”its: .975:
9

‘3. r. 47»
3:. .95....
:x...
3.3:,

F..x.=..i.€x.. E;
{12.45. I r
i! {new
.0 .6:

t 0’. z
5...}. 1’.
.3 71
z.

I.)
liasl . :1.

l.
39 .v:
:5:

35%.},
13'.\ l
V “3.5.5.,
.. t a .
3mm. 53...:

. r . . any:
«as. .31.... % Vii; .2. {a7}...
3: :23. .. .5... . _ ... L
a)! an i umiJaivth .u
I his: 3 .ylezfi.o
3.. .3 i511: 4
I1 2:5!=x§~

M11!
: :ll.
.S.t.\!v3.:
ll...-

\t..¢) n
I. .2

1A“. 3.
e h.

3 .
:Is

:2). :0: cu
s-
:6

 

THESmx

Q Dbl

 

LIBRARY
Mlchtgan State
University

 

‘m.

 

 

This is to certify that the
dissertation entitled
The Role of Individual Differences, Structural Factors
and Goal Importance in Allocating Resources to Group

Tasks: The Case of Parallel Teams

presented by

Kok Yee Ng

has been accepted towards fulfillment
of the requirements for

PhD degree in Management

 

 

i .
Nuns VA/Vx
Major professo

Date August 81 2001

 

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

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8/01 cJCIRCDateDuapes-ots

 

FOR MYSELF OR FOR THE GROUP:
THE ROLE OF INDIVIDUAL DIFFERENCES, STRUCTURAL FACTORS AND
GOAL COMMITMENT IN INDIVIDUALS’ GROUP TASK PERFORMANCE IN
PARALLEL TEAMS ‘
By

Kok Yee Ng

AN ABSTRACT OF A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Management

2001

Linnea Van Dyne

ABSTRACT
FOR MYSELF OR FOR THE GROUP:

THE ROLE OF INDIVIDUAL DIFFERENCES, STRUCTURAL FACTORS AND
GOAL COMMITMENT IN INDIVIDUALS’ GROUP TASK PERFORMANCE IN
PARALLEL TEAMS
By

Kok Yee Ng

This dissertation examined the reactions of parallel team members (parallel teams
are teams that run alongside the formal structure of the organization) to the dilemma of
Simultaneously pursuing an individual goal and a group goal. Adopting Kanfer’s (1990)
motivational framework, I examined how distal constructs (i.e., individual differences
and team structural factors) and proximal constructs (individual and group goal
commitment) influence parallel team members’ performance in a group task (in the
presence of an individual task).

Overall, results of this experimental study (n= 318) conducted in the United States
and Singapore demonstrated that individual differences (horizontal collectivism and
agreeableness) indirectly influenced members’ performance in the group task via group
goal commitment, while conscientiousness had a direct positive influence. Results,
however, demonstrated no direct effects of structural factors (task interdependence and
leader status) on goal commitment or performance. Rather, structural factors were more
important when considered jointly with individual differences. Specifically, task
interdependence interacted with horizontal collectivism and conscientiousness to
influence group goal commitment.

Results also indicated that while Singapore subjects discriminated between

individual and group goal commitment, U.S. subjects did not. This raises cultural
differences as an important question for future research on goal commitment in the
context of parallel teams. Post-hoc results also suggest that using a group goal priority
measure (which requires individuals to prioritize individual and group goal) is a
potentially promising approach.

In the discussion, I highlight implications of current findings and raise future
research questions that can help advance the field’s understanding of parallel teams.
Given the popularity of parallel teams in organizations, this research can confer

considerable practical significance to the field.

To my Good Shepherd, who led me here and gave me Life

iv

ACKNOWLEDGEMENTS

I am grateful to many individuals for their intellectual stimulation during the
course of attaining this degree. Two individuals in particular played a key role in my
training. I am especially indebted to Linn Van Dyne, my mentor and chair of this
dissertation, for her dedication in nurturing me for the past four years. She has taught me,
amongst many things, the importance of perserverance and rigor in conducting research. I
am also greatly indebted to Soon Ang, my mentor in Singapore, who is the first person to
stimulate my interest in research, and to inspire me with her own enthusiasm toward
research. Moreover, I am grateful to her for making my data collection in Singapore
possible. I also wish to express my utmost gratitude to my dissertation committee
members — Murray Barrick, John Hollenbeck, and Dan 1] gen, for their time spent in
ploughing through my dissertation, and for their critical insights in refining my thesis.
Finally, I am deeply appreciative of Chay Hoon Lee for her help in the logistics of data
collection in NTU (Singapore), without which this dissertation could not have been
completed smoothly.

Many others have also contributed indirectly to this dissertation. I want to
especially thank Barbara Strong for her prayers, and for showering me with the love and
warmth of a family here in East Lansing. Likewise, I thank my wonderful friends from
the University Reformed Church for their prayers and spiritual encouragement.

And perhaps the one person who has seen the worst of me during the stressful
periods in this process, and yet, is always there for me, is Edwin -— my husband. Thank

you, my dear, for your loving patience and support in these four years that we were apart.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. ix
LIST OF FIGURES ................................................................................. x
INTRODUCTION .................................................................................. 1
What are Parallel Teams? ..................................................................... 2
Gaps in Parallel Team Research .............................................................. 4
Research Framework and Conceptual Model ................................................ 5
Constructs ........................................................................................ 7
Individual Differences ..................................................................... 7
Team Structural Factors .................................................................. 10
Goal Commitment ........................................................................ 12
Performance Outcome .................................................................... 12
Potential Contributions of this Dissertation ................................................ 12
Boundary Conditions .......................................................................... 14
Outline of Dissertation ........................................................................ 15
CHAPTER 1
THE ROLE OF GOAL COMMITMENT IN PARALLEL TEAMS ....................... 18
Goal Commitment ............................................................................. 18
Multiple Goals and Goal Commitment ..................................................... 23
Goal Commitment in Parallel Teams ...................................................... 26
CHAPTER 2
ANTECEDENTS AND CONSEQUENCES OF GOAL COMMITMENT .............. 29
Individual Differences as Antecedents ..................................................... 29
Cultural Values ................................................................................ 31
Horizontal and Vertical Collectivism ................................................. 34
Personality ..................................................................................... 37
Conscientiousness ........................................................................ 38
Agreeableness ............................................................................. 40
Structural Factors .............................................................................. 42
Task Interdependence .................................................................... 43
Team Leader Status ....................................................................... 44
Person by Situation Interactions ............................................................ 45
Consequences of Goal Commitment ....................................................... 51
CHAPTER 3
METHOD .......................................................................................... 54
Setting .......................................................................................... 54
Power Analysis ................................................................................ 54
Participants .................................................................................... 55

vi

Experiment ................................................................................... 56

