IMPLICATIONS OF REGULATORY MODE AND FIT FOR GOAL COMMITMENT,
IDENTITY AND PERFORMANCE IN THE DOMAIN OF MULTITEAM SYSTEMS
By
Robert B. Davison

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Organizational Behavior – Human Resource Management
2012

ABSTRACT
IMPLICATIONS OF REGULATORY MODE AND FIT FOR GOAL COMMITMENT,
IDENTITY AND PERFORMANCE IN THE DOMAIN OF MULTITEAM SYSTEMS
By
Robert B. Davison
Individuals resident in multiteam systems face complex task environments which tax the
resources at their disposal. Performance is determined, in no small way, by how these limited
resources are allocated and by the degree to which they are brought to bear. This study
investigated the role played by regulatory mode, goal commitment and identity in determining an
individual’s incremental contribution to performance across the three levels of a multiteam
system. Findings from a laboratory study of 20 multiteam systems and 179 participants indicate
that a strong commitment to goals at the component team level and strong identification at that
level positively affect performance across all levels. Results also suggest that too strong an
identification with the multiteam system and too strong a commitment to the goals of the system
may serve, ultimately, to undermine performance. Implications for leaders of multiteam systems
and suggestions for future research are also discussed.

Copyright by
Robert B. Davison, PhD.
2012

DEDICATION

To my loving wife Maria, my children Rebecca, Sean and Teresa, and my grandchildren Oden,
Eulalia, Livinia, Jack, Ophelia and those yet to be: You are the wind in my sails, my reason for
being and my joy; eres mi vida. To John Hollenbeck my mentor, teacher and my friend: I doubt I
can ever really thank-you enough or truly repay you for all you have done. (But I will try.) To
Dan Ilgen: like John, my debt to you is large; thank you for everything. To Joy Elizabeth and
Nikos (and Ed): un pour tous, tous pour un. To my committee, Brent Scott, Joe Cesario and Russ
Johnson: Send me back to the beginning of this journey and I’d still pick each and every one of
you. To Olinda and Lou: Nothing I can say would ever suffice; I love you.

iv

ACKNOWLEDGEMENTS

I would like to gratefully acknowledge and thank Michael Howe for his assistance training the
participants in this study. In addition, I would like to thank the faculty, my fellow doctoral
students and the staff of the Department of Management at Michigan State for all the support and
encouragement they provided throughout my doctoral studies. I wish them only the very best in
their personal and professional lives; they deserve nothing less.

v

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................................... ix
LIST OF FIGURES .................................................................................................................... xii
Introduction................................................................................................................................... 1
Contextual Background: “Systems of Teams” or “Teams of Teams” ..................................... 6
Multiteam System (MTS) ........................................................................................................... 7
Goal structure of a multiteam system ................................................................................. 8
Key Construct Theories.............................................................................................................. 10
Goals ......................................................................................................................................... 10
Goal setting .......................................................................................................................... 10
Goal commitment................................................................................................................ 11
Goal striving ........................................................................................................................ 15
Multiple goals ...................................................................................................................... 16
Identity...................................................................................................................................... 19
Activated identity ................................................................................................................ 21
Role-identity ........................................................................................................................ 24
Self-regulation.......................................................................................................................... 26
Goal shielding ...................................................................................................................... 27
Regulatory mode ................................................................................................................. 29
Locomotion (aka “Just do it”) ......................................................................................... 30
Assessment (aka “Do the right thing”) ........................................................................... 31
Key research findings....................................................................................................... 32
Regulatory fit (aka “It feels right”)................................................................................... 34
Key research findings....................................................................................................... 36
Model Development and Hypotheses ........................................................................................ 39
Model Overview........................................................................................................................ 39
Hypotheses Derived from Theories of Self-regulation ........................................................... 42
Goal commitment as criterion ........................................................................................... 42
Identity as criterion............................................................................................................. 48
Performance as criterion.................................................................................................... 52
Relationships between Goal Commitment and Performance................................................. 55
Relationships between Identity and Performance .................................................................. 58
Mediation.................................................................................................................................. 60

vi

Method Section............................................................................................................................ 63
Participant Sample................................................................................................................... 63
Leadership Development Simulator (LDS)............................................................................. 63
LDS task environment........................................................................................................ 63
Common Operational Picture (COP)................................................................................ 66
MTS assets ........................................................................................................................... 68
MTS objective and scoring................................................................................................. 69
Specifics of the Laboratory Manipulation .............................................................................. 70
MTS structure and participant roles ................................................................................ 70
Component teams............................................................................................................. 72
Information team ............................................................................................................. 73
Goal structure...................................................................................................................... 74
Training ............................................................................................................................... 78
LDS scenario specifics ........................................................................................................ 78
Time Period 1 ................................................................................................................... 79
Time Period 2 ................................................................................................................... 79
Surveys ................................................................................................................................. 80
Overview of Analysis................................................................................................................ 81
Measures .............................................................................................................................. 82
Objective performance (criterion variable at level-1)..................................................... 83
Chronic regulatory mode (predictor variables at level-1)............................................... 85
Chronic self concept (control variables at level-1) ......................................................... 85
Goal commitment (predictor variable at level-1) ............................................................ 86
Identity (predictor variables at level-1) ........................................................................... 87
Enacted role (state) regulatory mode (manipulation check).......................................... 89
Results .......................................................................................................................................... 91
Measurement Model Goodness-of-Fit..................................................................................... 91
Test of Study Conditions ........................................................................................................ 100
Lack of Independence Tests .................................................................................................. 101
Results Involving Goal Commitment .................................................................................... 105
Goal commitment as criterion ......................................................................................... 105
Performance as criterion.................................................................................................. 114
Results Involving Identity ...................................................................................................... 123
Identity as criterion........................................................................................................... 123
Performance as criterion.................................................................................................. 127
Discussion................................................................................................................................... 132
Study Hypotheses ................................................................................................................... 132
Goal commitment as criterion ......................................................................................... 132
Identity as criterion........................................................................................................... 135
Performance as criterion.................................................................................................. 136

vii

Post Hoc Analyses.................................................................................................................. 138
Use of multilevel models for analyses.............................................................................. 138
Goal commitment as criterion ......................................................................................... 141
Identity as criterion........................................................................................................... 147
Performance as criterion.................................................................................................. 150
Role of goal commitment ............................................................................................... 150
Curvilinear models of goal commitment ....................................................................... 159
Role of identity ............................................................................................................... 164
Combined models of performance .................................................................................. 169
Limitations.............................................................................................................................. 173
Contribution to the Literature ............................................................................................... 176
Appendix A: Regulatory Mode Questionnaire (RMQ) ......................................................... 179
Appendix B: Levels of Self-concept Scale (LSCS) ................................................................. 180
Appendix C: Post-training Survey .......................................................................................... 181
Appendix D: Goal Commitment Measures ............................................................................ 182
Appendix E: Individual-level Identity Measure..................................................................... 184
Appendix F: Group-level Identity Measures.......................................................................... 185
Appendix G: State Regulatory Mode Questionnaire (based on the RMQ)......................... 187
REFERENCES.......................................................................................................................... 190

viii

LIST OF TABLES

Table 1. Confidence Indicator Codes............................................................................................ 74
Table 2. LDS Objective Measures ................................................................................................ 77
Table 3. Scale Reliabilities ........................................................................................................... 83
Table 4. Descriptive Statistics and Correlations (both conditions)............................................... 92
Table 5. Descriptive Statistics and Correlations (high state assessment condition only)............. 94
Table 6. Descriptive Statistics and Correlations (high state locomotion condition only) ............ 96
Table 7. Confirmatory Factor Analysis of Scales Employed in this Study .................................. 98
Table 8. Component team-level Independent-samples Means Tests of Regulatory Mode ........ 100
Table 9. Proportion of Goal Commitment Variance at the MTS-level....................................... 102
Table 10. Proportion of Identity Variance at the MTS-level ...................................................... 103
Table 11. Proportion of Performance Variance at the MTS-level .............................................. 104
Table 12. Models Examining Regulatory Mode – Goal Commitment Relationships in High
State Assessment Condition (Hypothesis 1, 3, & 5)................................................. 106
Table 13. Models Examining Regulatory Mode – Goal Commitment Relationships Employing
State Assessment Measure (Hypothesis 1, 3, & 5) ................................................... 107
Table 14. Models Examining Regulatory Mode – Goal Commitment Relationships in High
State Locomotion Condition (Hypothesis 2, 4, & 6) ................................................ 113
Table 15. Effects of Goal Commitment on Individual Role Performance in High State
Assessment Condition (Hypothesis 9, 11, 13, & 17)................................................ 116
Table 16. Effects of Goal Commitment on Component Team Performance Contribution in High
State Assessment Condition (Hypothesis 9, 11, 13, & 17)....................................... 117
Table 17. Effects of Goal Commitment on Multiteam System Performance Contribution in High
State Assessment Condition (Hypothesis 9, 11, 13, & 17)....................................... 118
Table 18. Effects of Goal Commitment on Performance Contribution in High State Locomotion
Condition (Hypothesis 10, 12, 14, & 18).................................................................. 119
Table 19. Effects of Other Goal Commitment on Multiteam System Performance Contribution
in High State Assessment Condition (Hypothesis 9, 11, 13, & 17).......................... 122
ix

Table 20. Models Examining Regulatory Mode – Identity Relationships in High State
Assessment Condition (Hypothesis 7) ...................................................................... 124
Table 21. Effects of Regulatory Mode and Identity on Individual Role Performance in High
State Locomotion Condition (Hypothesis 8, 16, & 20) ............................................ 125
Table 22. Effects of Regulatory Mode and Identity on Component Team Performance
Contribution in High State Locomotion Condition (Hypothesis 8, 16, & 20).......... 126
Table 23. Effects of Regulatory Mode and Identity on Component Team Performance
Contribution in High State Assessment Condition (Hypothesis 15, & 19) .............. 129
Table 24. Effects of Regulatory Mode and Identity on Multiteam System Performance
Contribution in High State Assessment Condition (Hypothesis 15, & 19) .............. 130
Table 25. Interrater Agreement Statistics for Contextual (Level 2) Variables (post hoc).......... 141
Table 26. Models Examining Regulatory Mode – Goal Commitment Relationships (post hoc)144
Table 27. Models Examining Regulatory Mode – Identity Relationships (post hoc) ................ 149
Table 28. Effects of Regulatory Mode and Goal Commitment on Individual Role Performance
(post hoc) .................................................................................................................. 154
Table 29. Effects of Regulatory Mode and Goal Commitment on Component Team Performance
Contribution (post hoc) ............................................................................................. 155
Table 30. Effects of Regulatory Mode and Goal Commitment on Multiteam System Performance
Contribution (post hoc) ............................................................................................. 156
Table 31. Exploratory HLM Models Examining Curvilinear Influence of Goal Commitment on
Individual Role Performance (post hoc) ................................................................... 161
Table 32. Exploratory HLM Models Examining Curvilinear Influence of Goal Commitment on
Component Team Performance Contribution (post hoc).......................................... 162
Table 33. Exploratory HLM Models Examining Curvilinear Influence of Goal Commitment on
Multiteam System Performance Contribution (post hoc) ......................................... 163
Table 34. Effects of Regulatory Mode and Identity on Individual Role Performance (post hoc)
................................................................................................................................... 166
Table 35. Effects of Regulatory Mode and Identity on Component Team Performance
Contribution (post hoc) ............................................................................................. 167
Table 36. Effects of Regulatory Mode and Identity on Multiteam System Performance
Contribution (post hoc) ............................................................................................. 168
Table 37. Models Combining Determinants of Performance (post hoc) .................................... 172
x

Table 38. Regulatory Mode Questionnaire (RMQ) .................................................................... 179
Table 39. Levels of Self-concept Scale (LSCS) ......................................................................... 180
Table 40. Post-training Survey ................................................................................................... 181
Table 41. Goal Commitment Measures ...................................................................................... 183
Table 42. Individual-level Identity Measure............................................................................... 184
Table 43. Group-level Identity Measures ................................................................................... 185
Table 44. State Regulatory Mode Questionnaire........................................................................ 187

xi

LIST OF FIGURES

Figure 1. Hypothesized Model...................................................................................................... 41
Figure 2. LDS Game Grid............................................................................................................. 64
Figure 3. Sample Common Operational Picture (COP)................................................................ 67
Figure 4. LDS Team Structure...................................................................................................... 71
Figure 5. Hierarchy of Goal Measures (with Performance Targets) ............................................ 76
Figure 6. Plot of Trait Locomotion x State Assessment Interaction........................................... 109
Figure 7. Plot of Trait Assessment x Trait Locomotion x State Assessment Interaction ........... 111
Figure 8. Plot of Trait Locomotion x Trait Assessment x State Assessment Interaction ........... 112

xii

Introduction
A sales team from Indigo Computer is preparing a proposal for a major new network
installation and the network engineer becomes so focused on the technical details that he fails to
properly brief the sales representative before the client meeting. The sales representative is
unprepared to answer all the questions raised by the client and they lose the sale. A sub-team
from Gulf Consulting redesigns a client’s supply chain optimizing it for cost but fails to align the
proposed changes with another sub-team developing a new business unit strategy based on
differentiation. The new supply chain is successful at reducing costs but fails to support the new
service-focused business model. The client loses market share. These are two examples of
countless business situations where individuals can become so committed to goals and identify
so strongly with their role at one level that they ignore the requirements of higher order goals and
thus fail to achieve them. The purpose of this dissertation is to investigate this phenomenon to
gain a better understanding of the factors involved.
Multiple goals and priorities is a not so uncommon fact of life in complex organizations,
yet little is known about the determinants of goal prioritization in these settings (Schmidt &
DeShon, 2007) and their subsequent effect on performance contribution. The hierarchy of goals
associated with complex organizational systems in combination with resource limitations (e.g.,
internal psychological resources such as cognition and external resources such as time) often
creates a situation whereby individuals can become so focused on a particular goal or set of goals
that they fail to focus on or outright ignore other important goals (Shah & Kruglanski, 2008).
This highlights an important research question: When faced with a hierarchy of goals, where
each level in the hierarchy is associated with a specific level in the organization (e.g., individual,
work group or department, function or business unit) and goals at each level are an

1

amalgamation of goals originating at that level and at levels lower in the hierarchy, what
determines an individual’s goal focus and, ultimately, their performance contribution? The
answer, I contend, lies in understanding the self-regulation of goal commitment and activated
identity.
Goal commitment is of interest because it has been linked to an unwillingness to abandon
or to lower an original goal and has been shown to inhibit consideration of alternative or
competing goals (Campion & Lord, 1982; Hollenbeck & Klein, 1987; Shah, Friedman, &
Kruglanski, 2002). Research has also clearly shown that goal commitment increases persistence
at and performance on focal tasks, often at the expense of alternative tasks (Schmidt & DeShon,
2007; Shah et al., 2002). Thus a conundrum clearly exists – goal commitment leads to higher
performance but does so by inhibiting alternative goals, yet the success of the system overall is
dependent on effectively addressing multiple goals that compete for the same limited resources.
This is exactly the situation illustrated in the opening paragraph. Individuals became so
committed to one level of goals that they ignored, and thus failed to adequately address other,
higher order goals resulting in negative performance consequences for the organization. Goal
commitment has not been adequately studied in the context of complex organizational systems
with an interconnected goal hierarchy, thus extant research cannot speak to the conundrum
facing individuals in these systems.
Understanding identity in the context of complex organizations is important as well
because the specific identity that is active at a given moment in time affects behavior, and thus
performance, through its influence on both cognitive and motivational mechanisms (Brewer &
Gardner, 1996; Ellemers, Spears, & Doosje, 2002). Yet as Farmer and Van Dyne point out,
“quantitative knowledge is particularly limited concerning identity-based action in organizational

2

contexts” and “little if any quantitative research has been conducted into how this affects role
performance in organizations” (2010, p. 505). Individuals in complex organizations have
opportunity to identify with their role, their sub-team, the organization overall, or some
combination of these, and the behavior resulting from activating these different identities may
affect the performance of the organization in important ways. For example, too strong an
identification with the role of supply chain expert could lead individuals to a focus on the lower
order goals of the supply chain sub-team at the expense of the project’s higher-order, superordinate goals. Thus activated identity is implicated in the decision dilemma articulated earlier
through its direct affects on behavior and performance.
Theory and empirical work in the area of self-regulation have shown that these processes
play a central and critical role in goal-directed behavior and performance (e.g., Shah et al.,
2002). Self-regulation is the psychological process by which an individual brings thinking and
behavior in line with some preexisting and desired set of standards, rules, norms, ideals, and/or
goals (Forgas, Baumeister, & Tice, 2009). “It is not unrealistic to propose that psychology has
thus far identified only two main variables that contribute to human success across almost the
full range of human striving. These are intelligence and self-regulation” (Forgas et al., 2009, pp.
5-6).
Regulatory mode orientation concerns the psychological state of goal pursuit. The
assessment dimension is implicated in processes of evaluation and selection (resource targeting)
while the locomotion dimension is implicated in the commitment of limited psychological
resources (Kruglanski et al., 2000). Individuals high on the assessment dimension are most
concerned with evaluating alternative goals and available means to decide which are best to
pursue while individuals high on the locomotion dimension are most concerned with movement

3

irrespective of outcome valence (Kruglanski et al., 2000). For example, the network engineer
may very well have been a high locomotion oriented individual, thus singularly focused on the
goal of network design because it offered ample movement despite the fact that the expected
value of this activity was nil unless the team closed the sale. Had this individual been stronger on
the assessment dimension, he or she may have chosen to allocate their limited resources
differently.
Regulatory fit affects an individual’s attainment expectations, the value that individuals
attribute to an activity, and the value individuals attach ex-ante to outcomes (Brockner,
Paruchuri, Idson, & Higgins, 2002; Higgins, 2000, 2005; Higgins, Idson, Freitas, Spiegel, &
Molden, 2003). Individuals experience regulatory fit when the means of goal pursuit available to
them match their regulatory orientation. Thus, enactment of a role oriented toward a particular
regulatory orientation is likely to “feel right” to individuals high in that orientation, causing them
to experience fit and thus identify more strongly with it. Importantly, the concept of value
employed here goes beyond the hedonic experience of pleasure; it includes the experience of
motivational force (Higgins, 2006). Motivational force is the combined effect of the expectation
of goal attainment and the attractiveness of goal attainment (Vroom, 1964), two primary
determinants of goal commitment (Erez & Kanfer, 1983; Hollenbeck & Klein, 1987; Locke,
Saari, Shaw & Latham, 1981). Collectively, these arguments strongly suggest that processes of
self-regulation will influence an individual’s commitment to goals and their activated identity.
Finally, multiteam systems are a recently identified organizational form, and only a
smattering of empirical studies have investigated the unique mechanisms at work in these
complex systems (Davison, Hollenbeck, Barnes, Sleesman, & Ilgen, 2012). Multiteam systems
(MTSs) are complex organizational entities organized as highly specialized component teams

4

whose members have a distinct skill set and expertise. Component teams are highly
interdependent, and pursue a unique set of proximal goals that collective aim at achieving a set
of higher order, and more distal, super-ordinate system goals. Thus a key attribute of multiteam
systems is that goals are hierarchically organized with individual goals nested in component
team goals which are nested in the overall goals of the system (Mathieu, Marks, & Zacarro,
2001). This is a perfect context in which to study the phenomenon of interest in this dissertation.
This dissertation sought to take a critical step toward understanding the psychological and
social mechanism that determine the focus and contribution of individuals situated in a complex
organizational setting with hierarchically arranged goals by gaining an understanding of the
regulatory process driving activated identity and goal commitment in the context of multiteam
systems. The ultimate objective of this dissertation was to gain an understanding of the role
played by regulatory mode, identity and goal commitment in driving or inhibiting incremental
performance contribution across the goal levels of a multiteam system.

5

Contextual Background: “Systems of Teams” or “Teams of Teams”
The complexity and dynamism of today’s business environment is increasingly forcing
organizations to shift from structures of work centered around more traditional stand alone teams
to newer organizational forms consisting of systems of teams. This unique organizational form,
which includes teams-of-teams (Hoegl, Weinkauf, & Gemuenden, 2004) and multiteam systems
(Mathieu et al., 2001), representatively consists of a network of highly interdependent teams (or
higher level organizational entities such as strategic business units), organized as a tightly
coupled activity system capable of ‘real time’ mutual adjustment, where each team possesses
specialized skills, capabilities, and functions that distinctly contribute to achieving a shared,
super-ordinate goal. Systems of teams are, in fact, ubiquitous in today’s business environment.
For example, one could successfully argue that extant organizational forms that cross ever more
permeable organizational boundaries, for example strategic alliances and joint ventures, are
systems of teams. It could also be argued that every collective of sub-organizations within a
corporation whose representatives come together as a task force or that is led by a program
manager or program management team or that is based on a matrix structure is a system of
teams.
Organizations have long been conceptualized as multilevel systems where individuals are
nested in teams, teams are nested in organizational sub-units, and these sub-units are nested in
organizations (Klein & Kozlowski, 2000; Thompson, 1967). While there has been a great deal of
research targeted at individuals, stand alone teams, and organizations, several scholars have
noted the dearth of research on large organizational sub-units. For example, Gist and colleagues
called upon researchers to “extend beyond the study of groups in isolation to the study of groups
as part of a system of organizational activity” (Gist, Locke, & Taylor, 1987, p. 253).

6

Multiteam System (MTS)
Mathieu and colleagues extended extant organizational theory and advanced our
conception of complex, organizational systems through the articulation of “a new ‘teams-ofteams’ organizational form that [they] refer to as multiteam systems” (2001, p. 289). A
multiteam system is a network of two or more non-reducible and uniquely distinct component
teams consisting of interdependent members, where all component teams exhibit input, process,
and output interdependence with at least one other team in the system. This network of teams
forms a tightly coupled activity system, with each member team possessing specialized skills,
capabilities, and functions that uniquely contribute to achieving a (set of) shared, super-ordinate
goal(s). For example, an emergency response team, consisting of teams from the fire department,
the police department, the rescue or ambulance squad (i.e., emergency medical technicians or
EMTs), an emergency room team, other specialized hospital based teams such as surgery, postop, and critical care units, is an example of a network of highly specialized component teams that
work together in a network fashion to accomplish a set of super ordinate goals (see Mathieu et
al., 2001 for a more complete discussion).
The specific division of labor and organizational structure of an MTS stems from both
their target task domain (i.e., goal set) and the technologies employed. This means that a
multiteam system can be wholly contained within a single legal entity (e.g., corporation), for
example a task force or project team hastily formed to address an urgent product or market issue
such as a product recall, or it can be comprised of teams from several different legal entities as
might be the case for the emergency response example outlined above (Mathieu et al., 2001).
Importantly, due to the high degree of interdependence across teams, issues regarding the sharing
of information and other resources (Ancona & Caldwell, 1988), goal prioritization, and the

7

ability to flexibly adapt to the shifting demands of the task environment often arise. Thus while
KSAOs are important prerequisites to accomplishing work effectively at all levels, and while
intra-team coordination is an additional important aspect at the level of the work group or team,
what is especially critical and distinctive at the level of the MTS is the need for effective interteam coordination (i.e., boundary spanning activities; Davison & Hollenbeck, 2011) and real
time, mutual adjustment.
Goal structure of a multiteam system. A critical feature of multiteam systems
articulated by Mathieu and colleagues is the notion of goal hierarchies. Mathieu and colleagues
stipulate that “Goal hierarchies within MTSs give rise to multiple kinds of functional process
interdependencies among component teams” (2001, pg. 294). In fact, it is arguably the other way
around.
The structure of goals in a multiteam system is an artifact of the division of labor that is
embodied in the structure of the system and thus it is specific to each MTS. Since the division of
labor is contingent on the demands of the task environment (Lawrence & Lorsch, 1967) and the
technology employed (Keller, 1994), the goal structure of a multiteam system is fundamentally
driven by these factors as well. Because of functional specialization, each of the component
teams in the system has associated with it a set of proximal goals, some exclusive and some
interdependently shared with other component teams in the system. Goals are nested
hierarchically, with goal interdependence increasing from lower to higher levels such that the
MTS super-ordinate goal is at the apex of the hierarchy (Mathieu et al., 2001).
The objectives of a multiteam system are pursued and accomplished over the course of a
series of both overlapping and serial performance episodes, thus timing (i.e., temporal order) and
priority (i.e., importance) are crucial attributes that must also be incorporated into the goal

8

structure. In essence, the goal structure is a rule set; a roadmap that serves to inform and guide
the activities of a multiteam system. By depicting the linkages among component team
objectives, their priority and importance, the goal structure of a multiteam system represents the
beginning foundation of a shared mental model of the task domain that serves to clarify how the
objectives of the individual component teams come together to accomplish the super-ordinate
goals of the system.
Failure to attend to the needs of the other component teams in the system can result in the
sub-optimization of the overall system and an inability of the system to achieve super-ordinate
goal. Too exclusive a focus on the goals at one level can lead to a lack of focus on the goals of
another level. Thus while a strong commitment to goals is prototypically depicted as good for
performance, too strong a commitment to a sub-set of goals in the hierarchy can be problematic
in the multiteam system context.

9

Key Construct Theories
Goals
There is certainly no shortage of research on goal constructs (Austin & Vancouver,
1996), and topics as far ranging as motivation (Locke, Latham, & Erez, 1988), learning (Kanfer
& Ackerman, 1989), performance, and personality (Cropanzano, James, & Citera, 1993) have all
been included, in one way or another, in the goals literature. Recently, however, goals research
has sub-divided into two focused topic areas, goal setting and goal striving. “Goal setting refers
to antecedents to action such as goal choice and goal acceptance (Klein, Austin & Cooper, 2008).
Goal striving refers to striving to meet the goal (Kanfer & Kanfer, 1991; Lord & Levy, 1994)”
(Mitchell, Harman, Lee & Lee, 2008, p. 198). Goal striving, or self-regulation, entails the
psychological processes and behaviors that occur between setting a goal and accomplishing or
failing to accomplish it, including the regulation of multiple goals vying for the same limited
resources.
Goal setting. The pioneering work of Locke and colleagues (Locke, 1968; Locke &
Latham, 1990a; Locke, Latham, & Erez, 1988), Hollenbeck and colleagues (Hollenbeck & Klein,
1987; Hollenbeck, O’Leary, Klein, & Wright, 1989), and several others provides the theoretical
foundation upon which theories of goal setting in the organizational sciences sit. Goal setting
research focuses mainly on the content of particular goals and how goals are chosen (Mitchell et
al., 2008). Goals establish the standards by which attainment is measured, with discrepancies
between current performance and desired end states serving to initiate and fuel motivational
processes (Locke & Latham, 1990a). As aptly put by Kruglanski, “Goals energize our behavior
and guide our choices; they occupy our thoughts and dominate our reveries” (1996, p.599).

10

Desired end states or goals can be either externally imposed or intrinsically generated
standards of performance, and these standards can be either hopes, wishes, and aspirations (i.e.,
‘ideals’) or duties, obligations, and responsibilities (i.e., ‘oughts’) (Higgins, 1997). Research has
found that difficult yet clear and specific goals are more motivating than vague or easy goals
(Hollenbeck, Williams, & Klein, 1989; Locke & Latham, 1990b), and that commitment to a goal
is positively related to the amount of resources (e.g., cognitive, emotional, self-regulatory)
allocated to it (Diefendorff & Lord, 2008). Once a goal is accepted (committed to), performance
feedback provides the mechanism by which progress toward a goal is assessed (Erez, 1977), and
by which discrepancies are identified. Performance discrepancies, in turn, serve to stimulate both
emotional reactions and cognitive evaluations of competence, and serve as an impetus to action
(i.e., motivation to close the gap) (Locke & Latham, 1990b). While most of the goal setting
literature has focused on antecedents to action enacted to achieve an established and static goal,
some research has considered the process of goal revision which is an alternative taken as a
result of either failing to achieve a goal or upon the realization that a pursued goal is
unachievable (Klein et al., 2008).
Goal commitment. Goals are the object or aim of actions, and the relationship between
goals and performance is strongest when people are committed to their goals (Locke & Latham,
2002). Thus goal commitment is critical because it is a necessary condition for goals to influence
behavior (Klein & Wright, 1994). It can have a main effect on performance, and research has
consistent shown that “highly committed individuals exert more effort, and are more persistent
toward goal attainment, than individuals who are less committed to the goal” (Seijts & Latham,
2000a, p. 316). Further, “the beneficial effect of goal setting on performance has been replicated
across a wide variety of participants, tasks, criterion measures, settings, and countries using a

11

multitude of research designs (Bandura, 1997; Locke & Latham, 1990b)” (Seijts & Latham,
2000a, p. 315). As a central component of goal setting theory, these same characteristics can be
attributed to goal commitment.
As the review by Tubbs and Dahl (1991) demonstrates, a plethora of conceptualizations
and definitions of goal commitment have been proposed. The construct of goal commitment, as
conceptualized in task goal theory, is defined as “one’s attachment to or determination to reach a
goal, regardless of the goal’s origin” (Locke, Latham, & Erez,1988, p.24), “implies the extension
of effort, over time, toward the accomplishment of an original goal and emphasizes an
unwillingness to abandon or to lower the original goal (Campion & Lord, 1982)” (Hollenbeck &
Klein, 1987 p.212). Similar definitions have been put forward by Naylor and Ilgen (1984),
Kernan and Lord (1988), and most recently by DeShon and Landis (1997). Importantly, choice,
effort, and persistence are common elements of all of these definitions (Seijts & Latham, 2000b).
It is important to note that goal commitment and goal acceptance are similar but
distinguishable concepts, although they are often used interchangeably in the literature. The
former implies that the focal individual is psychologically bound to a goal without concern for its
originating source (i.e., regardless of whether it is self-set or assigned), which is not necessarily
the case in the latter situation (Campion & Lord, 1982; Hollenbeck & Klein, 1987; Hollenbeck,
O’Leary et al., 1989; Locke et al., 1988). As articulated by Locke and colleagues, “Goal
commitment implies a determination to try for a goal (or to keep trying for a goal), but the source
of the goal is not specified. It could be an assigned goal or a participatively set goal or a goal that
one set on one's own. Goal acceptance implies that one has agreed to commit oneself to a goal
assigned or suggested by another person” (Locke et al., 1981, p. 143).

12

Locke and colleagues suggested that "goal acceptance or commitment can be considered
a form of choice, (i.e., the choice between accepting or rejecting a goal that was assigned or set
participatively),” and that the factors underlying an individual’s goal choice and goal
commitment “fit easily into two major categories, which are the main components of expectancy
theory" (Locke et al., 1981, p. 144); i.e., expectations of success (i.e., expectancy) and perceived
value of attaining or trying for the goal (i.e., valence). These two components are often expressed
as a quasi-mathematical model:
Force (F) = Expectancy x Valence.
This model essentially states that the strength of the tendency of a focal individual to act in a
certain way, denoted “Force” (a throw back to Lewin, 1951), depends on the strength of their
expectancy that the act will result in a set of outcomes and on the value of these outcomes to
them.
Expectancy theory (Vroom, 1964; Lawler, 1971) is based on the supposition that
individuals decide to act in a certain way because they are motivated by their expectations
regarding the outcomes likely to result from that specific set of behaviors. It predicts that
“individuals will be motivated to put forth effort if they believe that their effort will result in
good performance (expectancy); that this performance will lead to secondary outcomes, such as
rewards, recognition, or satisfaction (instrumentality); and if they assign a high positive valence
to the secondary outcomes” (Erez & Isen, 2002, p. 1055). Expectancy and instrumentality are
cognitive representations of an individual’s subjective perceptions of reality based on their own
experiences, their self-construal, and their observations of others, whereas valence is a
representation of what the individual values (i.e., their value system).

13

An important consequence of Locke and colleagues (1981) perspective is that it
established expectancy theory as the theoretical underpinning for task goal theory’s conception
of goal commitment. Researchers in this tradition employ expectancy theory to argue that
strength of goal commitment is primarily influenced by the degree to which an individual
believes that they can attain the goal (i.e., self-efficacy) and by factors that make goal attainment
important to the individual such as the outcomes that they expect as a result of pursuing and
attaining the goal (e.g., Erez & Kanfer, 1983; Hollenbeck & Klein, 1987; Locke et al., 1981).
The combination of these factors is sometimes referred to as Motivation Force, represented by
the quasi-mathematical model:
Motivational Force (MF) = Expectancy x Attractiveness,
where Attractiveness = Instrumentality x Utility.
Theorists argue that individuals select the option with the greatest motivational force when
deciding among behavioral options. In other words, behavior selection is determined by the
desirability of the outcomes in their totality; i.e., where both primary outcomes such as
performance on the task and secondary outcomes such as rewards, recognition, and satisfaction
are considered.
Empirical research has provided considerable support for this expectancy theory based
conceptualization (Klein & Wright, 1994). A meta-analysis conducted by Klein (1991) found the
average weighted effect size for the relationship between expectation of goal attainment and goal
commitment to be 0.19, and between attractiveness of goal attainment and goal commitment to
be 0.43. A later meta-analysis conducted by Klein and colleagues found the weighted mean
effect size between goal commitment and expectancy, attractiveness, and motivational force to
be 0.36, 0.29, and 0.33 respectively, when corrected for measurement error (Klein et al., 1999).

14

Further, research has consistently demonstrated that personal and situational factors have a main
as well as an interactional effect on both expectation of goal attainment and the attractiveness of
goal attainment as hypothesized by Hollenbeck and Klein in their expectancy theory model of
goal commitment (Klein et al., 1999).
Goal striving. Goal striving encompasses those cognitive, emotional, and behavioral
processes involved in attaining one or more pre-established goals (Pervin, 1989). Known in the
psychology literature under the auspices of self-regulation (the active control of psychological
processes, both cognitive and emotional, for the purpose of attaining goals; Diefendorff & Lord,
2008; Gollwitzer & Bargh, 1996), goal striving has been conceptualized and studied in the
organizational sciences literature as three distinct yet interrelated areas. The first of these, how
individuals learn and acquire new skills over time, is the focus of Kanfer and colleagues (Kanfer
& Ackerman, 1989; Kanfer & Heggestad, 1997; Kanfer, 1996). Research in this area is
particularly interested in how individual differences in personality and ability influence an
individual’s allocation of resources across time (i.e., at different points on of the learning curve).
Based on concepts originally articulated by Powers (perceptual control theory; 1973),
researchers in the control theory tradition conceptualize goal striving as a dynamic, goal
discrepancies driven process that influences behavior differentially over time (Campion & Lord,
1982; Klein, 1989; Lord & Hanges, 1987; Lord & Levy, 1994). Control system theories of selfregulation have been criticized by Bandura and Locke (2003) as being overly mechanical
however, research findings support the validity of these models as one component of a more
complex system involving cognitive, emotional and other psychological dimensions.
The third area, action theory, conceptualizes goal striving as a sub-part of an “action
cycle” (Frese & Zapf, 1994; Gollwitzer, 1990, 1996; Kuhl, 1984). Gollwitzer (1990) described

15

an action cycle as consisting of the four phases goal setting (mind-set: deliberative), planning
(mind-set: implemental), goal pursuit(mind-set: actional), and goal attainment / revision (mindset: evaluative), whereas Lord and colleagues categorize these four phases as referent
establishment, planning and goal maintenance, action, and evaluation and feedback (Lord,
Diefendorff, Schmidt, & Hall, 2010). Importantly, the planning- and action-phases are thought to
be relatively automatic.
Cognitions and emotions change within phases and can differ substantially between
phases. For example, Kuhl (1984, 1986) argues that a key function of the planning phase is to
prepare oneself to guard against and outright ignore disruptive thoughts and emotions that may
occur during the subsequent goal pursuit phase (i.e., goal shielding). Activated during the
planning and goal maintenance phase, goal shielding, “the processes by which active goals and
related information are protected from interference from information related to competing goals”
(Lord et al., 2010, p.551), occurs prototypically without conscious thought or attention.
Importantly, Shah et al., (2002) found that goal shielding plays a central role in the relationship
between goal commitment and both task persistence and task performance.
Multiple goals. Employing theory imported from the control theory of self-regulation
(Carver & Scheier, 1998), expectancy theory (Vroom, 1964), and goal setting theory (Locke &
Latham, 1990), researchers in the traditions of motivation, goal setting, and goal striving are
increasing turning their focus to understanding the determinants of goal prioritization and goaldirected behavior when individuals are faced with alternative and competing goals. Despite the
richness of theory – or perhaps because of it as this theory was developed, for the most part,
based on single, isolated goals – multiple goal studies have obtained varying results suggesting a

16

level of complexity yet to be revealed (Schmidt & DeShon, 2007). Four recent studies are
illustrative of the direction of current research in this area.
A laboratory study conducted by Schmidt and DeShon (2007) investigated the effects that
goal discrepancies, monetary incentives, and time remaining to complete a goal have on goal
prioritization. Facing comparable tasks with equivalent goals they found that greater goal
discrepancies influenced goal-directed behavior with greater attention given to goals with the
largest discrepancies early in time, but this focus shifts to the goals with the highest expectancy
of completion as the deadline to complete the goals approaches. Further, monetary incentives
focused attention on the goals offering the highest payout. These findings are consistent with the
research emerging out of both the expectancy theory and the goal commitment literatures
(Hollenbeck & Klein, 1987; Vroom, 1964).
Building on concepts from the control theory of self-regulation (Carver & Scheier, 1998)
and the expectancy theory literatures, Vancouver and colleagues (Vancouver, Weinhardt, &
Schmidt, 2010) developed a computational model to investigate factors that might explain the
shift in resource allocation focus over time reported in the Schmidt and DeShon (2007) study.
Unlike traditional uses of expectancy theory that employ static conceptions of valence and
expectancies, the model employs dynamic valence and expectancies which allows them to
investigate the effect of changes in these two variables on resource allocation over time. The
model indicates that differences in valence outweigh differences in expectancies throughout most
of the allocated time for task completion thus accounting for choice of task based on
discrepancy. However as the completion deadline approached expectancies play an ever greater
role in choice of task.

