your performance. If you keep these goals in mind, and try to meet them, it should
help your team perform well.

These goals are as follows:

You should:
1. Never allow enemy targets to remain in the restricted zone for more than 3
nrinutes. .

2. Try and engage enemy targets residing in the highly forbidden zone within 1
minute after entry.

3. Try not to shoot down more than 1 friendly target during the 45-minute task.

If you focus on doing the things I just mentioned, it will increase your chances of
earning the bonus prize of $100 each.

Second, as discussed earlier, team members will have new specialty sheets for the
tasks that count. Unlike the practice task, each team member will now have a
speciﬁc area of expertise. To create areas of expertise within the team, the vehicles
and knowledge about the targets will be split up among the team members.
Regarding the vehicles, each team member will be responsible for one type of
vehicle. That means that one team member will have four tanks, one team member
will have four helicopters, one team member will have four jets, and one team
member will have four AWACS planes stationed at their base. Regarding the four
targets, the A, B, C, and D targets will still correspond to G0, G1, G3, and G5, but
not in that order. Each team member will know the power level of exactly one
target. This information is contained on each of your specialty sheets.

224

 

Pass out the correct specialty sheets depending on condition (in the low shared
information conditions, they only get 1 piece of information regarding the wave targets)

Although each team member will individually know less about the task, collectively
the team will possess the same set of vehicles and will know the power of the A, B, C,
and D targets.

Finally, your specialty sheets contain information regarding the wave targets that
will be entering the screen at various points during the task. Your specialty sheets
tell which DM will be receiving the wave of targets, when and where the wave will
enter the screen, and what targets the wave will consist of. The wave targets
represent a signiﬁcant portion of the total number of targets during the task.
Therefore, how you deal with the wave targets will have a large influence on your
scores. The information contained in your specialty sheets can help you prepare for
the wave targets and should help you increase your score. You may be the only one
in possession of the wave target information provided on your specialty sheet or you
may share that information with your teammates.

That means that information and expertise is split up between the team members.
At the beginning of the task, you will not know what your teammates’ specialties
consist of. So you will not know who has what vehicle and who knows what each
target corresponds to. Therefore, you need to work together in order to perform
well as a team and have a chance at the bonus money.

You CANNOT write anything down during the task. That means that your
communication skills are of utmost importance. That also means that there will be a
lot of information that, as a team, you need to remember. However, each team
member does go_t need to remember all the information themselves. You should try
and utilize your teammates’ knowledge so that you don’t get confused during the
task. For example, let’s say that you are DM2 and you get a C target in your
forbidden zone. Maybe you don’t know what the C target represents, but you know
DM4 has it on their specialty sheet. Then ask DM4 speciﬁcally for the information
you need. “DM4, this is DM2. What was the power level of the C target?” That way,
you don’t have to overload your brain with information. You only need to have a
general idea of who knows what. If you try and retrieve information just by yelling
things out to the whole team, no one will know who you are talking to. In other
words, your comments will get lost or ignored because your teammates are
concentrating on their own duties. In addition, if you don’t specify who you are
talking to, all the team members will have to think about whether they have the
information you need, distracting them from what they are doing. This will really
hurt the efﬁciency and effectiveness of the team as a whole.

Let me give you another example. Let’s say you are DM3 and you have a target with

power 5 in your quadrant, but you only have the jets. In this situation, most teams
are very inefﬁcient. They ask “Who has the tanks?.” Then they have to say “Can

225

 

you come and help me. I’m DM3.” This takes time and blocks the lines of
communication within the team. Instead, you should know exactly who to direct the
request to. If DM4 has the tanks, say something like “DM4, this is DM3. I have an
enemy with power 5 in my forbidden zone.” If you can always use the same
communication format by giving your station number, the station number of the
person you are talking to, and very speciﬁc information, you will have a much better
chance at earning the $100 cash prize. The more you communicate this way, the
better you will do as a team. That means that you must try to get to know what each
team member can and cannot do. The only way you will learn this information is by
specifying who you are, who you are talking to, and exactly what you need.

 

Finally, remember that during both tasks, you will encounter U targets. When you
ﬁrst detect the U targets, they will look like any other target. In other words, they
will appear as a diamond with a question mark in the middle. However, when you
identify them, they will not appear as A, B, C, or D. They will appear as U+, U-, U#,
and UX. You must ﬁnd out, through trial and error learning, what each U target’s
power is if you hope to do well in the task. Each U target corresponds to a G0, G1,
G3, or G5. Their power remains the same across both tasks, so once you ﬁgure out
what they are, you should have no trouble dealing with them.