Tasks ..................................................................................... 56
Manipulations ........................................................................... 57
Procedure ................................................................................ 58
Goals ..................................................................................... 59
Measures ..................................................................................... 60
Individualism-Collectivism ........................................................... 61
Conscientiousness and Agreeableness ............................................... 63
Goal Commitment ...................................................................... 63
Group Goal Priority ..................................................................... 64
Performance Outcomes ................................................................ 65
Control Variable ........................................................................ 66
Manipulation Checks ................................................................... 66
Data Analyses ................................................................................ 67
CHAPTER 4
RESULTS ........................................................................................ 70
Measures ..................................................................................... 7O
H-C and V-C ............................................................................ 7O
Conscientiousness and Agreeableness ............................................... 71
Goal Commitment ...................................................................... 73
Group Goal Priority ..................................................................... 77
Descriptive Statistics ........................................................................ 78
Manipulation Checks ........................................................................ 82
Test of Hypotheses .......................................................................... 83
Main Effects of Individual Differences (H1 — H3) ................................. 83
Main Effects of Structural Factors (H4 — H5) ....................................... 83
Interactions between Individual and Structural Factors (H6 — H7) ................ 83
Effects of Goal Commitment .......................................................... 90
Post-Hoe Analyses ........................................................................... 91
Post-Hoe Results ............................................................................. 94
H1-H8 using Original Scales .......................................................... 94
Group Goal Priority ..................................................................... 94
Perceptual Measures of Structural Factors .......................................... 98
Tests of Mediation ..................................................................... 103
Country Differences ................................................................... 106
CHAPTER 5
DISCUSSION .................................................................................... 108
Discussion of Empirical Results .......................................................... 110
Goal Commitment ...................................................................... 111
Group Goal Priority .................................................................... 113
Individual Differences (H1 - H3) .................................................... 114
Task Interdependence and Team Leader Status (H4 — H5) ....................... 117
Interactions between Structural and Individual Factors (H6-H7) ................ 120
Implications .................................................................................. 123

vii

On Theory ............................................................................... 123

On Practice ............................................................................. 128
Limitations ................................................................................... 130
Future Directions ........................................................................... 131
Conclusion .................................................................................... 136

APPENDICES ................................................................................... 137
Appendix A: Experimental Materials ..................................................... 138
Appendix B: Measures ...................................................................... 144

REFERENCES .................................................................................... 147

viii

LIST OF TABLES

Table l. Confirmatory Factor Analysis for H-C and V-C Measures ........................ 72

Table 2. Confirmatory Factor Analyses for Individual and Group Goal Commitment
Measures ................................................................................. 75

Table 3. Confirmatory Factor Analyses for Individual Goal Commitment, Group
Goal Commitment and Group Goal Priority ....................................... 76

Table 4. Descriptive Statistics, Cronbach’s Alpha, and Correlations of Variables in
Study ................................................................................. 79-80

Table 5. Hierarchical Regression Results for Individual Differences and Goal
Commitment (H1 — H3) ............................................................... 85

Table 6. Hierarchical Moderated Regression Results for Individual Differences and
Task Interdependence on Group Goal Commitment (H4, H6a—H6c) ............ 86

Table 6a. Regression Results for H-C X Task Interdependence and Conscientiousness
X Task Interdependence .............................................................. 87

Table 7. Hierarchical Moderated Regression Results for Individual Differences,
Leader Status and Group Goal Commitment (H5, H7a—H7d) ................... 88

Table 8. Hierarchical Moderated Regression Results for Individual and Group Goal
Commitment on Performance in Group Task (H8) ................................ 90

Table 9. Post Hoe: Regression Results for Effects of Individual Differences and
Task Interdependence on Group Goal Priority ..................................... 96

Table 10. Post-Hoc: Regression Results for Effects of Individual Differences and
Leader Status on Group Goal Priority .............................................. 97

Table 11. Post-Hoe: Regression Results for Individual Differences on Perceived Task
Interdependence and Perceived Leader Status .................................... 100

Table 12. Post-Hoe: Regression Results for Effects of Individual Differences and
Perceived Task Interdependence on Group Goal Priority ...................... 101

Table 13. Post-Hoc: Regression Results for Effects of Individual Differences and
Perceived Leader Status on Group Goal Priority ................................. 102

ix

LIST OF FIGURES

Figure 1a. A Conceptual Model ................................................................... 8
Figure lb. An Empirical Model .................................................................. 16
Figure 2. A general fiamework depicting Mischel’s situational strength argument ......46
Figure 3. Proposed interaction between V-C and leader status on group goal
commitment ........................................................................... 50
Figure 4. Proposed interaction between individual and group goal commitment on
individual performance in the group task .......................................... 52
Figure 5. Interaction between H-C and Task Interdependence on Group Goal
Commitment ........................................................................... 89
Figure 6. Interaction between Conscientiousness and Task Interdependence on Group
Goal Commitment ..................................................................... 89
Figure 7. Post-Hoc: Interaction between Conscientiousness and Task Interdependence
On Group Goal Priority ............................................................... 98
Figure 8. Post-Hoe: Interaction between H-C and Perceived Leader Status on Group
Goal Priority ........................................................................... 104
Figure 9. Post-Hoe: Interaction between Agreeableness and Perceived Leader Status

On Group Goal Priority .............................................................. 104

Figure 10. Post-Hoe: Interaction between Conscientiousness and Perceived Leader

Status on Group Goal Priority ..................................................... 105

INTRODUCTION

Consider the following scenario. X is working under tight deadlines on two tasks:
the first is an individual task for which he is accountable alone for his performance, while
the second is a group task for which he and his group members (belonging to a task
force) are jointly accountable for their performance. Given the limited effort and time,
how will X perform in the two tasks?

Although having multiple goals that compete for common resources is a dilemma
faced by employees frequently at the workplace (Sniezek, May & Sawyer, 1990),
relatively little research, whether theoretical or empirical, has addressed this topic (Austin
& Vancouver, 1996; Locke & Latharn, 1990). This represents an important gap in the
literature because the increasingly team-based nature of organizations (Ilgen & Pulakos,
1999) creates an environment in which a specific form of dilemma - a tension between
individual and collective interests, is likely to occur (Sheppard, 1993; Sniezek et al.,
1990)

Parallel teams (e.g., employee involvement programs, problem-solving groups,
task forces, quality circles) for instance, provide a classic example where group members
are responsible for both individual and group tasks, and hence, face the dilemma between
individual and collective interests (Colquitt, 1999; Cotton, 1993). This dissertation uses
parallel teams as a context to examine one important research question: what factors
affect individuals’ performance in their group task when they are simultaneously faced
with demands from their personal tasks? To address this question, I adOpt Kanfer’s
(1990) distal-proximal framework of motivation to explain how distal factors (e.g.,

individual differences and Situational factors) influence parallel team members’

performance outcomes via the proximal mechanism of goal commitment (Hollenbeck &
Klein, 1987; Locke, Shaw, Saari, & Latham, 1981). Before elaborating on this
framework, however, I describe some distinctive features of parallel teams in the ensuing
section in order to provide a clearer understanding of my research context and the
motivation underlying my research questions.

Mt Are Pagllel Tems?

Parallel teams are teams that exist alongside the formal structure of the
organization (Cohen & Bailey, 1997). Unlike work teams which are permanent and
stable, parallel teams pull people from different work units or jobs to perform functions
that are not under the jurisdiction of specific departments or teams (such as special task
forces). Recent results of a long term study conducted by the Center for Effective
Organizations indicated that ahnost every organization uses some form of parallel teams,
and many use more than one (Lawler, 1999). The popularity of parallel teams may be
largely attributed to the potential benefits of employee participation (Cotton, 1993). In
addition, the immense success of quality circles in Japan also fueled the adoption of
parallel teams by US. firms in various industries, ranging from manufacturing firms to
service organizations.

Typical purposes of parallel teams include enhancing employee participation
without disrupting existing work practices and organizational structure, and resolving a
single-time, non-recurring issue with maximum input from a wide range of people
(Cotton, 1993). Parallel teams are also similar to the advice/involvement teams
characterized by Sundstrom, DeMeuse, & Futrell (1990) in that the output of the team
typically includes selection, suggestions, proposals and recommendations. Two features

are common to parallel teams (or advice/involvement teams): low differentiation and low

external integration. Differentiation refers to the degree of specialization, independence
and autonomy of a work team in relation to other work units (Lawrence & Lorsch, 1969);
external integration refers to the degree of coordination and synchronization with
suppliers, managers, peers and customers (Sundstrom et al., 1990). Parallel teams
typically have low differentiation in that membership is often broadly representative,
working time is limited, and the group has a limited life span. External integration is also
minimal because parallel teams require little synchronization with other work units
(Sundstrom et al., 1990).