17

A study conducted by Schmidt and Dolis (2009) investigated how the allocation of
limited psychological resources is influenced over time by goal difficulty, progress toward goal
achievement, and goal attainment expectancy. A key contribution of this article is the
introduction of a new construct, dual-goal expectancy, defined by these authors as “individuals’
beliefs regarding the likelihood of meeting both goals within the allotted time” (p. 680). They
posit and show in a laboratory study that when dual-goal expectancy is high, resources are
allocated to the goal furthest from attainment. Conversely, when dual-goal expectancy is low (all
else being equal), resources are allocated to the goal that is closest to being met. In addition, this
study found that dual-goal expectancies change over time which, the authors argue, could
account for shifts in the allocation of resources such as those reported in the Schmidt and
DeShon (2007) study.
Ethiraj and Levinthal (2009) employed a computerized model to investigate the effects
that three different decision prioritization strategies have on multiple performance goals at the
organizational level: (1) goal myopia, focus on a singular goal; (2) spatial differentiation, the
assignment of different goals to different departments; and (3) temporal differentiation, goal
myopia within a limited time period where a different goal is given singular focus in each time
period. They found that lacking a prioritization strategy, the pursuit of multiple goals leads to a
“lock-in to the status quo” (p. 19); i.e. future performance that, at best, matches existing
performance. They also found that each of the three goal priority strategies serve to mitigate
what they call the “status quo bias”, although to differing degrees and in different circumstances.
This study clearly has important implications for multiteam systems that both employ a goal
hierarchy akin to the spatial differentiation strategy articulated in the article, and face a dynamic
task environment that requires the adoption of a temporal differentiation strategy. That said, the

18

research question in this study – When faced with a hierarchy of goals, where each level in the
hierarchy is associated with a specific level in the organization (e.g., individual, work group or
department, function or business unit) and goals at each level are an amalgamation of goals
originating at that level and at levels lower in the hierarchy, what determines an individual’s goal
focus? – differs from that addressed by this research.
While this line of research clearly has important implications for multiteam systems, the
proposed study differs in some key and significant ways. Although one study by Wright and
colleagues did investigate the effects of goals and incentives on helping behaviors (Wright,
George, Farnsworth, & McMahon, 1993), an important feature that almost all of the multi goal
studies have in common is that the multiple goals facing individuals in these studies are all
performance goals directly related to a job or a task for which the focal individual is personally
responsible and incentivized to achieve. In cases such as these the decision is one of maximizing
personal performance or gain independent of the influence or impact on others. While closely
related and informative, this is a different decision scenario than the one underlying this study,
the decision to allocate limited resources to the achievement of more distal goals of a collective
versus the allocation of these same resources to the achievement of proximal goals for which the
focal individual is directly accountable.
Identity
As defined by Tajfel (1972) social identity represents “the individual’s knowledge that he
belongs to certain social groups together with some emotional and value significance to him of
this group membership” (p. 292). A primary motivational premise underlying social identity
theory (Turner, 1975; Tajfel & Turner, 1979) is simply that individuals are driven to develop and
maintain a positive self-concept or positive sense of self. The theory argues that individuals

19

possess a basic desire to view their social groups and thus themselves more favorably, and to
gain the respect and admiration of salient others (collectively referred to as the need for positive
regard; Hetts, Sakuma, & Pelham, 1999), thus people are driven by self-enhancement motives to
undertake behaviors that maximize the value of the groups they belong to. Individuals seek to
accomplish this by enhancing the status of the groups to which they belong (in-group bias)
and/or denigrating the groups they do not belong to (out-group derogation). Increasing the value
of the groups to which an individual belongs increases the value of the individual’s social self
which influences, in turn, feelings of self-esteem, self-respect, and self-worth. Finally, social
identities can be threatened by challenging the individual’s membership in the group, the
uniqueness of the individual in relation to other members of the group, or the clarity of the
boundary between the group and other groups.
Fundamental to social identity theory, but often overlooked, is the notion that the selfconcept, or identity of an individual, consists of both a personal identity and a social identity. In
this view, self-concept consists of both an individual or personal self reflecting all the
idiosyncratic aspects of the self (personalized identity, “what makes me I”, different from
everyone else), and a social self reflecting prototypical attributes of social categories and groups
to which the individual belongs that are assimilated into the individual’s self-concept
(depersonalized social identity, “what makes me We”, part of something beyond myself).
Further, social identities are categorizations of the self into ever more socially inclusive units that
depersonalize the self-concept, “a shift toward the perception of self as an interchangeable
exemplar of some social category and away from the perception of self as a unique person”
(Turner, Hogg, Oakes, Reicher, & Wetherall, 1987, p. 50).

20

Activated identity. The personal identity is an attribute of the individual that is relatively
stable over time whereas the social identity is much more dynamic, shifting over time. Cues in
the immediate environment (i.e., social context) in combination with an individual’s knowledge,
understanding, and theories about the social world influence which identity prototypes – selfand group- categorization schemas (or prototypes) stored in memory (Hogg & Terry, 2000;
Turner, 1982, 1985; Turner et al., 1987)– are accessed and what aspects of these identities are
made active as heuristics for perceiving and understanding the behavior of individuals, including
one’s own behavior (Ellemers, Spears, & Doosje, 2002). Importantly, identity affects behavior
through its influence on both cognitive and motivational mechanisms, and differences in
behavioral patterns can be explained by contextually driven shifts in the social identity portion of
the self-concept.
Also of importance due to the proposed laboratory context of this study, minimal group
studies have demonstrated that individuals behave in accordance with group identity even when
there is no preexisting basis for the identity (Tajfel & Turner, 1979). Research on minimal
groups indicates that even when categorization is based on random assignment, the mere
recognition that one belongs to a social category suffices to produce in-group identification
(Hogg & Abrams, 2003). Further, members of artificial groups created for experimental purposes
– i.e., with neither a history, nor a future, nor meaningful interaction with other group members –
often perceive their group as better than other groups (i.e., in-group bias), and arbitrary
distinctions between “us and them” suffice to elicit commitment (Roccas, Sagiv, Schwartz,
Halevy, & Eidelson, 2008).
Within a given social context an individual can be categorized along a continuum of
attributes ranging from total uniqueness (individuation) on one end to total submersion in the

21

socially relevant context on the other (deindividuation). Brewer (1991) viewed identity as a
compromise between the need to validate the self-concept by being categorized as similar to
significant (salient) others (assimilation need) and the need to be seen as different and unique
(differentiation need). She argued that individuals avoid being identified with social categories
that are either too inclusive or too differentiating and seek, instead, an optimal group identity that
simultaneously satisfies the need for assimilation or inclusion in a group and the need for
differentiation through the distinctions between the focal group and other groups. She called this
optimal distinctiveness defined as a state of equilibrium between the need for assimilation and
the need for differentiation (Brewer, 1991).
Brewer and Gardner (1996) argued that an individuals self-representation (i.e., selfconstrual or simply self) is “associated with corresponding transformations of the bases for
content of the self-concept, the frame of reference for evaluations of self-worth, and the nature of
social motivation” (p. 84). By “bases for content of the self-concept” Brewer and Gardner were
referring to the concept of the extended self-concept whereby an individual incorporates aspects
of salient others into their self-concept (Aron, Aron, & Smollan, 1992; Tropp & Wright, 2001).
“Frame of reference for evaluations of self-worth” shifts from the self-evaluation of personal
traits and characteristics based on direct comparisons to relevant others (Pelham, 1995; Pelham
& Swann, 1989; Suls & Wills, 1991), to evaluations of role behavior in specific contexts
(Markus & Kitayama, 1991; Stryker, 1991), to status of the in-group in comparison to salient
other groups (Turner at al., 1987) as activated identity shifts from individual to relational to
group. Finally, “the nature of social motivation” refers to the change in the basic goals of social
interaction.

22

Brewer and Gardner use this argument to delineate the self into the self as an individual,
the interpersonal or relational self, and the collective or group self. In this conception, the
individual whose activated identity is individualistically oriented is driven to promote and
advance self-interests, and conceives of self for purposes of comparison based on individual
traits and characteristics. The individual whose activated identity is relationally oriented is
concerned with promoting the interest of and procuring benefit for important (significant) others,
and conceives of self in terms of roles in the social network (Markus & Kitayama, 1991). The
individual whose activated identity is group oriented is primarily concerned with ensuring the
welfare and status of their in-group in relation to relevant other out-groups, and conceive of self
in terms of the group prototype (see Brickson, 2000 Table 1 on page 85).
While relational and group identities are both social extensions of the self, the
fundamental social premise underlying them differs. As the moniker implies, the relational (or
interpersonal) self is based on social connections best described as personal bonds or attachment
whereas group (or collective) identities are derived from the depersonalized identification with a
symbolic group or social category (e.g., stereotype or exemplar) and thus do not require personal
contact or personal relationships, although these can exist (Brewer & Gardner, 1996; Turner et
al., 1987). This perspective is consistent with earlier work carried out by Prentice and colleagues
(Prentice, Miller, & Lightdale, 1994). who found that two types of group identities exist, one
based on attachments to other group members, which they termed common-bond groups, and one
based on collective or prototypical group identities, which they termed common-identity groups.
Finally, research based on the conceptual work of Brewer and Gardner (1996) conducted by
Johnson and others has explored the degree to which a focal individual is dispositionally oriented
to be more or less individual, relational, or collective in their chronic self-concept (Johnson &

23

Chang, 2008; Johnson, Chang, & Rosen, 2010; Johnson & Saboe, in press; Johnson, Selenta, &
Lord, 2006).
Specific to the context of a multiteam system, within component team relations is best
conceptualized as a common-bond group because individuals would typically have some degree
of interpersonal relationship with the other members of their component team, whereas between
component team relations is best conceptualized as a common-identity group because
relationships between members of different component teams would typically be much less
interpersonal, although this would vary to an extent based on degree of task interdependence.
Role-identity. This perspective is also consistent with earlier theorizing carried out in
sociology by Stryker (Identity Theory; 1980; 2008) and McCall and Simmons (Role-identity
Theory; 1966). Stryker (1980; 2008) envisioned a multifaceted self with an identity salience
structure composed of multiple identities arranged hierarchically. Salience is based on
commitment to the identity, and the more salient an identity, the more likely it is to be invoked in
an interactional situation (Owens, Robinson, & Smith-Lovin, 2010). Interestingly, a recent study
conducted by Stryker, Serpe, and Hunt (2005) showed that intermediate-level social structures
such as neighborhood and school, and not more proximal social structures such as fellow
workers who are relatives, influenced commitment most by fostering in-group identity-based
relationships.
Based on Foote’s (1951) conception of situated motivation, McCall and Simmons’s
(1966) Role-identity Theory argues that social structure and one’s position in it serve to define
one’s situation-induced social identity, and that role-identity serve as the primary determinant of
an individual’s behavior (Owens et al., 2010). As described by McCall and Simmons, role
identity is the “character and the role that an individual devises for himself as an occupant of a

24

particular social position” (1966, p. 67). These scholars argue that individuals harbor many roleidentities organized into a hierarchy of prominence reflecting the role’s relative value to the
individual’s conception of an ideal self, and that commitment to any particular role-identity is
intimately related to how much their self-esteem is bound to its successful performance.
Importantly, the concept of ideal self as employed in Role-identity Theory incorporates duties
and obligations as well as desires and wishes (i.e., a combination of what Higgins refers to as
ought -self and ideal-self; 1987; 1989).
Research by Farmer and colleagues (e.g., Farmer, Tierney, & Kung-McIntyre, 2003;
Farmer & Van Dyne, 2010), and others, has shown that concepts of role-identity extend to the
workplace and that workplace role-identities “are practically important in the workplace because
they influence work-related performance” and employee work behaviors (Farmer & Van Dyne,
2010, p. 503). Farmer and Van Dyne go on to stress, however, that “the joint importance of self
and situation together has been neither well elucidated nor well explored in role identity
research. While a healthy stream of qualitative investigation exists (e.g., Ashforth, Kreiner,
Clark, & Fugate, 2007; Kreiner, Hollensbe, & Sheep, 2006), quantitative knowledge is
particularly limited concerning identity-based action in organizational contexts. For example,
even though researchers have proposed that psychologically important identities are more likely
to be activated by contextual factors (Ashforth, 2001; Stryker & Serpe, 1994), little if any
quantitative research has been conducted into how this affects role performance in organizations”
(2010, p. 505). Finally, because sharing an activated group identity can elicit cooperative and
shared problem solving behaviors among group members even in the absence of interpersonal
communication, the degree to which members of a multiteam system share activated group
identities is conceivably an extremely important determinant of system-wide performance.

25

Self-regulation
“Self-regulation is the process in which people seek to bring themselves (their behaviors
and self-conceptions) into alignment with relevant goals and standards” (Brockner et al., 2002, p.
7). However, the presence of alternative and competing goals often creates an “approachapproach” conflict which, in turn, hampers progress toward any of the desired end states (Lewin,
1935, 1951; Shah et al., 2002; Zeigarnik, 1938). Even so, most individuals are adept at regulating
(i.e., self-control) to avoid the temptation of these alternative goals in order to accomplish some
focal task. In fact, there is strong evidence that the self-regulatory process of cognitive intergoal
inhibition, alternatively known as goal shielding, is a habitual and autonomous mental control
process that develops early on in the life cycle of humans (Shah et al., 2002). (Note: See Wegner
& Wenzlaff, 1996 for a recent review of the mental control literature). While a singular focus
and persistence on a task is often touted as a positive attribute because of its positive relationship
to task performance, it can be problematic in the context of a multiteam system.
Regulatory mode (Higgins, Kruglanski, & Pierro, 2003; Kruglanski et al., 2000) and
regulatory fit (Higgins, 2000; 2002) are two aspects of self-regulatory motivational orientations
that are key to understanding motivated cognition and goal directed behavior. Regulatory mode
denotes two orthogonal dimensions of self-regulation, one concerned with the critical evaluation
of alternative goals or means to decide which are best to pursue, as well as the appraisal of
performance, and a second concerned with movement from state to state, including commitment
of psychological resources to initiate and maintain such movement (Kruglanski et al., 2000).
Attributes of these dimensions serve to characterize individuals and contextual features (such as
focal activities or roles). Regulatory fit connotes the degree of fit between a goal-directed
regulatory orientation (e.g., locomotion or assessment) and the means of goal pursuit available to

26

a focal actor. Studies have shown that fit increases the value individuals ascribe to an activity
and that it is a key determinant of individual behavior and performance outcomes.
Goal shielding. One aspect of self-regulation that appears of particular relevance is goal
shielding, a self-regulatory process that has been shown to inhibit consideration of alternative or
competing goals (Shah at al., 2002). The fundamental premise underlying the concept of goal
shielding is that the inhibition of alternative goals increases concentration on a focal goal by
amassing limited cognitive and self-regulatory resources, focusing them on current pursuits, and
through this mechanism both task persistence and task performance are improved. Although
intergoal inhibition is thought to be an autonomous process, it is not indiscriminately applied but,
rather, is directed by both internal and external (to the individual) cues. As Shah and colleagues
explain, “goal shielding theory assumes that the automatic inhibition of alternative goals is
sensitive to (a) the characteristics of the goals themselves, and (b) to the motivational and
emotional context in which the self-regulatory activity is unfolding” (2002, p. 1262).
Shah and colleagues (Shah et al., 2002) conducted a series of six empirical studies that
investigated “how the activation of focal goals to which the individual was committed inhibits
the accessibility of alternative goals” (Shah et al., 2002, p. 1261). Importantly, this research
provides convergent evidence that goal commitment increases the degree to which alternative
goals are inhibited. “All six studies exhibited the theoretically predicted, positive relation
between the degree of goal commitment and the inhibition of alternative goals, replicating these
effects across a range of methodological approaches …Our findings suggest that the inhibition of
alternative goals increases with individuals’ level of commitment to the focal goal” (Shah et al.,
2002, p. 1277-1278).

27

Abandoning a goal is most often discussed in the literature under the more encompassing
heading self-regulatory failure. This is unfortunate as it leads to the misguided perspective that
goal abandonment is universally a bad thing. However, a steadfast commitment to a failing
course of action in the face of a changing environment (e.g., escalation of commitment) is often
quite dysfunctional (Klein et al., 1999). For example, in the case of dynamic task environments
such as those encountered by multiteam systems, persistence at a task when that task’s goal
priority has fallen relative to alternative tasks’ goals (or lost meaningful priority at all) can be
detrimental to system performance. Goal abandonment in such situations represent the better
course of action. Karoly (1993) argued that failures of self-regulation can be attributed to any of
three broad causes: (1) failing to start an action in a timely manner; (2) stopping an activity
prematurely; (3) failing to stop goal striving when one should.
Interestingly, researchers studying escalation of commitment phenomena observed three
decades or more ago that “individuals may selectively filter information so as to maintain their
commitment to a policy or course of action (Caldwell & O'Reilly, 1982; Lord, Ross, & Lepper,
1979)” (Staw, 1981, p. 580). With little alteration, these same comments could be used to
roughly describe commitment to the activated goal and the key role goal shielding plays in
maintaining task persistence (Shah et al., 2002).
Failures of self-regulation can be the result of either conscious or unconscious processes
(Diefendorff & Lord, 2008). Unfortunately, conflicts among multiple goals and the problems
resulting from failures to regulate appropriately in response to shifting task environment
demands “will be confounded when the content of goals suggested by task or environmental cues
is inconsistent with chronic goal orientations or higher-level personality structures (e.g.,
Kozlowski & Bell, 2006). Such incongruities make it difficult to sufficiently activate goals so

28

that they become conscious (gain access to the GNW [global neuronal workspace]) and active in
the PFC [prefrontal cortex]” (Diefendorff & Lord, 2008, p. 183). Note: the PFC is the region of
the brain indicated in the control of thoughts and behaviors (Fuster, 2002), whereas the GNW is
indicated in conscious processing, enacting the role of “boundary spanner” between key brain
systems (Dehaene & Naccache, 2001). Finally, emotions are also indicated as contributing to
self-regulatory success and failure. Kuhl and Kazen (1999) suggest that inaction can stem from
an individual’s inability to escape negative affect when it occurs, their inability to generate
positive affect when needed, or some combination of the two.
Regulatory mode. Unlike classic perspectives of self-regulation such as control theory
(e.g., Carver & Scheier, 1990) and the rubicon model (e.g., Gollwitzer, 1990) which theorized
locomotion and assessment as two sides of the same coin (i.e., inseparable), Kruglanski, Higgins,
and colleagues posited and demonstrated that these are two distinct and orthogonal dimensions
of regulation (Kruglanski et al., 2000). As explained by Kruglanski and colleagues,
“[a]ssessment constitutes the comparative aspect of self-regulation that critically evaluates
alternative goals or means to decide which are best to pursue and appraises performance.
Locomotion constitutes the aspect of self-regulation concerned with movement from state to
state, including commitment of psychological resources to initiate and maintain such movement”
(2000, p. 793). Thus assessment is implicated in goal selection and prioritization, the evaluation
of different goals and the means available to pursue them in order to prioritize and select which
to pursue and in what order, whereas locomotion involves the control of action whereby the
individual is driven to move from their current state to some other alternative state (Higgins et
al., 2003).

29

Locomotion and assessment can work independently or interdependently to affect
behaviors and outcomes, and at any particular moment one mode may be emphasized over the
other. Importantly, the nature and consequences of a goal pursuit will predominantly be
influenced by whichever regulatory mode orientation is dominant at that moment. In addition,
the regulatory modes of locomotion and assessment vary across individuals (chronic trait) and
can be situation induced by specific activities (Higgins et al., 2003).
Locomotion (aka “Just do it”). To a high locomotion oriented individual, change of state
is its own reward; it is the journey and not the destination that is important. High locomotors are
likely to engage in whatever activity is currently most accessible to them that offers movement
irrespective of direction or valence, and without much regard for specific goal pursuit. It is the
state of immobility that high locomotors loath. Thus locomotors are mastery oriented (Dweck,
1991; Elliott & Dweck, 1988) and more intrinsically than extrinsically motivated (Deci & Ryan,
1985, 1991), are highly decisive, and prefer forceful, more autocratic leadership styles; for
example, coercive, legitimate, and directive (Bass, 1990; Kruglanski, Pierro, & Higgins, 2007;
Raven, Schwarzwald, & Kozlowski, 1998) and transformational (Bass, 1990; Benjamin & Flynn,
2006). Of interest, because locomotors are concerned with change, per se, they tend to pay
particular attention to the attainment expectancy of goals (i.e., attainable goals promise quick
movement) and thus it is possible for resources to be committed to the pursuit of a goal that is
not worth the resources allocated to it (Kruglanski et al, 2000; Higgins et al., 2003).
Locomotion is positively related to attentional control, conscientious, emotional stability,
vitality, and the Behavioral Activation System (BAS; Carver & White, 1994), especially the
‘Drive’ sub-scale which measures an individual’s motivation to pursue things and their
willingness to take risks while pursuing them. During goal pursuit, high locomotors enhance

30

experiential involvement (i.e., flow; Csikszentmihalyi, 1975) by increasing intrinsic motivation
and maintenance of commitment to action independent of intrinsic motivation, and by increasing
activity involvement (for example effort investment: Brown & Leigh, 1996; job involvement:
Kanungo, 1982) and commitment to goal pursuit independent of activity involvement. Finally,
locomotion is negatively associated with sensitivity to social criticism and feeling anxious in
social interactions, and thus high locomotors are not high self-evaluators or self-monitors
(Higgins et al., 2003).
Assessment (aka “Do the right thing”). Assessment is the dimension of self-regulation
that is concerned with making value comparisons among available alternatives in order to select
the best one. Individuals independently assess the value or importance of the current and end
states, and the value or utility of the means by which they might move toward a desired or away
from an undesired end state. Assessors are performance oriented (Dweck, 1991; Elliott &
Dweck, 1988), seek to avoid errors, and are more extrinsically than intrinsically motivated (Deci,
Koestner, & Ryan, 1999; Deci & Ryan, 1991). Because they seek the best alternative, high
assessors are not comfortable with ambiguity and thus look at more background information
before making a choice than do low in assessment orientation. In other words, assessors tend to
adopt a full comparison or compensatory decision making strategy. Also, assessors prefer more
democratic and consultative leadership styles, for example, expert, referent, and participative
(Bass, 1990; Kruglanski et al., 2007; Raven et al., 1998).
The assessment dimension is positively related to emotional instability, public selfconsciousness, private self-consciousness, need for social comparison, and the Behavioral
Inhibition System (BIS; Carver & White, 1994) which is consistent with a wait and see attitude.
It is negatively related to expectancy. Assessors are influenced by social norms and normative

31

standards, and thus are concerned with doing the right thing. In contrast to low assessors, high
assessors are sensitive to social criticism and feel anxious during social interactions. Concerns
with standards and social criticism makes individuals high on the assessment dimension both
high self-evaluators and high self-monitors (Mabe & West, 1982; Snyder, 1974) which, in turn,
makes them vulnerable to negative affect, lower self-esteem, and lower optimism (Higgins et al.,
2003).
Key research findings. The keystone article in the regulatory mode literature is
Kruglanski et al., 2000. This article established locomotion and assessment as theoretically
distinct self-regulatory imperatives. In addition, this article reports the development of two scales
to independently measure locomotion and assessment, along with the results of psychometric
tests attesting to their convergent validity, divergent validity, and temporal stability.
In a series of studies conducted in both the laboratory and in organizational settings in
Europe, Pierro, and colleagues investigated the relationship between regulatory mode and several
phenomena of particular interests to organizational scholars. For example, the results of one of
two studies published in 2006 indicate that locomotion is positively related to effort invested in
work activities and to both job involvement and successful performance, as assessed by a selfreport measure and by manager ratings (Pierro, Kruglanski, & Higgins, 2006b). Further, this
study found that job involvement partially mediated the relationship between locomotion and
effort, and that effort investment mediated the relationship between locomotion and
performance. The second study investigated the relationship between regulatory mode and task
motivation (Pierro, Kruglanski, & Higgins, 2006a). As predicted, this study found that
locomotion is positively related to intrinsic task motivation and assessment is positively related
to extrinsic motivation. This study also reported that locomotion is positively related to effort

32

investment which is positively related to goal attainment. Finally, this study found that
individuals are most likely to attain their goals if they are high on both regulatory mode
dimensions.
Kruglanski et al. (2007) investigated locomotion and the need for cognitive closure in the
context of organizational change. They found that need for closure was negatively related to
coping with change whereas locomotion orientation was positively related to coping with
change. They also found that an organizational climate supportive of change attenuated the
negative relationship between need for cognitive closure and coping with change, and that postchange work attitudes are determined by degree of success coping with change. Another recent
study examined the influence of regulatory modes on immediate versus delayed monetary reward
choices (Mannetti et al., 2009). As predicted, this study found that choices in an induced
assessment state condition were less impulsive and more far-sighted than those in an induced
locomotion state condition, suggesting that it might be possible to induce economically rational
inter-temporal choices by prescriptive means that induce an assessment state orientation. Pierro
et al. (2008) conducted three laboratory studies to examine the affects of regulatory mode
orientations on engaging in counterfactual thinking and experiencing post-decisional regret.
They found, as predicted, that when contemplating a decision with a negative outcome high
(versus low) locomotion was associated with less counterfactual thinking and less regret,
whereas the opposite was true for high (versus low) assessment. These results held for both
chronic and experimentally induced mode orientations.
An innovative study undertaken by Mauro and colleagues (Mauro, Pierro, Mannetti,
Higgins, & Kruglanski, 2009) varied the regulatory mode composition of 4-member teams to
study the affect of regulatory mode composition on performance speed and accuracy. One third

33

of the teams were composed of exclusively high locomotion oriented individuals, one third were
composed of exclusively high assessment oriented individuals, and one third consisted of a mix
of regulatory mode oriented individuals. As expected, teams consisting of exclusively high
locomotion oriented individuals performed assigned tasks faster than teams consisting of
exclusively high assessment oriented individuals, and teams consisting of exclusively high
assessment oriented individuals out performed teams consisting of exclusively high locomotion
oriented individuals on measures of accuracy. Echoing findings at the individual level, the mixed
teams performed tasks as fast as the teams consisting of only locomotors and as accurate as the
teams consisting of only assessors.
Finally, regulatory mode has also been investigated from the perspective of regulatory fit
(Avnet & Higgins, 2003; Higgins, 2000). I have included the studies and articles that
investigated fit employing concepts drawn from regulatory mode theory in the section on
regulatory fit that immediately follows this section.
Regulatory fit (aka “It feels right”). It is a fundamental reality of human nature that
humans are motivated to approach pleasure and avoid pain. Known as the hedonistic principle,
this theoretical concept has underpinned much of the scholarship in the area of motivation.
Higgins (1997) argued that scholars need to look beyond this enduring principle to gain an
understanding of the distinct principles that underlie it. His theory of self-regulatory focus
introduced the concept of regulatory focus as a distinct principle of hedonistic self-regulation
based on an individual’s desire to move their current states closer to some desired end-states.
This theory posits that how the hedonic principle of approaching pleasure and avoiding pain
operates is contingent upon the nature of the accessible and currently activated goals, where
goals are broadly defined as “representational structures that guide the system in its pursuit of a

34

reference or end state” (Markman & Brendl, 2000, p. 98); i.e., the standards and needs an
individual seeks to satisfy.
The theory of regulatory fit emerged as a corollary of regulatory focus theory. Dubbed
“feels right” in the literature (e.g., Cesario, Grant, Higgins, 2004; Cesario & Higgins, 2008;
Higgins, 2000; 2005), theory suggests and studies have shown that individuals experience
regulatory fit when they pursues a goal in a manner that sustains their regulatory orientations
(Higgins et al., 2003; Higgins, 2005). Regulatory fit has been shown to influence judgments and
decision making, feelings and emotions, attitude and behavior change, and task performance
(Higgins, 2005).
The vast majority of research into the concept of regulatory fit conducted to date has
investigated fit from the perspective of regulatory focus orientation (Higgins, 1997). However,
one of the first studies to look at regulatory fit from the perspective of mode and not focus
employed the findings from previous research based on regulatory focus fit to predict outcomes
based on regulatory mode fit (Avnet & Higgins, 2003). Avnet and Higgins argue that their study
generalizes fit beyond regulatory focus to other motivational domains, and that the phenomenon
of value from fit transfers specifically to fit based on regulatory mode orientations. This
theoretical extension to the original premise of regulatory fit theory is further established by
Higgins when he stipulates that “fit effects can be found for other orientations as well” (2005, p.
210). The specific example that Higgins uses to illustrate this point is the Avnet and Higgins
study involving regulatory mode.
Individuals experiencing regulatory fit attribute greater value to what they are doing and
this value from fit increases (is added on top of) the value expect or resulting from the outcomes
of the activity (Higgins, 2000, 2002). As a result of fit, an individual’s interest in activities is

35

greater (Higgins, Cesario, Hagiwara, Spiegel, & Pittman, 2010) and their motivation during goal
pursuit is stronger. This, in turn, leads to increased effort at goal pursuit (Pierro et al., 2006a;
Spiegel, Grant-Pillow, & Higgins, 2004). In addition, because activities and tasks vary in being
more or less related to locomotion or assessment, the effort invested by an individual in an
activity or task depends on whether the mode that is emphasized fits the requirements of the
activity undertaken (Pierro et al., 2006b).
Key research findings. Empirical studies conducted by Higgins and associates have
identified and demonstrated five value consequences of regulatory fit: “(a) people are more
inclined toward goal means that have higher regulatory fit (Crowe & Higgins, 1997; Higgins et
al., 1994), (b) people's motivation during goal pursuit is stronger when regulatory fit is higher
(Forster, Higgins, & Idson, 1998; Shah, Higgins, & Friedman, 1998), (c) people's (prospective)
feelings about a choice they might make is more positive for a desirable choice and more
negative for an undesirable choice when regulatory fit is higher (Idson, Liberman, & Higgins,
2000; 2004), (d) people's (retrospective) evaluations of past decisions or goal pursuits is more
positive when regulatory fit was higher (Freitas & Higgins, 2002; Higgins et al., 2003; Idson et
al., 2004), and (e) people assign higher value to an object that was chosen with higher regulatory
fit (Higgins et al., 2003)” (Higgins, 2000, pp. 1217-18; citations not in the original). Due to its
value from fit precept, regulatory fit concepts appear in a number of recent marketing studies
investigating influences on customer choice and satisfaction (e.g., Aaker & Lee, 2006; Avnet &
Higgins, 2006; Higgins, 2002; Higgins & Scholer, 2009; Mogilner, Aaker, & Pennington, 2008).
A substantial amount of fit research has been conducted recently in the areas of
information processing and perception, persuasion, and social influence, and several studies have
shown that regulatory fit plays a key role in these areas. Much of this work, however, involved

36

regulatory focus based fit (e.g., Cesario & Higgins, 2008; Friedman & Forster, 2001; Lee &
Aaker, 2004; Hong & Sternthal, 2010; Koenig, Cesario, Molden, Kosloff, & Higgins, 2009).
Research has investigated the affects of fit in a plethora of other areas as well. Motivation during
goal pursuit (Spiegel et al., 2004), emotions related to anticipating desirable versus undesirable
choices (Idson et al., 2004), decision confidence (Chernev, 2009), proximal versus distal
purchase decisions made by consumers (Mogilner et al., 2008), and leadership styles and leader
– follower relations (De Cremer, Mayer, van Dijke, Schouten, Bardes, 2009; Kark & Van Dijk,
2007; Neubert, Kacmar, Carlson, Roberts, & Chonko, 2008) for example.
One of the first studies to look at regulatory fit from the perspective of mode and not
focus was the study reported by Avnet and Higgins (2003). Drawing on previous research that
found the promotion and prevention based regulatory fit increases an individual’s perception that
a decision they made was "right," and thus value is transferred to the decision outcome (Higgins,
2000, 2002; Higgins et al., 2003), these authors predicted that the same monetary value effect of
regulatory fit could be generalized to locomotion and assessment orientations. Results confirmed
the authors’ predictions; participants that employed a decision strategy that fit their regulatory
mode orientation (i.e., a compensatory strategy for assessors and a non-compensatory strategy
for locomotors) were willing to pay upwards of 40% more for a book light than were their
counterparts in the non-fit condition.
Recently, Henderson and associates (Henderson, de Liver, & Gollwitzer, 2008)
investigated whether an implemental (locomotion oriented) mind-set fostered stronger attitudes.
They found that participants who made a decision about how to act, or who planned the
implementation of a decision, expressed more extreme attitudes toward an issue unrelated to the
decision, and exhibited more accessible and less ambivalent attitudes toward a variety of objects

37

unrelated to the decision, than did participants who held off from making a decision. Another
recent study looked at relationship between the assessment dimension and negotiator behavior
(Appelt, Zou, & Higgins, 2010). These authors found that while low assessors were responsive to
feelings of fit and non-fit, fit effects were eliminated and in some cases even reversed for high
assessors and replaced, instead, by correctness concerns even to the point of overcorrecting.
A third recent study of persuasive communications employing printed advertisements
investigated fit from the perspective of regulatory mode. This study found that locomotion
orientated individuals experienced fit and thus were more persuaded by exposure to "dynamic"
versus "static" visual images and, conversely, assessment orientation individuals experienced fit
and thus were more persuaded by "static" versus "dynamic" images (Mannetti, Giacomantonio,
Higgins, Pierro, & Kruglanski, 2010).
The concept of regulatory fit has also been used to explain evaluations of significant
others, both group members and leaders, and their influence on focal actors. Benjamin and Flynn
(2006) found that transformational leadership (Bass, 1990) is more effective at increasing
motivation and at eliciting positive evaluations from locomotion oriented individuals relative to
assessment oriented individuals. Studies conducted by Kruglanski and colleagues in four diverse
organizational contexts “consistently show that individuals high in locomotion prefer a
‘‘forceful’’ leadership style, represented by ‘‘coercive’’, ‘‘legitimate’’, and ‘‘directive’’ kinds of
strategic influence, whereas individuals high in assessment prefer an ‘‘advisory’’ leadership
style, represented by ‘‘expert’’, ‘‘referent’’, and ‘‘participative’’ kinds of strategic influence”
(Kruglanski, Pierro, & Higgins, 2007, p. 137).

38

Model Development and Hypotheses
Model Overview
Multiteam systems are a recently identified organizational form, and only a smattering of
empirical studies have investigated the unique mechanisms at work in these complex systems.
The structuring of labor into highly specialized component teams means that a particular
component team is best capable of addressing a particular set of task demands and thus each
component team pursues a somewhat unique set of proximal goals that collective aim at
achieving a set of higher order, and more distal, super-ordinate system goals (Davison et al.,
2012). In other words, goals are nested hierarchically with goal interdependence increasing from
lower to higher levels such that the MTS super-ordinate goal is at the apex of the hierarchy
(Mathieu et al., 2001).
This means that individuals in a multiteam system also face a hierarchy of multiple goals
and priorities. As discussed in the Introduction, the hierarchy of goals associated with complex
organizational systems in combination with resource limitations (e.g., internal psychological
resources such as cognition and external resources such as time) often creates a situation
whereby individuals can become so focused on a particular goal or set of goals that they fail to
focus on or outright ignore other important goals (Shah & Kruglanski, 2008). This highlights an
important research question: When faced with a hierarchy of goals, where each level in the
hierarchy is associated with a specific level in the organization (e.g., individual, work group or
department, function or business unit) and goals at each level are an amalgamation of goals
originating at that level and at levels lower in the hierarchy, what determines an individual’s goal
focus? This is, at its core, a question of self-regulation, goal commitment, and activated identity.

39

Research conducted by Shah and colleagues (2002) indicates that commitment to the
current goal increases task persistence and inhibits the consideration of alternative goals, and that
degree of commitment to this goal determines the willingness of individuals to redirect actions to
the pursuit of alternative or competing goals. This research and research conducted by Higgins
and others (Higgins, 2000, 2005; Kruglanski et al., 2000) suggests that process of self-regulation
are key determinants of goal commitment and behavior. Further, research by Brewer and
colleagues (Brewer, 1991; Brewer & Gardner, 1996) indicates that activated identity also plays a
key role in behavior. To gain insight into the willingness of individuals in the context of
multiteam systems to redirect actions to the pursuit of alternative or competing goals, this
dissertation investigates the affect of self-regulation on identity and goal commitment, and the
influence of these variables on (incremental) performance at the three levels of a multiteam
system: individual, component team, and MTS. The proposed model is illustrated in Figure 1.