(The experimenter should then instruct the participants to press START in the report area
of the screen. Once all participants verify that they have REFRESH on their screens, the
experimenter should then press START at DMO to begin the task. The experimenter
should stay in the room to provide assistance in case there is a problem with the
computers as the participants complete the 30-minute task.)

During the ﬁrst task, the experimenter should remain seated in ﬁont of and watch the
DMO screen as to get a sense as to how the participants are handling the task. It is very
important that the experimenter appears to be watching the participants’ performance
since the experimenter will be provide performance feedback to one of the participants in
the experimental conditions. The participants need to believe that the feedback is
veridical. The experimenter also needs to stay in the room in order to record any
communication between the team members.

They CANNOT write anything down during either task, so take away any writing
utensils they may have on their desks.

226

 

Part 7. After the First lS-Minute Task

Immediately after the ﬁrst task, the experimenter should instruct the participants not
discuss their scores with each other as he or she records them. The experimenter should
check to ensure that all of the participants’ team scores match, the participants in the
north group have matching group scores, and the participants in the south group have
matching group scores. The experimenter should then press "TERMINATE
EXPERIMENT” at the controller window.

DO NOT read the following manipulations. Try to do it ﬁ'om memory.
You have now ﬁnished the ﬁrst lS-minute task. Before we start the ﬁnal 30-minute

task, there are a couple of things we need to discuss. The ﬁrst concerns your
progress toward the goals we set before the ﬁrst task.

 

Individual Feedback Manipulation (conditions 2, 4, 6, 8)

DM2, after watching your performance during the ﬁrst 15 minutes, it appears that
you have been having some difficulty with the goals we set earlier. In particular, you
allowed a number of targets to remain in m restricted zone (i.e., your portion of
the green box) for more than 3 minutes and several targets remained in m highly
restricted zone (i.e., your portion of the red box) for well over 1 minute, which is
much longer than we discussed. In addition, you have aheady shot down more than
1 friendly target in m quadrant. By not doing these things, 1% performance has
probably hurt the team’s chances at earning the bonus money.

Feedback Manipulation (conditions 1, 3, 5, 7)

After watching your team’s performance during the ﬁrst 15 minutes, it appears that
the team as a whole has been having some difﬁculty with the goals we set earlier. In
particular, the team as a whole allowed a number of targets to remain in the
restricted zone for more than 3 minutes and several targets remained in the highly
restricted zone for well over 1 minute, which is much longer than we discussed. In
addition, the team as a whole has already shot down more than 1 friendly target on
the screen. By not doing these things, your team’s collective performance has
probably hurt the team’s chances at earning the bonus money.

Acute Stressor Manipulation (conditions 1-4)

Each semester, we notice that there are a few teams that seem to lag far behind the
others. We feel that this is due to a lack of motivation among certain students.
Everyone gets credit for participating, so some students feel that they can slack off
during their 3-hour session in the lab. When one person doesn’t pull their weight in
the team, the other team members need to pick up the slack for them. If one person
isn’t taking care of his or her duties, the team’s performance will suffer. Therefore,
in order to make sure you give us your full attention and effort for the next 30

227

 

minutes, we are going to put a little pressure on you. Right now, there are almost
600 students enrolled in your MGT 315 course. We assume that you would prefer
not to be displayed in front of all your friends if you aren’t pulling your weight in
the lab today. Therefore, we are going to videotape your team’s performance. If
your team is one of the three lowest performers, Dr. Morgeson will show the tape to
the entire MGT 315 class the last week of the semester as an example of ineffective
team behavior. I don’t think any of you want everything you say picked apart by
Dr. Morgenson during class. So just remember that the camera will be on you for
the next 30 minutes.

Pass out fake release form and have them sign

We also may be putting some additional pressure on you by warning you at
frequent intervals if time is getting short and you have too many enemy targets in
the forbidden zones and you really need to hurry up. Often team members lose
track of the time and end up with large numbers of targets in their forbidden zones.
A little time pressure should help you remain extremely vigilant during the entire
30-minute task.

Finally, before we start the second task, I should let you know that each team
member will have the same area of expertise and will be responsible for the same
knowledge regarding the targets and the same set of vehicles. In addition, the U
target symbols will remain the same.