One distinctive feature of parallel teams that is particularly germane to this
dissertation is that assignment to the team is not full-time, and members have their own
daily tasks to manage in addition to their team responsibilities (Cotton, 1993). Hence,
parallel team members face a resource allocation dilemma because they need to decide
how to devote their effort and time between their own tasks and their group tasks
(Colquitt, 1999; Cotton, 1993; Lawler & Cohen, 1992). The dilemma arises because time
and effort are finite resources, and commitment of resources to one task usually leads to
reduced commitment to the other tasks (N aylor, Pritchard & 1] gen, 1980). This in turn has
important ramifications for performance outcomes. Indeed, Lawler and Cohen (1992)
observed that the effectiveness of parallel teams is often hindered by competition for the
time and resources of their members. Moreover, since parallel team members have
limited interaction with their group members, they are often less compelled to contribute
to the team compared to members of full-time work teams (Cotton, 1993). This problem
is exacerbated by the fact that parallel teams usually possess little organizational
legitimacy (Cohen & Lawler, 1992). For instance, the team leader usually has. limited

mandate over team members, either in terms of setting performance expectations, or

rewarding performance outcomes.

Several field studies have alluded to the problems associated with parallel teams.
In a field study of a tearn-based employee involvement program, May and Schwoerer
(1994) found that due to production pressures from their regular jobs, members in the
employee involvement team could not find enough time to work on the team’s project.
Adam (1991) found that one reason for the decline in the size of quality circles in two
industrial firms was that employees were “too busy.” Adam’s results also showed that
members of quality circles did not possess better job attitudes (e.g., general satisfaction,
job design) compared to employees who were in the control groups, suggesting that the
additional group commitments may have more adversarial effects than benefits on
employees.
Gag in Pa_r§llel Team Resear_c_l_i

Despite the widespread use of parallel teams in organizations, research has paid
considerably less attention to parallel teams than to firll-time work teams (Cohen &
Bailey, 1997). The few studies conducted with quality circles and employee involvement
programs were primarily interested in the success of these initiatives, and provided some
useful insight for the design and implementation of parallel teams. For instance, Magjuka
and Baldwin (1991) found that structural factors such as accessibility to information,
heterogeneity of job functions and a large team size facilitated the performance of team-
based employee involvement programs. In addition to these factors, May and Schwoerer
(1994) suggested team autonomy, leadership and compensation as important variables in
implementing employee involvement programs. However, theSe recommendations reside
at the structural level and omit the psychological processes of individuals placed in

parallel teams. This lack of focus on the individuals in parallel teams gives rise to two

[1

problems.

First, ignoring individual differences implies that everyone will react positively to
such interventions —- an assumption that contradicts interactional psychology, which
asserts that different situations present different psychological meanings and behavioral
potential to different individuals as a result of the differences in individuals’ cognitions,
abilities and motivations (Schneider, 1983; Terborg, 1981). Second, there is little
understanding of the theoretical mechanisms that explain parallel team members’ reaction
to the resource dilemma presented to them by the individual and group tasks, thus
rendering the existing literature on parallel teams relatively a—theoretical. This in part is
due to the fact that existing studies did not highlight the resource dilemma inherent in
parallel teams, even though this represents one central feature that distinguishes parallel
teams from other types of teams.

Hence, one objective of this dissertation is to provide a more in-depth
understanding of the factors that influence parallel team members’ performance in their
group task when faced simultaneously with an individual task to complete. In the next
section, I present an overarching framework that guides my theoretical development, and
a conceptual model that summarizes my arguments in this dissertation.

Research F rarnework_and Conceptual Model

The research question posed in this dissertation mirrors closely an important
question posed in motivation research: what explains the choices that individuals make
amidst an array of alternative acts or responses (V room, 1964)? Motivation, according to
Vroom, is the intra- and inter-individual variability in behavior that is not due solely to
individual differences in ability, or to overwhelming environmental demands that coerce

action.

Kanfer’s (1990) framework of motivation provides an elegant integration of the
numerous motivational theories that have been advanced in the field. Kanfer proposed
that motivation constructs and theories can be ranged on a distal—proximal continuum. On
one end, distal theories are those that posit indirect effects on action, and the emphasis is
on motivational constructs that affect goal choice and intended future effort. Examples
include theories that involve individuals’ needs, personality and interests. On the other
end of the continuum, proximal theories are those that posit direct effects on action, with
an emphasis on motivational mechanisms that control the initiation and execution of
actions. Examples include goal-setting and self-regulation theories.

Following Kanfer’s (1990) framework, I examine both distal and proximal factors
that may influence parallel team members’ performance outcomes. The distal set of
determinants includes both personal factors (cultural values and personality) and
situational factors (task interdependence and leader status), and hence, is consistent with
the perspective of interactional psychology. Besides, including both personal and
situational factors is also aligned with the literature on goal commitment -— the proximal
motivational construct examined in this dissertation.

Goal commitment has a more proximal relationship to behavior since it describes
the amount of determination one has to strive for a goal (Locke et al., 1981), thus
suggesting a more direct impact on behavior than personality or Situational factors. To
understand how goal commitment is influenced by the distal factors in my model, I adopt
Hollenbeck and Klein’s (1987) expectancy model of goal commitment. Essentially,
Hollenbeck and Klein applied Vroom’s (1964) expectancy theory to propose that
personal and situational factors influence goal commitment via either the expectancy

component (the belief that one’s acts will be followed by a particular outcome) or the

valence component (how attractive an outcome is to an individual).

The ultimate outcome of interest in this model is parallel team member’s
performance in the group task, given that they are simultaneously faced with an
individual task. This design therefore captures the dilemma which parallel team members
face when they have only limited resources to accomplish both their individual and group
tasks. It should be noted that although performance is a joint function of ability and
motivation (Gagne & Fleishman, 1959; Maier, 1955), my focus in this dissertation is on
the motivational rather than the ability component. As such, the tasks I employed in this
dissertation are resource-sensitive in nature (Kanfer & Ackerman, 1989), so that
performance is closely related to the amount of personal resources (such as time and
effort) individuals choose to expend in the task. This is opposed to resource-insensitive
tasks, where performance is limited by task characteristics (e.g., task difficulty) rather
than the level of resources devoted by individuals.

So far, I have provided an overview of this dissertation by highlighting an
overarching framework, followed by specific theories that I have chosen to address my
research question. Figure 1a illustrates a conceptual model that summarizes the
relationships of the various constructs examined in this dissertation. Next, I elaborate on
these constructs as well as the rationale for examining them in the context of parallel
teams.

Constructs

Individual Differences. The individual differences examined in this dissertation
are values and personality. Values describe broad tendencies to “prefer certain states of
affairs than others” (Hofstede, 1984: 18), while personality traits refer to a person’s stable

Characteristics that account for consistent patterns of response to situations (Pervin,

 

 

 

 

x3... 9.9.0
:. oucascotom
_«:E>_u:_

 

525.8500
=30 9.9.0

 

 

 

 

llllllllllllllllll

 

magnum .33.. Each
oncogene—9.3:. sun...

m10h0<u
Agahuzmhm

 

 

 

.25

.3232. e.

oocwgotom
.3222.

 

aEGF—gEEOU
.80 .2232.

 

 

r
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I...

.30.... 3:33:00 < :2. 2:9“.