40

MTS
Goal-Directed
Performance
(incremental
contribution)

Multi-team System
• Goal Commitment
• Identity Strength
Trait Regulatory
Mode Orientation
• Assessment
• Locomotion

(–)

(–)

Component Team
• Goal Commitment
• Identity Strength
State Regulatory
Mode Orientation
(Role Requirement)
• Assessment
• Locomotion

Component Team
Goal-Directed
Performance
(incremental
contribution)

Individual Role
• Goal Commitment
• Identity Strength

Individual Role
Goal-Directed
Performance

Figure 1. Hypothesized Model
Notes: All variables measured at the individual level
Dashed lines represent an effect hypothesized for goal commitment only.
For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this
dissertation.
41

Hypotheses Derived from Theories of Self-regulation
Goal commitment as criterion. The presence of multiple goals often creates an
“approach-approach” conflict which, in turn, hampers progress toward any of the desired end
states (Lewin, 1935, 1951; Shah et al., 2002; Zeigarnik, 1938). Even so, most individuals are
adept at regulating (i.e., self-control) to avoid the temptation of alternative goals in order to
accomplish some focal task. While a singular focus and persistence on a task is often touted as a
positive attribute because of its positive relationship to task performance, it can be problematic in
the context of a multiteam system because system performance is dependent on satisfactory
performance at all levels in the hierarchy.
Self-regulation is the psychological process by which an individual brings thinking and
behavior in line with some preexisting and desired set of standards, rules, norms, ideals, and/or
goals (Forgas, Baumeister, & Tice, 2009). To this point, theory and empirical work in the area of
self-regulation strongly suggest that these processes play a central role in goal-directed behavior
and performance (e.g., Shah et al., 2002). In addition, goal commitment, defined as “one’s
attachment to or determination to reach a goal, regardless of the goal’s origin” (Locke, Latham,
& Erez,1988, p.24), is a necessary condition for goals to influence behavior (Klein & Wright,
1994). As the psychological process through which individuals align their thinking and behavior
with pre-existing goals, it stands to reason that self-regulation is involved in the process of goal
commitment.
Locke and colleagues suggested (Locke et al., 1981), and research has shown (Klein &
Wright, 1994), that expectations of success (i.e., expectancy) and perceived value of attaining or
trying for a goal (i.e., valence) are the main factors underlying the degree to which individual’s
commit to the goal. The concept of motivational force suggests that when these two factors are

42

high, goal commitment will be strong. Individuals experiencing regulatory fit attribute greater
value to what they are doing and this value from fit increases the value of expect outcomes
(Higgins, 2000, 2002). Thus valence should be high for individuals experiencing regulatory fit.
In addition, Koenig and colleagues (2009) found that information processing is more or less
contingent upon the experience of incidental fit or non-fit. Thus expectancy should be high for
individuals experiencing regulatory fit as well because their information processing is positively
biased (i.e., they will accept success cues and reject failure cues). These arguments strongly
implicate regulatory fit in the process of goal commitment.
Regulatory fit occurs when individuals pursue goals in a manner that sustains their
regulatory orientations (Higgins et al., 2003; Higgins, 2005), thus individuals are more inclined
toward goal means that offer higher regulatory fit (Crowe & Higgins, 1997; Higgins et al., 1994).
Fit has been shown to influence judgments and decision making, feelings and emotions, attitude
and behavior change, and task performance (Higgins, 2005). Mode is the regulatory orientation
concerned with psychological state during goal pursuit (Higgins, Kruglanski, & Pierro, 2003;
Kruglanski et al., 2000). Locomotion is the dimension of regulatory mode that is concerned with
the commitment of psychological resources to initiate and maintain movement. Assessment is
the dimension that is concerned with making value comparisons among available alternatives in
order to select the highest value alternative.
Because activities and tasks vary in being more or less related to locomotion or
assessment, the effort invested by an individual in an activity or task depends on whether the
mode that is emphasized fits the requirements of the activity undertaken (Pierro et al., 2006b).
Individuals enacting a role whose activities provide the means of goal pursuit that match their
trait regulatory mode orientation experience regulatory fit. This leads to greater interest in

43

activities, stronger motivation during goal pursuit, and increased effort at goal pursuit (Higgins,
Cesario, Hagiwara, Spiegel, & Pittman, 2010; Pierro et al., 2006a; Spiegel, Grant-Pillow, &
Higgins, 2004). These same effects are known to result from strong goal commitment (Locke &
Latham, 1990, 2002).
Finally, research that has explored the impact of regulatory orientation on persuasion has
found that individuals enacting a role whose activities matches their trait regulatory mode
orientation are more apt to attend to situational cues that match and confirm this orientation, and
thus are more susceptible to being influenced by them. (e.g., Cesario et al., 2004; Cesario &
Higgins, 2008; Friedman & Forster, 2001; Hong & Sternthal, 2010; Koenig et al., 2009; Lee &
Aaker, 2004; Mannetti et al., 2010). In other words, individuals enter a situation with a
dispositional regulatory orientation and cues emerging from the situation that match (or not) this
orientation serve to reinforce (or not) this orientation; i.e., situations that match an individual’s
dispositional regulatory orientation “just feel right.” Goal commitment research has consistently
demonstrated that personal and situational factors have an effect on both expectation of goal
attainment and the attractiveness of goal attainment. Thus the regulatory fit findings based on the
combination of trait and situational cues align nicely with the expectancy theory model of goal
commitment hypothesized by Hollenbeck and Klein (1987).
Collectively, theory and research findings strongly suggest that regulatory fit is positively
related to goal commitment; thus I posit,
Hypothesis 1a. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with commitment to goals
at the individual level.

44

Hypothesis 1b. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with commitment to goals
at the component team level.
Hypothesis 1c. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with commitment to goals
at the MTS level.
Hypothesis 2a. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with commitment to goals
at the individual level.
Hypothesis 2b. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with commitment to goals
at the component team level.
Hypothesis 2c. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with commitment to goals
at the MTS level.
The arguments in support of a positive relationship between fit and goal commitment
suggest that the relationship between fit and goal commitment will be adversely affected when
the means available through an enacted role do not fit the individual’s regulatory orientation (i.e.,
a regulatory non-fit). They suggest that individuals experiencing regulatory non-fit will not
attribute greater value due to fit to what they are doing or to expect outcomes. This suggests that
valence will be negatively impacted. In addition, they suggest that expectancy will be negatively
impacted in a non-fit condition as well because information processing will be negatively biased
(i.e., individuals will accept failure cues and reject success cues). Thus I further posit,

45

Hypothesis 3a. For individuals enacting a situation-induced assessment role, trait
locomotion orientation will exhibit a negative relationship with commitment to goals
at the individual level.
Hypothesis 3b. For individuals enacting a situation-induced assessment role, trait
locomotion orientation will exhibit a negative relationship with commitment to goals
at the component team level.
Hypothesis 3c. For individuals enacting a situation-induced assessment role, trait
locomotion orientation will exhibit a negative relationship with commitment to goals
at the MTS level.
Hypothesis 4a. For individuals enacting a situation-induced locomotion role, trait
assessment orientation will exhibit a negative relationship with commitment to goals
at the individual level.
Hypothesis 4b. For individuals enacting a situation-induced locomotion role, trait
assessment orientation will exhibit a negative relationship with commitment to goals
at the component team level.
Hypothesis 4c. For individuals enacting a situation-induced locomotion role, trait
assessment orientation will exhibit a negative relationship with commitment to goals
at the MTS level.
Despite the predictions that regulatory mode non-fit will negatively impact goal
commitment, the study conducted by Pierro and colleagues (2006a) found that individuals are
most likely to attain their goals if they are high on both regulatory mode dimensions. These
authors use the argument that mode influences motivation and effort invested which then leads to
goal attainment. This mediation argument is then employed as the basis for their prediction that

46

individuals are most likely to reach their goals if they are high on both dimensions. This suggests
that the interaction effect is on motivation and effort invested, which suggests the effect is on
goal commitment. Therefore I posit,
Hypothesis 5a. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with commitment to goals
at the individual level, and this effect will be stronger for individuals high in trait
locomotion orientation relative to those low in trait locomotion orientation.
Hypothesis 5b. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with commitment to goals
at the component team level, and this effect will be stronger for individuals high in
trait locomotion orientation relative to those low in trait locomotion orientation.
Hypothesis 5c. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with commitment to goals
at the MTS level, and this effect will be stronger for individuals high in trait
locomotion orientation relative to those low in trait locomotion orientation.
Hypothesis 6a. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with commitment to goals
at the individual level, and this effect will be stronger for individuals high in trait
assessment orientation relative to those low in trait assessment orientation.
Hypothesis 6b. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with commitment to goals
at the component team level, and this effect will be stronger for individuals high in
trait assessment orientation relative to those low in trait assessment orientation.

47

Hypothesis 6c. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with commitment to goals
at the MTS level, and this effect will be stronger for individuals high in trait
assessment orientation relative to those low in trait assessment orientation.
Identity as criterion. A primary motivational premise underlying social identity is that
individuals are driven to develop and maintain a positive self-concept or positive sense of self
(Turner, 1975; Tajfel & Turner, 1979). This has also been termed the need for positive regard
(Hetts, Sakuma, & Pelham, 1999). Social identity not only denotes group membership, it denotes
an attribution of “value significance” to group membership (Tajfel, 1972). Individuals are driven
by self-enhancement motives to undertake behaviors that maximize the value of the groups they
belong to because group value has important implications for feelings of self-esteem, selfrespect, and self-worth. Perceived value of attaining or trying for a goal (i.e., valence) are the
main factors underlying the degree to which individual’s commit to a goal (Locke et al., 1981;
Hollenbeck & Wright, 1987). This suggests that goal commitment and identity are positively
related.
In addition, McCall and Simmons (1966) argue that individuals harbor many roleidentities organized into a hierarchy of prominence reflecting the role’s relative value to the
individual’s conception of an ideal self, and that commitment to any particular role-identity is
intimately related to how much their self-esteem is bound to its successful performance. Goal
commitment research has repeatedly shown that commitment to a goal is a key determinant of
performance (Klein & Wright, 1994; Locke & Latham, 1990, 2002), thus the argument made by
McCall and Simmons also suggests a positive relationship between identity and goal
commitment.

48

“Self-regulation is the process in which people seek to bring themselves (their behaviors
and self-conceptions) into alignment with relevant goals and standards” (Brockner et al., 2002, p.
7). To this point, Brewer (1991) argues that individuals respond in terms of activated identity
when faced with a conflict between goals. As discussed previously, regulatory fit occurs when
individuals pursue goals in a manner that sustains their regulatory orientations (Higgins et al.,
2003; Higgins, 2005), thus individuals are more inclined toward goal means that offer higher
regulatory fit (Crowe & Higgins, 1997; Higgins et al., 1994). Fit has been shown to influence
judgments and decision making, feelings and emotions, attitude and behavior change, and task
performance (Higgins, 2005). In addition, differences in behavioral patterns can be explained by
contextually driven shifts in the social identity portion of the self-concept (Brewer & Gardner,
1996; Hogg & Terry, 2000). This suggests that regulatory fit influences identity activation and is
consistent with the earlier argument for a regulatory fit influence on goal commitment.
Collectively these arguments suggest that identity and goal commitment will be moderately and
positively correlated at the same level in a MTS.
Research based on Brewer’s (1991) concept of optimal distinctiveness indicates that
individuals facing a social situation seek a self-concept that provides for an optimal balance
between similarity to the other members of the group (the ‘We’ in identity) and uniqueness in
comparison to salient others (the ‘I’ in identity). Importantly, this varies across individuals
situated in the same social context which makes it exceedingly difficult to predict the population
level relationship across identity levels in a MTS. While one individual in a MTS context may
find optimal distinctiveness at the component team level of identity, being similar to others due
to the equivalence of roles yet different across many other attributes of salience to the focal
individual, another individual may find optimal distinctiveness at the level of the MTS, being

49

similar to others due to the mutual commitment to a set of shared super-ordinate goals yet unique
due to differences in role enacted in the respective component teams and other attributes of
salience to the focal individual.
This argues for a positive relationship within individuals across adjacent levels at the
levels deemed to be inclusive and unique enough by the individual, and a negative relationship
between distal levels. It suggests that an individual in the first example would exhibit both a
strong component team identity and individual identity, and the individual in the latter example
would exhibit both a strong MTS identity and component team identity. In the first example, the
uniqueness attribute of most salience to them might come from a factor unrelated to the MTS
social context specifically; for example, the fact that they are the only female or person of color
in an otherwise all male or white collective of individuals. The individual in the latter example
may or may not also exhibit a strong individual identity depending upon whether or not
identification at the component team level provide a strong enough sense of distinctiveness.
Research has shown that the two dimensions of regulatory mode are each related to a
distinctly different set of personal attributes (Higgins et al., 2003; Kruglanski et al., 2000). Thus
there are important differences across the two dimensions of regulatory mode that offer
predictive potential. The assessment dimension is positively related to public self-consciousness,
private self-consciousness, and need for social comparison. Individuals high on the assessment
dimension are influenced by social norms and normative standards, and thus are concerned with
doing the right thing. Their concerned for doing the right thing can be influenced by important or
significant others; for example, by those with whom they share an identity. High assessors are
sensitive to social criticism and feel anxious during social interactions. Concerns with standards

50

and social criticism makes individuals high on the assessment dimension both high selfevaluators and high self-monitors.
All three levels of identity in a MTS provide a basis for some combination of social
comparison, self-evaluation and/or self-monitoring suggesting that assessors might have some
degree of identification at all these levels. However, Brewer (1991) argued that individuals need
to validate their self-concept and do this by seeking identities that render them similar to others
yet different and unique as well. Thus an individual avoids being identified with categories that
they find to be either too inclusive or too differentiating which suggests that assessors will not
identify strongly with all levels. Their concern for social norms, normative standards, and doing
the right thing suggests that their social motivation is focused on group welfare, which is
indicative of a collective identity bias (Brewer & Gardner, 1996; Brickson, 2000). These
arguments suggest that individuals high on the assessment dimension will identify most strongly
with groups. Therefore I posit,
Hypothesis 7a. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with component team
identity.
Hypothesis 7b. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with MTS identity.
Brewer’s (1991) argument that an individual seeks an optimal identity, one that
simultaneously satisfies the need to be similar to salient others and the need to be different
and unique, suggests that locomotors will not identify strongly with all levels. Research has
shown that locomotion is negatively associated with sensitivity to social criticism and
feeling anxious in social interactions and thus high locomotors are not high self-evaluators

51

or self-monitors (Higgins et al., 2003). High locomotors also tend to be intrinsically as
opposed to extrinsically motivated. These attributes imply that their social motivation is
focused much more on self-interests than on the welfare of the group (Brickson, 2000).
This suggests strong identification at the individual level. Theories of identity argue that
members of a component team will tend to form relational attachments to each other, to
some degree, and a prototypical role-based group identity at a minimum (Brewer &
Gardner, 1996; Prentice et al., 1994; Turner et al., 1987). Collectively these arguments
suggest that locomotion oriented individuals will exhibit a strong individual identity and
some degree of identification at the component team level. Therefore I posit,
Hypothesis 8a. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with individual identity.
Hypothesis 8b. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with component team
identity.
Performance as criterion. As signified by the hierarchical structure of goals in
multiteam systems (Mathieu et al., 2001), successful goal attainment at lower levels in a MTS is
a necessary prerequisite for goal attainment at higher levels. Because of this, under norms of
rationality (Thompson, 1967) individual’s highly committed to the goals of a focal level will be
committed to the goals of the level below. The reverse, however, is not necessarily true. Drawing
on expectancy theory (Vroom, 1964) and the model of goal commitment put forth by Hollenbeck
and Klein (1987), lack of commitment to the goals at a focal level will negatively impact
expectancy of goal attainment at levels above and thereby goal commitment at those higher
levels.

52

Goals are the object or aim of actions, and the relationship between goals and
performance is strongest when goals are deemed difficult but attainable and people are
committed to their goals (Locke & Latham, 2002). As articulated by Seijts and Latham, “highly
committed individuals exert more effort, and are more persistent toward goal attainment, than
individuals who are less committed to the goal” (2000a, p. 316). Goal commitment is a necessary
condition for goals to influence behavior (Klein & Wright, 1994), and research has repeatedly
shown that goal commitment is a key determinant of performance. Because goal commitment is
a key determinant of performance, much like goal commitment, successful performance at the
lower levels in a MTS is strongly implicated in performance success at higher levels.
Performance at a given level in a multiteam system is an amalgamation of performance resulting
from goals originating at that level and performance resulting from goals originating at lower
levels. In other words, like goals performance is nested.
Individual role directed performance exists at the lowest level in the hierarchy and
constitutes performance against goals originating exclusively at that level. Component team
directed performance exists at the middle or component team level in the hierarchy and
constitutes performance against goals originating exclusively at the component team level. Thus
total individual contribution to performance at the component team level consists of two parts,
Individual role directed performance and Component team directed performance. Illustrated
mathematically,
Total individual contribution
to performance at
=
Component Team level

Individual role
directed performance

+

Component team
directed performance

Likewise, MTS directed performance exists at the top or MTS level in the hierarchy and
constitutes performance against goals originating exclusively at the MTS level. Thus total

53

individual contribution to performance at the MTS level consists of three parts, Individual role
directed performance, Component team directed performance, and MTS directed performance.
Illustrated mathematically,
Total individual
contribution to
performance at
MTS level

=

Individual role
directed
performance

+

Component team
directed
performance

+

MTS
directed
performance

Although the vast majority of research into the concept of regulatory fit conducted to date
has investigated fit from the perspective of regulatory focus orientation (Higgins, 1997), one of
the first studies to look at regulatory fit from the perspective of mode and not focus employed the
findings from previous research based on regulatory focus fit to predict outcomes based on
regulatory mode fit (Avnet & Higgins, 2003). Avnet and Higgins argue that their study
generalizes fit beyond regulatory focus to other motivational domains, and that the phenomenon
of value from fit transfers specifically to fit based on regulatory mode orientations. This
theoretical extension to the original premise of regulatory fit theory is further established by
Higgins when he stipulates that “fit effects can be found for other orientations as well” (2005, p.
210).
Very little if any extant research has specifically investigated the relationship between
regulatory fit based on mode and task performance. However, as previously articulated, previous
research based on regulatory focus based fit has demonstrated that regulatory fit effects task
performance (Higgins, 2005). Thus I posit,
Hypothesis 9a. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with individual role
performance.

54

Hypothesis 9b. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with component team
performance.
Hypothesis 9c. For individuals enacting a situation-induced assessment role, trait
assessment orientation will exhibit a positive relationship with MTS performance.
Hypothesis 10a. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with individual role
performance.
Hypothesis 10b. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with component team
performance.
Hypothesis 10c. For individuals enacting a situation-induced locomotion role, trait
locomotion orientation will exhibit a positive relationship with MTS performance.
Relationships between Goal Commitment and Performance
As discussed in detail previously, while goal commitment at a focal level is a necessary
condition, it is not a sufficient condition for commitment to the goals of higher levels. Drawing
on expectancy theory and the Hollenbeck and Klein model once again, attaining the higher level
goals may not be attractive to a focal individual; i.e., the individual may not attribute any (or
enough) value to trying for or attaining the incremental or more super-ordinate goals at the
higher level. Further, expectancy may be lower at the higher level because the performance of
other individuals or component teams becomes a factor. Thus under norms of rationality
(Thompson, 1967) an individual highly committed to the goals of the MTS will be committed to
the goals of their component team and their individual goals, but an individual might be highly

55

committed to their individual goals, or individual goals and component team goals, and not to the
goals of the level or levels above. These arguments suggest that goal commitment at a focal level
will be positively related to incremental performance at that level regardless of whether the focal
individual is enacting an assessment oriented role or a locomotion oriented role.. Therefore I
posit,
Hypothesis 11a. For individuals enacting a situation-induced assessment role,
individual goal commitment will exhibit a positive relationship with individual role
directed performance.
Hypothesis 11b. For individuals enacting a situation-induced assessment role,
component team goal commitment will exhibit a positive relationship with
component team directed performance.
Hypothesis 11c. For individuals enacting a situation-induced assessment role, MTS
goal commitment will exhibit a positive relationship with MTS directed performance.
Hypothesis 12a. For individuals enacting a situation-induced locomotion role,
individual goal commitment will exhibit a positive relationship with individual role
directed performance.
Hypothesis 12b. For individuals enacting a situation-induced locomotion role,
component team goal commitment will exhibit a positive relationship with
component team directed performance.
Hypothesis 12c. For individuals enacting a situation-induced locomotion role, MTS
goal commitment will exhibit a positive relationship with MTS directed performance.
As previously discussed, the goal set both broadens and increases additively at
successively higher levels in the MTS hierarchy. Too exclusive a focus on the goals at one level,

56

however, can lead to a lack of focus on the goals of a higher level, a phenomenon the goal
shielding and multiple goals literature suggest is exceedingly possible (Schmidt & DeShon,
2007; Shah et al., 2002). Goal shielding theory asserts that commitment to an accessible, focal
goal deters attention away from competing goals by inhibiting the recognition of distracting or
non-confirmatory information thereby focusing cognitive resources on the focal goal related
task(s) at hand.
Shah and colleagues (Shah et al., 2002) conducted a series of empirical studies that
investigated the determinants of goal shielding and the role of goal shielding in task persistence
and task performance. More specifically, they investigated “how the activation of focal goals to
which the individual was committed inhibits the accessibility of alternative goals” (Shah et al.,
2002, p. 1261), and further, how this dynamic was exhibited vis-à-vis task persistent behavior
and performance. Several of these studies tested and showed that strong commitment to a
currently active goal set both inhibited the accessibility of alternative goals, and was positively
and significantly related to persistence at the current task. In other words, goal commitment
generally increases an individual’s tendency to inhibit alternatives.
In addition, multiteam systems are collectives of teams where all component teams
exhibit input, process, and output interdependence with at least one other team in the system
(Mathieu et al., 2001). Thus individuals must, at times, attend to the needs of other component
teams. Failure to attend to these needs can result in the sub-optimization of the overall system
and an inability of the system to achieve super-ordinate goals. This brings to mind the old adage,
“winning the battle while losing the war.”
Because of resource limitations (e.g., internal psychological resources such as cognition,
and external resources such as time) members of a multiteam system facing difficult goals

57

cannot do it all even if they wanted to. Collectively these arguments suggest that contributions to
performance at higher levels in the MTS goal structure, above and beyond (excluding) the direct
contribution made by performance impacting (measured against) lower level goals, will be
negatively related to goal commitment at the lower levels. Further, because of component team
specialization (a focal individual’s goals are more clearly aligned with other individuals on the
same component team) I would expect this relationship to only be exhibited between the two
lower levels in the hierarchy (i.e., the component team and individual levels) and the MTS level.
Therefore I posit,
Hypothesis 13a. For individuals enacting a situation-induced assessment role,
individual goal commitment will exhibit a negative relationship with MTS directed
performance.
Hypothesis 13b. For individuals enacting a situation-induced assessment role,
component team goal commitment will exhibit a negative relationship with MTS
directed performance.
Hypothesis 14a. For individuals enacting a situation-induced locomotion role,
individual goal commitment will exhibit a negative relationship with MTS directed
performance.
Hypothesis 14b. For individuals enacting a situation-induced locomotion role,
component team goal commitment will exhibit a negative relationship with MTS
directed performance.
Relationships between Identity and Performance
As discussed previously, a primary motivational premise underlying social identification
is the desire to develop and maintain a positive self-concept or positive sense of self (Turner,

58

1975; Tajfel & Turner, 1979). Thus individuals are driven by self-enhancement motives to
undertake behaviors that maximize the value of the groups they belong to because group value
has important implications for feelings of self-esteem, self-respect, and self-worth. One way that
individuals can realize value is by attaining goals (i.e., performance).
Identity affects behavior through its influence on both cognitive and motivational
mechanisms, and differences in behavioral patterns and thus performance can be explained by
contextually driven shifts in the social identity portion of the self-concept (Ellemers et al., 2002).
In addition, salience of an identity is directly related to commitment, and the more salient an
identity, the more likely it is to be invoked in an interactional situation and thus to influence
performance (Owens, Robinson, & Smith-Lovin, 2010).
Finally, perceived value of attaining or trying for a goal (i.e., valence) is one of the two
main factors underlying the degree to which individual’s commit to a goal (Locke et al., 1981;
Hollenbeck & Wright, 1987) which suggests that identity and goal commitment are positively
related. This suggests, in turn, that identity will exhibit a similar relationship with performance
as goal commitment (i.e., a positive relationship). However, while identity and goal commitment
are related, they are distinct constructs as well. Thus identity and goal commitment may
influence incremental performance differently at different levels.
As discussed in detail previously, each dimensions of regulatory mode is related to a
distinctly different set of personal attributes (Higgins et al., 2003; Kruglanski et al., 2000) and
these differences offer predictive potential. Individuals high on the assessment dimension are
concerned with doing the right thing, and their concerned for doing the right thing can be
influenced by important or significant others; for example, by those with whom they share an
identity. Attributes associated with the locomotion dimension imply that the social motivation of

59

individuals high on this dimension is focused much more on self-interests than on the welfare of
the group suggesting a strong bias toward identification at the individual level. Therefore I posit,
Hypothesis 15a. For individuals enacting a situation-induced assessment role,
component team identity will exhibit a positive relationship with component team
directed performance.
Hypothesis 15b. For individuals enacting a situation-induced assessment role, MTS
identity will exhibit a positive relationship with MTS directed performance.
Hypothesis 16a. For individuals enacting a situation-induced locomotion role,
individual identity will exhibit a positive relationship with individual role directed
performance.
Hypothesis 16b. For individuals enacting a situation-induced locomotion role,
component team identity will exhibit a positive relationship with component team
directed performance.
Mediation
The predicted relationships between regulatory fit and performance, between regulatory
fit and goal commitment, and between goal commitment and performance suggest that goal
commitment intervenes in the relation between regulatory fit and performance, and thus might
account for the relation between them. This is precisely the definition of mediation given by
Baron and Kenny: “In general, a given variable may be said to function as a mediator to the
extent that it accounts for the relation between the predictor and the criterion” (1986, p. 1176).
Therefore I posit,

60

Hypothesis 17a. For individuals enacting a situation-induced assessment role, the
relationship between trait assessment orientation and individual role directed
performance will be mediated by individual goal commitment.
Hypothesis 17b. For individuals enacting a situation-induced assessment role, the
relationship between trait assessment orientation and component team directed
performance will be mediated by component team goal commitment.
Hypothesis 17c. For individuals enacting a situation-induced assessment role, the
relationship between trait assessment orientation and MTS directed performance will
be mediated by MTS goal commitment.
Hypothesis 18a. For individuals enacting a situation-induced locomotion role, the
relationship between trait locomotion orientation and individual role performance will
be mediated by individual goal commitment.
Hypothesis 18b. For individuals enacting a situation-induced locomotion role, the
relationship between trait locomotion orientation and component team directed
performance will be mediated by component team goal commitment.
Hypothesis 18c. For individuals enacting a situation-induced locomotion role, the
relationship between trait locomotion orientation and MTS directed performance will
be mediated by MTS goal commitment.
The predicted relationships between regulatory fit and performance in combination
with the predicted relationships between regulatory fit and identity and between identity
and performance also suggest that identity intervenes in the relation between regulatory fit
and performance, and thus might account for the relation between them. Therefore I posit,

61

Hypothesis 19a. For individuals enacting a situation-induced assessment role, the
relationship between trait assessment orientation and component team directed
performance will be mediated by component team identity.
Hypothesis 19b. For individuals enacting a situation-induced assessment role, the
relationship between trait assessment orientation and MTS directed performance will
be mediated by MTS identity.
Hypothesis 20a. For individuals enacting a situation-induced locomotion role, the
relationship between trait locomotion orientation and individual role performance will
be mediated by individual identity.
Hypothesis 20b. For individuals enacting a situation-induced locomotion role, the
relationship between trait locomotion orientation and component team directed
performance will be mediated by component team identity.

62

Method Section
Participant Sample
A total of 179 participants recruited from the undergraduate population at Michigan State
University were involved in this study. Each of 20 multiteam systems was composed of nine
members, four on each of two component teams and one information liaison enacting a role that
was independent of either component team. (Note: one multiteam system was short a participant
thus the researcher enacted the role of information liaison for this system.) Participants were
randomly assigned to roles.
Leadership Development Simulator (LDS)
This study was conducted as a laboratory manipulation utilizing the Leadership
Development Simulator (LDS) to enact a task environment requiring use of a multiteam system
organizational form. The Leadership Development Simulator (LDS), jointly developed by the
United States Air Force and Michigan State University, was specifically designed to present a
complex multiteam system task environment in which participants must collaborate to
effectively manage a large number of resources in a short amount of time. In LDS, teams of up to
14 people, subdivided into two component teams and a coordination team, integrate multiple
sources of information and coordinate activities with the common objective of finding and
engaging targets located in a predefined but active environment. Importantly, LDS is
programmed to capture objective measures associated with individual behaviors and decisions,
coordinated action and collective cognitive, as well as performance at the individual, component
team, and MTS level. (See Davison et al., 2012 for a detailed description of LDS.)
LDS task environment. The task environment in the LDS consists of a grid, 16 rows (116) by 16 columns (A-P), totaling 256 cells (see Figure 2).

63

Figure 2. LDS Game Grid

64

The LDS task environment is further divided into Far West (Columns A-D), Mid West
(Columns E-H), Mid East (Columns I-L), and Far East (Columns M-P) regions, as well as
Southern (Rows 1-8) Northern (Rows 9-16), Central (Rows 5-12) and Border (Rows 1-4 and
Rows 13-16) sectors. At the start of the simulation teams were presented with a blank grid,
however, hidden throughout the 256 distinct cells of the task environment were a variety of
targets of differing characteristics and point values. Some targets, opportunities, were purely
offensive in nature and gained the team points when effectively engaged (thus successfully
engaging opportunities attacks represented a promotion oriented task), while others, threats,
attacked the team’s RPAs and cost the team points when not properly engaged (thus avoiding the
loss of RPAs represented a prevention oriented task). In addition, some targets were large, some
small, some were stationary throughout the game while others were mobile, and some required
coordinated effort to be engaged (i.e., more than one RPA) while others did not. Mobile targets
moved one space in any direction between rounds except between rounds three and four, rounds
six and seven, and rounds nine and ten when they became momentarily stationary. Thus these
rounds (four, seven, and ten) were critical action phases when the team was best able to
successfully engage mobile targets. Point values varied across targets, and targets were
permanently removed from a scenario’s task environment once they destroyed. Finally, the
team’s bases could be attacked by mobile targets that moved into Row 1, and the team lost points
when this occurred.
Prior to engaging in a scenario, teams were provided with an initial intelligence report
and given 5 minutes to engage in a planning session to determine goals, mission priorities, and
initial tactics. Teams then engaged the simulation in an episodic or round-based fashion
involving overlapping action and transition phases (Marks, Mathieu & Zacarro, 2001) delineated

65

as three sequential sub-phases: Sub-team Action (2 minute 30 seconds in length), Execution
(under a minute in length), and Analysis (1 minute in length). Each round began with the Subteam Action sub-phase during which time component team staff members processed intelligence
feedback from the last round and deployed their respective assets (each deployment is denoted a
mission). The deployed assets then interacted with the task environment during the Execution
sub-phase of a round, and feedback detailing the results of these interactions (i.e., based on the
deployment of their assets) was provide at the start of the final sub-phase, Analysis. Feedback
was used by members of the multiteam system to develop a common representation of the
environment and to develop a plan of action for the next round. Importantly, feedback needed to
be processed correctly in order to be useful – incorrect processing of feedback is a key source of
subsequent decision errors in LDS – and there was only a very short period of time to synthesize
this information into a common representation of the environment (denoted the Common
Operational Picture or COP for short) before the start of the next round.
Common Operational Picture (COP). The Common Operational Picture (COP) was a
computerized tool that enabled the team to build a graphically displayed representation of the
task environment as they collectively perceived it (Figure 3 is an illustration of a sample COP).
A member of the MTS not specifically assigned to either of the component teams, the
information liaison, acted in the capacity of team scribe building and updating the COP. The
information liaison placed icons representing the target type indicated onto the COP at the
location specified by component team staff members; they were only allowed to utilize
information as provided by component team members. Thus, component team members were
ultimate responsible for the accuracy and thoroughness of the COP.

66

Figure 3. Sample Common Operational Picture (COP)

67

Importantly, no single team member could acquire or process all of the environmental
information in LDS and thus all members had to collaborate to build an effective COP. The COP
was continuously displayed on three monitors mounted on the walls, in clear view of all
participants, and was thus available for reference throughout the exercise.
MTS assets. The assets available to the MTS consisted of both operations assets capable
of engaging targets and intelligence assets capable of passively observing the environment.
Operations assets directly engaged the environment, and thus were capable of destroying and be
destroyed by enemy targets in the environment. There were four different types of operations
assets: Strike RPAs (Remotely Piloted Aircraft), Escort RPAs, Refuel RPAs, and Info RPAs.
Strike RPAs were the only asset capable of engaging opportunities, while Escort RPAs were the
only asset capable of destroying threats. Refuel RPAs enabled other operations assets to reach
distant portions of the environment (the top half of the grid, the ‘North’). Info RPAs were used to
gather information about the environment from three adjacent locations, the cell to which they
were deployed and the cells immediately to the north and to the south of this location. All other
assets, whether an operations or an intelligence asset, could only gather information from the
location to which they were deployed. In addition, it took two Strike RPAs to capitalize on a
large opportunity, one Strike RPAs to capitalize on a small opportunity, two Escort RPAs to
destroy a large threat, and one Escort RPAs to destroy a small threat. Operations assets also
gathered intelligence information and the information gathered by operations assets was
perfectly accurate (in contrast to intelligence assets whose information accuracy was
probabilistically determined based on location as will be described next). There were four RPAs
of each type for a total of 16 operations assets.

68

Intelligence asset were capable of being deployed to a single location during a round, and
were used to gather information regarding that location in the environment. They passively
observed the environment and could not be destroyed, thus intelligence asset did not directly
contribute to the team’s score, although the information they gathered was a critical enabler of
the overall success of the MTS. There are four different types of intelligence assets labeled
Communications, Human, Allied, and Visual, and the accuracy of the information gathered by
intelligence assets was probabilistically determined based on the accuracy of the deployed asset
in that particular location. Each of the four types was 95% accurate and 5% inaccurate in one of
the four sectors of the environment, and only 5% accurate (i.e., 50% inaccurate) outside this
sector. For example, a particular scenario could be programmed such that all Communications
assets are 95% accurate (5% inaccurate) in the north sector (Rows 9-16), all Human assets are
95% accurate (5% inaccurate) in the central sector (Rows 5-12), all Allied assets are 95%
accurate (5% inaccurate) in the border sector (Rows 1-4 and Rows 13-16), and all Visual assets
are 95% accurate (5% inaccurate) in the south sector (Rows 1-8). Correspondingly, all
Communications assets would thus only be 5% accurate (95% inaccurate) in the south sector, all
Human assets would only be 5% accurate (95% inaccurate) in the border sector, all Allied assets
would only be 5% accurate (95% inaccurate) in the center sector, and all Visual assets would
only be 5% accurate (95% inaccurate) in the north sector. Intel asset – sector accuracies were a
manipulation that was predetermined for each scenario and programmed in LDS. There were
eight intelligence assets of each type for a total of 32 intelligence assets.
MTS objective and scoring. The super-ordinate objective for the MTS was to maximize
total score. Score was influenced by three types of events: points were gained with the
destruction of an opportunity, and teams lost points when their RPAs were destroyed or their

69

bases were attacked. In other words, teams could only gain points by successfully engaging
opportunities and they lost points when an RPA was destroyed or their base was attached; they
did not gain points by destroying threats. Teams were awarded four points for capitalizing on a
small opportunity and 16 points for capitalizing on a large opportunity. Teams lost eight points
for each asset destroyed and eight points for each attack on a base. Destroyed assets were
replenished at the start of each round in order to insure round by round comparability of
resources and targets are removed from the environment once they are destroyed adding to the
dynamic nature of the task environment. Targets were placed in the task environment in a
predetermined fashion at the start of a scenario thus insuring that the task environment facing
every MTS in the study was equivalent at the start of a scenario.
Specifics of the Laboratory Manipulation
MTS structure and participant roles. The nine-person MTS consisted of a four-person
operations component team and a four-person intelligence component team, with the ninth
member of the MTS serving in the capacity of information liaison, a role that was independent of
either component team (see Figure 4). The entire MTS was located in a small room with each
component team sitting at their own table at opposite ends of the room. The information liaison
sat at a table in the middle of the room between the two component teams. Thus, there was
nothing that physically prevented any member of the MTS from talking to any other member.
Importantly, the time allocated to each round was also purposefully set short to negate the
possibility of protracted negotiation or discussion and to keep cognitive load high; i.e., to force
each individual to take decisions in ‘real time’ and, thereby, mimic or portray as closely as
possible in a lab setting, the environment and situation facing a typical ‘real’ multiteam system.

70

Information
Liaison
(COP)

Information
Team
Far West

Far East
Far West

Far East

Operations
Component Team
Mid East

1 – Strike
1 – Escort
1 – Refuel
1 – Info
(each)

Intelligence
Component Team

Mid West

Mid East

Figure 4. LDS Team Structure

71

2 – Allied
2 – Comm
2 – Human
2 – Visual
(each)

Mid W est

To motivate active participation, subjects were able to earn a bonus of $10. There were
multiple ways that a participant could earn the bonus – a high individual performance score, a
high unit performance score, a high organization performance score, or high performance on a
score calculated as the equal combination of the three. Participants did not know performance
results until after the exercise was completed thus their actions were driven by their perceptions
of performance and not feedback of actual performance. This approach was based on the
findings of the multiple goals literature (e.g., Schmidt & DeShon, 2007).
Component teams. Each member of a component team had the same resources and
capabilities as all other members of that component team. Each member of the Operations
component team had control of one of each type of operations asset (i.e., one Strike RPA, one
Escort RPA, one Refuel RPA, and one Info RPA), and each member of the intelligence
component team had control of two of each type of intelligence asset (i.e., two Allied, two
Communications, two Human, and two Visual. Members of the operations component team
made 4 decisions per round while each intelligence component team member made 8 decisions
per round.
Importantly, the two dimensions of regulatory mode, locomotion and assessment, were
assumed to be prescriptively manipulated by the roles enacted in LDS. By the nature of the task
and responsibilities, operations component team members were expected to be induced into a
locomotion orientation while intelligence component team members were expected to be induced
into an assessment orientation. Note that this manipulation only addressed two of four possible
state conditions: high locomotion and high assessment. This is a limitation of the study, and an
opportunity for future research, that is discussed in the Discussion Section.