Call up the second task using the steps below.

a. At the dos prompt at LO, type “con.” A blue controller window will appear. In
this blue con window:

b. Select the appropriate Team Name based on list in the log book.

c. Select the ﬁrst task (aleks30) from the Expt # section. All teams will
complete this task ﬁrst.

d. Click on “Start New Game.”

e. Highlight the xterm window, you should see “Start Five Locals, Please.”

f. Go to each station and enter LO-IA at the command prompt.

(The experimenter should then instruct the participants to press START in the report area
of the screen. Once all participants verify that they have REFRESH on their screens, the
experimenter should then press START at DMO to begin the task. The experimenter '
should stay in the room to record communication between the team members)

For conditions 1-4, read the following statement (you can alter it a bit so it doesn’t seem
so static) at the following times: 10, 15, 20, 22, 24, 26, 28

You now have only xx minutes left to work on the task, which is not a lot of time. In

order to perform well, you need to hurry up and work harder at keeping your
forbidden zones free from enemy targets.

228

Part 8. After the Second 30—Minute Task

You are done working on the task now. The last thing you need to complete is the
post-game questionnaire. Please ﬁll in your name and PID on the front and begin
answering the questions until it says stop. The ﬁrst set of questions asks you about
how you felt during the task and how you feel now after the task is completed.
Please take your time and answer all of the questions honestly. Only the
experimenters will see your answers to the survey questions.

While the participants are ﬁlling in the acute stressor manipulation checks, go ahead and
record their scores in the logbook. Once they all ﬁnish the ﬁrst section, take them through
the cognitive maps portion of the questionnaire.

Ok, the next portion of the questionnaire is a little more difﬁcult. Please turn the
page and I’ll explain what you need to do. As you can see, there are a number of
empty bubbles connected to one another. The objective of the team is to clear DMl’s
zone of enemy target. Your job is to indicate the steps each team member needs to
go through in order to reach your objective by ﬁlling in the bubbles for each team
member using the options provided. So you need to think about each team
member’s area of expertise and what they need to do in this situation. For example,
in DMl’s ﬁrst bubble, you could write in “Attack A targets” if you think that’s what
DMl should do ﬁrst. Does everyone understand? ...... There are several of these
sheets that you have to complete. After you ﬁnish, there are a number of other
questions that you need to answer. The instructions for those questions are
provided, but if you have any questions, let me know.

The experimenter should now allow the team the remaining time to complete the post-
game questionnaire. When they ﬁnish, they can leave the lab. However, before they
leave, the experimenter should thank each participant and ask them not to talk with
anyone about the experiment. The experimenter should also debrief the participants as
follows:

(In the acute stressors condition)

Before you leave, I just want to let you know that you will not see your team’s
performance replayed during the last class of the semester. After looking at your
performance across both tasks, there is almost no chance that your team will be one
of the three at the bottom of the list. You all put forth a lot of effort and that is
reflected in your scores, which, at the end of the semester, will likely fall well above
the average score for the class.

I also want to let you know that during the experiment, you may or may not have

received some information about how you or your team performed on the task. This
information may or may not have reflected your actual performance on the task.

229

 

However, only the actual performance of the team will be used to determine which
teams will be awarded the cash prizes.

Also, we would greatly appreciate it if you could do us a favor and not discuss the
experiment in detail with any of your classmates or anyone else that may be coming
into the laboratory during this semester to participate in the experiment. For us,
your discretion helps maintain the integrity of the experiment study. For you, you
will be competing against others who will work on the task, so the more insight you
provide others about the experiment, the more you reduce your chances for winning
the cash prize.

Thanks. We really appreciate your participation.

The experimenter should also address any questions/concerns participants have in regards
to the experiment.

After the participants leave, the experimenter should:
1. check to make that the scores have been recorded
2. collect all measures/scantrons and make sure all the necessary identifying
information has been provided
3. ﬁle all the information in the drawer labeled Aleks’ Dissertation.

230

 

APPENDIX D: COMNIUNICATION CODING

Team: Coder: Date/1‘ irne:

 

Time DMl DM3 Green DM4 ellow

0-2
2-4
4-6
6-8

 

Directory Updating

SE- sharing information about one’s own target or vehicle specialty (e.g., “I’m DM2 and
I have the tanks”)

RE- requesting information about someone else’s target or vehicle specialty (e. g., “Who
knows what the D target is?”)