 

 

mmoccfimo05<
wmocmaozcofimcoo

gficoflcn.
_8_em> - 5.0.3.6260
.9535: - Em_>_fio__oo
no. :_u> .2330

mmOhU<m 4<zOmmmm

 

 

 

 

1980). Specifically, I examine two cultural values (horizontal collectivism and vertical
collectivism) and two personality traits (conscientiousness and agreeableness) because of
their relevance to performance in parallel team settings.

Collectivism, defined as the relationship between the self and the group
(Hofstede, 1984), has received heightened attention as a result of increasing globalization
and the concomitant surge in cross-cultural research. Consistent with recent development
in individualism-collectivism (I-C) theory, I adopt a more fine-grained conceptualization
that distinguishes between the horizontal and vertical components of collectivism (Chen,
Meindl & Hunt, 1997; Probst, Camevale & Triandis, 1999; Singelis, Triandis, Bhawuk &
Gelfand, 1995; Triandis, 1995; Triandis & Gelfand, 1998). Specifically, the horizontal
component emphasizes equality, while the vertical component incorporates an element of
inequality.

Conscientiousness and agreeableness are two personality traits that have also
received extensive attention since the introduction of the Five Factor Model of
personality. Conscientiousness involves planning, persistence, and purposeful striving
toward goals, while agreeableness revolves around interpersonal relations that includes
thoughtful and conscience-governed concern for others (Digrnan & Inouye, 1986). Both
have been found to be important predictors of individual and team performance (e. g.,
Barrick, Stewart, Neubert & Mount, 1998; Neuman, Wagner, & Christiansen, 1999;
Neuman & Wright, 1999).

The criterion for the choice of personality dimensions to include in this
dissertation is based on the extent to which my research context will invoke the
Characteristics of the respective personality dimensions. Conscientiousness and

agreeableness are relevant personality dimensions because they relate to goal-

striving/dependability and concern for others respectively -- characteristics that should
have theoretical relevance to individuals’ reactions to the dilemma of pursuing individual
versus collective interests. Other dimensions of the big five seem less applicable. For
example, neuroticism and openness relate more to individuals’ response tendencies
toward uncertain and novel situations, and hence, seem to have little relevance in social
dilemmas. Extraversion describes traits such as being sociable, gregarious, assertive,
talkative and active (McCrae & Costa, 1985; Barrick & Mount, 1991), and therefore on
the surface, may appear to have some relevance to my dissertation. Arguably, extraverts
may be more attracted to group work since it offers more external stimuli than working
alone. Yet, unlike agreeable people, extraverts are motivated by their personal need for
external stimulation rather than by an altruistic concern for others. As such, the relevance
of extraversion to social dilemmas appears to be marginal.

Team structural factors. Although existing work on group design offers many
group characteristics that can impact group effectiveness, such as group composition
(Hackman, 1987), reward structure (Shea & Guzzo, 1987), and group resources
(Gladstein, 1984), few of these factors have been examined in the context of parallel
teams. One exception is the study conducted by Magjuka and Baldwin (1991), in which
group composition factors such as heterogeneity and group size were found to influence
the effectiveness of employee-involvement programs (a form of parallel teams). In this
dissertation, I choose to focus on two group characteristics that have not been examined
in parallel teams - task interdependence and leader status.

Task interdependence, defined as the degree of connectedness between jobs such
that performance of one depends on the efforts, skills, and successful performance of

others (Kiggundu, 1981; Wageman & Baker, 1997), has been widely advocated as an

10

integral factor in team effectiveness models (Campion, Medsker, & Higgs, 1993; Guzzo
& Shea, 1992). However, since most studies were conducted with full-time work teams,
little is known about the effectiveness of task interdependence in eliciting cooperative
behavior from parallel team members. Hence, one objective of this dissertation is to
assess the effects of task interdependence in a different type of team, where members
have their personal tasks to manage in addition to the group task.

The second structural factor is team leader status, which refers to the position that
the team leader occupies in the organizational hierarchy. Compared to task
interdependence, team leader status has received relatively less theoretical and empirical
attention to date. Given that hierarchical relationships occur in practically all types of
organizations, it is important for research to pay greater attention to the phenomenon.
Besides, in the context of parallel teams, the formal status of the team leader is an
important design feature as it reflects the way in which the team is being managed, and
possibly, the team’s legitimacy in the organization. Thus, in response to the general lack
of organizational legitimacy faced by parallel teams (Cohen & Lawler, 1992), it is
possible that leader status may offer a solution to the problem.

In addition, both task interdependence and leader status have the merit of being
“manipulable,” which lend them to be feasible interventions that organizations can
implement. For instance, organizations can choose between training parallel team
members so that they can individually complete the task with minimal interaction (i.e.,
low task interdependence), versus assigning tasks according to members’ expertise so
that they need to rely on one another to complete the entire job (i.e., high task
interdependence). Likewise, organizations can choose between appointing a senior

company executive versus a junior executive to lead the team.

11

Ev— —.' git.

Goal commitment. Goal commitment, the proximal motivational construct in my
model, is an important goal dimension that has not been examined in multiple-goal
settings such as parallel teams. Given that parallel team members are faced with two
goals: one for their individual task and one for their group task, I argue in this dissertation
that parallel team members develop two types of goal commitment: individual goal
commitment and group goal commitment. Applying Locke et al.’s (1981) definition,
individual goal commitment refers to the amount of determination one has for achieving
the goal assigned to the individual task, while group goal commitment refers to the
amount of determination one has for achieving the goal assigned to the group task.

Performance outcome. Goal commitment is an important construct because it
directly influences individuals’ task performance outcomes (e.g., Klein, Wesson,
Hollenbeck, & Alge, 1999; Latharn & Lock, 1991). In this dissertation, I am specifically
interested in parallel team members’ performance in their group task Given the
motivational perspective I have adopted, this outcome reflects individuals’ willingness to
contribute to their group task at the expense of their individual task.

I do not focus explicitly on performance in the individual task for several reasons
(hence the dotted lines and box). First, this is consistent with my primary interest, which
is the willingness of individuals to contribute to the collective in the presence of
competing personal interest. Second, the limited amount of time available for the
completion of two tasks with similar requirements necessarily implies a strong negative
relationship between the two performance outcomes. As such, it will be redundant to
include performance in both the individual and group tasks since they should be
complementary of each other.

Potential Contributions of this Dissertation

 

12

The research context and questions addressed in this dissertation contribute to
existing knowledge in several ways. First, parallel teams have been under-represented in
existing research (Cohen & Bailey, 1997) and little is known about the factors
influencing parallel team members’ performance in their individual and group tasks when
they are faced with limited resources. Even though existing research has established
several structural and dispositional factors as instrumental in eliciting cooperative
behavior from members in the more traditional types of team, such as full-time work
teams (for reviews, see Bettenhausen, 1991; Cohen & Bailey, 1997), these findings may
not generalize to parallel teams. For instance, Magjuka and Baldwin (1991) found that
the size of parallel teams, contrary to expectations, was positively related to group
effectiveness. Hence, it is important to assess existing findings concerning group and
member characteristics in a non-traditional group setting such as that of parallel teams. It
may be that the presence of individual responsibilities faced by parallel team members
could reduce or neutralize the benefits of structural interventions on cooperation within
teams. Or, it may be that individual differences in personality and cultural values are
particularly important in parallel teams, given that different people may be differentially
committed to the achievement of their individual and group goals. Therefore, this
dissertation provides a theoretical model to understanding the influence of various
individual differences and situational factors on individuals’ reactions to the dilemma
posed by competing individual and group tasks.

In addition, I highlight the role of goal commitment in a novel setting such as
parallel teams. Specifically, I conceptualize goal commitment as a more proximal
mechanism that explains the influences of dispositional and situational characteristics on

individuals’ group task performance. This approach provides a more in-depth

l3

 

understanding of the role of distal factors on behavior, and thus, can help advance
theoretical development in the relevant domains of research such as values, personality
and team interventions. Personality research, for instance, has typically neglected to
explain the mechanisms underlying personality effects on behavior (cf. Barrick, Mount &
Strauss, 1993; Graziano, Hair & Finch, 1997).