72

Information team. Information exchange was facilitated through the use of an
information team. This team was comprised of the information liaison, a member of the MTS not
specifically assigned to either of the component teams, and a randomly selected representative
from each of the two component teams. The two component team representatives assisted the
information liaison by double checking that the information relayed from their respective subteam appeared on the COP correctly. The information team had no formal authority and thus was
not capable of editing or changing asset deployment decisions made by component team staff
members (i.e., no team or member was tasked with leadership or coordination).
Two steps were taken to minimize the influence and impact of the information liaison
role. First, direction given to the information liaison during training was intended to insure they
understood their role was a passive one. As the following statement taken directly from the
training script illustrates, the direction given the information liaison was crystal clear: “only
information that is provided by the sub-team member is to be placed on the COP … nothing
more … nothing less. It is their responsibility to provide target type, location, and a rating of
confidence regarding the accuracy of the information. Never prompt them for information.”
Second, the researcher sat next to the information liaison thoroughout the whole of the
performance episode to insure that the information liaison did not deviate from this instruction.
Updating of the COP could continue through the Sub-team Action sub-phase, and the
time allocated to the Analysis sub-phase was purposely set short to force overlap of the COP
update and asset deployment activities. This constraint was an important part of the manipulation
as it required component team members to make a decision each round regarding how to best
spend their limited time during the Sub-team Action sub-phase – deploying their assets

73

effectively, which directly impacted lower level goals, or updating the COP, which had greater
implications for higher level goals.
Based on the information received from the members of the component teams, the
information liaison was also able to indicate the degree of accuracy of the information on the
COP through the use of confidence indicators that appear like “cell phone strength of reception
indication bars” at the side of each icon displayed on the COP (see the example of a COP above).
Table 1. Confidence Indicator Codes
No indicator bar

N/A

1 indicator bar

Low confidence

2 indicator bars

Medium confidence

3 indicator bars

High confidence

4 indicator bars

Ground truth

No indication provided
Target identified by intel; accuracy unverified
Not used
Target identified by intel; accuracy verified
Target identified by operations

Every MTS in the study was trained on the meaning of the confidence indicators, and a chart
containing the descriptions of the indicators was prominently displayed on the wall in plain sight
of all participants to facilitate use of a common language and to minimize confusion.
Goal structure. The super-ordinate goal of the MTS is to maximize total score (net
points). This is accomplished by accurately identifying a large number of targets, the key
contribution of the intelligence component team, and by deploying the appropriate RPAs to
destroy the targets, the key contribution of the operations component team. In support of this, the
Operations component team was tasked with maximizing the number of targets destroyed (i.e.
both large and small) while minimizing the number of assets lost to enemy attack (i.e., both
RPAs and bases). The Intelligence component team was tasked with maximizing the amount of
target information resulting from intelligence queries displayed on the COP while minimizing

74

the communication of incomplete information. At the individual performer level, members of the
Operations component team were tasked with the goal of destroying targets independently (i.e.,
small targets that do not require coordinated action) without the loss of assigned assets (i.e.,
RPAs) while Intelligence component team members were tasked with the goal of accurately
identifying targets and avoiding the underutilization of assets (i.e., non-deployments and bad
queries).
Figure 5 provides further details of the goal hierarchy and associated goal targets. The
goals and goal targets were clearly explained during the training and this information was
prominently displayed on the wall in plain sight of all participants. Specific measures of
performance against goals, documented in Table 2, are described in detail in the Measures
Section.

75

Operations Role
Operations Role
++ Small targets destroyed (count)
Small targets destroyed (count)
–– RPAs lost (count)
RPAs lost (count)

43 ne
43 nett

Intellig nce Sub-t t am
Intelligeence Sub-eeam
++ Targets displayed on COP
Targets displayed on COP
resultin f from ueries
resultingg rom qqueries
–– Targets displayed on COP
Targets displayed on COP
with ut a conf fi ence level
withoout a coniddence level

Organization

Sub-team

t
tt
ne
ne
neet
neet
0
00 n
00 n

110
0 n
nee
tt

Operations Sub-team
Operations Sub-team
++ Large & small targets
Large & small targets
destroyed (count)
destroyed (count)
–– RPAs lost & Base
RPAs lost & Base
attacks (count)
attacks (count)

Goal Level

tt
tt
ne
ne
neet
nnet
1
1
188
118

33 7
7 n
nee
tt

183 points
183 points

Super-ordinate = Maximize Total Score
+ Large & small targets destroyed (points)
+ Targets displayed on COP at correct
confidence level
– RPAs lost & Base attacks (points)
– Targets displayed on COP at incorrect
confidence level

Intelligence Rolee
Intelligence Rol
++ Targets accuately iddentified count) )
Targets accur rately i entified ( (count
–– Bad queries noot a sweet spo)t) (count)
Bad queries ( (n t a sweet spot (count)

Figure 5. Hierarchy of Goal Measures (with Performance Targets)

76

Individual

Table 2. LDS Objective Measures
Goal level
MTS*

Measure
Total Points (T5)

MTS

Targets accurately displayed on COP =
2 * (VC27 + VC28 + VC30 + VC17 +
VC18 + VC20) – (VC1 + VC2)
Targets destroyed total
(T1 + T2)
RPAs lost and Base attack
(T3 + T4)
Targets displayed on COP resulting
from queries (V9 + V14)
Targets displayed on COP at unknown
confidence resulting from queries
(V61 + V62)
Small targets destroyed
(Extracted from LDS Scores Table)

Component team
– Operations*
Component team
– Operations*
Component team
– Intelligence
Component team
– Intelligence
Individual – Ops*
Individual – Ops*
Individual – Intel
Individual – Intel

RPAs lost
(Extracted from LDS Scores Table)
True positive queries (V59)
Bad queries (i.e., not deployed to a
sweet spot) (V1 – V5)

Description of measure
Points from target destruction minus both assets lost points and bases
attacked points
Count of the number of icons displayed on the COP with accurate
confidence level indication (3 or 4 bars for verified, 1 bar for
unverified) minus icons displayed with an inaccurate indication
Count of the number of Ops missions resulting in destruction of a
target (threat or opportunity)
Count of the number of Ops missions resulting in loss of an RPA or a
Base attack
Count of accepted Intel queries resulting in an any type of icon
displayed on the Common Operating Picture
Count of accepted Intel queries resulting in an any type of icon
displayed on the Common Operating Picture at unknown confidence
(i.e., 0 bars)
Count of the number of Ops missions by RPAs assigned to the
individual resulting in destruction of a small target (threat or
opportunity)
Count of the number of Ops missions resulting in loss of an RPA
assigned to the individual
Count of Intel queries where a target is displayed (Intel Results) and
it is the actual (true) target
Count of the total number of Intel queries minus count of Intel
queries where an intel asset was sent to a sweet spot (cell where asset
type accuracy > 80%)

* - Individual contribution at this goal level extracted from LDS Scores Table

77

Training. All teams received the same comprehensive training for LDS consisting of a
20 minute illustrated slide presentation followed by 2 rounds of hands-on scripted training and
practice using a scenario specifically designed for training purposes. As identity was an
important dimension of this study, the training consisted of both organization level training
delivered to the group as a whole (e.g., slide presentation) and sub-team specific training
delivered independently, but at the same time, to the two component teams (e.g., how to deploy
their assets). However, so as not to bias identity activation, the training instructions repeatedly
emphasized the multiple ways that participants were able to earn bonus money.
To insure the consistency of training received by all teams the illustrated slide
presentation was automated and used pre-recorded voice over, and the hands-on training was
conducted utilizing a written protocol and script. The hands on training instructed participants on
the ins and outs of asset deployment, how to place target icons on the COP and assign a
confidence indicator (CI) to them, how to read intelligence feedback reports, and the different
methods they might use to ascertain the accuracies (“sweet spot” sectors) of their intelligence
assets (e.g., using a fixed target identified by an operations asset in a previous round as a test
target for intel asset types in later rounds). Finally, a survey to check the effectiveness of the
training was completed by all participants prior to the start of the simulation (see Appendix C).
LDS scenario specifics. After the training, teams participated in a total of 15 rounds of
LDS broken into two time periods. The first time period comprised five rounds and the second
time period comprised ten rounds. This longitudinal approach allowed for the temporal
separation of measures, and resetting of the task environment for Time Period 2 to insure
performance measurement was consistent. As previously delineated, each round was around 4

78

minutes in length thus Time Period 1 took approximately 30 minutes to complete all in (i.e.,
including pre-game intelligence briefing and planning), and Time Period 2 ran under an hour.
Time Period 1. The information provided in the pre-scenario intelligence briefing for
Time Period 1 was as follows: “cluster of large targets somewhere in Rows 13 thru 16, middle of
the grid contain a mixture of small, mobile and fixed targets, and Rows 1 thru 4 are empty.”
Further, in this time period, all Visual and Communications assets were 95% accurate (5%
inaccurate) in the border sectors (rows 1-4 and rows 13-16), and all Allied and Human assets
were 95% accurate (5% inaccurate) in the central sector (rows 5-12).
Time Period 2. As previously explained, the task environment was reset for time period 2
to insure performance measurement was consistent (i.e., every RP / MTS faces the same
environment at the beginning of the performance measurement period). The information
provided in the pre-scenario intelligence briefing for Time Period 2 differed slightly from the
Time Period 1 briefing: “large targets in the top few rows of the grid, middle of the grid contain a
mixture of small, mobile and fixed targets, some targets in Rows 5 through 8 are headed south,
and Rows 1 and 2 are empty.” In addition, Time Period 2 included an ‘unknown’ target as a
distraction to create additional cognitive load. Specifically, the icon for one of the known target
types had been replaced by the letter ‘X’ to hide its identity. Thus, unknown target ‘X’ was the
known target type it replaced, only the visual representation on the game grid had been changed.
Teams were briefed on the unknown target as part of intelligence briefing given just prior to the
last 10 rounds. They were also told that intelligence accuracies may or may not be different in
this time period than they were in the time period just completed (they actually were), thus, they
needed to be on guard for this and adapt accordingly. This was intended to increase the
assessment salience and involvement of the intelligence roles (i.e., need to constantly analyze)

79

while resetting every multiteam system in the study to the same baseline state at the start of time
period 2.
In Time Period 2, all Visual assets were 95% accurate (5% inaccurate) in the central
sector (rows 5-12), all Communications assets were 95% accurate (5% inaccurate) in the border
sectors (rows 1-4 and rows 13-16), all Allied assets were 95% accurate (5% inaccurate) in the
south sector (rows 1-8), and all Human assets were 95% accurate (5% inaccurate) in the north
sector (rows 9-16).
Finally, commitment to goals is only a truly meaningful construct when goals are difficult
(Hollenbeck, Williams, & Klein, 1989). Thus stretch goals were assigned – overall MTS goals,
goals for each of the component teams, and individual role goals – based on extant normative
data on performance from teams that had previously experienced LDS. (Figure 5 illustrates the
goal hierarchy in LDS and documents the assigned goals.) Goals were discussed and delineated
during both the training and the intelligence briefing held prior to each time period, and
participants documented their goal targets in the survey conducted prior to start of the first round.
Surveys. At the time of registration, a week or more prior to arriving for the study,
research participants completed a survey to measure chronic individual differences in regulatory
mode and self-concept. Trait regulatory mode was measured this using the 24-item Regulatory
Mode Questionnaire (RMQ) developed by Kruglanski and colleagues, the most widely used
measure of regulatory mode in the literature (Kruglanski et al., 2000; included as Appendix A).
Twelve of the items are specific to assessing the assessment dimension of regulatory mode, and
twelve items are specific to assessing the locomotion dimension of regulatory mode. Trait selfconcept was measured using the 15-item Levels of Self-concept Scale (LSCS) developed by
Selenta and Lord (2005; included as Appendix B). The LSCS comprises three subscales

80

measuring self-concept across three levels; the comparative identity subscale measures selfconcept at the individual-level, the concern for others subscale measures self-concept at the
relational-level, and the group achievement focus subscale measures self-concept at the
collective-level.
After the intelligence briefing for Time Period 2 but prior to actual game play (“Time 1”),
RPs completed a survey consisting of six scales to measure their goal commitment and degree of
identification at each of the three levels in the multiteam system (i.e., individual, component
team, and multiteam system levels). Three of these scales are adapted from the nine item
Hollenbeck and colleagues goal commitment scale (Hollenbeck, O’Leary et al., 1989; see
Appendix D). Individual identity was measured utilizing a 5-item scale adapted from the
comparative identity subscale of the Levels of Self-concept Scale, LSCS (Johnson et al., 2006;
see Appendix E). Two eight item scales adapted from Roccas and colleagues group identification
scale (Roccas, Sagiv, Schwartz, Halevy, & Eidelson ,2008) were employed to measure identities
at the component team and the multiteam system levels (see Appendix F).
As a check the effectiveness of the enacted role regulatory orientation manipulation, RPs
completed a short survey consisting of the RMQ items altered to reference current time (i.e.,
“right now”) at the end of Time Period 2 (see Appendix G). Lastly, prior to the start of Time
Period 1 (“Time 0”), but after the pre-scenario intelligence briefing, RPs completed a survey
instrument developed to provide feedback regarding the effectiveness of the training (see
Appendix C). This latter survey was collected solely to check the proficiency of the training and
was not specifically used in any analyses.
Overview of Analysis

81

This study investigated phenomena at the individual level in the context of a multi-team
system. Thus group membership effects (i.e., lack of independence) were a potentially important
source of variance for the phenomenon of interest. I conducted ANOVAs to test for lack of
independence and where the evidence supported lack of independence (i.e., meaningful variance
exists at the group levels of the model) I grand mean centered study variables before utilizing
OLS regression to test the hypotheses. Interaction terms employed to test moderation were
created from the main effect variables grand mean centered to minimize issues associated with
nonessential multicollinearity (per Cohen, Cohen, West and Aikens, 2003, p. 267). The
regression coefficients resulting from this methodology represent a combination of individuallevel effects and the effects of membership in a group. Finally, the hypothesized relationships are
all directional predictions therefore single tailed tests of significance were employed where
required to alleviating concerns related to power.
Measures. The model of the hypothesized relationships (see Figure 1) investigated in this
study contained three criterion variables and eight predictor variables, all measured at the
individual level, and two situation-induced regulatory mode conditions resulting from the
requirements of specific roles. Cronbach’s Alpha was calculated for all self-report scales used in
this study; the internal consistency of these scales ranged from acceptable to excellent (See Table
3).

82

Table 3. Scale Reliabilities
Scale

Items

Cronbach’s
Alpha

Trait assessment (RMQ)

12

.85

Trait locomotion (RMQ)

12

.84

Comparative identity subscale (Trait individual identity)

5

.76

Concern for others subscale (Trait relational identity)

5

.74

Group achievement focus subscale (Trait collective identity)

5

.68

Individual level goal commitment

5

.84

Component team level goal commitment

5

.87

MTS level goal commitment

5

.88

Individual role identity

5

.85

Component team identity

8

.88

Multiteam system identity

8

.92

State assessment (RMQ-based)

12

.82

State locomotion (RMQ-based)

12

.87

Objective performance (criterion variable at level-1). In order to mirror the goal
hierarchy in a MTS, this study employed a three tiered approach to performance measurement.
As this study was interested in investigating the determinants of performance focus, it was
important that performance contribution from behaviors directed primarily at each specific level
in the goal hierarchy be delineated. Therefore, the incremental performance targeted at the
specific goals of a focal level was used in the respective analyses of that level. In addition, to
allow for the temporal separation of predictor (e.g., goal commitment and identity) and criterion

83

variables, performance was measured over the ten rounds of Time Period 2 only. The end of this
time period is denoted “Time 2.”
Specifically, performance contribution by an Operations component team member at the
individual level was measured as the number of small targets destroyed by the focal member
minus the number of RPAs under their control that were destroyed, and incremental performance
contribution at the component team level was measured as the number of large targets destroyed
with the involvement of RPAs under a focal member’s control minus the number of base attacks.
Operations component team member incremental performance contribution at the MTS level was
measured as the points resulting from the deployment of assigned assets minus the number of
targets destroyed plus the number of RPAs destroyed and base attacks. (Note: incremental
performance contribution at the MTS level represents the weighted point value net of the activity
oriented measure of performance at the component team level.) Performance of Intelligence
component team members at the individual level was measured as the number of targets
accurately identified by assigned assets (i.e., true positives) minus the number of bad queries
(i.e., missions that do not target the intelligence asset’s sweet spot zone), and incremental
performance contribution at the component team level was measured as the number of targets
resulting from a focal member’s queries displayed on the COP minus the number of targets
displayed on the COP without a confidence level indication. Finally, Intelligence component
team member incremental performance contribution at the MTS level was measured as the
number of targets displayed on the COP with the correct confidence indication (i.e., accurately
displayed) minus the number of targets displayed on the COP with an incorrect confidence
indication (i.e., inaccurately displayed). Figure 5 presents this information in graphic form and
Table 2 presents it in table format.

84

Chronic regulatory mode (predictor variables at level-1). Chronic regulatory mode
orientation was measured at the time that participants register for the study thus providing
temporal separation between this measure and participation in the experiment. As discussed
previously, I plan to measure chronic regulatory mode orientation using the 24-item Regulatory
Mode Questionnaire (RMQ) developed by Kruglanski and colleagues (Kruglanski et al., 2000;
Appendix A). Thorough psychometric testing of these scales was conducted and reported in
Kruglanski et al. (2000):
“Overall, the correlation between the two indexes was quite small, but it was
reliable in the large omnibus sample, r(4256) = .11, p < .001. The very small
amount of overlapping variance (about 1%) is consistent with the proposition that
the locomotion and assessment scales are measuring distinct psychological
dimensions.” (p. 802)
“… we found that our Assessment and Locomotion Scales are unidimensional and
possess satisfactory degrees of internal consistency and temporal stability. These
properties were demonstrated across numerous replications, including a crosscultural replication with an Italian sample. Our Locomotion and Assessment
Scales satisfactorily distinguished between groups that on a priori grounds may be
expected to emphasize one tendency more so than another.
Our Locomotion and Assessment Scales related in a theoretically predicted way
to several individual difference constructs and demonstrated discriminant validity
in regard to other constructs …” (p. 812).
Chronic self concept (control variables at level-1). Three levels of chronic self-concept
were also measured at the time that participants register for the study thus providing temporal

85

separation between this measure and participation in the experiment. As discussed previously,
chronic self-concept was measured using the 15-item Levels of Self-concept Scale (LSCS)
developed by Selenta and Lord (2005; included as Appendix B). The LSCS contains multiple
subscales at each of the three levels of self-concept, however in accordance with the technique
employed by Johnson and colleagues, only the first scale at each level was used in this study as
these are “most indicative of their respective self-concept level” (Johnson et al., 2006, p. 180).
As discussed by Johnson et al. (206, p. 180):
LSCS items are “based on a set of constructs reflected in the literature on selfconcept levels (e.g., Brewer & Gardner, 1996). Evidence for the validity of the
LCSC was established using a two-step process. First, factor analytic evidence
established the distinction among the three self-concept levels and specific item–
factor associations. Second, regression analyses involving values (Schwartz,
1992), self-consciousness (Scheier & Carver, 1985), masculinity–femininity
(Spence & Helmreich, 1978), and individuals’ sex (Gabriel & Gardner, 1999)
demonstrated the convergent and discriminant validity of the self-concept
subscales.”
Goal commitment (predictor variable at level-1). Three scales were used, one per
hierarchical level in a multiteam system (i.e., one per goal level), adapted from the Hollenbeck
and colleagues goal commitment scale (Hollenbeck, O’Leary et al., 1989; see Appendix D).This
scale has been described as “a general, flexible measure in that it can be used to assess goal
commitment regardless of goal origin or timing” (Seijts & Latham, 2000, p. 320), and is one of
the most commonly used measures of goal commitment (Klein et al., 1999; Wright, O’Leary,
Cortina, Klein, & Hollenbeck, 1994).

86

The psychometric properties of the original scale were thoroughly tested in a manner that
is consistent with recommendations for construct validity analysis (see Hollenbeck, O’Leary et
al., 1989). Analyses using a single sample (Hollenbeck, Williams, & Klein, 1989) and a
collection of three independent samples (from Hollenbeck, Williams, & Klein, 1989; Klein,
1987, and Wright, 1987) strongly suggest use of either a four or seven item version of the scale.
Both of these versions reveal uni-dimensionality when factor analyzed, and good internal
consistency reliabilities (Cronbach’s alpha of .71 for the four item scale in the multi-sample
analysis and alphas of .88 and .80 in the two studies for the seven item scale).
An analysis conducted by Klein and colleagues (1999) combined the results of 17
independent samples and 2,918 respondents. Utilizing meta-analytic techniques and
confirmatory factor analysis, this study found that a five item version of the scale maximized fit
of the uni-dimensional goal commitment model. Based on this result, I plan to employ the five
item scale. Further, as was used in the construct validation studies, I will employ a 7-point Likert
response scale anchored by strongly agree / strongly disagree, with negative items recoded so
that a high score on the scale is indicative of high goal commitment. This scale was used to
measure both degree of commitment to component team goals and to MTS goals.
Identity (predictor variables at level-1). As with goal commitment I used three scales to
measure identities at the different hierarchical levels in a multiteam system (see Appendices D
and E). At the individual level I employed a 5-item scale adapted from the comparative identity
subscale of the Levels of Self-concept Scale (LSCS; Johnson et al., 2006). Comparative identity
emphasizes one’s abilities, performance, and general standing relative to that of others. The
LSCS is based on the constructs reflected in the literature on self-concept levels (e.g., Brewer &
Gardner, 1996). Factor analysis was used to establish the distinction among the three self-

87

concept levels in the LSCS, and regression analysis involving several different theoretically
related and unrelated constructs was used to demonstrate the convergent and discriminant
validity of the self-concept subscales (Johnson et al., 2006). Lastly, Johnson and colleagues
reported that the comparative identity subscale exhibited a high degree of internal consistency (α
= .90).
Two eight item scales adapted from Roccas and colleagues group identification scale
(Roccas et al.,2008) were employed to measure identities at the component team and the
multiteam system levels (see Appendix F). The Roccas et al. scale integrates conceptions from
four closely related perspectives – social identity, individualism–collectivism, nationalism–
patriotism, and identification with organizations – into a conceptualization of identity consisting
of four distinct modes (sub-dimensions): “importance (how much I view the group as part of
who I am), commitment (how much I want to benefit the group), superiority (how much I view
my group as superior to other groups), and deference (how much I honor, revere, and submit to
the group’s norms, symbols, and leaders)” (Roccas et al., 2008, p. 280). The scale utilized a 7point response scale anchored by strongly agree / strongly disagree.
The eight items that were used encompass two of the four sub-dimensions of the scale,
Importance (four items) and Commitment (four items). Importance represents the degree to
which an individual perceives the group as an important part of their self-definition, and the
degree to which they define themselves in terms of the group (i.e., think of the group in terms of
“we” rather than “they”). “It follows directly from Tajfel’s definition of social identity and is
consistent with the emphasis on the cognitive aspects of identity in self-categorization theory
(e.g., Turner et al., 1987). This mode also captures an element of Triandis’s definition of
collectivism (“closely linked individuals who see themselves as parts of one or more

88

collectives”). It is especially related to horizontal collectivism, which refers to shared goals and a
sense of similarity to other group members.” (Roccas et al., 2008, p. 283).
Commitment refers to the desire to contribute to the welfare of the group. Like
Importance, it is closely linked to Tajfel’s definition of social identity, thus studies from the
social identity perspective frequently emphasize it, and it refers specifically to horizontal
collectivism. In addition, “This mode of identification also captures a key element in Triandis’s
definition of collectivism (“willing to give priority to the goals of these collectives over their
own personal goals”). … Finally, this mode is central to organizational research, especially to
studies on organizational commitment (e.g., Meyer & Allen, 1997; Mowday, 1998; Mowday et
al., 1979)” (Roccas et al., 2008, p. 284).
In a series of studies conducted by Roccas and colleagues (Roccas et al., 2008) on both
work and military samples, this scale was shown to fit the proposed four mode model better than
alternative one factor and two factor models. For example, three samples drawn from work
2

2

organizations yielded the following fit indices: χ / df = 2.87 (χ = 849.65, df = 296), CFI = .92,
and RMSEA = .052. Although the eight items to be adapted for use in this study represent two of
these four factors, Importance and Commitment, I expect that a one factor model will adequately
fit the data because these two factors correlated highly in all of the reported studies (e.g., r = .79,
.76, .77).
Enacted role (state) regulatory mode (manipulation check). The structure employed in
this study was purposely chosen because it places half of the participants in a prescriptively
locomotion oriented task role (i.e., “Operations team member”) and the other half in a
prescriptively assessment oriented task role (i.e., “Intelligence team member”). As a check of
this manipulation, after completing all 15 rounds RPs completed a short survey consisting of the

89

RMQ items adjusted to reflect current state orientation to check the effectiveness of the
regulatory mode manipulation (see Appendix G). Sample survey questions are: “During the
scenario just completed, I didn’t mind doing things even if they involved extra effort” and “I am
a "workaholic" will read “I could easily have continued on playing the first game” (two questions
from the Locomotion scale). In addition, the instructions for the RMQ were altered and read:
“Take a few moments to reflect back on your participation in the scenario of LDS which you
have just completed. Then, please answer the following question with only these reflections in
mind; i.e., how you feel right now based on your participation in the scenario.”

90

Results
Table 4 presents descriptive statistics and correlations for all key study variables across
both situation-induced conditions, Table 5 presents the same for the situation-induced assessment
condition only, and Table 6 for the locomotion condition only. Scale reliability (Cronbach’s
Alpha) was calculated for all self-reported scales used in this study and documented along the
diagonal of Tables 4, 5 and 6. Table 7 contains goodness-of-fit statistics resulting from
confirmatory factor analysis employing LISREL 8.8 (Jöreskog & Sörbom, 1996) for the scales
used in this study, and Table 8 contains the results of tests conducted to determine the situationinduced regulatory mode state of research participants in the two study conditions. Tables 9, 10
and 11 contain the results of one-way random effects ANOVAs conducted to determine the
proportion of the variance explained at the level of the multiteam system (i.e., degree of
interdependence). Tables 12 through 24 contain the results of the OLS regression tests of the
study hypotheses. All reported point estimates are standardized and tests of significance are twotailed unless otherwise specified.
Measurement Model Goodness-of-Fit
Prior to conducting the analyses the goodness-of-fit properties of the measurement
models for all latent constructs were investigated utilizing confirmatory factor analysis (CFA);
Table 7 presents a summary of the results. All analyses were conducted using LISREL 8.8
(Jöreskog & Sörbom, 1996). The chi-squared goodness of fit statistic is sensitive to sample size
and to violations of multivariate normality assumptions, therefore, four other fit statistics
recommended in the measurement literature were used (see Bollen & Long, 1993;
http://davidakenny.net/cm/fit.htm).

91

Table 4. Descriptive Statistics and Correlations (both conditions)
Variables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.

MTS-level performance contribution
Component team-level performance contribution
Individual role performance contribution
State assessment (T2)
State locomotion (T2)
MTS goal commitment (T1)
Component team goal commitment (T1)
Individual goal commitment (T1)
MTS identity (T1)
Component team identity (T1)
Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

Mean

s.d.

1

0.00
0.00
0.00
3.60
4.14
4.75
4.95
4.77
4.76
5.40
2.57
4.05
4.42
3.34
4.02
4.44

1.00
1.00
1.00
0.74
0.84
1.15
1.12
1.20
1.14
0.97
0.92
0.52
0.42
0.70
0.70
0.58

0.62
0.40
0.06
0.21
0.04
0.04
-0.08
0.17
0.14
-0.05
-0.04
-0.12
0.02
-0.04
-0.04

2
**
**

0.17
0.09
** 0.18
0.12
0.11
-0.04
*
0.20
†
0.24
-0.07
0.18
0.04
0.08
-0.05
0.03

3

*
*

*
**
*

Notes. N = 159 individual-level observations from 20 MTSs. Scale reliabilities are on the diagonal.
** p < .01 (two-tailed tests).
* p < .05 (two-tailed tests).
† p < .10 (two-tailed tests).

92

0.07
0.20 *
-0.08
-0.03
0.01
-0.07
-0.02
0.05
0.02
-0.05
0.01
0.07
-0.08

4

0.82
0.40
0.31
0.43
0.35
0.19
0.26
0.27
0.24
0.14
0.30
0.32
0.14

5

**
**
**
**
*
**
**
**
†
**
**
†

0.87
0.46
0.50
0.36
0.37
0.44
-0.04
0.21
0.08
0.07
0.02
0.23

**
**
**
**
**
**

**

Table 4 (cont’d)
Variables

11.
12.
13.
14.
15.
16.

s.d.

6

MTS goal commitment (T1)
Component team goal commitment (T1)
Individual goal commitment (T1)
MTS identity (T1)
Component team identity (T1)
Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

4.75
4.95
4.77
4.76
5.40
2.57
4.05
4.42
3.34
4.02
4.44

1.15
1.12
1.20
1.14
0.97
0.92
0.52
0.42
0.70
0.70
0.58

0.91
0.84
0.62
0.65
0.54
-0.02
0.27
0.16
-0.01
-0.04
0.16

Variables

6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.

Mean

Mean

s.d.

11

2.57
4.05
4.42
3.34
4.02
4.44

0.92
0.52
0.42
0.70
0.70
0.58

Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

93

7
**
**
**
**

0.87
0.60
0.54
0.56
0.10
** 0.25
* 0.13
0.11
0.06
* 0.20

8

**
**
**
**
†

12

0.85
0.14 † 0.68
0.13
0.45 **
0.38 ** 0.28 **
0.36 ** 0.20 *
0.09
0.51 **

*

9

10

0.87
0.31 ** 0.92
0.27 ** 0.67 ** 0.88
0.13
0.28 ** 0.22 ** 0.33 **
0.17 * 0.10
0.12
0.14 † 0.07
0.12
-0.05
0.10
0.15 † 0.10
0.16 *

13

14

15

0.74
0.22 ** 0.76
0.18 * 0.58 ** 0.85
0.36 ** 0.23 ** 0.20 * 0.84

Table 5. Descriptive Statistics and Correlations (high state assessment condition only)
Variables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

MTS-level performance contribution
Component team-level performance contribution
Individual role performance contribution
MTS goal commitment (T1)
Component team goal commitment (T1)
Individual goal commitment (T1)
MTS identity (T1)
Component team identity (T1)
Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

Mean

s.d.

1

-5.53
4.53
-24.62
4.51
4.71
4.61
4.66
5.29
2.56
4.04
4.38
3.34
4.01
4.39

9.70
5.31
14.66
1.15
1.13
1.21
1.08
0.97
0.93
0.53
0.42
0.73
0.74
0.60

0.47
0.22
0.00
0.05
-0.17
0.20
0.08
-0.15
-0.36
-0.29
-0.10
-0.13
-0.24

2
**
*

†

**
*

*

0.29 **
0.08
0.09
-0.16
0.18
0.29 *
-0.15
0.04
0.02
-0.04
-0.13
-0.12

3

4

-0.08
0.06
0.13
-0.07
-0.06
0.13
-0.02
-0.04
0.11
0.15
-0.17

0.88
0.83
0.60
0.59
0.39
-0.02
0.18
0.10
-0.05
-0.17
0.17

**
**
**
**

Notes. N = 79 individual-level observations from 20 4-member Intelligence component teams. Scale reliabilities are on the diagonal.
** p < .01 (two-tailed tests).
* p < .05 (two-tailed tests).
† p < .10 (two-tailed tests).

94

Table 5 (cont’d)
Variables

10.
11.
12.
13.
14.

s.d.

5

Component team goal commitment (T1)
Individual goal commitment (T1)
MTS identity (T1)
Component team identity (T1)
Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

4.71
4.61
4.66
5.29
2.56
4.04
4.38
3.34
4.01
4.39

1.13
1.21
1.08
0.97
0.93
0.53
0.42
0.73
0.74
0.60

0.86
0.55
0.54
0.50
0.15
0.21
0.05
0.11
-0.02
0.24

Variables

5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

Mean

Mean

s.d.

10

4.04
4.38
3.34
4.01
4.39

0.53
0.42
0.73
0.74
0.60

0.65
0.41
0.36
0.31
0.47

Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

95

6
**
**
**
†

*

0.83
0.25
0.21
0.20
0.25
0.05
0.14
-0.10
0.23

11
**
**
**
**

7

*
†
†
*

*

8

9

0.90
0.54 ** 0.86
-0.02
-0.01
0.85
0.03
0.28 * 0.17
0.03
-0.04
0.21 †
0.01
0.09
0.50 **
-0.07
0.07
0.47 **
0.01
0.11
0.31 **

12

13

0.78
0.21 † 0.78
0.14
0.69 ** 0.87
0.30 ** 0.35 ** 0.21

14

†

0.85

Table 6. Descriptive Statistics and Correlations (high state locomotion condition only)
Variables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

MTS-level performance contribution
Component team-level performance contribution
Individual role performance contribution
MTS goal commitment (T1)
Component team goal commitment (T1)
Individual goal commitment (T1)
MTS identity (T1)
Component team identity (T1)
Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

Mean

s.d.

1

3.38
1.46
1.14
4.99
5.19
4.93
4.87
5.50
2.57
4.07
4.46
3.35
4.03
4.48

26.31
2.61
2.72
1.10
1.08
1.17
1.19
0.97
0.91
0.52
0.41
0.68
0.66
0.56

0.77
0.58
0.07
0.04
0.01
0.15
0.20
0.05
0.27
0.05
0.14
0.06
0.18

2
**
**

†
*

0.06
0.16
0.13
0.07
0.21
0.19
0.02
0.31
0.07
0.21
0.04
0.20

3

†
†
**
†
†

4

-0.09
-0.14
-0.12
-0.08
0.03
-0.03
0.07
-0.05
-0.10
-0.02
0.02

0.88
0.83
0.62
0.71
0.68
-0.02
0.36
0.19
0.02
0.10
0.13

**
**
**
**
**
†

Notes. N = 80 individual-level observations from 20 4-member Operations component teams. Scale reliabilities are on the diagonal.
** p < .01 (two-tailed tests).
* p < .05 (two-tailed tests).
† p < .10 (two-tailed tests).

96

Table 6 (cont’d)
Variables

10.
11.
12.
13.
14.

s.d.

5

Component team goal commitment (T1)
Individual goal commitment (T1)
MTS identity (T1)
Component team identity (T1)
Individual role identity (T1)
Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

5.19
4.93
4.87
5.50
2.57
4.07
4.46
3.35
4.03
4.48

1.08
1.17
1.19
0.97
0.91
0.52
0.41
0.68
0.66
0.56

0.86
0.64
0.54
0.60
0.05
0.29
0.19
0.11
0.16
0.13

Variables

5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

Mean

Mean

s.d.

10

4.07
4.46
3.35
4.03
4.48

0.52
0.41
0.68
0.66
0.56

Trait collective identity (pre-lab)
Trait relational identity (pre-lab)
Trait individual identity (pre-lab)
Trait assessment (pre-lab)
Trait locomotion (pre-lab)

97

6
**
**
**
*
†

7

8

9

0.86
0.35 ** 0.93
0.32 ** 0.79 ** 0.89
0.05
-0.01
0.00
0.30 ** 0.38 ** 0.38 **
0.28 *
0.14
0.27 *
0.13
0.13
0.15
0.00
0.02
0.12
0.05
0.17
0.20 †

11

12

13

0.72
0.49 ** 0.70
0.19 † 0.24 * 0.74
0.22 † 0.45 ** 0.83
0.08
0.54 ** 0.43 ** 0.10
0.19

0.85
0.11
0.05
0.26 *
0.24 *
-0.14

14

†

0.83

Table 7. Confirmatory Factor Analysis of Scales Employed in this Study
X2 / df

RMSEA

90% Confidence
Interval

Closeness of Fit

SRMR

NNFI

Trait Regulatory Mode Questionnaire (2 factor)

1.60

0.062

(0.050, 0.073)

p = 0.05

0.084

0.93

Multiteam System Goal Commit at T1
Component Team Goal Commit at T1
Individual Role Goal Commit at T1

1.56
1.17
1.30

0.060
0.033
0.044

(0, 0.100)
(0, 0.081)
(0, 0.088)

p = 0.32
p = 0.67
p = 0.54

0.031
0.033
0.034

0.99
0.99
0.99

Identity at T1 (3 factor)

1.63

0.063

(0.045, 0.080)

p = 0.10

0.066

0.97

State Regulatory Mode Questionnaire (2 factor)

1.44

0.053

(0.038, 0.066)

p = 0.36

0.071

0.97

98

Steiger's (1990) Root Mean Square Error of Approximation (RMSEA) and Hu and
Bentler’s (1999) Standardized Root Mean Square Residual (SRMR) are measures of the average
standardized residuals of the predicted covariance matrix from the observed covariance matrix.
As absolute measure of (bad) fit they presume that the best fitting model has a fit statistic of
zero, thus, lower values for these statistics are better. MacCallum, Browne and Sugawara (1996)
suggest that RMSEA values less than .01, .05, and .08 indicate excellent, good, and acceptable fit
to the data, respectively while Browne and Cudeck (1993) set the rule-of-thumb for acceptable
fit at less than .10. Further, the closeness of fit measure is a one-sided test of the null hypothesis
that the population RMSEA is .05; a non-significant p (> .05) indicates a close fitting model. A
value less than .08 for SRMR is generally considered a good fit (Hu & Bentler, 1999).
The Tucker-Lewis (1993) or Non-normed Fit Index (TLI or NNFI) is an incremental
measure of fit selected because it is less affected by sample size than many of the other measures
(Bentler, 1990; Bollen, 1990; Hu & Bentler, 1995, 1998). In addition, more parsimonious (i.e.,
more constrained) models are rewarded by an increase in the fit index because NNFI takes the
degrees of freedom of the specified model and the independence model into consideration.
2

Analogous to R , a value of one indicates a prefect fit, values over .97 indicate excellent fit,
values between .95 and .97 indicate good fit, and values between .90 and .95 indicate acceptable
fit (Hu & Bentler, 1999).
As reported in Table 7, all scales employed in this study exceed both the RMSEA and the
SRMR metric for acceptable fit with the exception of trait regulatory mode (SRMR = .084),
several scales exceed the RMSEA and SRMR metrics for good fit, and all the 90% confidence
intervals for RMSEA include .05. In addition, the NNFI statistic indicates that all scales have
good to excellent fit except for trait regulatory mode which has an acceptable fit (NNFI = .93).