Information Allocation

IA- allocating information to the person with the correct target or vehicle specialty (e. g.,
“DM3, I have several C targets in my restricted zone”)

Retrieval Coordination

RC- requesting information that is known to be part of someone’s target or vehicle
specialty (e.g., “DM3, what is the C target again?”)

23]

APPENDIX E: SHARED MENTAL MODELS MEASURE

Sﬂation #1: DMl is getting attacked by a wave of eight unidentiﬁed targets, some of
which are enemies. All eight ground targets slowly move through BM] ’5 restricted area
and stop inside DMl ’s highly restricted area (i.e., the red box). They are the only targets
on the screen. What are the actions that each team member needs to go through in order
to clear the enemy targets from the forbidden zones?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DM] |———_’
J L Enemy
J I wave
‘ targets
[ 1 f cleared
DM2 ﬁ'om
J L l a.
I I restrict
zones
DM3
J l
Identify Send Send
Tracks Helicopters Tanks
Send Attack Attack
Jets A Targets B Targets
Send Attack Attack
AWACS Planes D Targets C Targets

 

 

 

232

Situation #2: There are a number of targets on the screen. DM2 has 2 B targets inside his
or her restricted zone (i.e., the green box). DM3 has 2 C targets in his or her restricted
zone. DMl has 2 D targets in his or her restricted zone. DM4 has a wave of eight
unidentiﬁed targets about to enter his or her restricted zone. What are the actions that
each team member needs to go through in order to maximize the team’s score?

 

 

 

 

. 1H :55}
H l:
}_,

 

 

 

 

restrict
zones

 

 

 

DM3

1H H

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Send tanks Send helicopters Transfer Attack any D
to DM4 to DM4 identity of targets targets in DM4
Send jets Attack 2 Attack any A Ask for help
to DM4 C targets targets in DM4
Send Attack 2
AWACS planes D targets Attack any B
to DM4 targets in DM4
Identify Attack 2 Attack any C
targets B targets targets in DM4

 

 

 

233

APPENDD( F: AREAS OF EXPERTISE

DMl Role Description

Vehicle Specialty:

You will be in charge of four AWACS planes during both games

Target Specialty:

The D target represents a G0, or a friendly ground target, and this
will remain consistent across both games.

Task Information:

You will receive a wave of eight targets, two GOs, two G1 3, two
G33, and two G53, during the ﬁrst ten minutes of the second 30-
minute game. These targets will enter at the lower right-hand
comer of the screen and will proceed to the highly restricted area.

DM3 will receive a wave of eight targets, two GOs, two Gls, two
G3 8, and two G5s, during the last ten minutes of the second 30-
minute game. These targets will enter at the lower left-hand comer
of the screen and will proceed to the highly restricted area.

234

 

.83 00.0.08. 0...“... 0... 0. 00000.0 ...3 000 000.8 0... ..0 .08
0000-8. .0000 0... .0 8.00 ...3 0.0».0. 080... .080» 0.00.8.8 00008
0... .0 0.00. 0.00.8.8 0... 00003 .5 03. 000 do 03. .mm 03. .0... 03.
:0... 00».0. .0».0 .0 033 0 0>.000. 0. 08.00 0.0.. 0... 0.. ...3 ~20 .8

.83 00.0.08. 0...»... 0... 0. 00000.0 :00 000 000.8 0... .0 .008
0000-8. .030. 0... .0 .0.8 .03 0.0»3. 08...... .080» 0.00.8.8 00008
0... .0 .38 0.00.8-2 0... 00003 .8. 03. 000 .6 03. am. 03. .m< 03.

:0... 00»3. .0».0 .0 0.33 0 0.800. 0. 08.00 00008 0... 0.. :2. ms... .8

.83 00.0.08. 0...»... 0... 0. 00000... ...3 000 000.8 0... .0 .008
000..-.0»... .030. 0... .0 8.00 :2. 0.080. 080... .080» 0.00.8.8 00008
0... .0 x30. 0.00.8-3 0... 0000.0 .8. 03. 000 .00 03. .mm 03. .00. 03.