Further, the popularity of parallel teams in organizations confers considerable
practical significance to research that seeks to understand what factors influence parallel
team members’ performance in their group task. Given that parallel team members need
to juggle between their individual tasks and group tasks, there are two potential risks
posed by parallel teams to organizations. First, as pointed out by existing research (e.g.,
Adam, 1991; May & Schwoerer, 1994), the task assigned to parallel team members may
not receive adequate attention due to their concurrent individual workload. The
consequences of such neglect can be serious if the parallel team is formed to address a
critical issue. On the contrary, problems may also arise if parallel team members over-
emphasize their team assignments at the expense of their individual workload. This
ironical downside of teamwork can occur if members fail to comprehend the significance
of their individual tasks vis-a-vis the significance of their group tasks to the
organizations. Hence, understanding what factors influence parallel team members’
motivation to contribute to competing group task and individual task can provide
important insight to organizations on how to design parallel teams in order to promote an
optimal division of employees’ resources.

Boundgy Conditions
As with all research, my conceptual model is based on several assumptions which

will limit the generalizability of the model to other contexts. First, consistent with my

14

motivational focus, I have assumed that the two tasks do not demand different skills or
abilities. Second, I have also assumed a “zero-sum” game (i.e., a strong negative
relationship between performance outcomes in both tasks) which negates the possibilities
of any expansion of one’s resources (e. g., working over-time).

Third, I have restricted the structural variables to those relating to the design of
the parallel team, such as task interdependence and team leader status. There are
potentially many structural features of the organization that can influence members’
commitment to their individual goal (e.g., compensation system), but are not within the
scope of this thesis. In part, this is to maintain the group context as the focal point of my
research questions. Moreover, organizations are likely to have greater ease and flexibility
in altering the design of parallel teams than to change existing organizational structures.
Indeed, one common objective of parallel teams is to elicit employee participation
without disrupting existing work practices and organizational structure (Cotton, 1993).
As such, investigating parallel team interventions is likely to provide greater practical
insight than interventions aimed at changing individual work structure.

Outline of Dissertation

To summarize, this dissertation examines factors that influence parallel team
members’ performance in their group tasks in the presence of competing demands from
their personal tasks. I propose that individual and structural factors affect members’
individual goal commitment and group goal commitment, which in turn, influence their
performance in the group task. The remaining chapters in this dissertation are organized
to provide theoretical and empirical support for my hypotheses. Figure lb presents an
empirical model depicting my hypotheses.

Chapter 1 presents a literature review on goal commitment, and elucidates its role

15

 

 

x3... 95.5
E ooggotom

@6365 oilfilln

 

 

mm

 

 

25.5.8800
BOO 955

 

 

 

 

 

25.5.8800
How 36365

 

3E - «E an.
38m .8va 8.8...

 

Sm - 8: .Ev
ooqouaomowufifi xmfl.

 

9.323 3.85253

 

 

 

/ Ammv mmoaofimooumao.
lama 30532330800
/€.Efimm>uoo=oo Home»;
12:. 83382.8 Seacom
meanest:— 13231:—

 

 

 

 

 

\ 3N5 383330.880
SE. 83.3.8 35>

 

.\/\

 

 

[3.5 S5023 .288:
828.85: 132.55—

 

 

Enos. .mo_.._nEm :< 2:. 2:2“.

 

 

l6

in a multi-goal environment, such as that of parallel teams. Chapter 2 proposes the
antecedents and consequences of goal commitment relevant to the current dissertation.
The antecedents are individual differences of team members and structural features of
parallel teams. The outcome of interest is parallel team members’ performance in their
group task. Chapter 3 describes the methodology and statistical analyses designed to test
these propositions. Chapter 4 reports the results. Chapter 5 concludes this thesis with a

discussion of the implications of the findings.

17

CHAPTER 1

THE ROLE OF GOAL COMMITMENT IN PARALLEL TEAMS

A unique feature of parallel teams is that parallel team members have two sets of
organizationally-relevant acts to accomplish (individual and group tasks), compared to
either non-team members (only individual tasks) or regular, full-time team members
(only group tasks). This chapter explicates the role of goal commitment in such multiple
goal environments. As shown in Figure 1a, 1 theorize that goal commitment serves as a
proximal construct in explaining the distal influence of individual differences and
structural factors on parallel team members’ performance in their group task.

This chapter is organized in three sections. First, I review the literature on goal
commitment. Next, I discuss the role of goal commitment in the broader context of a
multiple-goal environment. I then apply goal commitment to parallel teams and discuss
its conceptualization in this dissertation.

Goal Commitment

Goal commitment, defined as the determination to try for a goal (Locke at al.,
1981), involves the extension of effort, over time, toward the accomplishment of a goal.
In essence, it emphasizes an unwillingness to abandon or to lower the original goal.
(Campion & Lord, 1982).

During the early stages of its inception, goal commitment was viewed primarily
as a necessary condition for the setting of difficult goals to lead to higher performance.
According to Locke (1968), people who are not committed to attaining a goal, when
confronted with a hard task, are likely to decide that the goal is impossible to reach and

therefore, stop trying for that goal. Hence, Locke attributed the inconsistent findings of

18

the goal difficulty-performance relationship in the existing literature to the potential
moderating effect of goal commitment. Yet, despite the central role that goal commitment
is purported to play in the goal-setting literature, it has received relatively little attention
compared to other goal dimensions such as goal specificity and goal difficulty. For
instance, Hollenbeck and Klein (1987) noted that 61% of the 109 empirical studies they
reviewed did not mention goal commitment; in another 12% of the studies, goal
commitment was mentioned but not empirically assessed. In the remaining studies that
did examine goal commitment, the authors found the treatment of the construct to be
inconsistent with Locke’s (1968) original formulation of goal commitment.

Subsequent to several reviews of goal commitment that unanimously pointed for
more work to be done on the construct (e.g., Hollenbeck & Klein, 1987; Locke, Latham,
& Erez, 1988), a more systematic program of research began to emerge in the field.
These studies may be further categorized as having at least one of the two themes:
establishing the measurement properties and/or establishing the nomological network of
goal commitment. While both types of studies are important in the advancement of goal
commitment research, my focus in this chapter is on the latter. I defer the discussion of
measurement issues to Chapter 3 when I present the goal commitment measures used in
this dissertation.

Two theoretical models articulating the nomological network of goal commitment
were proposed in the 805, one by Hollenbeck and Klein (1987), and the other by Locke et
a1. (1988). The two models differ in several ways. First, while Hollenbeck and Klein
(1987) adopted a specific theory to explain the antecedents of goal commitment, Locke et
a1. adopted a more general cognitive processing model. Second, the specific antecedents

proposed in the two models differ from each other. Nonetheless, the two models are

19

similar in that these antecedents can be categorized broadly into situational (Locke et al.
labeled as external and interactive) and personal (Locke et al. labeled as internal) factors.

Of particular relevance to this dissertation is Hollenbeck and Klein’s (1987)
elaboration of the expectancy model of goal commitment. This model is built upon
Vroom’s (1964) expectancy theory, which asserts that the force on a person to perform an
act is a function of the sum of products of the valences of all outcomes and the strength of
the person’s expectancies concerning these acts. Valence refers to the person’s
anticipated satisfaction with a particular outcome, or the degree of attractiveness the
outcome holds for that individual. Expectancy refers to the belief that the act will be
followed by the attainment of these outcomes, or in other words, the expectation that
effort will lead to the attainment of a goal.