99

Finally, Hu and Bentler (1999) recommend a 2-index strategy using a cutoff value close to .95
for NNFI in combination with a cutoff value close to .09 for SRMR. All scales employed in this
study exceed this metric with the exception of trait regulatory mode which is arguably close.
Therefore, it is reasonable to employ these scales as representations of their respective
underlying latent constructs.
Test of Study Conditions
The hypotheses in this study are dependent upon the state regulatory mode of the research
participants and, thus, the study was designed to induce one half of the participants into a high
assessment state and one half of the participants into a high locomotion state. Specifically, the
design expectation was that placing individuals into Intelligence component team roles would
induce a high assessment state (situation-induced assessment condition) and that placing
individuals into Operations component team roles would induce a high locomotion state
(situation-induced locomotion condition). Independent samples means t-tests were employed to
compare participant’s role (state) regulatory modes across the two component teams. As
documented in Table 8, there was no significant difference in state assessment between the two
component teams (difference in means = -.071, t(157) = -.604, p = n.s.) but there was a
significant difference in state locomotion (difference in means = -.664, t(157) = -5.44, p < .01).
Table 8. Component team-level Independent-samples Means Tests of Regulatory Mode
Trait
Component Team
Mean: Intelligence*
Operations*
Difference

Assessment
4.011
4.030
-.020

Locomotion
4.394
4.484
-.090

Significance
(2-tailed)

t(157) = -.176
p = .860

t(157) = -.975
p = .331

State (enacted role)
Assessment
Locomotion
3.564
3.805
3.635
4.470
-.071
-.664
t(157) = -.604
p = .547

* - Levene's Test for Equality of Variances was not significant.
100

t(157) = -5.440
p = .000

These findings suggest that assignment to an Operations component team role induced a
high locomotion state whereas assignment to an Intelligence component team role may not have
induced a high assessment state. Therefore, only the hypotheses for the situation-induced high
locomotion condition were tested solely under the assumption that the manipulation was
successful. The hypotheses for the situation-induced high assessment condition were tested
twice. The situation-induced high assessment condition hypotheses were first tested under the
assumption that the manipulation worked as intended; i.e., that the manipulation check was
erroneous. The hypotheses were then tested employing the assumption that the manipulation
check was accurate; i.e., that the manipulation had not worked as intended. The second set of
tests utilized interaction terms created by multiplying centered study variables with the centered
manipulation check variable for state assessment. (Note: post hoc analyses employing both of the
self-reported measures of regulatory mode state condition are included in the discussion.)
Lack of Independence Tests
Table 9 documents that 0.2% of the variance in individual role goal commitment, 12.6%
of the variance in component team goal commitment, and 13.5% of the variance in MTS-level
goal commitment occurred at the group (MTS) level in Intelligence component teams. In
Operations component teams, 7.2% of the variance in individual role goal commitment, 19.5%
of the variance in component team goal commitment, and 15.7% of the variance in MTS-level
goal commitment occurred at the group level. These results suggest lack of independence of the
goal commitment data with the possible exception of individual role goal commitment.

101

Table 9. Proportion of Goal Commitment Variance at the MTS-level

Dependent Variable

Individual-level
(Level-1)
2
Variance (σ )

MTS-level
(Level-2)
Variance (τ00)

Variance at
MTS-level
2

τ00 / (σ + τ00)

Intelligence component team (Situation-induced assessment)
Individual role goal commitment

1.460

0.003

0.2%

Component team goal commitment

1.119

0.162

12.6%

MTS goal commitment

1.147

0.179

13.5%

Operations component team (Situation-induced locomotion)
Individual role goal commitment

1.275

0.100

7.2%

Component team goal commitment

0.937

0.227

19.5%

MTS goal commitment

1.026

0.190

15.7%

Note. The variance components were obtained from one way ANOVA with random effects HLM
2

models (Raudenbush & Bryk, 2002). σ = individual-level (level-1) variance in the
dependent variable. τ00 = team-level (level 2) variance in the dependent variable. Portion
of variance at the level of the group is computed as τ00

/ (σ2 + τ00).

As documented in Table 10, 0.2% of the variance in individual role identity, 0.3% of the
variance in component team identity, and 7.6% of the variance in MTS-level identity occurred at
the group level in Intelligence component teams. In Operations component teams, 3.4% of the
variance in individual role identity, 31.2% of the variance in component team identity, and 2.4%
of the variance in MTS-level identity occurred at the group level. These results suggest that lack

102

of independence of the identity data may be of minimal concern with the possible exception of
component team identity for members of the Operations component teams.
Table 10. Proportion of Identity Variance at the MTS-level

Dependent Variable

Individual-level
(Level-1)
2
Variance (σ )

MTS-level
(Level-2)
Variance (τ00)

Variance at
MTS-level
2

τ00 / (σ + τ00)

Intelligence component team (Situation-induced assessment)
Individual role identity

0.859

0.002

0.2%

Component team identity

0.938

0.003

0.3%

MTS identity

1.081

0.089

7.6%

Operations component team (Situation-induced locomotion)
Individual role identity

0.806

0.028

3.4%

Component team identity

0.649

0.294

31.2%

MTS identity

1.387

0.034

2.4%

Note. The variance components were obtained from one way ANOVA with random effects HLM
2

models (Raudenbush & Bryk, 2002). σ = individual-level (level-1) variance in the
dependent variable. τ00 = team-level (level 2) variance in the dependent variable. Portion
of variance at the level of the group is computed as τ00

/ (σ2 + τ00).

Finally, Table 11 documents that 60.9% of the variance in individual role performance
contribution, 62.3% of the variance in component team performance contribution, and 73.7% of
the variance in MTS performance contribution occurred at the group level in Intelligence
component teams. In Operations component teams, 0.3% of the variance in individual role
103

performance contribution, 68.1% of the variance in component team performance contribution,
and 45.2% of the variance in MTS performance contribution occurred at the group level.
Table 11. Proportion of Performance Variance at the MTS-level

Dependent Variable

Individual-level
(Level-1)
2
Variance (σ )

MTS-level
(Level-2)
Variance (τ00)

Intelligence component team (Situation-induced assessment)
Individual role performance
87.658
136.532
contribution

Variance at
MTS-level
2

τ00 / (σ + τ00)

60.9%

Component team performance
contribution

10.836

17.915

62.3%

MTS performance contribution

25.419

71.385

73.7%

Operations component team (Situation-induced locomotion)
Individual role performance
contribution

7.369

0.019

0.3%

Component team performance
contribution

2.238

4.778

68.1%

386.183

317.977

45.2%

MTS performance contribution

Note. The variance components were obtained from one way ANOVA with random effects HLM
2

models (Raudenbush & Bryk, 2002). σ = individual-level (level-1) variance in the
dependent variable. τ00 = team-level (level 2) variance in the dependent variable. Portion
of variance at the level of the group is computed as τ00

/ (σ2 + τ00).

Thus lack of independence of the performance contribution data appears to be a substantial
concern with the possible exception of individual role performance contribution in the situationinduced locomotion condition. As a consequence, study hypotheses were tested first employing

104

OLS regression and then a second time employing hierarchical linear modeling. These latter tests
did not provide information that materially impacted the findings, thus, only the HLM results for
supported hypotheses (i.e., with significant OLS results) are reported in the Discussion Section.
Results Involving Goal Commitment
Goal commitment as criterion. The results of the regressions testing the relationships
between trait regulatory mode and goal commitment in the assumed (i.e., manipulation was
successful) situation-induced assessment condition, Hypotheses 1, 3 and 5, are reported in Table
12. Table 13 reports the results of the regressions testing the relationships between trait
regulatory mode and goal commitment, Hypotheses 1, 3 and 5, under the assumption that the
manipulation did not work. Hypothesis 1 predicted that trait regulatory assessment would exhibit
a positive relationship with all three levels of goal commitment in the situation-induced high
assessment state. As reported in Table 12 Step 1, trait regulatory assessment does not exhibit a
statistically significant relationship with either individual role goal commitment or component
team goal commitment (β = -.159, p = n.s. and β = -.074, p = n.s., respectively). Trait regulatory
assessment does exhibit a marginally significant relationship with multiteam system goal
commitment (β = -.212, p < .10) but the direction is opposite to that posited in Hypothesis 1c.

105

Table 12. Models Examining Regulatory Mode – Goal Commitment Relationships in High State Assessment Condition
(Hypothesis 1, 3, & 5)

Step 1

Component team
goal commitment (β)
Model 2
Step 1
Step 2

.022

Predictor \ Criterion

Individual role
goal commitment (β)
Model 1
Step 1
Step 2

-.008

Intercept
Trait assessment
Trait locomotion

Step 2

-.159
.260*

Trait assessment x
Trait locomotion

-.149
.281*

-.074
.260*

-.108

-.078
.252*

Multiteam system
goal commitment (β)
Model 3
Step 1
Step 2
-.004
-.212†

-.213†

.216†

.213†

.041

.018

2

.075†

.017

.065†

.003

.072†

.000

2

.075†

.093†

.065†

.067

.072†

.073

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

106

Table 13. Models Examining Regulatory Mode – Goal Commitment Relationships Employing State Assessment Measure
(Hypothesis 1, 3, & 5)

Step 1

Component team
goal commitment (β)
Model 2
Step 1
Step 2
Step 3

.052

Predictor \ Criterion

Individual role
goal commitment (β)
Model 1
Step 1
Step 2
Step 3
.053

.009

.015

-.268*

-.269*

-.184†

-.183†

-.166

-.301**

-.307**

Intercept

Multiteam system
goal commitment (β)
Model 3
Step 1
Step 2
Step 3
.033

.027

Trait assessment
Trait locomotion

Step 3

.232*

.303**

.321*

.227*

.218*

.111

.190†

.213†

.187

State assessment
Step 2

-.255*
.391**

.479**

.478**

.449**

.466**

.461**

.367*

.425**

.430**

Trait assessment x
State assessment
Trait locomotion x
State assessment

-.310**

-.093

-.089

-.050

-.081

-.093

-.097

-.231*

-.205

.030

-.043

-.075

-.144

Trait assessment x
Trait locomotion
Trait assessment x
Trait locomotion x
State assessment

-.016

.003

.060

-.031

.176†

.049

2

.217**

.056†

.001

.252**

.003

.031

.197**

.015

.007

2

.217**

.274**

.275**

.252**

.255**

.286**

.197**

.213**

.220*

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

107

As reported in Step 2 of Table 13 trait regulatory assessment exhibits a statistically
significant mean (weighted average across all observed values of state assessment) negative
relationship with goal commitment at all three levels (β = -.268, p < .05, β = -.183, p < .10 and β
= -.307, p < .05, for individual role, component team and multiteam system goal commitment,
respectively). These results are opposite to the hypothesized direction and do not vary across
levels of assessment state (β = -.093, p = n.s., β = -.050, p = n.s. and β = -.093, p = n.s., for
individual role, component team and multiteam system goal commitment, respectively).
Hypotheses 1a, 1b, or 1c are not supported.
Hypothesis 3 predicted that unlike trait regulatory assessment, trait regulatory locomotion
would exhibit a negative relationship with all three levels of goal commitment in the situationinduced assessment condition. As reported in Table 12 Step 1, trait regulatory locomotion
exhibits statistically significant positive relationships with individual role goal commitment (β =
.260, p < .05), component team goal commitment (β = .260, p < .05), and multiteam system goal
commitment (β = .216, p < .10). The direction of the relationship is opposite to that predicted in
Hypotheses 3a, 3b, and 3c, however. Table 13 Step 2 indicates that trait regulatory locomotion
exhibits a statistically significant mean (weighted average across all observed values of state
assessment) positive relationship with goal commitment at all three levels (β = .303, p < .01, β =
.218, p < .05 and β = .213, p < .10 respectively for individual role, component team and
multiteam system goal commitment). These results are opposite to the hypothesized direction as
well. Further, these models suggest that the relationship between trait regulatory locomotion and
individual role goal commitment is the only one to vary with state assessment (β = -.231, p <
.05). The simple slope does not turn negative until state assessment exceeds the mean plus 1.31
s.d., however; it is still slightly positive at the normative value for high state assessment equal to

108

the mean plus 1 s.d. (see Figure 6). Collectively, these results suggest that Hypotheses 3a, 3b,
and 3c are not supported.

Individual Role Goal Commitmen

5.4

High State
Assessment
4.9

4.4
Low State
Assessment
3.9

3.4
-1 s.d.

+1 s.d.
Trait Locomotion

Figure 6. Plot of Trait Locomotion x State Assessment Interaction
Hypothesis 5 predicted that the interaction of trait regulatory assessment with trait
regulatory locomotion would exhibit a positive relationship with all three levels of goal
commitment in the situation-induced assessment condition. As reported in Table 12 Step 2, the
interaction of trait regulatory assessment with trait regulatory locomotion does not exhibit a
relationship with goal commitment at any of the three levels (β = -.108, p = n.s. for individual
role goal commitment; β = .041, p = n.s. for component team-level goal commitment; β = .018, p

109

= n.s. for multiteam system-level goal commitment). The results under the assumption that the
manipulation was unsuccessful, reported in Table 13 Step 3, support these findings for individual
role goal commitment (β = -.016 for the mean effect and -.031, p = n.s. for the moderated effect)
and multiteam system goal commitment (β = .060 for the mean effect and .049, p = n.s. for the
moderated effect). Thus, Hypotheses 5a and 5c are not supported.
The results reported in Table 13 Step 3 for component team goal commitment (Model 2)
suggest a relationship between the interaction of trait regulatory assessment with trait regulatory
locomotion that varies across levels of state assessment (β = .176 p < .10). Figures 7 and 8
contain plots of these interactions. The top plot in Figure 7, representing the high state
assessment condition, indicates that trait assessment orientation exhibits a negative relationship
with commitment to goals at the component team level, and that this effect is stronger for
individuals low in trait locomotion orientation relative to those high in trait locomotion
orientation. (Note: Figure 8 is an alternative illustration of this relationship.) This indicates that
Hypothesis 5b is not supported.

110

5.8

Component Team Goal Commitment

5.6
5.4

High Trait
Locomotion

5.2

High State
Assessment

5.0
4.8
4.6
4.4
4.2
4.0
3.8

High Trait
Locomotion

Low Trait
Locomotion

Low State
Assessment

Low Trait
Locomotion
-1 s.d.

+1 s.d.
Trait Assessment

Figure 7. Plot of Trait Assessment x Trait Locomotion x State Assessment Interaction

111

5.8

Component Team Goal Commitment

5.6
5.4
5.2

Low Trait
Assessment

High State
Assessment

5.0
4.8
4.6

High Trait
Assessment

Low Trait
Assessment
Low State
Assessment

4.4
4.2

High Trait
Assessment

4.0
3.8
-1 s.d.

+1 s.d.
Trait Locomotion

Figure 8. Plot of Trait Locomotion x Trait Assessment x State Assessment Interaction
The results of the regressions testing the relationships between trait regulatory mode and
goal commitment in the situation-induced locomotion condition, Hypotheses 2, 4 and 6, are
reported in Table 14. Hypothesis 2 predicted that trait regulatory locomotion would exhibit a
positive relationship with all three levels of goal commitment. As reported in Table 14 Step 1,
trait regulatory locomotion does not exhibit a relationship with goal commitment at any of the
three levels (β = .052, p = n.s. for individual role goal commitment; β = .109, p = n.s. for
component team-level goal commitment; β = .115, p = n.s. for MTS-level goal commitment).
Hypotheses 2a, 2b, and 2c are not supported.

112

Table 14. Models Examining Regulatory Mode – Goal Commitment Relationships in High State Locomotion Condition
(Hypothesis 2, 4, & 6)

Individual role
goal commitment (β)
Model 1
Step 1
Step 2

Predictor \ Criterion
Step 1

Intercept

Component team
goal commitment (β)
Model 2
Step 1
Step 2

.014

Multiteam system
goal commitment (β)
Model 3
Step 1
Step 2

-.009

.001

Trait assessment

.000

.136

.128

.076

.077

Trait locomotion
Step 2

-.012
.052

.057

.109

.106

.115

.116

Trait assessment x
Trait locomotion

-.074

.049

-.007

2

.003

.006

.036

.003

.022

.000

2

.003

.008

.036

.039

.022

.022

ΔR

R

Notes. N = 80 observations from 20 Operations component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

113

Hypothesis 4 predicted that unlike trait regulatory locomotion, in the situation-induced
locomotion condition trait regulatory assessment would exhibit a negative relationship with all
three levels of goal commitment. As reported in Table 14 Step 1, trait regulatory assessment
exhibits no relationship with individual role goal commitment (β = -0.12, p = n.s.), no
relationship with component team goal commitment (β = .136, p = n.s.), and no relationship with
multiteam system goal commitment (β = .076, p = n.s.). These results do not support Hypotheses
4a, 4b, and 4c.
Hypothesis 6 predicted that the interaction of trait regulatory assessment with trait
regulatory locomotion would exhibit a positive relationship with all three levels of goal
commitment in the situation-induced locomotion condition. As reported in Table 14 Step 2, and
equivalent to the findings for the situation-induced assessment condition, the interaction of trait
regulatory assessment with trait regulatory locomotion does not exhibit a relationship with goal
commitment at any of the three levels (β = -.074, p = n.s. for individual role goal commitment; β
= .049, p = n.s. for component team-level goal commitment; β = -.007, p = n.s. for MTS-level
goal commitment). Hypotheses 6a, 6b, and 6c are not supported.
Performance as criterion. Hypothesis 9 predicted that trait regulatory assessment would
exhibit a positive relationship with all three levels of incremental performance in the situationinduced assessment condition. The results of the regressions testing these relationships are
reported in Tables 15, 16 and 17 for the individual level, component team level and multiteam
system level, respectively. Model 1, Step 1 reports the results under the assumption that the
manipulation worked as intended and Model 1, Step 3 reports the results under the assumption
that the manipulation did not worked as intended. As reported in Model 1, Step 1, trait regulatory
assessment does not exhibit a statistically significant relationship with any level of performance

114

contribution (β = .149, p = n.s., β = -.128, p = n.s. and β = -.132, p = n.s. for individual
performance contribution, component team performance contribution and multiteam system
performance contribution, respectively). The results from the alternate test reported in Model 1,
Step 3 produce equivalent findings. Trait regulatory assessment does not exhibit a statistically
significant relationship with any level of performance contribution (β = .106, p = n.s. for the
mean effect and .001, p = n.s. for the moderated effect for individual performance contribution, β
= -.111, p = n.s. for the mean effect and -.072, p = n.s. for the moderated effect for component
team performance contribution and β = -.117, p = n.s. for the mean effect and .006, p = n.s. for
the moderated effect for multiteam system performance contribution). Hypotheses 9a, 9b, and 9c
are not supported.
The results of the regressions testing the relationships between trait regulatory mode and
performance in the situation-induced locomotion condition, Hypothesis 10, are reported in Step 1
of Table 18. Model 1 reports results for individual role performance contribution, Model 2 for
component team performance contribution and Model 3 for multiteam system performance
contribution. Hypothesis 10 predicted that trait regulatory locomotion would exhibit a positive
relationship with all three levels of performance contribution in the situation-induced locomotion
condition. As reported in Table 18, trait regulatory locomotion does not exhibit a statistically
significant relationship with either individual performance contribution (β = .016, p = n.s.) or
multiteam system performance contribution (β = .179, p = n.s.). Thus, Hypotheses 10a and 10c
are not supported. Trait regulatory locomotion does exhibit a positive relationship with
component team performance contribution that is marginally significant, however (β = .197, p <
.10). A one-tailed test for significance based on directional prediction indicates support for
Hypothesis 10b.

115

Table 15. Effects of Goal Commitment on Individual Role Performance in High State Assessment Condition (Hypothesis 9,
11, 13, & 17)

Individual role performance (β)
Predictor \ Criterion
Step 1

Model 1
Step 2

Step 1 Intercept

Step 1

.000

Trait assessment
Individual level
goal commitment

.149

.106

.162

Trait assessment x
State assessment
Individual level
goal commitment x
State assessment

Step 3

Model 3
Step 1 Step 2 Step 3

.003

.106

.004
.164

.133

Step 2 State assessment
Step 3

Step 3

Model 2
Step 2

.162

.075
.164

.074

.128

.131

.150

.104

.103

.120

.107

.161

.001

.011
-.008

-.019

2

.022

.025

.000

.018

.023

.000

.044

.011

.001

2

.022

.047

.047

.018

.041

.041

.044

.056

.056

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

116

Table 16. Effects of Goal Commitment on Component Team Performance Contribution in High State Assessment Condition
(Hypothesis 9, 11, 13, & 17)

Component team performance contribution (β)
Predictor \ Criterion
Step 1

Model 1
Step 2

Step 1 Intercept

Step 1

.019

Trait assessment
Component team level
goal commitment

-.128

-.110

-.067

Trait assessment x
State assessment
Component team level
goal commitment x
State assessment

Step 3

Model 3
Step 1 Step 2 Step 3

.021

-.111

.026
-.126

.091

Step 2 State assessment
Step 3

Step 3

Model 2
Step 2

.163

-.165

-.034

.162

-.089

-.087

.088

.147

.145

-.135

-.108

-.167

-.072

-.061
-.049

-.025

2

.016

.004

.006

.008

.022

.003

.024

.014

.005

2

.016

.020

.026

.008

.030

.033

.024

.038

.043

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

117

Table 17. Effects of Goal Commitment on Multiteam System Performance Contribution in High State Assessment Condition
(Hypothesis 9, 11, 13, & 17)

Multiteam system performance contribution (β)
Predictor \ Criterion
Step 1

Model 1
Step 2

Step 1 Intercept

Step 1

-.001

Trait assessment
Multiteam system level
goal commitment

-.132

-.117

-.056

Trait assessment x
State assessment
Multiteam system level
goal commitment x
State assessment

Step 3

Model 3
Step 1 Step 2 Step 3

.014

-.117

.007
-.135

.001

Step 2 State assessment
Step 3

Step 3

Model 2
Step 2

.037

-.096

-.059

.031

-.117

-.111

-.021

-.001

.005

-.056

-.074

-.120

.006

.004
-.046

-.027

2

.017

.003

.000

.000

.008

.002

.018

.003

.000

2

.017

.020

.020

.000

.008

.011

.018

.020

.021

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

118

Table 18. Effects of Goal Commitment on Performance Contribution in High State Locomotion Condition (Hypothesis 10, 12,
14, & 18)

Predictor \ Criterion

Step 1 Trait locomotion
Step 2

Individual role
performance contribution
(β)
Model 1
Step 1 Step 2
Step 3
.022

.016

Individual level
goal commitment

-.117

Component team
performance contribution
(β)
Model 2
Step 1
Step 2 Step 3
.197†

.183

Multiteam system
performance contribution (β)
Step 1
.179

.010

.131

.106

.036

MTS level goal
commitment

.073

2

.000

2

.000

R

.000
.014

Step 4

.173

-.118

Component team
level goal
commitment

ΔR

Model 3
Step 2 Step 3

.039†

.014

.039†

.011
.017

.032

.050

.032

Notes. N = 80 observations from 20 Operations component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

119

.051
.003

.005

.035

.000 .001

Hypotheses 11 predicted that all three levels of goal commitment would exhibit a positive
relationship with their respective level of performance contribution in the situation-induced
assessment condition. The results of the regressions testing these relationships are reported in
Tables 15, 16 and 17 for the individual level, component team level and multiteam system level,
respectively. Model 2, Step 1 reports the results under the assumption that the manipulation
worked as intended and Model 2, Step 3 reports the results under the assumption that the
manipulation did not worked as intended. As reported in Model 2, Step 1, individual role goal
commitment does not exhibit a relationship with individual role performance contribution (β =
.133, p = n.s.), component team goal commitment does not exhibit a relationship with component
team performance contribution (β = .091, p = n.s.), and multiteam system goal commitment does
not exhibit a relationship with multiteam system performance contribution (β = .001, p = n.s.).
The results from the alternate test reported in Model 2, Step 3 produce equivalent findings. Goal
commitment does not exhibit a statistically significant relationship with any level of performance
contribution (β = .074, p = n.s. for the mean effect and -.008, p = n.s. for the moderated effect at
the individual level, β = .163, p = n.s. for the mean effect and -.049, p = n.s. for the moderated
effect at the component team level and β = .037, p = n.s. for the mean effect and -.046, p = n.s.
for the moderated effect at the multiteam system level). Hypotheses 11a, 11b, and 11c are not
supported.
The results of the regressions testing the relationships between goal commitment and
performance in the situation-induced locomotion condition, Hypothesis 12, are reported in Step 2
of Table 18. Hypothesis 12 predicted that all three levels of goal commitment would exhibit a
positive relationship with their respective level of performance contribution in the situationinduced locomotion condition. As reported in Table 18, individual role goal commitment does

120

not exhibit a relationship with individual role performance contribution (β = -.117, p = n.s.),
component team goal commitment does not exhibit a relationship with component team
performance contribution (β = .131, p = n.s.), and multiteam system goal commitment does not
exhibit a relationship with multiteam system performance contribution (β = .073, p = n.s.).
Hypotheses 12a, 12b, and 12c are not supported.
Hypotheses 13 and 14 predicted that individual role and component team goal
commitment would exhibit negative relationships with multiteam system performance
contribution in both the high state assessment condition and the high state locomotion condition.
As reported in Table 19, individual level goal commitment (Model 1) and component team level
goal commitment (Model 2) do not exhibit a relationship with multiteam system performance
contribution under either the situation-induced assessment manipulation was successful
assumption (Step 1: β = -.166, p = n.s. and β = .055, p = n.s., respectively) or the manipulation
was unsuccessful assumption (Step 3: β = -.151, p = n.s. for the mean effect and .022, p = n.s. for
the moderated effect at the individual level, and β = .117, p = n.s. for the mean effect and -.108, p
= n.s. for the moderated effect at the component team level). As reported in Step 4 of Model 3 in
Table 18, individual role goal commitment and component team goal commitment also do not
exhibit a relationship with multiteam system performance contribution in the situation-induced
locomotion condition (β = .010, p = n.s. and β = .036, p = n.s., respectively). Hypotheses 13a,
13b, 14a and 14b are not supported.
All hypotheses of mediation were tested utilizing the procedure described by Baron and
Kenny (1986). The results of the models testing goal commitment as a mediator of the
relationship between trait assessment orientation and performance, Hypothesis 17, are reported
in Tables 15, 16 and 17 for the individual level, component team level and multiteam system

121

level, respectively. Step 1 reports the results under the assumption that the manipulation worked
as intended and Step 3 reports the results under the assumption that the manipulation did not
worked as intended. The results of the models testing goal commitment as a mediator of the
relationship between trait locomotion orientation and performance, Hypothesis 18, are reported
in Table18 Models 1, 2 and 3 for the individual level, component team level and multiteam
system level, respectively. As already discussed, none of the hypothesized relationships between
goal commitment and performance contribution (Hypotheses 11 and 12) were statistically
significant thus mediation by goal commitment is disproved. Thus, Hypotheses 17a, 17b, 17c,
18a, 18b, and 18c are not supported.
Table 19. Effects of Other Goal Commitment on Multiteam System Performance
Contribution in High State Assessment Condition (Hypothesis 9, 11, 13, & 17)
Multiteam system performance contribution (β)
Model 1
Model 2
Step 1
Step 2
Step 3
Step 1
Step 2
Step 3

Predictor \ Criterion
Step 1 Intercept
Individual level
goal commitment
Component team level
goal commitment

-.008
-.166

-.154

.046

-.151
.055

Step 2 State assessment

-.032

Individual role level
Step 3 goal commitment x
State assessment
Component team level
goal commitment x
State assessment

.117

-.135

-.025

.113

-.140

.022
-.108

2

.027

.001

.001

.003

.015

.013

2

.027

.028

.029

.003

.018

.031

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).
122

Results Involving Identity
Identity as criterion. Hypothesis 7 predicted that trait regulatory assessment would
exhibit a positive relationship with component team level identity and multiteam system identity
in the situation-induced assessment condition. The results of the regressions testing these
relationships are reported in Table 20 Step 2 under the assumption that the manipulation worked
as intended and Step 4 under the assumption that the manipulation did not worked as intended.
Trait regulatory assessment does not exhibit a relationship with either component team identity
(β = -.030, p = n.s.) or multiteam system identity (β = -.147, p = n.s.) under the assumption that
the manipulation worked as intended. The tests conducted under the assumption that the
manipulation did not worked as intended yield equivalent results (β = -.032, p = n.s. for the mean
effect and -.006, p = n.s. for the moderated effect at the component team level, and β = -.108, p =
n.s. for the mean effect and -.185, p = n.s. for the moderated effect at the multiteam system
level). Hypotheses 7a and 7b are not supported.
The results of the regressions testing the relationships between trait regulatory mode and
identity in the situation-induced locomotion condition, Hypothesis 8, are reported in Model 1 of
Tables 21 and 22. Hypothesis 8 predicted that trait regulatory locomotion would exhibit a
positive relationship with individual level identity and component team level identity in the
situation-induced locomotion condition. As reported in Step 2, trait regulatory locomotion
exhibits a significant but negative relationship with individual level identity (β = -.318, p < .05)
after controlling for trait identity, and no relationship with component team identity (β = -.048, p
= n.s.). Hypotheses 8a and 8b are not supported.

123

Table 20. Models Examining Regulatory Mode – Identity Relationships in High State Assessment Condition (Hypothesis 7)

Predictor \ Criterion
Step 1

Intercept
Trait individual identity
Trait relational identity
Trait collective identity

Step 2

Trait assessment
Trait locomotion

Step 3

Component team identity (β)
Model 1
Step 1
Step 2
Step 3 Step 4
.061
.003
.023
.022
-.026
-.187
-.188
-.187
-.221†
.359**
.364*
.341*
.339*

State assessment

Step 4

-.030
-.006

-.052
-.005

-.032
.039

.113

Multiteam system identity (β)
Model 2
Step 1
Step 2
Step 3
Step 4
.056
-.010
.091
.091
.016
.020
.018
.018
.005
.029
.048
.037
-.001

.174

Trait individual identity x
State assessment
Trait relational identity x
State assessment
Trait collective identity x
State assessment
Trait assessment x
State assessment
Trait locomotion x
State assessment

-.147
-.017

-.158
-.017
.057

-.108
.068
.243†

-.284

-.387*

.158

.258*

-.061

.285†

-.006

-.185

-.009

-.209

2

.109*

.000

.012

.079

.002

.011

.003

.274**

2

.109*

.110

.121

.201

.002

.013

.016

.289*

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests); * p < .05 (two-tailed tests); † p < .10 (two-tailed tests).
124

Table 21. Effects of Regulatory Mode and Identity on Individual Role Performance in High State Locomotion Condition
(Hypothesis 8, 16, & 20)

Model 1
Step 1
Step2

Individual role
performance contribution (β)
Model 2
Step 1
Step 2
Step3
Step 4

.261*

.163

-.104

-.104

-.100

-.100

-.015

-.095

-.093

-.095

-.092

.134

.138

.136

.142

Individual identity (β)
Predictor \ Criterion
Step 1

Trait individual identity
Trait relational identity
Trait collective identity

Step 2

-.053
.085

.240†
.206†

Trait locomotion
Step 3

Trait assessment

-.318*

-.009

Individual identity

-.014
-.013

-.016

2

.076

.086*

.025

.000

.000

.000

2

.076

.161*

.025

.025

.025

.025

ΔR

R

Notes. N = 80 observations from 20 Operations component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

125

Table 22. Effects of Regulatory Mode and Identity on Component Team Performance Contribution in High State Locomotion
Condition (Hypothesis 8, 16, & 20)

Component team
identity (β)
Model 1
Step 1
Step2

Predictor \ Criterion
Step 1

Component team
performance contribution (β)
Model 2
Step 1
Step 2
Step3
Step 4

.062

.033

.179

.181

.174

.176

Trait relational identity

.102

.103

-.135

-.152

-.143

-.161

Trait collective identity
Step 2

Trait individual identity

.320*

.346*

-.048

.307*

.316*

.280†

.065

Trait locomotion
Step 3

Trait assessment

.341**

.077

Component team identity

.080
.080

.082

2

.159**

.004

.133*

.004

.005

.004

2

.159**

.163*

.133*

.137*

.139*

.143*

ΔR

R

Notes. N = 80 observations from 20 Operations component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

126

Performance as criterion. Hypothesis 15a predicted that component team identity
would exhibit a positive relationship with component team performance contribution in the
situation-induced assessment condition while Hypothesis 15b predicted that multiteam system
identity would exhibit a positive relationship with multiteam system performance contribution.
The results of the regressions testing the relationships between identity and performance
contribution in the situation-induced assessment condition, Hypotheses 15a and 15b, are reported
in Model 1 of Tables 23 and 24, respectively. As reported in Step 2, under the assumption that
the manipulation worked as intended component team identity exhibits a positive and statistically
significant relationship with component team performance contribution (β = .307, p < .05), and
multiteam system identity exhibits a relationship with multiteam system performance
contribution (β = .218, p < .05).
These results hold up under the assumption that the manipulation did not worked as
intended. As reported in Step 4, component team identity exhibits a positive and statistically
significant mean effect relationship with component team performance contribution (β = .302, p
< .05) and multiteam system identity exhibits a positive and statistically significant mean effect
relationship with multiteam system performance contribution (β = .261, p < .05). Neither of these
relationships appear to be affected by state assessment level, however (-.178, p = n.s. for the
moderated effect at the component team level and .012, p = n.s. for the moderated effect at the
multiteam system level); therefore they are positive in the high state assessment condition. Thus,
Hypotheses 15a and 15b are supported.
The results of the regressions testing the relationships between identity and performance
contribution in the situation-induced locomotion condition, Hypotheses 16a and 16b, are
reported in Model 2 of Tables 21 and 22, respectively. Hypothesis 16a predicted that individual

127

identity would exhibit a positive relationship with individual role performance contribution in the
situation-induced locomotion condition while Hypothesis 16b predicted that component team
identity would exhibit a positive relationship with component team performance contribution. As
reported in Step 3, neither individual identity nor component team identity exhibit a relationship
with performance contribution at their respective level (β = -.013, p = n.s. and β = .080, p = n.s.,
respectively). Hypotheses 16a and 16b are not supported.
Hypotheses of mediation were tested utilizing the procedure described by Baron and
Kenny (1986). The results of the models testing identity as a mediator of the relationship
between trait assessment orientation and performance for the situation-induced assessment
condition are reported for the component team level (Hypothesis 19a) in Table 20 Model 1 and
Table 23, and Table 20 Model 2 and Table 24 for the multiteam system level (Hypothesis 19b).
As previously discussed, trait assessment orientation does not exhibit a relationship with either
component team level (Hypothesis 7a) or multiteam system level performance (Hypothesis 7b)
in the situation-induced assessment condition. Therefore, there is no relationship to mediate and,
thus, Hypotheses 19a and 19b are not supported.

128

Table 23. Effects of Regulatory Mode and Identity on Component Team Performance
Contribution in High State Assessment Condition (Hypothesis 15, & 19)
Component team performance contribution (β)
Model 1
Model 2
Step 1 Step2 Step 3 Step 4
Step 1
Step 2

Predictor \ Criterion
Step 1 Intercept

.022

Trait individual identity

-.060 -.060

.022

-.041

-.079

.069

.068

Trait relational identity

.000

.058

.059

.099

.052

.104

Trait collective identity

.065 -.045

-.021

-.052

-.027

-.042

.323**

.302*

-.143

-.119

-.105

.019

-.026

-.171

-.199

.083

.080

-.178

-.193

Step 2 Component team identity

.307*

Step 3 State assessment
Step 4

Trait individual identity x
State assessment
Trait relational identity x
State assessment
Trait collective identity x
State assessment
Component team identity
x State assessment

Step 5 Trait assessment
Step 6

.304*

-.195

Trait assessment x
State assessment

.296*

-.230
.060

2

.005

.084*

.019

.030

.104*

.054

2

.005

.089

.108

.138

.109

.163

ΔR

R

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

129

Table 24. Effects of Regulatory Mode and Identity on Multiteam System Performance
Contribution in High State Assessment Condition (Hypothesis 15, & 19)
Multiteam system performance contribution (β)
Model 1
Model 2
Step 1 Step2 Step 3 Step 4 Step 1 Step 2

Predictor \ Criterion
Step 1

Intercept

-.051

-.038

Trait individual identity

.051

.053

.055

.079

.091

.065

Trait relational identity

-.170

-.175

-.175

-.169

-.176

-.177

Trait collective identity

-.306*

-.312*

-.310*

-.301*

-.307*

-.301*

.218*

.218*

.261*

.213*

.236†

-.012

-.062

-.021

.133

.228

.046

.031

.076

.101

.012

-.017

Step 2

Multiteam system identity

Step 3

State assessment

Step 4

Trait individual identity x
State assessment
Trait relational identity x
State assessment
Trait collective identity x
State assessment
Multiteam system identity
x State assessment

Step 5

Trait assessment

-.057

Step 6

Trait assessment x
State assessment

-.006
-.141

2

.153** .047*

2

.153** .200** .201**

ΔR

R

.000

.027

.049

.035

.228*

.202**

.237†

Notes. N = 79 observations from 20 Intelligence component teams. All coefficients standardized.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).
The results of the models testing identity as a mediator of the relationship between trait
locomotion orientation and performance for the situation-induced locomotion condition are
reported in Step 1 of Table 18 Model 1 and Model 2 of Table 21 for the individual role level

130

(Hypothesis 20) and Step 1 of Table 18 Model 2 and Model 2 of Table 22 for the component
team level (Hypothesis 20), respectively. Per Table 18, trait regulatory locomotion does not
exhibit a relationship with individual role performance contribution (Hypothesis 10a) and only a
marginally significant relationship with component team performance contribution (Hypothesis
10b) in the situation-induced locomotion condition. This latter relationship disappears entirely,
however, when controls for trait identity are present in the regression (see Table 22, Step 2). In
addition, identity does not exhibit a relationship with performance at either level (Hypotheses
16a and 16b). Hypotheses 20a and 20b are not supported.

131

Discussion
Study Hypotheses
Goal commitment as criterion. The confluence of the expectancy theory model of goal
commitment (Hollenbeck & Klein, 1987) and the regulatory fit theory (Higgins, 2000, 2002)
form the basis of the theoretical argument supporting the goal commitment as criterion
hypotheses in this study. The expectancy theory model of goal commitment argues that
attractiveness of goal attainment (i.e., valence) and expectancy of goal attainment (i.e.,
expectancy) are the key determinants of the degree to which individuals commit to a goal. In
addition, the theory stipulates that there are personal and situational factors that influence
valence and expectancy, and that these factors work both independently and interdependently to
influence valence and expectancy.
Regulatory fit theory suggests that individuals experiencing regulatory fit increase the
value of expect outcomes (i.e., valence) and that outcome attainment expectations (i.e.,
expectancy) increase because information processing is positively biased toward accepting
success cues and rejecting failure cues (Koenig et al., 2009). Pursuing this reasoning further,
individuals experiencing regulatory misfit should discount the value of expect outcomes (i.e.,
valence) and information processing should be positively biased toward accepting failure cues
and rejecting success cues thus decreasing outcome attainment expectations (i.e., expectancy).
Stated more directly, regulatory fit should exhibit a positive influence on goal commitment while
misfit should exhibit a negative influence on goal commitment. Notably, this is one of only a
handful of studies to apply the concept of fit to regulatory mode; most studies of regulatory fit
are regulatory focus based in the literature.