:0... 0.0»3. .0».0 .0 033 0 03000. 0. 08.00 .00. 0... 0.. ...>.. .20 .8
.83 00.0.08. .20»... 0... 0. 0800.0 .03 000 000.8 0... .0 80.00
0000.8. .0000 0... .0 .0.8 ...>. 00»3. 080... .080» 0.00.8.8 00008
0... .0 x30. 0.00.8-8 0... 00003 .8. 03. 000 .80 03. .mm 03. .0... 03.
:0... 0.0»3. .0».0 .0 033 0 0.800. 0. 08.0.. 00... 0... 0.. ...3 N29 .8

035

 

.83 00.0.08. 8%... 0.... 0. 00000... ...3 000 000.8 0... .0 .08
000..-...»0 .030. 0... .0 8.00 ...>. 0.0»3. 080... .080» 0.00.8.8 00008
0... .0 x38 0.00.8-8 0... 00003 .8. 03. 000 .mU 03. .mm 03. .0... 03.
:0... 0.0»3. 0.».0 .0 0.33 0 0.800. 0. 08.00 .00. 0... 0.. ...3 .35 .8
.83 00.0.08. .20»... 0... 0. 00000.0 .03 000 000.8 0.... .0 .008
000..-.0».. .0000 0... .0 .0.00 ...3 00»3. 08...... .080» 0.00.8.8 2.88
0... .0 0.08 0.008.... 0... 00003 .8. 03. 000 .00 03. .mm 03. .0... 03.
:0... 0.0»3. .0».0 .0 033 0 0>.000. 0. 08.00 8.... 0... 0.. ...3 02m. .8

.003 00.0.08. 30w... 0... 0. 00000.3...» 000 000.8 0... .0 .08
0000-8»: .030. 0... .0 8.00. ...>. 00».0. 80....- .080» 0.00.8.8 00008
0... .0 x38 0.00.8-8 0... 00003 .8. 03. 000 .00 03. .mm 03. .00. 03.

:0... 0.0»3. .0».0 .0 033 0 0>.000. 0. 08.00 .00. 0... 0.. ...3 .35 .8
.83 00.0.08. .3»... 0... 0. 00000.0 ...3 000 000.8 0... .0 .08
0000-8. .0000 0... .0 8.8 ...B 0.0»3. 080... .080» 0.00.8.8 00008
0... .0 0.38 0.00.8.8 0... 00003 .5 03. 000 .00 03. .mm 0B. .3. 03.
:0... 0.0»3. 0.».0 .0 0.33 0 0.800. 0. 08.00 0...: 0... 0.. ...3 N20 .8

$29

 

.83 00.0.08. D.W.: 0... 0. 00000.0 :05 000 000.8 0... .0 .08
0000.8. .030. 0... .0 8.8 ...B 0.0»3. 080... .003» 0.00.8-8 00008
0... .0 038 0.00.8-2 0... 00003 .09 03. 000 do 03. am 03. .00. 03.
:0... 890. .0».0 .0 0>03 0 03000. 0. 08.00 00008 0... 0.. ...>. min. .8
.83 00.0.08. 0...»... 0... 0. 00000.0 ...3 23 000.8 0... .0 .008
000..-.0».. .0000 0... .0 .300 ...3 0030. 08....- .080» 0.00.8.8 00008
0... .0 0.08 0.00.8... 0... 00003 .5 03. 000 .00 03. am 03. .0... 03.
:0... 0.0».0. .0».0 .0 033 0 0.800. 0. 08.00 8.... 0... 0.. ...3 055 .8

.83 00.0.08. 30m... 0... 0. 00000... ...>. 000 000.8 0... .0 80.00 2.0..
-00. .030. 0... .0 8.00 ...3 0.0»3. 800... .003» 0.00.8.8 00008 0... .0
0.38 0.00.8-2 0... 00003 .8. 03. 000 .00 03. .mm 03. .3. 03. :0...
0.0»3. .0»... .0 0.33 0 0>.000. 0. 08.00 00008 0... 0.. ...3 min. .8
.83 00.0.08. .3»... 0... 0. 00000.0 ...3 000 000.8 0... .0 .008
000..-...».. .0000 0... .0 .080 .03 0.0»3. 080... .080» 0.00.8.8 00008
0... .0 038 0.00.8... 0... 00003 .8. 03. 000 .80 03. .mm 03. .00. 03.
:0... 0.0».0. .0».0 .0 903 0 0>.000. 0. 08.00 .00.. 0... 0.. ...>.. 035 .8

N35

 