Although early research has suggested integrating Vroom’s (1964) expectancy
theory into the goal-setting literature (e.g., Dachler & Mobley, 1973; Locke et al., 1981;
Oldham, 1975; Steers, 1975), few studies have explicitly explained goal effects using the
theory of expectancy. Using the expectancy model of goal commitment, Hollenbeck and
Klein (1987) provided some compelling reinterpretation of past research in the goal-
setting literature. In essence, the authors proposed that both situational and personal
factors can influence individual’s goal commitment via either the attractiveness or the
expectancy of attaining the goal.

For goal attractiveness, Hollenbeck and Klein (1987) proposed that situational
factors such as the nature of the goal (whether it is public, self-set or explicit), the type of
reward structure, and the presence of competitive pressure, should influence the
attractiveness of goal attainment. Further, the authors argued that individuals who are

high in need for achievement, endurance, Type A personality, organizational

20

 

commitment and job involvement are likely to attach greater values to the outcomes
associated with goal attainment.

For expectancy of goal attainment, Hollenbeck and Klein proposed that
situational factors such as the presence of social influence (e.g., knowledge of other’s
goals, commitment and performance), complex tasks, performance constraints and
supervisor supportiveness can influence one’s beliefs about attaining the goals. In
addition, personal factors such as ability, past experiences of success, self-esteem and
locus of control are posited to influence expectancy beliefs. Specifically, individuals who
possess more of such factors are likely to have greater expectancy beliefs than those who
possess less.

Besides proposing main effects of situational and personal factors on goal
attractiveness and goal expectancy, Hollenbeck and Klein also argued for interactive
effects, both within as well as across, the situational and personal categories. For
instance, volition (a situational factor) may be related to commitment only where there
are not substantial constraints on performance (also a situational factor). Or, a person’s
self-esteem may be related to goal commitment only when that person also experiences a
high level of job involvement (i.e, a person-person interaction). Finally, personal variable
can also interact with situational variable to influence goal commitment. One example is
that involving Type A personality and the presence of competitive pressure. Competition
may enhance the goal commitment of individuals with Type A personality, but not those
with Type B personality.

Several meta-analyses have examined the cumulative empirical evidence for the
proximal (i.e., goal valence and expectancy) and distal predictors of goal commitment

(e.g., Klein, 1991; Klein et al., 1999; Wofford, Goodwin, & Premack, 1992). In a recent

21

 

meta-analysis by Klein et al. (1999), goal attractiveness and expectancy (proximal
predictors) were found to be significantly related with goal commitment. For distal
predictors, Klein et al. found situational factors such as volition, goal specificity (or
explicitness), feedback, task information, incentives, social influence, supervisor
supportiveness and publicness to be positively associated with goal commitment. Task
complexity was found to be negatively related to goal commitment. Personal factors that
were positively associated with goal commitment included ability/past performance,
affect, need for achievement, and Type A personality.

While the antecedents of goal commitment have been relatively well validated,
the consequences of goal commitment are perhaps more ambivalent. Earlier research in
goal-setting has conceptualized goal commitment as an important boundary condition to
the relationship between goal difficulty and performance (Locke, 1968) — difficult goals
increase performance only when there is commitment to attaining the goal. Empirical
research, however, did not provide consistent support for this moderating fimction of goal
commitment (e.g., Erez & Zidon, 1984; Frost & Mahoney, 1976; Wright, O’Leary-Kelly,
Cortina, Klein, & Hollenbeck, 1994; Yukl & Latham, 1978). Hollenbeck and Klein
(1987) noted that because these studies measured goal commitment differently from one
another, it is difficult to attribute the inconsistent result to any specific cause.

Besides serving as a moderator to the goal difficulty-performance relationship,
goal commitment can also impact performance directly (Klein et al., 1999; Latharn &
Locke, 1991). In other words, goal commitment can exert a main effect on performance,
averaged across all levels of goal difficulty. This is because the persistence and
determination of committed individuals imply greater amount of effort being expended,

which holding other factors constant, should result in better performance. Meta-analytic

22

 

results from Klein et al.’s (1999) study supported this direct effect of goal commitment
on performance, and reported a mean weighted correlation of .23, corrected for
unreliability.

In summary, goal commitment is a critical construct in motivation research that is
receiving increasing attention. Several meta-analyses have largely substantiated
theoretical propositions of some proposed antecedents and consequences of goal
commitment. However, despite the vast literature on goal commitment, virtually no study
has applied it to a multiple-goal environment. In the next section, I discuss the
implications of goal commitment in multiple-goal settings.

Multiple Goals and Goal Commitment

Notwithstanding the fact that having multiple goals is a common occurrence in
organizations, relatively little empirical research has addressed how people cope with
their multiple goal striving (Austin & Vancouver, 1996; Keman & Lord, 1990; Locke &
Latham, 1990), and fewer still have explicitly examined the role of goal commitment.
Nonetheless, the results from some of these studies suggest important implications of
goal commitment on individuals’ prioritization of the multiple goals.

One study is that by Keman & Lord (1990), where they compared motivational
processes under single and multiple goal conditions. Specifically, they tested two
competing theories that could potentially predict the determination of goal priority:
cybernetic control theory (Carver & Scheier, 1981) and expectancy-valence theory
(V room, 1964). Cybernetic control theory predicts that the task that has the greater
discrepancy will be assigned a greater priority. On the other hand, expectancy-valence
theory predicts that the task that has the greater expectancy of success and valence will be

accorded greater priority. Keman and Lord’s (1990) results supported the expectancy-

23

 

valence theory that goal priority was determined largely by valence of the outcome (i.e.,
monetary reward). More specifically, discrepancies between goal and performance, as
well as subjects’ expectancies had an effect on goal priority only when achieving that
goal resulted in a valued outcome.

Another study is that by Schmidt, Kleinbeck and Brockmann (1984), where they
conducted an experiment that required subjects to engage simultaneously in a tracking
task and a reaction time task. The authors provided feedback and assigned a specific
performance goal for one task, but merely encouraged subjects not to reduce the
performance of the other task. Their results showed that subjects performed better in the
task with feedback and specific goals at the expense of the second task. Schmidt et al.
concluded that the task with feedback and goals received a higher priority, and as a result,
was allocated more resources.

The results from these two studies may be interpreted in light of the expectancy
model of goal commitment discussed earlier (Hollenbeck & Klein, 1987; Locke et al.,
1981). Between two competing tasks, the one that provides greater expectancy beliefs
(e. g., induced by feedback) and more valued outcomes (e.g., induced by monetary
rewards) is likely to result in greater commitment in the individual to pursue that goal.
Since greater commitment means greater unwillingness to abandon or lower the original
goal, greater amount of resources will be channeled to meet that goal, thus leading to
higher levels of performance.

Hence, applied to parallel teams, I argue that individuals’ commitment to their
individual goal and their group goal will influence their performance in the individual
and group tasks respectively. However, before I discuss the specific conceptualization of

goal commitment in this dissertation, it is worthwhile to highlight some differences

24

 

between existing multiple-goal studies and the parallel team setting examined in this
dissertation. The majority of studies that examined multiple goals were interested in the
performance outcomes of two concurrent individual tasks (e. g., Erez, Gopher & Arzi,
1990; Keman & Lord, 1990; Schmidt et. al., 1984), rather than contrasting between an
individual task and a group task. Only two published studies have operationalized
multiple goals as having both individual and group goals (Matsui, Kakuyama &
Onglatco, 1987; Mitchell & Silver, 1990). Both studies, however, did not create a
dilemma between achieving the individual goal and the group goal because both goals
related to only one task. For instance, in the Matsui et al. (1987) study, subjects in the
group goal condition were paired in twos and asked to set both individual and group
goals for a perceptual speed task which they worked independently on, while subjects in
the individual goal were asked to set individual goals only. The results showed that at the
individual level, subjects in the group goal condition performed significantly better than
subjects in the individual goal condition. Similarly, the results in a study by Mitchell and
Silver (1990) showed that subjects assigned with both individual and group goals
performed better in a “tower-building” task, than subjects assigned with only individual
goals. In both studies, since subjects essentially worked on one task and group
performance was measured by aggregating individuals’ performance, there was no
tension between achieving individual and group goals.