132

The regulatory fit argument suggests that trait assessment should exhibit a positive
relationship with all levels of goal commitment in the high state assessment condition, and that
trait locomotion should exhibit a negative relationship with all levels of goal commitment in the
case of Intelligence component team membership as this is a non- or misfit condition. As
reported in Table 12, trait assessment exhibited a marginally significant relationship with
multiteam system level goal commitment in the case of Intelligence component team
membership, although in the opposite (negative) direction (β = -.212, p < .10). The results
reported in Table 13 indicate that trait assessment exhibited a relationship with all levels of goal
commitment although in the direction opposite to that hypothesized (negative) once again.
Finally, trait locomotion was found to be positively related to all three levels of goal commitment
by both tests (i.e., assuming the manipulation worked and assuming it did not).
Similar arguments were made for the high state locomotion condition. Hypothesis 2
predicted that trait regulatory locomotion would exhibit a positive relationship with all three
levels of goal commitment and Hypothesis 4 predicted that trait regulatory assessment would
exhibit a negative relationship with all three levels of goal commitment. As documented in Table
14, neither of the trait regulatory mode dimensions exhibited a relationship with goal
commitment in this condition. While none of the fit argument based hypotheses were supported,
the analyses produced some interesting and potentially meaningful results.
Model 1, Table 13 indicates that the relationship between trait locomotion and individual
role goal commitment is influenced (moderated) by state assessment. This relationship, plotted in
Figure 6, indicates a positive relationship between trait locomotion and individual role goal
commitment in the high state assessment condition, not the hypothesized negative relationship.
The difference in the relationship in the low state assessment condition versus the high state

133

assessment condition suggests that inducing a high assessment state not only increases
commitment to individual goals (a main effect), it compensates for low trait behavioral drive
(e.g., low trait locomotion).
Perhaps the most interesting findings are depicted in the plots of the 3-way interaction of
trait assessment, trait locomotion and state assessment, Figures 7 and 8. Mirroring the individual
goal commitment finding, the difference in the relationship in the low state assessment condition
versus the high state assessment condition indicates a main effect; inducing a high assessment
state generally increases commitment to component team goals. However, commitment to goals
at the component team level is negatively influenced by trait assessment orientation. Locomotion
has a positive relationship with component team goal commitment when state and trait
assessment are matched (e.g., both high or both low), however, this relationship is increasingly
retarded to the point of turning negative as the degree of alignment between state and trait
assessment decreases. Thus, trait locomotion accentuates goal commitment at the component
team level when there is a strong fit between state and trait assessment, and plays little to no role
when state and trait assessment are mismatched.
Lastly, Figures 7 and 8 indicate that commitment to goals at the level of the component
team is lowest when all three (trait assessment, trait locomotion, and state assessment) are low,
as theory would indicate, and greatest when both trait regulatory orientations are low and state
assessment is high. This latter result suggests that trait and state assessment levels are driven by
different factors: trait assessment primarily by self-monitoring, self-evaluation and need for
social comparison concerns and state assessment primarily by the critical evaluation of
alternatives in order to select the highest value alternative to pursue. Collective goals inhibit selfmonitoring, self-evaluation and social comparison (within the collective) whereas the goals of a

134

collective in which the focal individual is a member may very well represent the highest value
alternative to pursue.
Notably, while state assessment influences the relationships between trait regulatory
orientations and goal commitment at lower and more proximal levels in the multiteam system
hierarchy it does not appear to affect them at the higher, most distal system level. It does exhibit
a main effect across all goal commitment levels, however. This suggests that managers might be
able to increase employee commitment to goals by enacting interventions that increase the
assessment state of their employees. Future research should investigate these extremely
interesting phenomena further to better understand the influence of state and trait regulatory
mode orientations on goal commitment.
Identity as criterion. The results documented in Tables 20, 21 and 22 provide the basis
for some insights. Hypothesis 7 predicted that trait regulatory assessment would exhibit a
positive relationship with component team level identity and multiteam system identity when
state assessment was high, and Hypothesis 8 predicted that trait regulatory locomotion would
exhibit a positive relationship with individual level identity and component team level identity
when state locomotion was high. Neither of these hypotheses was supported. Trait assessment
does not appear to exhibit the hypothesized relationship with identity at either the component
team or multiteam system level. Likewise, trait locomotion does not appear to exhibit the
hypothesized relationship with identity at either the component team or multiteam system level.
The results documented in Table 21 suggest the potential of a relationship between both
dimensions of regulatory mode and individual identity in the high state locomotion condition,
although the relationship between trait regulatory locomotion and individual identity was found
to be negative which is contrary to the prediction based on a regulatory fit argument (i.e.,

135

Hypothesis 8a which was not supported). While not hypothesized, the positive relationship
between trait assessment and individual identity (Table 21, Model 1, Step 2; β = .206, p < .10)
follows readily from regulatory mode theory. Assessors are high self-monitors and selfevaluators with a strong need for social comparison. This suggests that assessors will identify
more strongly at the individual level than will individuals low in trait assessment orientation.
Still, mirroring the conclusion emerging from the goal commitment as criterion analyses, these
findings provide little if any support for the regulatory fit theoretical argument.
Performance as criterion. Drawing once again on the fit theoretical argument discussed
previously Hypothesis 9 predicted that trait assessment would exhibit a positive relationship with
incremental performance contribution at all three levels in a high assessment state; no prediction
was made for trait locomotion. Likewise, Hypothesis 10 predicted that trait locomotion would
exhibit a positive relationship with incremental performance contribution at all three levels in a
high locomotion state and no prediction was made for trait assessment. The results reported in
Tables 15, 16, 17 and 18 suggest that trait assessment does not have a relationship with any level
of incremental performance contribution regardless of regulatory mode state and that trait
locomotion had a relationship with component team incremental performance contribution in the
high locomotion state condition (Hypothesis 10b). This finding did not hold up under the test
conducted post hoc employing hierarchical linear modeling techniques to account for the nesting
of individuals, however (γ = .077, p = n.s. for grand-mean centered trait locomotion and γ = .058,
p = n.s. for group-mean centered trait locomotion).
Somewhat of a surprise, goal commitment did not exhibit a relationship with
performance at any level or across levels in either condition (Hypotheses 11, 12, 13 or 14). It is
important to note, however, that Hollenbeck and Klein’s (1987) expectancy theory model of goal

136

commitment implicates goal commitment in the role of moderating the relationship between
goals and performance. Thus, these results may simply reflect the lack of a main effect; the
analysis does not address goal commitment in a moderating role. Alternatively, it could be that
the relationship between goal commitment and performance contribution across the levels of a
multiteam systems is more complex; i.e., curvilinear. Goal shielding research has shown a
positive relationship between myopic focus on a goal and goal attainment, however, this research
has also shown that too strong a commitment to a focal goal can be detrimental to performance
against alternative goals (Lord et al., 2010; Shah et al., 2002). This supposition was tested in the
post hoc analysis (discussed in the Curvilinear Models of Goal Commitment section later in this
document).
Tests of the relationship between identity and performance based on the fit theoretical
argument did yield some limited results (see Tables 21 through 24). Component team level
identity and multiteam system level identity were positively related to objective measures of
incremental performance contribution at the same level in the case of Intelligence component
team membership as posited by Hypotheses 15a and 15b (β = .307, p < .05 and β = .218, p < .05
for component team and multiteam system level incremental performance contribution,
respectively). However, identity did not exhibit a relationship with performance contribution in
the high state locomotion condition (Hypotheses 16a and 16b were not supported). In addition,
tests of Hypotheses 15a and 15b conducted post hoc employing hierarchical linear modeling
techniques failed to replicate these results. Component team identity did not exhibit a
relationship with component team performance contribution (γ = .120, p = n.s. for grand-mean
centering and γ = .073, , p = n.s. for group-mean centering) and multiteam system identity did
not exhibit a relationship with multiteam system performance contribution (γ = .012, , p = n.s. for

137

grand-mean centering and γ = -.028, , p = n.s. for group-mean centering) once clustering was
taken into account.
In totality, these findings provide little if any support for the regulatory fit based
theoretical arguments employed in this study. The lack of supporting results could be an artifact
of the experimental paradigm or indicative of a failure to induce the intended conditions, and
thus the lack of findings is not necessarily an indictment of the theory. Future research should
test the study hypotheses employing a different experimental paradigm.
Alternatives to the regulatory fit based theoretical arguments can be found in the
fundamental tenants of regulatory mode theory (Higgins et al., 2003; Kruglanski et al., 2000). In
addition, this study investigated phenomena at the individual level yet it was conducted in the
context of a multiteam system and the accuracy of the findings (data interpretation) is highly
dependent upon the validity of the two experimental manipulations for state regulatory mode
condition. Therefore, included below is a set of post hoc analyses that test suppositions based in
theoretical arguments emerging from regulatory mode theory. The post hoc analyses utilized
hierarchical linear modeling techniques to account for clustering effects and the self-reported
measures of regulatory mode state.
Post Hoc Analyses
Use of multilevel models for analyses. Members of a multiteam system are nested in a
task environment whereby they are interdependently linked with the other members in the
system. Lack of independence can lead to negatively biased standard errors and, thereby, liberal
tests of significance when traditional analysis techniques such as OLS regression are employed
(Kenny, Mannetti, Pierro, Levi & Kashy, 2002). The nesting of individuals in multiteam systems
can also produce cluster effects (Kish, 1965) that violate the OLS regression assumption of

138

statistically independent and normally distributed (i.i.d.) random errors of constant variance.
Finally, issues associated with non-independence are magnified as group size increases making it
a concern of greater salience when analyzing individual data collected in a multiteam system
context.
As the analyses documented in Tables 9, 10 and 11 clearly show, lack of independence
may be an issue in the data of this study. Hierarchical linear modeling (HLM) corrects for the
standard error bias associated with lack of independence of observations resulting in a more
statistically accurate and conservative test of predictions. Therefore, HLM techniques were
employed in the post hoc analyses to account for the nesting of individuals in a common
multiteam system context. Unfortunately, prototypical approaches to HLM analysis employ
either group-mean centered coefficients that only reflect within-group effects or grand-mean
centered coefficients that are weighted aggregates of within- and between-group effects and thus
difficult to interpret (Raudenbush & Bryk, 2002). The research question underlying this study
would be best informed by a disaggregation of effects into the pooled-within-group effect (βW)
and the contextual or compositional effect (βC); i.e., separating the individual or level-1
relationship from the effect of group membership.
Therefore, 2-level hierarchical models were constructed and analyzed in HLM 6.02
utilizing the procedure for disentangling person-level and compositional effects articulated by
Raudenbush and Bryk (2002, pp. 139 - 141). This procedure produces a pooled-within-group
effect, the person-level effect net of any group membership effect, and a contextual effect, the
estimated difference in an outcome variable due to membership in a group whose mean
(composition score) for a focal predictor variable differs by one unit from the mean of a
comparison group, controlling for the within person effect of the predictor variable (Raudenbush

139

& Bryk, 2002). This latter effect is “the extent to which the magnitude of the organization-level
relationship, βb, differs from the person-level effect, βW” (Raudenbush & Bryk, 2002, p. 139).
Measures from the two study conditions, Intelligence component team membership and
Operations component team membership, were combined and employed in all post hoc analyses.
Incremental performance contribution measures from the two study conditions were standardized
within condition before being merged into a single measure at each level. Control variables (e.g.,
trait identity) and predictor variables of interest were grand mean centered and entered into the
model at level-1. Within-group (multiteam system) means for these predictor variables were
calculated and included in the level-2 model for the intercept. All interrater agreement effect
sizes, reported in Table 25, are medium and support aggregation (LeBreton & Senter, 2008).
Significance of the coefficient associated with one of these level-2 variables suggests a
contextual effect above and beyond (after controlling for) the within person effect (Enders &
Tofighi, 2007). Finally, estimating the slopes and intercept separately for each group can be
problematic because group size is relatively small relative to the number of predictor variables in
the multilevel models. Therefore, the slopes of the predictor variables were not allowed to vary
per the advice of Kenny and colleagues (2002). The multilevel models had two random effects,
the intercept and the error.

140

Table 25. Interrater Agreement Statistics for Contextual (Level 2) Variables (post hoc)

Variable
Trait assessment

ICC(1)
.116

ICC(2)
.927

Trait locomotion

.115
.926

Individual role
goal commitment
Component team
goal commitment
Multiteam system
goal commitment

.103

Individual role
identity
Component team
identity
Multiteam system
identity

.103

State assessment

.117

.821
.103
.821
.102
.820
.822
.111
.889
.112
.890
.927

State locomotion

.115
.926

95% Confidence Intervals
Lower
Upper
F-test
.067
.227
13.624
.873
.966
.066
.225
13.491
.871
.965
.052
.687
.052
.687
.052
.685

.215
.916
.215
.916
.214
.916

.052
.689
.061
.807
.062
.808

.216
.917
.222
.948
.223
.948

.067
.873
.066
.872

.227
.966
.225
.965

Significance
p < .01
p < .01

5.582

p < .01

5.577

p < .01

5.550

p < .01

5.611

p < .01

9.018

p < .01

9.055

p < .01

13.685

p < .01

13.514

p < .01

Goal commitment as criterion. Regulatory mode theory posits that individuals high on
the assessment dimension are high self-evaluators, high self-monitors and are concerned with
appraising their individual performance relative to salient others. Further, individuals high on the
assessment dimension critically evaluate alternatives, in terms of importance and the utility of
means, in order to select the best alternative to pursue; i.e., the alternative that has the greatest
chance of reflecting positively on them. Thus high assessors are likely to focus on pursuing goals
whose successful attainment can be attributed to their individual actions and to reject goals that
cannot be attributed directly to them. Regulatory mode theory also posits that individuals high on
the locomotion dimension are concerned with movement irrespective of direction or valence, and
without much regard for specific goal pursuit. These arguments fit with the finding of a stronger

141

commitment to more proximal than to more distal goals in the situation-induced assessment
condition, and the findings of positive relationships between trait locomotion and goal
commitment in this condition. They also suggest that trait assessment will be negatively related
to multiteam system goal commitment and that trait locomotion will be positively related to goal
commitment regardless of level and condition.
These post hoc hypotheses were tested by regressing measures of individual, component
team and multiteam system level goal commitment on trait assessment, trait locomotion and a
dichotomous control variable denoting component team (0 denoting Operations component team
and 1 denoting Intelligence component team) and all tests were conducted utilizing HLM as
previously noted. Results reported in Table 26 Step 2 provide support for the supposition that
trait locomotion is positively related to goal commitment regardless of level (γ = .340, p < .05, γ
= .314, p < .05 and γ = .298, p < .05 for individual, component team and multiteam system level
goal commitment, respectively). Interestingly, the relationship between trait assessment and
multiteam system goal commitment is non-significant (γ = -.105, p = n.s.). Also of interest, the
coefficient for the component team variable is significant in all three regressions. This suggests a
contextual effect thus two additional exploratory analyses were conducted.
First, the two self-reported measures of state regulatory mode were included in the
regression to test if contextual influences on regulatory mode impact goal commitment. As
reported in Step 3 of Table 26, person-level state locomotion also exhibits a positive relationship
with goal commitment at all levels (γ = .251, p < .10, γ = .335, p < .01 and γ = .336, p < .01 for
individual, component team and multiteam system level goal commitment, respectively) as does
person-level state assessment (γ = .493, p < .01, γ = .448, p < .01 and γ = .353, p < .01 for
individual, component team and multiteam system level goal commitment, respectively). This
142

latter effect parallels the main effect discussed previously and illustrated in Figures 6, 7 and 8.
The collective regulatory state of the individuals across the focal individual’s multiteam system,
represented by the contextual variables, do not appear to influence the degree to which a focal
individual commits to goals except at the component team level (γ = .540, p < .10).
Second, measures of goal commitment for the two levels other than the focal level were
entered in the final step to partition out the interrelationship of the goals in a common hierarchy.
As reported in Step 4 of Table 26, goals across the three levels were positively and significantly
related at the person-level as expected. Of particular interest, different contextual factors are
indicated at each of the three goal levels once the relationship between goals in the hierarchy is
accounted for. The degree to which members of the multiteam system are committed to goals of
their component team is negatively and significantly related to an individual’s commitment to
goals at the individual level (γ = -1.219, p < .05). Likewise, the degree to which members of the
multiteam system are committed to their individual goals is negatively and significantly related
to an individual’s commitment to goals at the component team level (γ = -.279, p < .05). These
findings indicate that individuals perceive the degree of commitment of those around them and
act (adapt) accordingly.

143

Table 26. Models Examining Regulatory Mode – Goal Commitment Relationships (post hoc)

Person-level

Context

Predictor \ Criterion
(Intercept and context
are Level-2 variables)

Individual role
goal commitment (γ)
Step 1
Step 2
Step 3

Step 1 Intercept

4.770**

4.911**

Component team
goal commitment (γ)
Step 1
Step 2
Step 3

4.187**

Step 3 State assessment

5.173**

-.035

State locomotion

4.950**

.183

Step 2 Component team
(0 = Operations,
1 = Intelligence)

-.283*

Trait locomotion

.211

Step 3 State assessment

-.449**

3.581*
-.325
.593

-.211

-.460**

-.228

.052

-.071

-.105

-.187

.314*

.181

.298*

.161

.493**

.114*
1.322

.122*
1.277

.448**

.353**

.251†

State locomotion
Tau
Sigma-squared

4.980**

.540†

-.280*

.340*

4.751**

-.097

-.097

-.139

Trait assessment

3.169*

Multiteam system
goal commitment (γ)
Step 1
Step 2
Step 3

.335**

.336**

.072†
1.117

.179**
1.093

.173**
1.017

.046
.828

.189** .187**
1.133
1.062

Notes for Tables 26 thru 37. N = 159 observations from 20 multiteam systems.
Person-level variables are at level-1. Context are level-2 variables.
** p < .01 (two-tailed tests). * p < .05 (two-tailed tests). † p < .10 (two-tailed tests).

144

.099*
.920

Table 26 (cont’d)

Person-level

Context

Predictor \ Criterion
(Intercept and context
are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion
Individual goal
Step 4
commitment
Component team
goal commitment
Multiteam
system goal
Step 2 Component team
Trait assessment
Trait locomotion
Step 3 State assessment
State locomotion
Individual goal
Step 4
commitment
Component team
goal commitment
Multiteam
system goal
Tau
Sigma-squared

Individual role
goal commitment (γ)
Step 1
Step 3
Step 4
4.770** 4.187** 4.565**
-.035
.550
.183
.127

Component team
goal commitment (γ)
Step 1
Step 3
Step 4
4.950** 3.169*
4.255**
-.097
.222
.540†
.101

Multiteam system
goal commitment (γ)
Step 1
Step 3
Step 4
4.751**
3.581* 4.839
-.325
-.361†
.593
-.086

-.279*

.067

-1.219*

.258

.780
-.097
-.280*
.211
.493**
.251†

.175

.049
-.201†
.134
.222†
.014

-.211
-.071
.181
.448**
.335**

-.052
.074
.045
.157†
.083

-.228
-.187
.161
.353**
.336**

.139*

.166**

.340*

.710**

.341**
.114*
1.322

.072†
1.117

.066*
.770

145

-.060
-.076
-.017
-.051
.061

.644**
.179**
1.093

.046
.828

.000
.322

.189**
1.133

.099*
.920

.000
.357

Most interesting, perhaps, is the finding that neither the degree to which members of the
multiteam system are committed to goals of their component team (γ = .258, p = n.s.) nor the
degree to which members of the multiteam system are committed to their individual goals (γ =
.067, p = n.s.) influence an individual’s commitment to goals at the multiteam system level.
Rather, the degree to which an assessment orientation is active in peer actors across the system
has an influence on goal commitment at the multiteam system level; specifically, a negative
influence (γ = -.361, p < .10). Finally, note that component team affiliation is no longer
significant once these contextual effects are entered, as expected, and that the results documented
in Table 26 suggest that one or both of the non-focal level goal commitment variables may
mediate the relationship between state locomotion and goal commitment.
Collectively, these findings indicate that while locomotion is indicated in commitment to
goals irrespective of level in the goal hierarchy, context plays a central role in refining the degree
of commitment at each level. The degree to which a focal actor commits to individual and
component team goals is influenced by the commitment focus of their peer actors. The findings
reported above indicate that when an individual perceives an “all for one and one for all” attitude
individuals decrease their self-serving focus. Conversely, when they perceive an “everyone for
themselves” attitude in their team mates individuals tend to back away from committing to team
goals.
Further, the stronger the assessment state orientation collectively within the system the
weaker the commitment to the super ordinate goals of the system. Much like the affect of
collective commitment to component team goals on a focal individual’s commitment to their
individual role goals and the affect of collective commitment to individual role goals on a focal
individual’s commitment to component team goals, individuals may be influenced by their

146

perception of peer actors’ assessment of goal importance or conscious of peer actors’ social
comparison across the sub-units within the system. This may also be a subtle manifestation of ingroup bias or simply indicative of the predilection of individuals high on the assessment
orientation to identify more strongly at the individual level (as will be discussed next).
Regardless, these results clearly indicate that perceptions of collective orientations, once again,
drive individuals to adapt the degree to which they commit to goals accordingly. Future research
is needed to investigate these complex phenomena further to better understand the factors
leading to states of regulatory orientation that increase goal commitment so that they can be
influenced in ways that will facilitate the maximization of system performance.
Identity as criterion. As was the case for goal commitment, post hoc analyses were
conducted to test predictions developed directly from the fundamental tenants of regulatory
mode theory (Higgins et al., 2003; Kruglanski et al., 2000). High assessment oriented individuals
are concerned with self-evaluation, self-monitoring and with appraising their performance
relative to salient others. This suggests that assessors will identify strongly at the individual level
and abhor the anonymity associated with both component team level and multiteam system level
identities. Thus, both trait and state assessment orientation are expected to exhibit a positive
relationship with individual role identity.
High locomotion oriented individuals focus on activities that are accessible and that offer
a sense of movement. In addition, the locomotion dimension is negatively associated with
sensitivity to social criticism and feeling anxious during social interactions. These attributes
suggest that locomotors may discount the distinctiveness associated with individual identification
when in a social context such as a multiteam system preferring group identification instead.

147

Thus, trait and state locomotion orientation are expected to exhibit a positive relationship with
both component team level and multiteam system level identities.
The same analysis approach employed to test the post hoc goal commitment predictions
was adopted to test these hypotheses. Measures for individual, component team and multiteam
system level identity were regressed on a dichotomous variable denoting component team (0
denoting Operations component team and 1 denoting Intelligence component team), measures of
trait assessment, trait locomotion, state assessment, and state locomotion while controlling for
dispositional identities at the individual, relational and collective levels. Results of these tests are
reported in Table 27. As documented in Steps 2, the post hoc hypotheses predicted directly from
regulatory mode theory (Higgins et al., 2003; Kruglanski et al., 2000) were generally supported
for identity as criterion. Trait assessment was positively and significantly related to individual
identity (γ = .273, p < .05) but exhibited no relation with component team level identity or
multiteam system level identity (γ = .057, p = n.s. and γ = -.151, p = n.s.). Trait locomotion did
not exhibit the predicted relationship with component team level identity or with multiteam
system level identity, however (γ = -.014, p = n.s. and γ = -.037, p = n.s.). A subsequent test
removing the controls for trait identity from the regression yielded a significant relationship
between trait locomotion and multiteam system level identity (γ = .298, p < .05), but the
relationship with component team level identity remained elusive (γ = .211, p = n.s.).

148

Table 27. Models Examining Regulatory Mode – Identity Relationships (post hoc)

Person-level

Context

Predictor \ Criterion
(Intercept and context
are Level-2 variables)
Step 1 Intercept

Individual role
identity (γ)
Step 1
Step 2
Step 3
2.567**

2.566**

3.142**

Component team
identity (γ)
Step 1
Step 2
Step 3
5.401**

5.504** 5.212**

Multiteam system
identity (γ)
Step 1
Step 2
Step 3
4.764**

4.862**

5.169**

Step 3 State assessment

.041

-.167

-.074

State locomotion

-.165

.181

-.050

Step 2 Component team
(0 = Operations,
1 = Intelligence)
Trait individual
identity
Trait relational
identity
Trait collective
identity

.002

-.209

.086

-.198

.135

.337**

.027

.007

.080

.053

.074

-.168

-.065

.023

.155

.043

.018

.607**

.486**

.492*

.357†

.273*

Trait locomotion

.308*

.076

Trait assessment

.200†

.057

.068

-.151

-.136

-.014

-.125

-.037

-.164

-.050

Step 3 State assessment

-.011

.066

-.154
.000
.841

.000
.718

.000
.702

.142**
.805

149

.120**
.742

.061

.445**

.257*

State locomotion
Tau
Sigma-squared

-.080

.503**

.106**
.642

.135*
1.168

.126**
1.141

.145**
1.020

Mirroring trait, person-level state assessment was positively and significantly related to
individual identity (γ = .257, p < .05) but exhibited no relation with component team level
identity (γ = .066, p = n.s.) and no relation with multiteam system level identity (γ = .061, p =
n.s.). Unlike trait, person-level state locomotion exhibited a positive and significant relation with
both component team and multiteam system level identity as predicted (γ = .445, p < .01 and γ =
.503, p < .01 respectively), and no relationship with individual identity (γ = -.154, p = n.s.).
Unlike goal commitment, there was no contextual effect to be explained by component team
membership (γ = .002, p = n.s., γ = -.209, p = n.s. and γ = -.198, p = n.s. for individual role,
component team and multiteam system level identity respectively); neither state assessment nor
state locomotion exhibited a contextual effect (see Table 27, Step 3).
Of note, trait identity continues to explain a significant portion of variance in activated
(state) identity even after trait and state regulatory orientations are entered into the regression.
Although understandable in the case of individual and multiteam system identity, it is somewhat
surprising that trait relational identity does not explain any of the variance in identity at the
component team level. Collectively, these findings are in line with the expected relationships.
Performance as criterion. Post hoc analyses to investigate the relationship between state
regulatory mode and performance, and the incremental influence of both goal commitment and
identity on performance (i.e., variance explained by goal commitment and identity after
accounting for variance explained by state regulatory mode) were conducted to test predictions
based on the fundamental tenants of regulatory mode theory (Higgins et al., 2003; Kruglanski et
al., 2000).
Role of goal commitment. High assessment oriented individuals are high self-evaluators,
high self-monitors and are concerned with appraising their individual performance relative to
150

salient others. They evaluate alternatives in order to select the alternative that has the greatest
chance of reflecting positively on them. This suggests that assessors may utilize performance
information as feedback for input into the goal selection and commitment process. This is
exactly what the results reported in Table 26 (and previously discussed) indicate. State
assessment at the person-level was positively related to goal commitment for all three levels of
performance, and the relationship between trait assessment and individual goal commitment was
negative (γ = -.280, p < .05) once state assessment was accounted for. Theory does not implicate
assessment directly in performance, however. Thus neither trait nor state assessment is expected
to exhibit a relationship with performance at any level.
High locomotion oriented individuals are concerned with psychological movement
irrespective of direction or valence. As discussed previously, locomotion is positively related to
attentional control and the ‘Drive’ sub-scale of the Behavioral Activation System (BAS; Carver
& White, 1994) which essentially measures the degree to which an individual is engaged in the
pursuit of something. In addition, locomotion orientation is affected by salient cues in the
environment (i.e., the performance context). In combination, these arguments suggest that state
locomotion will exhibit a positively relationship with performance contribution irrespective of
level (i.e., at all levels).
Research has also consistent shown that individuals exert more effort and determination
to try for a goal (or to keep trying for a goal) when they are highly committed to a focal goal
relative to individuals who are less committed to the goal (Locke et al., 1981; Seijts & Latham,
2000a). However, goal shielding research argues that too strong a commitment to a goal will
inhibit consideration and focus on alternative or competing goals (Shah at al., 2002). Goals in
multiteam system are structured hierarchically (Mathieu et al., 2001) which means that

151

successful goal attainment at lower levels is often a prerequisite for goal attainment at higher
levels. Thus, there may be a set of goals that are not considered to be alternative or competing.
Under norms of rationality (Thompson, 1967), therefore, individuals highly committed to the
goals at a focal level higher in the hierarchy should be committed to the goals of the level below.
Finally, although Hollenbeck and Klein’s (1987) expectancy theory model of goal commitment
implicates goal commitment in the role of moderating the relationship between goals and
performance, research has shown that goal commitment can have a main effect on performance
as well. These arguments suggest that commitment to goals at a focal level of performance will
exhibit a positive relationship with performance contribution at that level and that commitment
to goals at levels above the focal level may also exhibit a positive relationship with performance
contribution at the focal level.
A recent investigation of the impact of coordination on performance in multiteam
systems suggests something different may occur in the specific context of a multiteam system.
This study found that subjugating one’s own performance and the performance of one’s
component team when it is in a supporting role, is necessary if the system is to achieve high
performance at the multiteam system level (Davison et al., 2012). This suggests that commitment
to goals at levels above the focal level may exhibit a negative relationship with performance
contribution at the focal level; most especially, commitment to the super-ordinate goals of the
system may adversely impact performance at the individual level.
Finally, performance in multiteam systems occurs in a social context thus the actions of
fellow actors in the system are bound to have an impact on a focal individual’s behavior; i.e., it is
highly likely that collective behavior or “bandwagon” effects will emerge. This suggests a
contextual effect; the performance of a focal individual at a given level may be influenced by the

152

degree of commitment of their fellow actors in the system. Further, strength of social ties may
play a role which suggests that the influence of focal actors will be strongest at the component
team level.
To test these predictions, measures representing individual role performance contribution,
component team performance contribution and multiteam system performance contribution were
regressed on measures of state assessment, state locomotion, individual goal commitment,
component team goal commitment and multiteam system goal commitment, controlling for trait
regulatory mode and component team (condition) assignment. All tests were conducted in HLM.
Results for individual role performance contribution are reported in Table 28. Table 29 contains
the results for component team performance contribution and the results for multiteam system
performance contribution are reported in Table 30.
The unconditional model (Model 1) was analyzed first to establish the base line for
variances. Model 2 is the unconditional model with only the control variables entered. The two
dimensions of state regulatory mode were included next in Model 3. The measure for goal
commitment at the focal performance level was then added to the model (Model 4a). The next
two models (4b and 4c) combine the measure for goal commitment at the focal performance
level with one of the two goal commitment measures for a non-focal performance level to
alleviate multicollinearity concerns. The final model (4d) contains all three goal commitment
measures.

153

Table 28. Effects of Regulatory Mode and Goal Commitment on Individual Role Performance (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion

Individual role performance (γ)
Model
Model
Model 1 Model 2 Model 3
4a
4b
.004
.004
-.401
-.343
-.515
.666
.685
.628
-.513
-.480
-.543

Step 4 Individual goal commitment

-.055

Component team goal commitment

-.148

-.190

-.200
.012

.276

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Step 3 State assessment
State locomotion
Step 4 Individual goal commitment
Step 2

Person-level

Model
4d
-.625
.674
-.558

.222

Multiteam system goal commitment
-.001

.261†

.258

.047
-.100

.082
-.205
-.168
.421**

.071
-.197
-.152
.428**
-.031

.226
.080
-.190
-.107
.474**
.059

.273

.214

.212

.053
-.188
-.122
.482**
.008

-.198**

Component team goal commitment
Multiteam system goal commitment
Tau
Sigma-squared

Model
4c
-.599
.669
-.560

-.043
-.245**

.195**
.809

154

.182**
.827

.186**
.766

.203**
.771

.225**
.753

.055
-.185
-.115
.485**
.094

.229**
.736

-.218†
.252**
.741

Table 29. Effects of Regulatory Mode and Goal Commitment on Component Team Performance Contribution (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion

Model 1
.004

Step 4 Individual goal commitment
Component team goal commitment
Multiteam system goal commitment

.674

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Step 3 State assessment
State locomotion
Step 4 Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Step 2

Person-level

Component team performance contribution (γ)
Model
Model
Model
Model 2 Model 3
4a
4b
4c
-.001
-.242
-.416
.781
.398
.278
.041
.245
-.332
-.205
-.761
-.580†
-.709

Tau
Sigma-squared

.012
-.126
.060

.189

.172

.171

.389**
.633

155

.435**
.603

-.131
-.023
.054
.306**

1.753†
-1.095

-.837*
1.074
-.148

.180

.179

-.133†
-.035
.052
.306**
.020
-.096

-.112
-.032
.054
.299**

-.084

.383**
.629

-.130
-.025
.017
.274**

-.864*
.966*

Model
4d
.837
.226
-.571

-.192†
.133

-.120
-.035
.058
.298**
-.002
-.190
.132

.409**
.602

.274**
.606

.395**
.600

.299**
.604

Table 30. Effects of Regulatory Mode and Goal Commitment on Multiteam System Performance Contribution (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion

Model 1
.004

Multiteam system performance contribution (γ)
Model
Model
Model
Model 2 Model 3
4a
4b
4c
-.003
-.918
-1.020
-.463
-.725
.155
.132
.343
-.094
.064
.096
.114
.003

Step 4 Individual goal commitment
Component team goal commitment
Multiteam system goal commitment

-.538
.013

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Step 3 State assessment
State locomotion
Step 4 Individual goal commitment
Component team goal commitment
Multiteam system goal commitment

Person-level

Step 2

Tau
Sigma-squared

.016
-.113
.097

.200
-.094
.016
-.057
.292**

.260

.653
-.477

.183

.186

.179

-.109
.029
-.030
.317**

-.077
.381**
.631

.398**
.633

156

.419**
.607

.452**
.606

-.124
.035
-.006
.322**
-.082
-.029
.414**
.605

-.100
.030
-.011
.328**
-.100
-.004
.475**
.607

Model
4d
-.458
.304
.100
-.538
.132
.163
.182
-.117
.038
.006
.329**
-.071
-.076
.020
.449**
.607

As reflected in Model 3 of all three analyses, and as predicted, there is no relationship
between state assessment and performance contribution at any level (γ = -.168, p = n.s., γ = .017,
p = n.s., γ = -.057, p = n.s. at the person-level and γ = .666, p = n.s., γ = .278, p = n.s., γ = .155, p
= n.s. at the contextual level for individual, component team and multiteam system level
performance contribution, respectively), however, state locomotion at the person-level is
positively and significantly related to performance contribution regardless of level (γ = .421, p <
.01, γ = .274, p < .01 and γ = .292, p < .01 for individual, component team and multiteam system
level performance contribution, respectively).
Surprisingly, and contrary to predictions, person-level goal commitment at the focal level
of performance (Model 4) did not exhibit a significant relationship with performance
contribution at any of the levels (γ = -.031, p = n.s., γ = -.084, p = n.s. and γ = -.077, p = n.s. for
individual, component team and multiteam system level performance contribution, respectively).
As discussed previously, these results could simply reflect the fact that goal commitment at the
person-level is acting in a moderating role between goals and performance within a focal level as
suggested by Hollenbeck and Klein’s (1987) expectancy theory model of goal commitment or
the relationship might be more complex in nature (e.g., curvilinear). Goal commitment does play
a cross-level role, as predicted, however. (Note: these results match the findings of the original
study.)
First, both component team and multiteam system level goal commitment at the personlevel (Table 28 Models 4b and 4c) exhibited a significantly negative relationship with individual
role performance contribution as expected (γ = -.198, p < .01 and γ = -.245, p < .01,
respectively). Second, as reported in Model 4b of Table 29, the collective individual goal

157

commitment of actors in the system exhibited a negative relationship with performance
contribution at the component team level (γ = -.864, p < .05) and the collective component team
goal commitment of actors in the system exhibited a positive relationship with performance
contribution at that level (γ = .966, p < .05). Contextual influence did not exhibit statistical
significance at the multiteam system performance contribution level, however (Table 30).
These results strongly suggest that multiteam system context matters materially to
performance contribution. The finding that the goal commitment of peer actors in the system
affects a focal actor’s performance at the component team level, both positively and negatively,
is a testament to this. The finding that commitment to higher level goals has a negative impact on
performance at the individual level attests to this as well and is consistent with the findings of
Davison and colleagues (2012) that superior multiteam system performance requires that
individuals and component teams sacrifice lower level metrics, at times (e.g., when in a
supporting role), for the greater good. There was no evidence of a negative impact on overall
system performance due to an in-group (component team) bias or too strong a focus on goals at
the individual level, however.
While this latter finding appears inconsistent with the findings reported in the goal
shielding literature (Shah at al., 2002), it may be due to the fact that multiteam systems goals are
hierarchically arrange and thus higher level goals encompass lower level goals. The former
finding may be an artifact of the laboratory setting; there may not have been enough time for ingroup biases to develop. These results may also be, simply, more indicative of a lack of power in
the analysis than of the true effect. Future research should investigate this phenomenon across
both a greater expanse of time and a larger number of multiteam systems. Also, practicing
managers should take note of the influence of peer actors and of “doing the right thing” on

158

performance with an eye to motivational and reward systems. Lastly, future research should
investigate the determinants of state locomotion in the multiteam system context as it is
implicated at all levels of performance contribution; understanding these may identify actions
that can be undertaken by leaders of these systems to increase performance.
Curvilinear models of goal commitment. Goal research has shown a positive relationship
between a determined focus on a goal and goal attainment, however, goal shielding research has
shown that too strong a commitment to a focal goal can be detrimental to performance against
alternative goals (Lord et al., 2010; Shah et al., 2002). This implies that there may be an optimal
degree of goal commitment in task environments involving multiple interdependent goals, and
that both greater and lesser commitment may lead to inferior performance in such circumstances.
This suggests an “inverted-U” shaped curvilinear relationship between goal commitment and
performance.
The task environment encountered by participants in a multiteam system adds an
additional dimension of complexity, however. Goals are organized hierarchically and goal
attainment at a focal level is interdependent on performance at other levels. (It should be noted
that the use of “performance” and not “goal attainment” in the last sentence was purposeful.)
Davison and colleagues (2012) showed that the task environment may require participants in a
multiteam system to underperform against some sub-set of goals in order to maximize superordinate system level performance. Further, the study reported here has shown that goal
commitment at levels other than the focal performance contribution level can influence
performance contribution at the focal level both positively (e.g., collective-level component team
goal commitment and component team performance contribution) and negatively (e.g., personlevel multiteam system goal commitment and individual role performance). This suggests that

159

the relationship between goal commitment and performance, while curvilinear, can be both
“inverted-U and U” shaped.
Models containing polynomial terms for goal commitment were introduced into the goal
commitment models for performance contribution in order to investigate whether these
suppositions are the case. Tables 31, 32 and 33 contain the results of this analysis for individual
role performance, component team performance contribution and multiteam system performance
contribution, respectively. As the results documented in these tables illustrate, the relationship
between goal commitment and performance contribution in multiteam systems is, indeed,
curvilinear and both “inverted-U and U” shaped.
The relationship between collective-level multiteam system goal commitment and
individual role performance (Table 31) exhibited a curvilinear inverted-U shape with a maximum
point of approximately 4.8 on the 7-point Likert response scale (assumes focal individual is at
the collective mean). In a similar vain, the relationship between collective-level component team
goal commitment and component team performance contribution (Table 32) exhibited an
inverted-U curvilinear shape. In addition, the relationship between person-level component team
goal commitment and component team performance contribution exhibited a curvilinear U
shape. Assuming a focal individual is at the collective mean for component team goal
commitment, the maximum point occurs at approximately 5.5 (on the 7-point Likert response
scale).