.003 00.0.08. .30»... 0... 0. 00000.0 ...3 000 000.8 0... .0 .008
0000-00. .030. 0.0 .0 .030 :2. 00»3. 800... .003» 0.00.8.8 00008
0.... .0 0.38 0.00.8-2 0... 00003 .8. 03. 000 .00 03. .mm 03. ......< 03.
:0... 0.0»3. .0».0 ..0 0.33 0 9.000. 0. 08.00 00008 0... 0.. ...3 min. .8
83 00.0.08. 0...»... 0... 0. 0080.0 :2. 000 000.8 0... .0 .008
0000-.0».. .0000 0.... 3 8.00 ...>. 0.0»3. 080... .080» 0.00.8.8 00008
0... ..0 2.08 0.00.8.0. 0... 0000.0 .8. 03. 000 .00 03. .mm 03. .0... 03.
:0... 0.0»3. .0».0 .0 0>03 0 2800. 0. 000.00 .0... 0... 0.. ....$ 0.20 .8

.83 00.0.08. 0...»... 0... 0. 00000.0 ...3 0.3 80.8 0... .0 .008 000..
-8. .030. 0... .0 8.8 ....$ 0.0».0. 08.5. .003» 0.00.8.8 00008 0... ..0
...08 0.00.8-2 0... 00003 .5 03. 000 .00 03. .mm 03. .2 03. :0...
0.0»3. .0»... .0 0>03 0 0>.000. 0. 08.00 00008 0... 0.. ...>. 9)... .8
.83 00.0.08. .2»... 0... 0. 00000.0 ...3 000 000.8 0... .0 .008
0000-.0».. .0000 0... .0 8.00 ...3 0.0»3. 08....- .080» 0.00.8.8 00008
0... .0 .38 0.088.... 0... 00003 .8. 03. 000 .00 03. an 03. .3. 03.
:0... 0.0».0. .0».0 .0 0>03 0 0>.000. 0. 000.00 .08 0... 0.. ....$ 035 .8

=20.

 

 

80008.... 020.0 .0 .93 33

 

80008.... .020... .o .25.. 00...

 

 

 

20:40.43; 20.90.5002. 923m .8 EQmeng.

235

APPENDIX H: ACUTE STRESSOR MAINPULATION CHECK
Pressure and Tension

For each of the following statements, please indicate how true it is for you, using the
following scale:

1 2 3 4 5 6 7
Not at all Somewhat Very
True True True

I felt very tense while doing this activity.

I was anxious while working on this task.

I felt pressured while doing this activity

I did not feel nervous at all while doing this task.
I was very relaxed while doing this task.

MPP’NE“

Affective (negative thoughts)

6. I became frustrated with my inability to improve my performance.
7. I thought about how poorly I was doing.

8. I was very satisﬁed with my overall performance on this task.

9. I got mad at myself during the task.

State Anxiety

A number of statements which people have used to describe themselves are listed below.
Reach each statement and then circle the appropriate number to the right of the statement
to indicate how you feel right now, that is, at this moment. There are no right or wrong
answers. Do not spend too much time on any one statement but give the answer which
seems to describe your present feelings best. Use the following scale:

 

Not at Very Much
All Somewhat Moderately So 80
< I I I I >
(1) (2) (3) (4)
10. I feel calm.

11. I feel secure.
12. I am tense.
13. I feel strained.
14. I feel at ease.
15. I feel upset.

236

 

16. I am presently worrying over possrble misfortunes.
17. I feel satisﬁed.

18. I feel frightened.
19. I feel comfortable.
20. I feel self—conﬁdent.
21. I feel nervous.

22. I am jittery.

23. I feel indecisive.

24. I am relaxed.

25. I feel content.

26. I am worried.

27. I feel confused.

28. I feel steady.

29. I feel pleasant.

Time Pressure and Threat
For each of the following 4 statements, please indicate how true it is for you, using the
following scale:

I I l l l l
I I I I I I

d—

1 2 3 4 5 6 7
Not at all Somewhat Very
True True True

30. I felt a lot of time pressure during this task

31. I felt stressed because there was not enough time to complete the task

32. The idea that other students may be aware of my performance on this task was
very stressful

33. I felt a lot of pressure to perform well on this task because there was a chance that
others could observe my behavior

 

 

 

Overall Stress
34. How stressed did you feel while playing the DDD command and control
simulation?
1 2 3 4 5 6 7
Not at all Somewhat Very
Stressed Stressed Stressed

237

.‘ﬁmka

 

‘19-?” ' I