Therefore, this dissertation differs from Matsui et al.’s (1987) and Mitchell and
Silver’s (1990) studies in that I examine individuals’ performance in their group task
when they are simultaneously faced with an individual task that competes for their time
and effort. This dissertation also extends beyond the few studies on multiple goals by

including the role of individual differences. Finally, this model examines the role of goal

25

EFF-f .fl-AJ

 

commitment as a proximal construct in influencing performance outcomes.
Goal Commitment in Parallel Teams

The majority of the goal commitment studies were focused on individual task and
performance, and fewer studies have examined goal commitment in a group setting
(Crown & Rosse, 1995; Weingart & Weldon, 1991). Even less attention has been paid to
goal commitment in less traditional group settings such as parallel teams.

Parallel teams present an important and interesting context for examining goal
commitment because of their prevalence in organizations, and the fact that unlike
traditional teams, there exists concurrently both individual goals and group goals. The
presence of multiple goals (individual and group) competing for individuals’ finite
resources (e.g., time and effort) may enhance the role of goal commitment in determining
performance, because unlike single-goal environments, individuals in multiple-goal
environments are likely to experience intra-individual goal conflict, defined as the
pressure exerted upon individuals to take incompatible actions or achieve incompatible
outcomes (Locke, Smith, Erez, Chah, & Schaffer, 1994). As a result, individuals faced
With such goal conflict must rely on some mechanism to resolve this conflict. I argue in
this dissertation that goal commitment is potentially one such mechanism, such that the
task which individuals are more committed to accomplishing will receive greater
attention and hence, produce better performance outcomes.

Besides, a multiple-goal setting such as parallel teams may also fit Locke et al.’s
(1981) recommendation to study goal commitment using designs that encourage a wide
range of goal commitment. For instance, the authors suggested that a within-subject
design that involves assigning participants with different goals under different conditions

may induce more varied commitment responses by providing a frame of reference.

26

 

Likewise, I would argue that presenting competing goals to individuals create conditions
that allow individuals to discriminate their commitment to different goals (i.e., intra-
individual differences), as well as to observe goal commitment differences between
individuals (i.e., inter-individual differences).

The literature review presented earlier suggests that when faced with both
individual and group tasks to accomplish, the level of commitment parallel team
members have toward achieving the individual goal (hereby termed individual goal
commitment) and the group goal (group goal commitment) will influence their
performance in the group task. This is illustrated in Figure 1a.

Conceptualizing goal commitment for the individual and group tasks as separate
constructs implies that members’ commitment to both tasks can be independently
influenced by other factors. Conceivably, factors that pertain to the group (e.g., group
task difficulty, group goal specificity) will have an impact on group goal commitment but
not necessarily on individual goal commitment. Likewise, factors that are relevant
specifically to the individual task (e.g., individual task difficulty, personal accountability
for individual task) will influence individual goal commitment but not necessarily group
goal commitment. Moreover, broader organizational factors such as employees’
performance appraisal and/or compensation scheme that influence the expectancy beliefs
and values associated with the attainment of both individual and group goals can
simultaneously affect both individual and group goal commitment.

Hence, I contend that there are four possible categories of individuals with
various combinations of individual and group goal commitment. Some individuals may
be equally highly committed to both their individual and group goals, while others may

be equally uncommitted to both. Although there are no intra-individual differences in the

27

 

level of commitment to both goals for all these individuals, there are inter-individual
differences between those who are equally committed and those who are equally
uncommitted to, the two goals. Consequently, although both groups of individuals are
likely to allocate resources equally between individual and group tasks, the total amount
of resources allocated to both tasks will differ between the two groups. On the other
hand, there are also individuals who may be more committed to the individual goal than
to the group goal, and vice versa. In these cases, the goal that individuals are more
committed to will receive a greater amount of resources vis-a-vis the other goal.

In this dissertation, my interest is to examine inter-individual differences in
parallel team members’ performance in their group task, when faced with competing
individual and group tasks. The central thesis, based on the expectancy model of goal
commitment, is that individuals should be more committed to the task that promises more
attractive outcomes and a greater possibility of attaining those outcomes.

In the next chapter, I propose dispositional and situational antecedents of parallel
team members’ individual goal commitment and group goal commitment, as well as the

impact of the two goal commitment constructs on performance in the group task.

28

-‘2 (1w fulfill

 

CHAPTER 2

ANT ECEDENTS AND CONSEQUENCES OF GOAL COMMITMENT

The antecedents of goal commitment examined in this dissertation, following
Hollenbeck and Klein’s (1987) model, are classified into individual differences and
structural factors. Individual differences are relatively stable characteristics that describe
individuals, while the structural factors refer to design features of the team that may be
altered by external interventions.

This chapter is organized in four sections. In the first section, I propose the
relationships between individual differences (cultural values and personality) and
individual and group goal commitment. In the second section, I focus on the relation
between structural factors (task interdependence and team leader status) and individual
and group goal commitment. In the third section, using a person-situation fiamework, I
propose the joint influence of individual differences and structural factors on group goal
commitment. Finally, in the fourth section, I propose the joint influence of individual and
group goal commitment on parallel team members’ performance in their group task.
Individual—Differences as Antecedents

Although a number of empirical studies have examined the effects of personal
factors on goal commitment, very few have used the more recent constructs advanced in
the domain of personality and values research. For instance, the value orientation of
individualism-collectivism (I-C; Hofstede, 1984; Markus & Kitayama, 1991; Triandis,
1995) and the Five Factor Model of personality (e.g., Costa & MaCrae, 1992) have rarely
been incorporated in goal commitment studies. The recent meta-analysis by Klein et al.

(1999) confirmed this observation: neither I-C nor the Big Five personality traits were

29

 

1 Ill-111m. .‘lu

included in their meta-analytic review. Instead, some examples of the more established
personal factors for goal commitment are ability/past performance, affect, need for
achievement, and Type A personality.

Therefore, one aim of this dissertation is to expand the range of individual
differences associated with goal commitment by including I-C and personality variables.
Specifically, I examine two types of I-C (horizontal and vertical collectivism) and two
personality characteristics (conscientiousness and agreeableness) for their influence on
individual goal commitment and group goal commitment of parallel team members.

The role of these individual differences in organizational behavior is widely
established. For instance, Kanfer’s (1990) distal-proximal framework of motivation states
that distal motivational theories (such as personality-based theories) can affect behavior
through proximal motivational constructs such as goal-setting. Borman, White, Pulakos,
and Oppler (1991) proposed that personality influence the “will-do,” rather than the “can-
do” aspects of the job, suggesting motivational effects of personality variables. Austin
and Vancouver (1996) remarked that individual characteristics cause individuals to
impute a relatively permanent amount of importance for specific goals (Austin &
Vancouver, 1996). Likewise, Cropanzano, James and Citera’s (1992) goal-based model

of personality argues that response tendencies at the high end of the hierarchy (such as

personality or values) influence specific task goals at the bottom of the hierarchy. Hence,

the authors advance that a lower-order goal is important to the individual insofar as it
fulfills a higher-order goal.