160

Table 31. Exploratory HLM Models Examining Curvilinear Influence of Goal
Commitment on Individual Role Performance (post hoc)

Person-level

Context

Predictor
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 2 State assessment
State locomotion
Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Individual goal commitment squared
Component team goal commitment squared
Multiteam system goal commitment
d
Component team (0 = Operations,
Step 2
1 = Intelligence)
Trait assessment
Trait locomotion
State assessment
State locomotion
Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Individual goal commitment squared
Component team goal commitment squared
Multiteam system goal commitment
d
Tau
Sigma-squared

161

Model 1
.004

Model 5
-4.919
.545
-.711
-4.018
6.380*
.400
-.636*
.213
.038
-.199
-.116
.482**
.088
-.246**

.195**
.809

.168**
.736

Table 32. Exploratory HLM Models Examining Curvilinear Influence of Goal
Commitment on Component Team Performance Contribution (post hoc)

Person-level

Context

Predictor
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 2 State assessment
State locomotion
Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Individual goal commitment squared
Component team goal commitment squared
Multiteam system goal commitment
d
Component team (0 = Operations,
Step 2
1 = Intelligence)
Trait assessment
Trait locomotion
State assessment
State locomotion
Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Individual goal commitment squared
Component team goal commitment squared
Multiteam system goal commitment
d
Tau
Sigma-squared

162

Model 1
.004

Model 5
-10.024
.219
-.777
-5.280
10.060**
.460
-.920*
.129
-.169†
-.067
.093
.253*
.326
-1.582**
-.032
.152**

.383**
.629

.183**
.581

Table 33. Exploratory HLM Models Examining Curvilinear Influence of Goal
Commitment on Multiteam System Performance Contribution (post hoc)

Person-level

Context

Predictor
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 2 State assessment
State locomotion
Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Individual goal commitment squared
Component team goal commitment squared
Multiteam system goal commitment
d
Component team (0 = Operations,
Step 2
1 = Intelligence)
Trait assessment
Trait locomotion
State assessment
State locomotion
Individual goal commitment
Component team goal commitment
Multiteam system goal commitment
Individual goal commitment squared
Component team goal commitment squared
Multiteam system goal commitment
d
Tau
Sigma-squared

163

Model 1
.004

Model 5
-13.414
-.010
-.091
5.498

-.543
.153
-.125
.014
.017
.297**
-.927†

.084†

.381**
.631

.443**
.594

Lastly, the results documented in Table 33 suggest that multiteam system performance
contribution may be influenced, in a curvilinear fashion, by person-level component team goal
commitment. Interestingly (and arguably not coincidentally), the maximum point occurs at
approximately 5.5 (on the 7-point Likert response scale), the same level of component team goal
commitment indicated for maximum influence on component team performance contribution.
These results should be viewed with caution, however, as the variance explained by this model is
arguably lower than other models documented in Table 30 (e.g., Model 4b).
Role of identity. As detailed previously, two closely related motivational premises
underlying identification are the desire to develop and maintain a positive self-concept (Turner,
1975; Tajfel & Turner, 1979) and the need for positive regard (Hetts et al., 1999). Attributions of
performance reflect on the individual or individuals involved which implies that identification at
a focal level may predict performance at that level. Alternatively, the reverse could be argued;
performance at a level will influence the degree to which individuals identify with that level.
Regardless of causality, this suggests that identification at a focal level will be positively related
to performance at that level.
Component teams within a multiteam system enact differentiated roles in order to best
meet the needs of the task environment (Davison et al., 2012). The primary performance
interface with the task environment is enacted by component teams in point roles while
supporting roles are enacted by the other component teams in the system. There are times during
a performance episode when maximizing the performance of the system requires members to
focus on the needs of others in the system at the expense of their own performance (due to limits
on resources such as time, etc.). This is especially true when a focal individual’s component team
is enacting a support role. In addition, individuals with strong identification at higher levels in

164

the system may be overly zealous and provide support to others to the detriment of their own
performance irrespective of system needs. This implies that component team level identity and
multiteam system level identity may be negatively related to individual level performance
contribution. This also implies that a collective (contextual) identification at the multiteam
system level may positively influence both component team and multiteam system performance.
These predictions were tested utilizing the same methodology as was employed to test the
goal commitment predictions. The combined measures of individual role performance
contribution, component team performance contribution and multiteam system performance
contribution were regressed on measures of state assessment, state locomotion, individual-level
identity, component team-level identity and multiteam system-level identity, controlling for
component team assignment, trait identities and trait regulatory mode. Results for individual role
performance contribution are reported in Table 34. Table 35 contains the results for component
team performance contribution and the results for multiteam system performance contribution
are reported in Table 36.
The unconditional model (Model 1) was analyzed first to establish the base line for
variances. Model 2 is the unconditional model with the control variables entered. The two
dimensions of state regulatory mode were included in Model 3 and the measure for identity at the
focal performance level was included in Model 4a. The next two models (4b and 4c) employ the
measures for identity at the non-focal performance level in combination with the measure at the
focal performance level except when concerns with multicollinearity were present. The final
model (4d) contains all three measures of identity.

165

Table 34. Effects of Regulatory Mode and Identity on Individual Role Performance (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion

Model 1

Individual role performance (γ)
Model
Model
Model
Model 2 Model 3
4a
4b
4c
.008
-.480
1.015
.377
.511
.678
1.055† 1.016†
1.023†
-.504
-.766
-.855†
-.888†

.004

Step 4 Individual role identity
Component team identity
Multiteam system identity

-.692

Person-level

-.682
.228

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Trait individual identity
Trait relational identity
Trait collective identity
Step 3 State assessment
State locomotion
Step 4 Individual role identity
Component team identity
Multiteam system identity
Step 2

Tau
Sigma-squared

-.071
.216

-.010

-.717
.048
.213

.195**
.809

.185**
.835

166

.141
-.200
-.114
-.030
.043
-.160
.415**

.190**
.777

.267†

.288†

.299†

.302†

.105
-.189
-.150
-.058
.033
-.197
.436**
.147

.120
-.211
-.151
-.071
.127
-.181
.538**
.152†
-.226*

.079
-.222
-.151
-.003
.101
-.184
.544**
.152†
-.210**

.104
-.112
-.107
-.135
.136

.256

Model
4d
.449
1.028†
-.901†

.091
-.224
-.153
-.021
.132
-.180
.568**
.154†
-.110
-.160†

.221**
.725

.241**
.726

.195**
.765

.214**
.739

Table 35. Effects of Regulatory Mode and Identity on Component Team Performance Contribution (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion

Model 1
.004

Component team performance contribution (γ)
Model
Model
Model
Model 2
Model 3
4a
4b
4c
.003
-.026
-1.965 -1.014
-1.688
.204
.170
.347
.157
-.191
-.590
-.677
-.520

Step 4 Individual role identity
Component team identity
Multiteam system identity

.686†

.669*

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Trait individual identity
Trait relational identity
Trait collective identity
Step 3 State assessment
State locomotion
Step 4 Individual role identity
Component team identity
Multiteam system identity

Person-level

Step 2

Tau
Sigma-squared

-.310
.607

Model
4d
-.680
.414
-.735
-.449
.150
.522

.001

.177

.177

.175

.185

.181

-.248*
-.091
.188
-.047
.320*

-.239*
-.158
.180
.033
.250
-.015
.267*

-.236*
-.154
.181
.020
.232
-.014
.267*

-.232*
-.153
.190
.023
.233
-.009
.263*
-.022
.002

-.235*
-.168
.173
.051
.260
-.013
.289*

-.239*
-.163
.185
.053
.242
-.009
.275*
-.021
.043
-.058

.002

-.039
.383**
.629

.372**
.615

167

.425**
.590

.351**
.595

.372**
.599

.342**
.593

.372**
.601

Table 36. Effects of Regulatory Mode and Identity on Multiteam System Performance Contribution (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 3 State assessment
State locomotion

Model 1
.004

Multiteam system performance contribution (γ)
Model
Model
Model
Model 2 Model 3
4a
4b
4c
.002
-.966
-2.329 -1.640
-.345
.145
.108
.134
.557
.085
-.183
-.055
-.475

Step 4 Individual role identity
Component team identity
Multiteam system identity

-.807
.547

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Trait individual identity
Trait relational identity
Trait collective identity
Step 3 State assessment
State locomotion
Step 4 Individual role identity
Component team identity
Multiteam system identity
Tau
Sigma-squared

Person-level

.479

-1.230†
-.816
.958*

.199

.199

.239

Step 2

.006
-.154
.088
.085
-.165
.081

.190
-.132
.019
.081
-.088
.020
-.063
.284**

.195
-.127
.011
.075
-.078
.019
-.062
.299**

.190
-.132
.021
.081
-.090
.013
-.062
.284**

-.142
.016
.064
-.091
.014
-.077
.307**
.059
-.024

-.142
.010
.061
-.070
.024
-.080
.303**
.058
.025
-.037

.354**
.625

.324**
.629

.002
-.023
.381**
.631

168

.391**
.642

.413**
.619

.365**
.623

Model
4d
1.740
.800
-.415

.429**
.623

None of the focal level identities exhibited the predicted relationship with performance
contribution at the person-level (Model 4a; γ = .147, p = n.s., γ = .002, p = n.s., and γ = -.023, p
= n.s. for individual, component team and multiteam system level performance contribution,
respectively) although component team identity did exhibit a marginally significant relationship
with component team performance contribution at the contextual level of analysis (γ = .686, p <
.10). This suggests that a focal individual’s performance contribution at the component team
level is positively influenced by the degree to which fellow actors in the system identify with
their component team. The results reported in Model 4c (Table 35) suggest that the degree to
which fellow actors in the system identify with the multiteam system as a whole also has a
positive influence on a focal individual’s performance contribution at the component team level
(γ = .669, p < .05).
These findings may represent the willingness of individuals to reach out across
component team boundaries to help (i.e., perform the support role). In fact, this explanation gains
further support when the results of Model 4b and 4c for individual role performance contribution
(Table 34) are considered. As predicted, performance at the individual level is negatively
impacted by the degree to which an actor independently (i.e., person-level of analysis) identifies
at higher levels in the multiteam system (γ = -.226, p < .05 and γ = -.210, p < .05 for component
team and multiteam system level identification, respectively). Finally, trait assessment exhibits a
negative relationship with component team performance contribution when trait identity is
controlled for (Table 35, Model 2) suggesting that concerns for self-evaluation, monitoring and
social comparison may lead individuals to contribute less to the team effort.
Combined models of performance. The results of this post hoc study implicate
regulatory mode, goal commitment, and identity in the determination of performance. Most

169

importantly, the results suggest that these factors have differential influence across the three
levels of performance contribution in a multiteam system. For example, trait assessment only
appears to have an influence on component team performance contribution whereas state
locomotion is implicated at every level of performance contribution. Multiteam system identity
exhibits a negative relationship with individual role performance contribution at the person-level
(Table 34, Model 4c) whereas it exhibits a positive relationship with component team
performance contribution at the contextual level (Table 35, Model 4c). These findings are the
result of separate analyses, however, leaving open the question: What is the combined influence
(or not) of these factors on performance contribution across the three levels of a multiteam
system?
An exploratory analysis was conducted to answer this question employing the previously
discussed findings as a guide. The results are reported as Model 2 in Table 37. As expected,
some factors lost significance when combined with other factors that exhibited equivalent
influences; for example, individually both of the multiteam system factors, goal commitment and
identity, exhibit a negative relationship with individual role performance contribution yet both
are rendered non-significant when entered simultaneously into the regression. In such cases the
factor resulting in the greatest decrease in variance, referenced to the unconditional model
(Model 1) was kept.
The combined models all have a greater impact on variance explained than do any of the
previous models (1 through 4d) with the possible exception of individual performance
contribution. Individual role identity and multiteam system goal commitment at the person-level
continue to influence individual role performance contribution; the former positively (γ = .142, p
< .10) and the latter negatively (γ = -.222, p < .05). As previously discussed, the negative

170

influence of multiteam system goal commitment may reflect the intention of the focal individual
to “do the right thing” by forgoing individual performance in order to support overall system
performance. The contextual influence of identification at the multiteam system level on
component team performance contribution is somewhat offset by individual role goal
commitment (γ = .721, p < .05 and γ = -.657, p < .05, respectively), and a similar dynamic is
apparent at the level of multiteam system performance contribution where the contextual
influence of identification at the multiteam system level on multiteam system performance
contribution is offset to some extent by multiteam system goal commitment (γ = 1.047, p < .01
and γ = -.779, p < .10, respectively).
The result at the component team performance level may simply reflect the negative
consequence of a collective void of team work (an individual performer climate or team norm) in
an environment constituting strong task performance interdependencies. The result at the
multiteam system performance level, in conjunction with the person-level result for individual
role performance contribution, may be a reflection of the negative consequences of goal
shielding in the context of a bandwagon effects. Specifically, when a collective gets too focused
on achieving the super-ordinate goal individuals may lose perspective and forgo the achievement
of the lower order goals necessary to realize high performance at the system level in order to
focus more exclusively on the super-ordinate goals. This raises two interesting questions: Is there
a point at which subjugating one’s own performance becomes detrimental to component team or
system performance? Likewise, is there a point at which the subjugation of the performance of
component teams in supporting roles becomes detrimental to system performance? Investigating
this phenomenon to better understand the dynamics involved would be an interesting and
valuable line of research to pursue in the future.

171

Table 37. Models Combining Determinants of Performance (post hoc)

Context

Predictor \ Criterion
(Intercept and context are Level-2 variables)
Step 1 Intercept
Step 2 State locomotion
Individual role identity
Multiteam system identity
Individual goal commitment
Multiteam system goal commitment

Person-level

Step 2

Performance Contribution (γ)
Individual role
Model 1
.004

Sigma-squared

Model 1
.004

Model 2
-.388
.001

Multiteam system
Model 1
.004

.721*
-.657*

Model 2
-2.964
.387
1.047**

.194

Component team (0 = Operations,
1 = Intelligence)
Trait assessment
Trait locomotion
Trait individual identity
Trait collective identity
State locomotion
Individual role identity
Multiteam system identity
Individual goal commitment
Multiteam system goal commitment
Tau

Model 2
.576
-.286
-.158

Component team

-.779†

.201

.169

.170

-.260*
-.175
.202†
.322*
.293**

-.137
-.160
.472**
.142†

-.018
.018
.302**

-.029
-.055

.028

-.222**

-.094

.195**

.273**

.383**

.218**

.381**

.266**

.809

.714

.629

.586

.631

.613

172

Lastly, component team is the only performance contribution level that exhibits a direct
relationship with traits. Reflecting the tendency of assessors to seek performance outcomes that
can be directly attributed to them, trait assessment is negatively related to component team
performance contribution (γ = -.260, p < .05). The finding that both individual (γ = .202, p < .10)
and collective (γ = .322, p < .05) identity are positively related to component team performance
contribution may be an aberration of Brewer’s (1991) optimal distinctiveness perspective.
Individuals are both a characterless part of a group, the component team, and a somewhat
uniquely distinctive individual performer at this level. Each component team is part of a broader
system, yet uniquely qualified and tasked. This may be an artifact of the situation, however. Each
multiteam system in the study performed in a vacuum; i.e., there was no comparison system from
which to be distinctive. This is often, but not always, the case for multiteam systems. Thus this
offers another fertile avenue for future research.
Limitations
Every study has limitations that offer opportunities for learning and improvement that
serve as a guide to future research. The intent of this study was to investigate the influences of
regulatory mode, goal commitment and identity on the three levels of performance contribution
in a multiteam system under specific, controlled conditions – high state assessment and high
state locomotion. These conditions were selected for investigation by design because prior
research has found that individuals are most likely to attain their goals if they are high on both
regulatory mode dimensions (Pierro, Kruglanski, & Higgins, 2006a). Thus, high state assessment
and high state locomotion were envisioned to be the two conditions under which maximum
multiteam system performance would emerge. While maximum performance is clearly an
important outcome, and thus of keen interest to management scholars and practicing managers

173

alike, it is not the only state of performance that offers valuable insights. Understanding why
some multiteam systems underperform their potential can be equally informative. Thus future
research should target other performance outcome levels and investigate the influences of
regulatory mode orientations other than the high state assessment and high state locomotion
conditions.
The assumption that Operations component team roles and Intelligence component team
roles would respectively place participants into a high locomotion and a high assessment
situation-induced state was untested and turned out to be somewhat tenuous. Fortunately, selfreported measures of state regulatory mode were collected. They proved invaluable to the
analyses, although they are not ideal for a couple of reasons. First, while the state regulatory
mode measurement tool was based on a psychometrically tested and sound survey instrument,
and exhibited a good fit when factor analyzed, the altered scale has never been rigorously tested
for construct validity. The scale did exhibit good fit when factor analyzed and the underlying
constructs it was intended to measure did perform as predicted by theory, two factors arguably in
support of validity. Still, scale validation is an opportunity for future research that would only
serve to strengthen the findings of this study should construct validity be demonstrated. Second,
the self-reported measures of state regulatory mode were collected immediately after the
performance episode and thus they may not be a true representation of the state regulatory mode
during the performance episode. Although participants had no knowledge of actual performance,
these measures may suffer from a retrospective measurement bias based on perceptions of
performance. Thus the findings of the analyses employing these measures, while informative and
illuminating, should be considered cautiously. Future research should be undertaken to validate
these findings.

174

Another limitation of this study is an unavoidable (and age old) consequence of the
nature of the research question, determinants of individual performance contribution in a
multiteam system context. It is hard, if not outright impossible, to completely disentangle
individual-level performance contribution at the level of a collective in a performance hierarchy
(e.g., component team- and multiteam system-level) comprising a system of highly
interdependent work from the data of one performance episode. While the objective measures
and analysis technique employed in this study afford some ability to do this, the degree to which
these measures do or do not accurately represent the intended underlying construct is a legitimate
concern.
In addition, the simulation exercise employed in this study is but one of a universe of
multiteam system performance situations, the performance episode was conducted in a
laboratory setting, and the participants in this study were all undergraduate students. While the
findings reported here represent a meaningful step toward understanding the complex issues
associated with performance contribution in a multiteam system, future research should be
conducted employing a variety of performance measurement tools across a series of performance
episodes and diverse populations in order to better understand individual performance
contribution in multiteam system contexts.
Finally, the use of hierarchical linear modeling techniques provides a more accurate
estimation of standard errors and thus a more conservative estimate of significance than does
OLS regression, however, the sample size at both levels of the models used in this study was
limited. While this represents an efficient use of research resources (a public good) the findings
reported here may not tell the full story as the power to detect statistical significant effects was
limited. Even so, statistical significance is arguably an overly simplistic yardstick with which to

175

measure the meaningfulness of a scientific inquiry. Theory informed inference and speculative
conjecture supported by empirical results are of potentially greater importance than mere
statistical significance, especially at the dawn of a new area of research as is the case for
multiteam systems. Against these criteria this study was a resounding success.
Contribution to the Literature
This study was motivated and conceived both to enhance our understanding of what
determines the goal focus of individuals situated in the complex organizational context of a
multiteam system by gaining an understanding of the regulatory process driving activated
identity and goal commitment, and to gain an understanding of the degree to which regulatory
mode, identity and goal commitment drive or inhibit incremental performance contribution
across goal levels in these contexts. The findings discussed previously are a testament to the
successful fulfillment of these intentions. In addition, while the study was undertaken with a
priori hypotheses in mind, the design was robust enough to withstand an erroneous assumption
and deliver insightful inferences from post hoc theorizing and analysis that meaningfully inform
the direction of future research.
This is the first study to join together theories of regulatory mode, identity, goal
commitment and multiteam systems into an informed perspective of their collective and
interdependent impact on performance. As such it serves to inform and expand the literature of
all these areas. Specifically, while finer details such as effect size are still in contention, this
study theorizes and demonstrates empirically that incremental performance contribution across
levels in the multiteam system goal hierarchy are differentially affected by context through its
selective influence on the key psychological processes of regulation, identity and goal
commitment. The findings reported herein are also of value to leaders of multiteam systems. For

176

example, while a layman’s perspective would lead to the conclusion that strong identification
with the organization and a strong commitment to the goals of the organization is always a good
thing, and the more the better, this research suggests that this is not always the case. The results
reported here indicate that strong identification with the multiteam system and a strong
commitment to the goals of the system may serve, ultimately, to undermine performance.
If teams are truly becoming the fundamental building block of organizations, then
multiteam systems will increasingly represent a structure through which organizations
accomplish their objectives. In contexts where the nature of the task environment requires
modular organization by specialization or broad representation from many different
constituencies, the multiteam system organizational form affords a performance advantage
(Davison et al., 2012). Yet individuals situated in multiteam systems face a hierarchy of
interdependent goals and responsibility associated with multiple organizational levels (e.g.,
individual, component team, system), embodied in a context fraught with resource constraints
(e.g., psychological, time). What factors determine their performance focus, and how focus
affects performance contribution are critical questions on the path to understanding the
determinants of superior multiteam system performance. Hopefully this study will motivate
future work in this important, but empirically under-developed area so that the operational
effectiveness of multiteam systems will someday attain their theoretical potential.

177

APPENDICES

178

Appendix A: Regulatory Mode Questionnaire (RMQ)
(Administered in pre-lab survey using Qualtrics)
Table 38. Regulatory Mode Questionnaire (RMQ)
Locomotion items
1.
2.
3.
4.
5.
6.

I don't mind doing things even if they involve extra effort.
I am a "workaholic."
I feel excited just before I am about to reach a goal.
I enjoy actively doing things, more than just watching and observing.
I am a "doer."
When I finish one project, I often wait awhile before getting started on a new one.
(reverse-scored)
7.
When I decide to do something, I can't wait to get started.
8.
By the time I accomplish a task, I already have the next one in mind.
9.
I am a "low energy" person, (reverse-scored)
10. Most of the time my thoughts are occupied with the task I wish to accomplish.
11. When I get started on something, I usually persevere until I finish it.
12. I am a "go-getter."
Assessment items
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
•

•

I never evaluate my social interactions with others after they occur. (reverse-scored)
I spend a great deal of time taking inventory of my positive and negative characteristics.
I like evaluating other people's plans.
I often compare myself with other people.
I don't spend much time thinking about ways others could improve themselves, (reversescored)
I often critique work done by myself or others.
I often feel that I am being evaluated by others.
I am a critical person.
I am very self-critical and self-conscious about what I am saying.
I often think that other people's choices and decisions are wrong.
I rarely analyze the conversations I have had with others after they occur. (reversescored)
When I meet a new person I usually evaluate how well he or she is doing on various
dimensions (e.g., looks, achievements, social status, clothes).

Kruglanski, A. W., Thompson, E. P., Higgins, E. T., Atash, M. N., Pierro, A., Shah, J. Y., &
Spiegel, S. 2000. To "do the right thing" or to "just do it": Locomotion and assessment as
distinct self-regulatory imperatives. Journal of Personality and Social Psychology, 79, 793815.
Also in: Higgins, E. T., Kruglanski, A. W., & Pierro, A. 2003. Regulatory mode: Locomotion
and assessment as distinct orientations. In M. P. Zanna (Ed.), Advances in Experimental
Social Psychology, Vol 35, 293-344. San Diego: Academic Press Inc.

179

Appendix B: Levels of Self-concept Scale (LSCS)
(Administered in pre-lab survey using Qualtrics)
Table 39. Levels of Self-concept Scale (LSCS)
Instructions: The following questions ask you to provide some information about yourself.
Please respond by marking the circle to the right of each statement that best represents your
answer.
Individual-level (Comparative identity subscale)
1.

I thrive on opportunities to demonstrate that my abilities or talents are better than those
of other people.
2.
I have a strong need to know how I stand in comparison to my coworkers.
3.
I often compete with my friends.
4.
I feel best about myself when I perform better than others.
5.
I often find myself pondering over the ways that I am better or worse off than other
people around me.
Relational-level (Concern for others subscale)
1.

If a friend was having a personal problem, I would help him/her even if it meant
sacrificing my time or money.
2.
I value friends who are caring, empathic individuals.
3.
It is important to me that I uphold my commitments to significant people in my life.
4.
Caring deeply about another person such as a close friend or relative is important to me.
5.
Knowing that someone I am close to acknowledges and values the role that I play in
their life makes me feel like a worthwhile person.
Collective-level (Group achievement focus subscale)
1.
2.
3.
4.
5.
•

Making a lasting contribution to groups that I belong to, such as a student organization
or club, is very important to me.
When I become involved in a group project, I do my best to ensure its success.
I feel great pride when my team or group does well, even if I’m not the main reason for
its success.
I would be honored if I were chosen by an organization or club that I belong to, to
represent them at a conference or meeting.
When I’m part of a team, I am concerned about the group as a whole instead of whether
individual team members like me or whether I like them.
Self-concept Scale (LSCS)
o Johnson, R. E., Selenta, C., & Lord, R. G. (2006). When organizational justice and
the self-concept meet: Consequences for the organization and its members.
Organizational Behavior and Human Decision Processes, 99, 175-201.
o Selenta, C., & Lord, R. G. (2005). Development of the levels of self-concept scale:
Measuring the individual, relational, and collective levels. Unpublished manuscript.

180

Appendix C: Post-training Survey
(Administered immediately after training using Qualtrics)
Table 40. Post-training Survey
Instructions: This questionnaire consists of statements about the training you have just
completed. Please read each statement that follows carefully, and take special care to mark the
circle that best represents your opinion.
Using the 1 to 6 scale to the right of these instructions, please indicate the degree to which you
agree or disagree with the following statements by marking the circle to the right of the
statement that best represents your opinion.
I know how to place (deploy) and move operations assets (Strike, Escort, Refuel, and Info
RPA icons) on the game grid.
I know how to place (deploy) and move intelligence assets (A, C, H, V) on the game grid.
Information obtained from intelligence assets placed in a location that is not a “sweet spot”
location (i.e., a cell where they are 90% inaccurate) is reliable (i.e., can be trusted). (Reverse
coded)
Strike, Escort, and Info missions to the bottom-half of the grid (the South, Rows 1-8) require a
Refuel RPA. (Reverse coded)
Strike RPAs can be used to destroy any target (i.e., both opportunities and threats). (Reverse
coded)
Only intelligence assets find targets. (Reverse coded)
Mobile targets identified in turns 3, 7, and 9 will be in the same location in the next turn.
Sending an intelligence asset type (A, C, H, V) to a location containing a target that is already
known to be there (i.e., was identified by an operations asset in a previous round) helps to
verify the asset type’s accuracy.
It takes 2 Escort RPAs to destroy a large mobile threat but only 1 Escort RPA to destroy a
large fixed threat. (Reverse coded)
Points are lost when intelliegence assets (A, C, H, V) are destroyed. (Reverse coded)
1 = Strongly disagree
2 = Disagree
3 = Mildly disagree
4 = Mildly agree
5 = Agree
6 = Strongly Agree

181

Appendix D: Goal Commitment Measures
(Administered as part of Time 1 survey using Qualtrics)
Individual-level Instructions: This questionnaire consists of statements about the goals assigned specifically to you (“destroying small
targets without losing an asset” if you are a member of the Operations sub-team or “accurately identifying targets without wasting
intel missions” if you are a member of the Intelligence sub-team). Please answer with respect to the goals assigned to you only.
Component Team-level Instructions: This questionnaire consists of statements about the goals assigned to your sub-team ("targets
destroyed" if you are on the Operations sub-team or "queries resulting in an accurate target icon on the COP" if you are on the
Intelligence sub-team). Please answer with respect to the goals assigned to your sub-team only.
The researcher has assigned a difficult but attainable goal to your sub-team that represents performance at or above 75% of teams who
have already completed this exercise. Please indicate the degree to which you agree or disagree with the following statements by
marking the circle to the right of the statement. Read each statement that follows carefully, and take special care to mark the circle that
best represents your opinion.
MTS-level Instructions: This questionnaire consists of statements about the goals of the organization as a whole. Please answer with
respect to the overall performance goals of the organization.
The researcher has assigned difficult but attainable goals for "Total Points Scored" and "COP Accuracy" that represent performance at
or above 75% of teams who have already completed this exercise. Please indicate the degree to which you agree or disagree with the
following statements by marking the circle to the right of the statement. Read each statement that follows carefully, and take special
care to mark the circle that best represents your opinion.

182

Appendix D (cont’d)
Table 41. Goal Commitment Measures
Individual level (Original items)

Component team

MTS

1 (R) It's hard to take this [my individual]
goal seriously.
2 (R) It's unrealistic for me to expect to reach
this [my individual] goal.
3 (R) It is quite likely that this [my individual]
goal may need to be revised, depending
on how things go.
4 (R) Quite frankly, I don't care if I achieve
this [my individual] goal or not.

It's hard to take the goals of the subteam seriously.
It's unrealistic for me to expect the subteam’s goals to be reached.
It is quite likely that the goals for the
sub-team need to be revised, based on
how things are going.
Quite frankly, I don't care if the subteam’s goals are achieved or not.

5

I am strongly committed to pursuing
the goals of the sub-team.
It wouldn't take much to make me
abandon the sub-team’s goals.
I think the goals of the sub-team are
good goals to shoot for.
I am willing to put forth a great deal of
effort beyond what I'd normally do to
achieve the sub-team’s goals.
There is not much to be gained by
trying to achieve the goals of the subteam.

It's hard to take the goals of the
organization seriously.
It's unrealistic for me to expect the
organization’s goals to be reached.
It is quite likely that the goals for the
organization need to be revised, based
on how things are going.
Quite frankly, I don't care if the
organization’s goals are achieved or
not.
I am strongly committed to pursuing
the goals of the organization.
It wouldn't take much to make me
abandon the organization’s goals.
I think the goals of the organization
are good goals to shoot for.
I am willing to put forth a great deal of
effort beyond what I'd normally do to
achieve the organization’s goals.
There is not much to be gained by
trying to achieve the goals of the
organization.

I am strongly committed to pursuing
this [my individual] goal.
6 (R) It wouldn't take much to make me
abandon this [my individual] goal.
7
I think this [my individual] goal is a
good goal to shoot for.
8
I am willing to put forth a great deal of
effort beyond what I'd normally do to
achieve this [my individual] goal.
9 (R) There is not much to be gained by
trying to achieve this [my individual]
goal.
Notes.

(R) – Reverse coded
Items 1, 4, 5, 6, and 7 form the 5-item scale (Klein, Wesson, Hollenbeck, Wright, & DeShon, 1999)
Hollenbeck, J. R., O’Leary, A. M., Klein, H. J., & Wright, P. M. (1989). Investigation of the construct validity of a selfreport measure of goal commitment. Journal of Applied Psychology, 74, 951-956.

183

Appendix E: Individual-level Identity Measure
(Administered as part of Time 1 survey using Qualtrics)
Instructions: Please indicate the degree to which you agree or disagree with the following statements by marking the circle to the right
of the statement. Read each statement that follows carefully, and take special care to mark the circle that best represents your opinion.

Table 42. Individual-level Identity Measure
Individual identity

Original items from Comparative identity subscale
1
2
3
4
5
•

I thrive on opportunities to demonstrate that my abilities or
talents are better than those of other people.
I have a strong need to know how I stand in comparison to
my coworkers.
I often compete with my friends.
I feel best about myself when I perform better than others.
I often find myself pondering over the ways that I am better
or worse off than other people around me.

I thrive on opportunities to demonstrate that my abilities are
better than those of the other people currently in the lab.
I have a strong need to know how I am doing in comparison
to the others in the lab right now.
I am competing with the others in the lab right now.
I feel best about myself when I perform better than others
currently in the lab.
I am often pondering the ways that I am better or worse off
than the other people in the lab right now.

Adapted from Levels of Self-concept Scale (LSCS)
o Johnson, R. E., Selenta, C., & Lord, R. G. (2006). When organizational justice and the self-concept meet: Consequences
for the organization and its members. Organizational Behavior and Human Decision Processes, 99, 175-201.
o Selenta, C., & Lord, R. G. (2005). Development of the levels of self-concept scale: Measuring the individual, relational,
and collective levels. Unpublished manuscript.

184

Appendix F: Group-level Identity Measures
(Administered as part of Time 1 survey using Qualtrics)
Table 43. Group-level Identity Measures
Team-level Instructions: This questionnaire consists of statements about how closely you identify with members of the sub-team you
were assigned to (i.e., you and the 3 individuals sitting at the same table with you). Please answer with respect to the individuals on
your sub-team only.
MTS-level Instructions: This questionnaire consists of statements about how closely you identify with members of the organization as
a whole (i.e., everyone in the room). Please answer with respect to the individuals on all sub-teams collectively.
Original item
1
2
3
4
5
6
7
8

Dimension

Component team

I feel strongly affiliated with this
Commitment I feel strongly affiliated with
this sub-team.
group.
Other groups can learn a lot from us.
Superiority
Belonging to this group is an important Importance Belonging to this sub-team is
an important part of my
part of my identity.
identity.
In times of trouble, the only way to know
Deference
what to do is to rely on the group leaders.
I am glad to contribute to this group.
Commitment I am glad to contribute to this
sub-team.
Compared to other groups of this kind,
Superiority
this group is particularly good.
It is important to me that I view myself Importance It is important to me that I
view myself as a member of
as a member of this group.
this sub-team.
All group members should respect the
Deference
customs, the institutions, and the leaders
of the group.

185

MTS
I feel strongly affiliated with
this organization.
Belonging to this organization
is an important part of my
identity.
I am glad to contribute to this
organization.
It is important to me that I
view myself as a member of
this organization.

Appendix F (cont’d)
Table 43 (cont’d)
9

I am strongly committed to this group.

10

Relative to other groups, we are a very
moral group.
It is important to me that others see me
as a member of this group.

11
12
13
14
15

16

It is disloyal to criticize this group.
I like to help this group.
This group is better than other groups in
all respects.
When I talk about the group members,
I usually say “we” rather than “they.”
There is usually a good reason for every
rule and regulation that the group leaders
propose.

Commitment I am strongly committed to
this sub-team.
Superiority
Importance

It is important to me that
others see me as a member of
this sub-team.

Deference
Commitment I like to help this sub-team.
Superiority
Importance

When I talk about the subteam members, I usually say
“we” rather than “they.”

I am strongly committed to this
organization.
It is important to me that
others see me as a member of
this organization.
I like to help this organization.
When I talk about the
organization members, I
usually say “we” rather than
“they.”

Deference

Note: The adapted scale to be used in this study will utilize the four Commitment items (1, 5, 9, 13) and the four Importance items (3,
7, 11, 15).
•

Roccas, S., Sagiv, L., Schwartz, S., Halevy, N., & Eidelson, R. (2008). Toward a unifying model of identification with groups:
Integrating theoretical perspectives. Personality and Social Psychology Review, 12, 280-306.

186

Appendix G: State Regulatory Mode Questionnaire (based on the RMQ)
(Administered as part of Time 2 survey using Qualtrics)
Table 44. State Regulatory Mode Questionnaire
Instructions: This questionnaire consists of statements about the role you were asked to perform
in today’s exercise. Please take a few moments to reflect back on your participation in the
exercise you have just completed. Then answer the following question with only these thoughts
in mind; i.e., how you feel right now based on your participation in this exercise.
Using the 1 to 6 scale to the right of these instructions, please indicate the degree to which you
agree or disagree with the following statements by marking the circle to the right of the
statement. Read each statement that follows carefully, and take special care to mark the circle
that best represents your opinion.
Locomotion items
1.

2.
3.
4.

5.
6.

7.
8.

9.

RMQ
I don't mind doing things even if they involve extra effort.
Revised I didn't mind doing the things I had to do, even when they took some extra
RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised

RMQ
Revised
10. RMQ
Revised
11. RMQ
Revised
12. RMQ
Revised

effort.
I am a "workaholic."
I could easily have played more rounds.
I feel excited just before I am about to reach a goal.
I felt excited every time I destroyed / identified a target (Ops / Intel).
I enjoy actively doing things, more than just watching and observing.
I enjoyed deploying my assets at the start of each round more than watching
others deploy their assets or waiting for the execution of my moves.
I am a "doer."
I was a “doer” in this exercise.
When I finish one project, I often wait awhile before getting started on a new
one. (reverse-scored)
I would have preferred more time between rounds. (reverse-scored)
When I decide to do something, I can't wait to get started.
When I decided to do something, I couldn’t wait to do it.
By the time I accomplish a task, I already have the next one in mind.
By the time a round ended, I already knew what I wanted to do in the next
round.
I am a "low energy" person. (reverse-scored)
I did not feel all that energized during the game. (reverse-scored)
Most of the time my thoughts are occupied with the task I wish to accomplish.
Most of the time I was thinking about what I wanted to accomplish.
When I get started on something, I usually persevere until I finish it.
Once I started something, I finished it.
I am a "go-getter."
I was a real "go-getter” in this exercise.

187

Appendix G (cont’d)
Table 44 (cont’d)
Assessment items
1.

RMQ
Revised

2.

3.
4.
5.

RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised

6.
7.
8.
9.
10.
11.

RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised
RMQ
Revised

12.

RMQ
Revised

•

I never evaluate my social interactions with others after they occur. (reversescored)
I did not evaluate the interactions I had with others after they occurred.
(reverse-scored)
I spend a great deal of time taking inventory of my positive and negative
characteristics.
I spent time taking stock of my good and my bad moves.
I like evaluating other people's plans.
I enjoyed evaluating what others on the team were doing.
I often compare myself with other people.
I would like to compare how I did to how others did.
I don't spend much time thinking about ways others could improve themselves,
(reverse-scored)
I didn’t spend much time thinking about ways others could do better. (reversescored)
I often critique work done by myself or others.
I spent time critiquing my moves and the moves others had made.
I often feel that I am being evaluated by others.
I often felt that other players were evaluating my moves and decisions.
I am a critical person.
I critically assessed things.
I am very self-critical and self-conscious about what I am saying.
I was very careful about what I said.
I often think that other people's choices and decisions are wrong.
I thought that many of the moves and decisions other people made were wrong.
I rarely analyze the conversations I have had with others after they occur.
(reverse-scored)
I rarely analyzed the conversations I had with others after they occurred.
(reverse-scored)
When I meet a new person I usually evaluate how well he or she is doing on
various dimensions (e.g., looks, achievements, social status, clothes).
I spent time evaluating others on various dimensions (e.g., looks, personality,
attitude, clothes).

Kruglanski, A. W., Thompson, E. P., Higgins, E. T., Atash, M. N., Pierro, A., Shah, J. Y., &
Spiegel, S. 2000. To "do the right thing" or to "just do it": Locomotion and assessment as
distinct self-regulatory imperatives. Journal of Personality and Social Psychology, 79, 793815.

188

REFERENCES

189

REFERENCES
Aaker, J. L., & Lee, A. Y. (2006). Understanding regulatory fit. Journal of Marketing Research,
43, 15-19.
Ancona, D. G., & Caldwell, D. F. (1988). Beyond task and maintenance: Defining external
functions in groups. Group & Organization (Management) Studies, 13, 468-494.
Anderson, J. C., & Gerbing, D. W. (1982). Some methods for respecifying measurement models
to obtain unidimensional construct measurement. Journal of Marketing Research, 19,
453-460.
Appelt, K. C., Zou, X., & Higgins, E. T. (2010). Feeling right or being right: When strong
assessment yields strong correction. Motivation and Emotion, 34, 316-324.
Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of other in the self scale and the structure
of interpersonal closenesss. Journal of Personality and Social Psychology, 63, 596-612.
Ashforth, B. E. (2001). Role transitions in organizational life: An identity-based perspective.
Mahwah, NJ: Erlbaum.
Ashforth, B. E., Kreiner, G. E., Clark, M. A., & Fugate, M. (2007). Normalizing dirty work:
Reducing the salience of occupational stigma. Academy of Management Journal, 50,
149–174.
Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organization. Academy of
Management Review, 14, 20-39.
Ashmore, R. D., Deaux, K., & McLaughlin-Volpe, T. (2004). An organizing framework for
collective identity: Articulation and significance of multidimensionality. Psychological
Bulletin, 130, 80-114.
Astley, W. G., & Zajac, E. J. (1990). Beyond dyadic exchange: Functional interdependence and
subunit power. Organization Studies, 11, 481-501.
Astley, W. G., & Zajac, E. J. (1991). Intraorganizational power and organizational design:
Reconciling rational and coalitional models of organization. Organization Science, 2,
399-411.
Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure, process, and
content. Psychological Bulletin, 120, 338-375.
Avnet, T., & Higgins, E. T. (2003). Locomotion, assessment, and regulatory fit: Value transfer
from "how" to "what". Journal of Experimental Social Psychology, 39, 525-530.
Avnet, T., & Higgins, E. T. (2006). How regulatory fit affects value in consumer choices and
opinions. Journal of Marketing Research, 43, 1-10.

190

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman.
Bandura, A., & Locke, E. A. (2003). Negative self-efficacy and goal effects revisited. Journal of
Applied Psychology, 88, 87-99.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.
Bass, B. M. (1990). Bass & Stodgill’s handbook of leadership: Theory, research, & managerial
applications (3rd ed.). New York: Free Press.
Benjamin, L., & Flynn, F. J. (2006). Leadership style and regulatory mode: Value from fit?
Organizational Behavior and Human Decision Processes, 100, 216-230.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107,
238−246.
Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of
covariance structures. Psychological Bulletin, 88, 588-606.
Bollen, K. A. (1989). A new incremental fit index for general structural equation models.
Sociological Methods and Research, 17, 303-316.
Bollen, K. A., & Long, J. S., eds. (1993). Testing structural equation models. Newbury Park,
CA: Sage.
Brewer, M. B. (1991). The social self: On being the same and different at the same time.
Personality and Social Psychology Bulletin, 17, 475-482.
Brewer, M. B., & Gardner, W. (1996). Who is this ''we''? Levels of collective identity and self
representations. Journal of Personality and Social Psychology, 71(1), 83-93.
Brickson, S. (2000). The impact of identity orientation on individual and organizational
outcomes in demographically diverse settings. Academy of Management Review, 25(1),
82-101.
Brockner, J., Paruchuri, S., Idson, L. C., & Higgins, E. T. (2002). Regulatory focus and the
probability estimates of conjunctive and disjunctive events. Organizational Behavior and
Human Decision Processes, 87, 5-24.
Brown, S. P., & Leigh, T. W. (1996). A new look at psychological climate and its relationship to
job involvement, effort, and performance. Journal of Applied Psychology, 81, 358-368.
Caldwell, D. F., & O' Reilly, C. A. (1982). Responses to failure: The effects of choice and
responsibility on impression management. Academy of Management Journal, 25, 121136.

191

Callero, P. L. (1985). Role-identity salience. Social Psychology Quarterly, 48, 203-215.
Campion, M. A., & Lord, R. G. (1982). A control systems conceptualization of the goal setting
and changing process. Organizational Behavior and Human Performance, 30, 265-287.
Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative affect: A
control process view. Psychological Review, 97, 19-35.
Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. New York, NY:
Cambridge University Press.
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective
responses to impending reward and punishment: The BIS/BAS scales. Journal of
Personality and Social Psychology, 67, 319-333.
Cesario, J., Grant, H., & Higgins, E. T. (2004). Regulatory fit and persuasion: Transfer from
"feeling right". Journal of Personality and Social Psychology, 86, 388-404.
Cesario, J., & Higgins, E. T. (2008). Making message recipients "Feel Right": How nonverbal
cues can increase persuasion. Psychological Science, 19, 415-420.
Chernev, A. (2009). Choosing versus rejecting: The impact of goal-task compatibility on
decision confidence. Social Cognition, 27, 249-260.
Cropanzano, R., James, K., & Citera, M. (1993). A goal hierarchy model of personality,
motivation, and leadership. In L. L. Cummings & B. M. Staw (Eds.), Research in
organizational behavior (Vol. 15, pp. 267-322). Greenwich, CT: JAI Press.
Crowe, E., & Higgins, E. T. (1997). Regulatory focus and strategic inclinations: Promotion and
prevention in decision-making. Organizational Behavior and Human Decision Processes,
69, 117-132.
Csikszentmihalyi, M. (1975). Beyond boredom and anxiety. San Francisco: Jossey-Bass.
Davison, R. B., & Hollenbeck, J. R. (2011). Boundary spanning in the domain of multiteam
systems. In S. J. Zaccaro, M. A. Marks & L. A. DeChurch (Eds.), Multiteam Systems: An
Organization Form for Dynamic and Complex Environments. New York: Routledge.
Davison, R. B., Hollenbeck, J. R., Barnes, C. M., Sleesman, D. J., & Ilgen, D. R. (2012).
Coordinated action in multiteam systems. Journal of Applied Psychology, 97, 808-824.
De Cremer, D., Mayer, D. M., van Dijke, M., Schouten, B. C., & Bardes, M. (2009). When does
self-sacrificial leadership motivate prosocial behavior? It depends on followers'
prevention focus. Journal of Applied Psychology, 94, 887-899.
Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments
examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin,
125, 627-668.

192

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human
behavior. New York: Plenum.
Deci, E. L., & Ryan, R. M. (1991). A motivational approach to self: Integration in personality. In
R. Dienstbier (Ed.), Nebraska symposium on motivation (Vol. 38). Lincoln, NE:
University of Nebraska Press.
Dehaene, S., & Naccache, L. (2001). Towards a cognitive neuroscience of consciousness: Basic
evidence and a workspace framework. Cognition, 79, 1-37.
DeShon, R. P., & Landis, R. S. (1997). The dimensionality of the Hollenbeck, Williams, and
Klein (1989) measure of goal commitment on complex tasks. Organizational Behavior
and Human Decision Processes, 70, 105-116.
Diefendorff, J. M., & Lord, R. G. (2008). Self-regulation and goal striving processes. In R.
Kanfer, G. Chen & R. D. Pritchard (Eds.), Work motivation: Past, present, and future
(pp. 151–196). New York: Routledge.
Dweck, C. S. (1991). Self-theories and goals: Their role in motivation, personality and
development. In R. Dienstbier (Ed.), Nebraska symposium on motivation (Vol. 38, pp.
199-235). Lincoln, NE: University of Nebraska Press.
Ellemers, N., Spears, R., & Doosje, B. (2002). Self and social identity. Annual Review of
Psychology, 53(1), 161-186.
Elliott, E. S., & Dweck, C. S. (1988). Goals: An approach to motivation and achievement.
Journal of Personality and Social Psychology, 54, 5-12.
Enders, C. K., &; Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel
models: A new look at an old issue. Psychological Methods, 12, 121-138. doi:
10.1037/1082-989X.12.2.121
Erez, A., & Isen, A. M. (2002). The influence of positive affect on the components of expectancy
motivation. Journal of Applied Psychology, 87, 1055-1067.
Erez, M. (1977). Feedback: A necessary condition for the goal setting-performance relationship.
Journal of Applied Psychology, 62, 624-627.
Erez, M., & Kanfer, F. H. (1983). The role of goal acceptance in goal setting and task
performance. Academy of Management Review, 8, 454-463.
Ethiraj, S. K., & Levinthal, D. (2009). Hoping for A to Z while rewarding only A: Complex
organizations and multiple goals. Organization Science, 20, 4-21.
Farmer, S. M., Tierney, P., & Kung-McIntyre, K. (2003). Employee creativity in Taiwan: An
application of role identity theory. Academy of Management Journal, 46, 618-630.
Farmer, S. M., & Van Dyne, L. (2010). The idealized self and the situated self as predictors of

193

employee work behaviors. Journal of Applied Psychology, 95, 503-516.
Foote, N. N. (1951). Identification as the basis for a theory of motivation. American Sociological
Review, 16, 14-21.
Forgas, J. P., Baumeister, R. F., & Tice, D. (2009). The psychology of self-regulation: An
introductory review. In J. P. Forgas, R. F. Baumeister & D. Tice (Eds.), The psychology
of self-regulation: Cognitive, affective, and motivational processes (pp. 1-17). New York:
Psychology Press.
Forster, J., Higgins, E. T., & Idson, L. C. (1998). Approach and avoidance strength during goal
attainment: Regulatory focus and the "goal looms larger" effect. Journal of Personality
and Social Psychology, 75, 1115-1131.
Freitas, A. L., & Higgins, E. T. (2002). Enjoying goal-directed action: The role of regulatory fit.
Psychological Science, 13, 1-6.
Frese, M., & Zapf, D. (1994). Action as the core of work psychology: A German approach. In H.
C. Triandis & M. D. Dunnette (Eds.), Handbook of industrial and organizational
psychology (2nd ed., pp. 271-340). Palo Alto, CA: Consulting Psychologists Press.
Friedman, R. S. (1999). The phenomenological correlates and consequences of distinct selfregulatory systems. Unpublished doctoral dissertation, Columbia University, New York.
Friedman, R. S., & Forster, J. (2001). The effects of promotion and prevention cues on creativity.
Journal of Personality and Social Psychology, 81, 1001-1013.
Fuster, J. M. (2002). Frontal lobe and cognitive development. Journal of Neurocytology, 31, 373385.
Gabriel, S., & Gardner, W. L. (1999). Are there ‘‘his’’ and ‘‘her’’ types of interdependence? The
implications of gender differences in collective versus relational interdependence for
affect, behavior, and cognition. Journal of Personality and Social Psychology, 77, 642–
655.
Gist, M. E., Locke, E. A., & Taylor, M. S. (1987). Organizational behavior: Group structure,
process, and effectiveness. Journal of Management, 13, 237-257.
Gladstein, D. L. (1984). Groups in context: A model of task group effectiveness. Administrative
Science Quarterly, 29, 499-517.
Gollwitzer, P. M. (1990). Action phases and mind-sets. In E. T. Higgins & R. M. Sorrentino
(Eds.), Handbook of motivation and cognition: Foundations of social behavior (Vol. 2,
pp. 53-92). New York: Guilford Press.
Gollwitzer, P. M. (1996). The volitional benefits of planning. In P. M. Gollwitzer & J. A. Bargh
(Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 287312). New York: Guilford.

194

Gollwitzer, P. M., & Bargh, J. A. (1996). The psychology of action. New York: Guilford Press.
Henderson, M. D., de Liver, Y., & Gollwitzer, P. A. (2008). The effects of an implemental mindset on attitude strength. Journal of Personality and Social Psychology, 94, 396-411.
Hetts, J. J., Sakuma, M., & Pelham, B. W. (1999). Two roads to positive regard: Implicit and
explicit self-evaluation and culture. Journal of Experimental Social Psychology, 35, 512559.
Higgins, E. T. (1997). Beyond pleasure and pain. American Psychologist, 52, 1280-1300.
Higgins, E. T. (2000). Making a good decision: Value from fit. American Psychologist, 55,
1217-1230.
Higgins, E. T. (2002). How self-regulation creates distinct values: The case of promotion and
prevention decision making. Journal of Consumer Psychology, 12, 177-191.
Higgins, E. T. (2005). Value from regulatory fit. Current Directions in Psychological Science,
14, 209-213.
Higgins, E. T. (2006). Value from hedonic experience and engagement. Psychological Review,
113, 439-460.
Higgins, E. T., Cesario, J., Hagiwara, N., Spiegel, S., & Pittman, T. (2010). Increasing or
Decreasing Interest in Activities: The Role of Regulatory Fit. Journal of Personality and
Social Psychology, 98, 559-572.
Higgins, E. T., Idson, L. C., Freitas, A. L., Spiegel, S., & Molden, D. C. (2003). Transfer of
value from fit. Journal of Personality and Social Psychology, 84, 1140-1153.
Higgins, E. T., & Kruglanski, A. W. (1995). A theory of regulatory modes: When locomotion
versus assessment is emphasized. Unpublished manuscript, Columbia University, New
York.
Higgins, E. T., Kruglanski, A. W., & Pierro, A. (2003). Regulatory mode: Locomotion and
assessment as distinct orientations. In M. P. Zanna (Ed.), Advances in Experimental
Social Psychology (Vol. 35, pp. 293-344). San Diego: Academic Press Inc.
Higgins, E. T., & Scholer, A. A. (2009). Engaging the consumer: The science and art of the
value creation process. Journal of Consumer Psychology, 19, 100-114.
Higgins, E. T., Shah, J., & Friedman, R. (1997). Emotional responses to goal attainment:
Strength of regulatory focus as moderator. Journal of Personality and Social Psychology,
72, 515-525.
Hoegl, M., Weinkauf, K., & Gemuenden, H. G. (2004). Interteam coordination, project
commitment, and teamwork in multiteam R&D projects: A longitudinal study.
Organization Science, 15, 38-55.

195

Hogg, M. A., & Terry, D. J. (2000). Social identity and self-categorization processes in
organizational contexts. Academy of Management Review, 25(1), 121-140.
Hollenbeck, J. R., & Klein, H. J. (1987). Goal commitment and the goal-setting process:
Problems, prospects, and proposals for future research. Journal of Applied Psychology,
72, 212-220.
Hollenbeck, J. R., O'Leary, A. M., Klein, H. J., & Wright, P. M. (1989). Investigation of the
construct validity of a self-report measure of goal commitment. Journal of Applied
Psychology, 74, 951-956.
Hollenbeck, J. R., Williams, C. R., & Klein, H. J. (1989). An empirical examination of the
antecedents of commitment to difficult goals. Journal of Applied Psychology, 74, 18-23.
Hong, J. W., & Sternthal, B. (2010). The Effects of Consumer Prior Knowledge and Processing
Strategies on Judgments. Journal of Marketing Research, 47, 301-311.
Hong, R. Y., Tan, M. S., & Chang, W. N. C. (2004). Locomotion and assessment: self-regulation
and subjective well-being. Personality and Individual Differences, 37, 325-332.
Hu, L., & Bentler, P. (1995). Evaluating model fit. In R. H. Hoyle (Ed.), Structural equation
modeling. Concepts, issues, and applications (pp. 76−99). London: Sage.
Hu, L., & Bentler, P. M. (1998). Fit indices in covariance structure analysis: Sensitivity to
underparameterized model misspecification. Psychological Methods, 3, 424−453.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1−55.
Idson, L. C., Liberman, N., & Higgins, E. T. (2000). Distinguishing gains from nonlosses and
losses from nongains: A regulatory focus perspective on hedonic intensity. Journal of
Experimental Social Psychology, 36, 252-274.
Idson, L. C., Liberman, N., & Higgins, E. T. (2004). Imagining how you'd feel: The role of
motivational experiences from regulatory fit. Personality and Social Psychology Bulletin,
30, 926-937.
Johnson, R. E., & Chang, C. H. (2008). Relationships between organizational commitment and
its antecedents: Employee self-concept matters. Journal of Applied Social Psychology,
38, 513-541.
Johnson, R. E., Chang, C. H., & Rosen, C. C. (2010). "Who I Am Depends on How Fairly I'm
Treated" Effects of Justice on Self-Identity and Regulatory Focus. Journal of Applied
Social Psychology, 40, 3020-3058.
Johnson, R. E., & Saboe, K. (in press). Measuring implicit traits in organizational research:
Development of an indirect measure of employee implicit self-concept. Organizational
Research Methods.

196

Johnson, R. E., Selenta, C., & Lord, R. G. (2006). When organizational justice and the selfconcept meet: Consequences for the organization and its members. Organizational
Behavior and Human Decision Processes, 99, 175-201.
Jöreskog, K. G., & Sörbom, D. (1996). Lisrel 8: User’s Reference Guide. Chicago: Scientific
Software.
Kanfer, R. (1996). Self-regulatory and other non-ability determinants of skill acquisition. In P.
M. Gollwitzer & J. A. Bargh (Eds.), The psychology of action: Linking cognition and
motivation to behavior (pp. 404-423). New York: Guilford Press.
Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative aptitude
treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74,
657-690.
Kanfer, R., & Heggestad, E. D. (1997). Motivational traits and skills: A person-centered
approach to work motivation. Research in Organizational Behavior, 19, 1-56.
Kanfer, R., & Kanfer, F. H. (1991). Goals and self-regulation: Applications of theory to work
settings. Advances in Motivation and Achievement, 7, 287-326.
Kanungo, R. N. (1982). Measurement of job and work involvement. Journal of Applied
Psychology, 67, 341-349.
Kark, R., & Van Dijk, D. (2007). Motivation to lead, motivation to follow: The role of the selfregulatory focus in leadership processes. Academy of Management Review, 32, 500-528.
Karoly, P. (1993). Mechanisms of self-regulation: A systems view. Annual Review of
Psychology, 44, 23-52.
Keller, R. T. (1994). Technology information-processing fit and the performance of R&D project
groups: A test of contingency theory. Academy of Management Journal, 37, 167-179.
Kenny, D. A., Mannetti, L., Pierro, A., Livi, S., & Kashy, D. A. (2002). The statistical analysis
of data from small groups. Journal of Personality and Social Psychology, 83, 126–137.
Kernan, M. C., & Lord, R. G. (1988). Effects of participative vs. assigned goals and feedback in
a multitrial task. Motivation and Emotion, 12, 75-86.
Kish, L. (1965). Survey sampling. John Wiley & Sons, New York.
Klein, H. J. (1989). An integrated control theory model of work motivation. Academy of
Management Review, 14, 150-172.
Klein, H. J. (1991). Further evidence on the relationship between goal setting and expectancy
theories. Organizational Behavior and Human Decision Processes, 49, 230-257.
Klein, H. J., Austin, J. T., & Cooper, J. T. (2008). Goal choice and decision processes. In R.

197

Kanfer, G. Chen & R. D. Pritchard (Eds.), Work motivation: Past, present, and future
(pp. 101–150). New York: Routledge.
Klein, H. J., Wesson, M. J., Hollenbeck, J. R., & Alge, B. J. (1999). Goal commitment and the
goal-setting process: Conceptual clarification and empirical synthesis. Journal of Applied
Psychology, 84, 885-896.
Klein, H. J., & Wright, P. M. (1994). Antecedents of goal commitment: An empirical
examination of personal and situational factors. Journal of Applied Social Psychology,
24, 95-114.
Klein, K. J., & Kozlowski, S. W. J. (2000). Multilevel theory, research, and methods in
organizations: Foundations, extensions, and new directions. San Francisco, CA: JosseyBass.
Koenig, A. M., Cesario, J., Molden, D. C., Kosloff, S., & Higgins, E. T. (2009). Incidental
experiences of regulatory fit and the processing of persuasive appeals. Personality and
Social Psychology Bulletin, 35, 1342-1355.
Kozlowski, S. W. J., & Bell, B. S. (2006). Disentangling achievement orientation and goal
setting: Effects on self-regulatory processes. Journal of Applied Psychology, 91, 900-916.
Kreiner, G. E., Hollensbe, E. C., & Sheep, M. L. (2006). Where is the “me” among the “we”?
Identity work and the search for optimal balance. Academy of Management Journal, 49,
1031–1057.
Kruglanski, A. W. (1996). Goals as knowledge structures. In P. M. Gollwitzer & J. A. Bargh
(Eds.), The psychology of action (pp. 599-618). New York: Guilford Press.
Kruglanski, A. W., Pierro, A., & Higgins, E. T. (2007). Regulatory mode and preferred
leadership styles: How fit increases job satisfaction. Basic and Applied Social
Psychology, 29, 137-149.
Kruglanski, A. W., Pierro, A., Higgins, E. T., & Capozza, D. (2007). "On the move" or "staying
put": Locomotion, need for closure, and reactions to organizational change. Journal of
Applied Social Psychology, 37, 1305-1340.
Kruglanski, A. W., Shah, J. Y., Fishbach, A., Friedman, R., Chun, W. Y., & Sleeth-Keppler, D.
(2002). A theory of goal systems. In M. P. Zanna (Ed.), Advances in experimental social
psychology (Vol. 34, pp. 331-378). San Diego, CA: Academic Press.
Kruglanski, A. W., Thompson, E. P., Higgins, E. T., Atash, M. N., Pierro, A., Shah, J. Y., et al.
(2000). To "do the right thing" or to "just do it": Locomotion and assessment as distinct
self-regulatory imperatives. Journal of Personality and Social Psychology, 79, 793-815.
Kuhl, J. (1984). Volitional aspects of achievement motivation and learned helplessness: Toward
a comprehensive theory of action control. In B. A. Maher & W. A. Maher (Eds.),
Progress in experimental personality research (Vol. 13, pp. 99-171). New York:
198

Academic Press.
Kuhl, J. (1986). Motivation and information processing:A new look at decision making, dynamic
change, and action control. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook of
motivation and cognition: Foundations of social behavior (pp. 404-434). New York:
Guilford Press.
Kuhl, J., & Kazen, M. (1999). Volitional facilitation of difficult intentions: Joint activation of
intention memory and positive affect removes Stroop interference. Journal of
Experimental Psychology: General, 128, 382-399.
Lawler III, E. E. (1971). Pay and organizational ef]ectiveness: A psychological view. New York:
McGraw-Hill.
Lawrence, P. R., & Lorsch, J. W. (1967). Differentiation and integration in complex
organizations. Administrative Science Quarterly, 12, 1-47.
LeBreton, J. M., & Senter, J. L. (2008). Answers to 20 questions about interrater reliability and
interrater agreement. Organizational Research Methods, 11, 815–852.
Lee, A. Y., & Aaker, J. L. (2004). Bringing the frame into focus: The influence of regulatory fit
on processing fluency and persuasion. Journal of Personality and Social Psychology, 86,
205-218.
Lewin, K. (1935). A dynamic theory of personality: Selected papers (D. E. Adams & K. E.
Zener, Trans.). New York: McGraw Hill.
Lewin, K. (1951). Field theory in social science. New York: Harper.
Lewis, K. (2003). Measuring transactive memory systems in the field: Scale development and
validation. Journal of Applied Psychology, 88, 587-604.
Locke, E. A. (1968). Toward a theory of task motivation and incentives. Organizational
Behavior and Human Performance, 3, 157-189.
Locke, E. A., & Latham, G. P. (1990a). Work motivation and satisfaction: Light at the end of the
tunnel. Psychological Science, 1, 240-246.
Locke, E. A., & Latham, G. P. (1990b). A theory of goal setting and task performance.
Englewood Cliffs, NJ: Prentice Hall.
Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and
task motivation: A 35-year odyssey. American Psychologist, 57, 705-717.
Locke, E. A., Latham, G. P., & Erez, M. (1988). The determinants of goal commitment.
Academy of Management Review, 13, 23-39.
Locke, E. A., Saari, L. M., Shaw, K. N., & Latham, G. P. (1981). Goal setting and task

199

performance: 1969-1980. Psychological Bulletin, 90, 125-152.
Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization:
Effects of prior theories on subsequently considered evidence. Journal of Personality and
Social Psychology, 37, 2098-2109.
Lord, R. G., Diefendorff, J. M., Schmidt, A. M., & Hall, R. J. (2010). Self-regulation at work.
Annual Review of Psychology, 61, 543-568.
Lord, R. G., & Hanges, P. J. (1987). A control system model of organizational motivation:
Theoretical development and applied implications. Behavioral Science, 32, 161-178.
Lord, R. G., & Levy, P. E. (1994). Moving from cognition to action: A control theory
perspective. Applied Psychology: An International Review, 43, 335-398.
Mabe, P. A., & West, S. G. (1982). Validity of self-evaluation of ability: A review and metaanalysis. Journal of Applied Psychology, 67, 280-296.
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and
determination of sample size for covariance structure modeling. Psychological Methods,
1, 130-149.
Mannetti, L., Giacomantonio, M., Higgins, E. T., Pierro, A., & Kruglanski, A. W. (2010).
Tailoring visual images to fit: Value creation in persuasive messages. European Journal
of Social Psychology, 40, 206-215.
Mannetti, L., Leder, S., Insalata, L., Pierro, A., Higgins, T., & Kruglanski, A. (2009). Priming
the ant or the grasshopper in people's mind: How regulatory mode affects inter-temporal
choices. European Journal of Social Psychology, 39, 1120-1125.
Markman, A. B., & Brendl, C. M. (2000). The influence of goals on value and choice. In
Psychology of Learning and Motivation: Advances in Research and Theory (Vol. 39, pp.
97-128). San Diego: Academic Press Inc.
Marks, M. A., Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and
taxonomy of team processes. Academy of Management Review, 26, 356-376.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion,
and motivation. Psychological Review, 98, 224-253.
Mathieu, J. E., Marks, M. A., & Zaccaro, S. J. (2001). Multiteam systems. In N. Anderson, D.
Ones, H. K. Sinangil & C. Viswesvaran (Eds.), Handbook of industrial, work and
organizational psychology. Volume 2: Organizational Psychology (Vol. 2, pp. 289-313).
Thousand Oaks, CA: Sage Publications.
Mauro, R., Pierro, A., Mannetti, L., Higgins, E. T., & Kruglanski, A. W. (2009). The Perfect
Mix: Regulatory Complementarity and the Speed-Accuracy Balance in Group
Performance. Psychological Science, 20, 681-685.

200

McCall, G. J., & Simmons, J. L. (1966). Identities and Interactions. New York: Free Press.
Meyer, J. P., & Allen, N. J. (1997). Commitment in the workplace: Theory, research and
application. London: Sage.
Mitchell, T. R., Harman, W. S., Lee, T. W., & Lee, D.-Y. (2008). Self-regulation and multiple
deadlines. In R. Kanfer, G. Chen & R. D. Pritchard (Eds.), Work motivation: Past,
present, and future (pp. 197-231). New York: Routledge.
Mogilner, C., Aaker, J. L., & Pennington, G. L. (2008). Time will tell: The distant appeal of
promotion and imminent appeal of prevention. Journal of Consumer Research, 34, 670681.
Mowday, R. T. (1998). Reflections on the study and relevance of organizational commitment.
Human Resource Management Review, 8, 387-401.
Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). Measurement of organizational
commitment. Journal of Vocational Behavior, 14, 224-247.
Naylor, J. C., & Ilgen, D. R. (1984). Goal Setting: A theoretical analysis of a motivational
technology. Research in Organizational Behavior, 6, 95-140.
Neubert, M. J., Kacmar, K. M., Carlson, D. S., Chonko, L. B., & Roberts, J. A. (2008).
Regulatory focus as a mediator of the influence of initiating structure and servant
leadership on employee behavior. Journal of Applied Psychology, 93, 1220-1233.
Owens, T. J., Robinson, D. T., & Smith-Lovin, L. (2010). Three faces of identity. Annual Review
of Sociology, 36, 477-499.
Pelham, B. W. (1995). Self-investment and self-esteem: Evidence for a Jamesian model of selfworth. Journal of Personality and Social Psychology, 69, 1141-1150.
Pelham, B. W., & Swann, W. B. (1989). From self-conceptions to self-worth: On the sources and
structure of global self-esteem. Journal of Personality and Social Psychology, 57, 672680.
Pervin, L. A. (1989). Goal concepts in personality and social psychology. Hillsdale, NJ:
Erlbaum.
Pierro, A., Kruglanski, A. W., & Higgins, E. T. (2006a). Regulatory mode and the joys of doing:
Effects of 'locomotion' and 'assessment' on intrinsic and extrinsic task-motivation.
European Journal of Personality, 20, 355-375.
Pierro, A., Kruglanski, A. W., & Higgins, E. T. (2006b). Progress takes work: Effects of the
locomotion dimension on job involvement, effort investment, and task performance in
organizations. Journal of Applied Social Psychology, 36, 1723-1743.
Pierro, A., Leder, S., Mannetti, L., Higgins, E. T., Kruglanski, A. W., & Aiello, A. (2008).

201

Regulatory mode effects on counterfactual thinking and regret. Journal of Experimental
Social Psychology, 44, 321-329.
Powers, W. T. (1973). Behavior: The control of perception. Chicago: Aldine.
Prentice, D., Miller, D., & Lightdale, J. (1994). Asymmetries in attachments to groups and to
their members: Distinguishing between common-identity and common-bond groups.
Personality and Social Psychology Bulletin, 20, 484-493.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models in social and behavioral
research: Applications and data-analysis methods (2nd ed.). Newbury Park, CA: Sage.
Raven, B. H., Schwarzwald, J., & Koslowski, M. (1998). Conceptualizing and measuring a
power-interaction model of interpersonal influence. Journal of Applied Social
Psychology, 28, 307–332.
Roccas, S., Sagiv, L., Schwartz, S., Halevy, N., & Eidelson, R. (2008). Toward a unifying model
of identification with groups: Integrating theoretical perspectives. Personality and Social
Psychology Review, 12, 280-306.
Roney, C. J. R., Higgins, E. T., & Shah, J. (1995). Goals and framing: How outcome focus
influences motivation and emotion. Personality and Social Psychology Bulletin, 21,
1151-1160.
Salancik, G. R. (1977). Commitment is too easy. Organizational Dynamics, 6, 62-80.
Schmidt, A. M., & DeShon, R. P. (2007). What to do? The effects of discrepancies, incentives,
and time on dynamic goal prioritization. Journal of Applied Psychology, 92, 928-941.
Schmidt, A. M., & Dolis, C. M. (2009). Something's got to give: The effects of dual-goal
difficulty, goal progress, and expectancies on resource Allocation. Journal of Applied
Psychology, 94, 678-691.
Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health: Assessment and
implications of generalized outcome expectancies. Health Psychology, 4, 219–247.
Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances
and empirical tests in 20 countries. In M. P. Zanna (Ed.). Advances in experimental social
psychology (Vol. 25, pp. 1–65). San Diego, CA: Academic Press.
Seijts, G. H., & Latham, G. P. (2000a). The construct of goal commitment: Measurement and
relationships with task performance. In R. Goffin & E. Helmes (Eds.), Problems and
solutions in human assessment (pp. 315-332). Dordrecht, the Netherlands: Kluwer
Academic.
Seijts, G. H., & Latham, G. P. (2000b). The effects of goal setting and group size on
performance in a social dilemma. Canadian Journal of Behavioural Science-Revue
Canadienne Des Sciences Du Comportement, 32, 104-116.

202

Shah, J., Higgins, E. T., & Friedman, R. S. (1998). Performance incentives and means: How
regulatory focus influences goal attainment. Journal of Personality and Social
Psychology, 74, 285-293.
Shah, J. Y., Friedman, R., & Kruglanski, A. W. (2002). Forgetting all else: On the antecedents
and consequences of goal shielding. Journal of Personality and Social Psychology, 83,
1261-1280.
Shah, J. Y., & Kruglanski, A. W. (2008). Structural dynamics: The challenge of change in goal
systems. In J. Y. Shah & W. L. Gardner (Eds.), Handbook of motivation science (pp. 217229). New York: The Guilford Press.
Snyder, M. (1974). Self-monitoring of expressive behavior. Journal of Personality and Social
Psychology, 30, 526-537.
Spence, J. T., & Helmreich, R. L. (1978). Masculinity and femininity: Their psychological
dimensions, correlates and antecedents. Austin, TX: University of Texas Press.
Spiegel, S., Grant-Pillow, H., & Higgins, E. T. (2004). How regulatory fit enhances motivational
strength during goal pursuit. European Journal of Social Psychology, 34, 39-54.
Staw, B. M. (1981). The escalation of commitment to a course of action. Academy of
Management Review, 6, 577-587.
Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation
approach. Multivariate Behavioral Research, 25, 173−180.
Stryker, S. (1980). Symbolic interactionism: A social structural version. Menlo Park, CA:
Benjamin/Cummings.
Stryker, S. (1991). Exploring the relevance of social cognition for the relationship of self and
society: Linking the cognitive perspective and identity theory. In J. Howard & P. Callero
(Eds.), The self-society dynamic: Cognition, emotion, and action (pp. 19-41). Cambridge,
England: Cambridge University Press.
Stryker, S. (2008). From Mead to a structural symbolic interactionism and beyond. Annual
Review of Sociology, 34, 15-31.
Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: Equivalent,
overlapping, or complementary constructs? Social Psychology Quarterly, 57, 16–35.
Stryker S, Serpe, R.T., & Hunt, M. O. (2005). Making good on a promise: The impact of larger
social structures on commitments. Advances in Group Process, 22, 93–124
Suls, J., & Wills, T. (1991). Social comparison: Contemporary theory and research. Hillsdale,
NJ: Erlbaum.
Tajfel, H. (1972). Social categorization. English manuscript of 'La categorisation sociale.' In S.

203

Moscovici (Ed.), Introduction a la Psychologie Sociale (Vol. 1, pp. 272-302). Paris:
Larousse.
Tajfel, H. (1982). Social psychology of inter-group relations. Annual Review of Psychology, 33,
1-39.
Tajfel, H., & Turner, J. C. (1979). The social psychology of intergroup conflict. In W. G. Austin
& S. Worchel (Eds.), The Social Psychology of Intergroup Relations (pp. 33-47).
Monterey, CA: Brooks/Cole.
Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill.
Tropp, L. R., & Wright, S. C. (2001). Ingroup identification as the inclusion of ingroup in the
self. Personality and Social Psychology Bulletin, 27, 585–600.
Tubbs, M. E., & Dahl, J. G. (1991). An empirical comparison of self-report and discrepancy
measures of goal commitment. Journal of Applied Psychology, 76, 708-716.
Tucker, L. R., & Lewis, C. (1993). The reliability coefficient for maximum likelihood factor
analysis. Psychometrika, 38, 1-10.
Turner, J. C. (1975). Social comparison and social identity: Some prospects for intergroup
behavior. European Journal of Social Psychology, 5, 5-34.
Turner, J. C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.),
Social identity and intergroup relations (pp. 15-40). Cambridge, England: Cambridge
University Press.
Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of
group behavior. In E. J. Lawler (Ed.), Advances in group processes: Theory and research
(Vol. 2, pp. 77-122). Greenwich, CT: JAI Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. C. (1987).
Rediscovering the social group: A self-categorization theory. New York: Blackwell
Publishers.
Vancouver, J. B., Weinhardt, J. M., & Schmidt, A. M. (2010). A formal, computational theory of
multiple-goal pursuit: Integrating goal-choice and goal-striving Processes. Journal of
Applied Psychology, 95, 985-1008.
Van Eerde, W., & Thierry, H. (1996). Vroom's expectancy models and work-related criteria: A
meta-analysis. Journal of Applied Psychology, 81, 575-586.
Vroom, V. H. (1964). Work and motivation. New York: Wiley.
Wegner, D. M., & Wenzlaff, R. M. (1996). Mental control. In E. T. Higgins & A. W. Kruglanski
(Eds.), Social psychology: Handbook of basic principles (pp. 466–492). New York:
Guilford Press.

204

Wright, P. M., George, J. M., Farnsworth, S. R., & McMahan, G. C. (1993). Productivity and
extra-role behavior: The effects of goals and incentives on spontaneous helping. Journal
of Applied Psychology, 78, 374–381.
Wright, P. M., O’Leary-Kelly, A. M., Cortina, J. M., Klein, H. J., & Hollenbeck, J. R. (1994).
On the meaning and measurement of goal commitment. Journal of Applied Psychology,
79, 795-803.
Zeigarnik, B. (1938). On finished and unfinished tasks In W. D. Ellis (Ed.), A source book of
Gestalt psychology (pp. 300–314). New York: Harcourt, Brace & World. (Reprinted and
condensed from Psychologische Forschung, 9, 1–85, by B. Zeigarnik, 1927).

205