In the ensuing sections, I provide a literature review of the values and personality
constructs included in my dissertation. This is followed by hypotheses for their

relationships with individual and group goal commitment.

30

‘I I.“

 

- ling-

Cultural Values

With the rapid globalization and diversification of businesses, cultural values play
an important role in determining the success of management initiatives such as teamwork
(Boyacigiller & Adler, 1991; Barley, 1989; 1993; Erez & Barley, 1993; Erez & Somech,
1996; Kirkman & Shapiro, 1997). Erez and Barley (1993) proposed that cultural values
influence the extent to which employees perceived managerial techniques to contribute to
their personal goals. Indeed, there have been numerous examples that demonstrate how
different motivational techniques and managerial practices emerge in different cultures,
and how their effectiveness changes when transferred across cultures.

Cross-cultural researchers assert that a theoretical approach to understanding
cultural influence is to specify the aspect of culture germane to the area of inquiry
(Hofstede, 1984; Lytle, Brett, Barsness, Tinsley, & Jarrssens, 1995). One dimension of
culture that has received extensive research attention is individualism-collectivism (I-C),
defined as “the relationship between the self and collectivity” (Hofstede, 1984: 148).
Other similar labels that revolve around this theme of “self versus others” include self-
orientation vs collectivity-orientation (Parsons & Shils, 1951), cooperation vs
individualism (Mead, 1967), and independent vs interdependent self-construals (Markus
& Kitayama, 1991). Although initial cross-cultural research focused on cultural
differences at the national or cultural level (e. g., Bochner & Hesketh, 1994; Chan,
Gelfand, Triandis & Tzeng, 1996; Hofstede, 1984), more recent studies have also
examined differences in the cultural values of individuals (e. g., Barley, 1989, 1993; Farh,
Barley & Lin, 1997; Wagner, 1995). I adopt the latter perspective and construe culture as
an individual-level construct.

Triandis (1995) characterizes an individualist as one who views the self as

31

independent of others, focuses on personal goals, acts upon personal beliefs and values,
and emphasizes task outcomes. A collectivist, on the other hand, construes the self as an
interdependent entity, adopts group goals, acts according to social norms, and stresses
interpersonal relationships.

The different emphasis that individualists and collectivists place on their personal
and group goals has many implications for group work. For instance, social loafing,
defined as the reduced performance of individuals who work as part of a group rather
than alone (Latane, Williams & Harkins, 1979), has been found to occur more in
individualists than collectivists (Barley, 1989, 1993; Karau & Williams, 1993). Results of
a meta-analysis of social loafing studies revealed that the magnitude of social loafing was
larger for subjects from Western cultures than for those from Bastem cultures (Karau &
Williams, 1993). In an experiment involving managerial trainees, Barley (1989) found
that social loafing was present only among individualists (primarily Americans) but not
among collectivists (primarily Chinese). Barley (1989) suggested several reasons for this
finding.

First, the interdependent nature of collectivists (Markus & Kitayama, 1991)
predisposes them to view their individual responsibility as indispensable in attaining
group outcomes. Individualists, being more concerned with their self interest, are more
likely to rely on others to attain collective goals and redirect effort to pursue personal
Outcomes. Second, since collectivists value group outcomes more than individualists,
they gain greater satisfaction and feelings of accomplishment from attaining group goals
than individualists. Third, because collectivists view their contributions as an important
and role-defined aspect of group membership, they anticipate other group members to

also contribute to group performance. However, the self-centered nature of individualists

32

I. VP “1". 1.”

 

results in less assurance that all members will contribute equally to the group outcome.
Consequently, individualists may contribute less because they do not want to appear as
the “sucker” who shoulders the burden of attaining collective goals for other group
members (Orbell & Dawes, 1981).

I-C also has significant relevance to cooperation with others, defined as the
“willful contribution of personal effort to the completion of interdependent jobs”
(Wagner, 1995: 152). Empirical evidence shows that collectivists are generally more
cooperative than individualists. For instance, Wagner (1995) found that the aspect of I-C
that concerns personal independence and self reliance had a direct effect on peer-
evaluated cooperative behavior - individualists who reported greater independence and
self-reliance were less cooperative than collectivists who reported greater
interdependence and reliance on groups.

In the above studies, LC was construed as a unidirnensional construct. Recently,
Triandis and his colleagues proposed that individualism and collectivism can be further
decomposed into vertical and horizontal components (Probst et al., 1999; Singelis et al.,
1995; Triandis, 1995; Triandis & Gelfand, 1998). This horizontal-vertical distinction
acknowledges that there are different emphases on horizontal and vertical social
relationships (Triandis, 1995). Generally, horizontal patterns stress the element of
equality, where one’s self is construed as more or less like every other self. Vertical
patterns, by contrast, consist of hierarchies in which the self is seen as different from
other selves. Collectivism emphasizes interdependence with others, while individualism
stresses the autonomous self. Crossing the horizontal-vertical dimension with I-C results
in four subdimensions: horizontal individualism (H-I), vertical individualism (V -l),

horizontal collectivism (H-C) and vertical collectivism (V -C.

33

 

Such a refined I-C scale has both advantages and disadvantages. The advantage is
that the horizontal-vertical distinction may potentially address deficiencies associated
with an overly abstract scale (i.e., the unidimensional approach). Chen et al. (1997)
suggested that using the horizontal and vertical dimensions can increase the precision of
I-C theory when the dimension is suitably matched to the behavior in question. For
instance, horizontal collectivism may be more appropriate when studying cooperative and
helping behaviors, while vertical collectivism may be more suitable for research on
intergroup competition and need for socialized power.

One disadvantage of the horizontal-vertical I-C distinction, however, is that it

 

! ("Whh I-“ --—-_

broadens a construct that is already diffuse (Barley & Gisbon, 1998). In particular, the
construal of individualism and collectivism as separate continua gives rise to
considerable confusion regarding the interpretation of LC. As an example, it is
ambivalent whether a person who has low collectivism (i.e., weak interdependent self-
concept) can be considered individualistic (i.e., independent self-concept).

For conceptual parsimony and clarity, I adopt a middle-ground approach proposed
by Chen et al. (1997), which included the horizontal-vertical distinction only for
collectivism. The authors reasoned that no such distinction was necessary for
individualism because the “self-centered” nature of individualists suggests that the nature
of their relationships with others (whether equal or unequal) is irrelevant. Hence, the
authors avoided the confusing nuances of the four-factor structure by proposing that
individuals who have low scores for vertical or horizontal collectivism are essentially
individualistic in nature.

Horizontal and Verticflollectivisg. As discussed earlier, a major common

attribute between H-C and V-C is the emphasis on interdependence with others (Singelis

34

—

   

et al., 1995; Triandis, 1995). Both horizontal and vertical collectivists emphasize the self
as part of the group. However, horizontal collectivism stresses equality among group
members, and is concerned with fostering common goals and social harmony.

Vertical collectivism, on the other hand, emphasizes inequality among members,
especially with regard to social status. Hence, members within the group view one
another as different, some having more status than others. This characterization of V-C is
similar to Hofstede’s (1984) power distance, defined as the extent to which those with
less power expect and accept that power is distributed unequally (Barley & Gibson,
1998). Thus, similar to the concept of power distance, individuals with high V-C are
more sensitive to authority pressure. For instance, Singelis et al. (1995) observed that in
Japan, a highly vertically collectivistic culture, it is important for people to know the
relative status of the speaker in order to decide on the use of an appropriate language.

Bmpirically, Ng and Van Dyne (2001) demonstrated that individuals with high V-
C reacted more positively to status and power than those with low V-C in a study on
minority influence. Specifically, the authors found that amongst individuals who were
exposed to a high-status minority influence agent (someone who advocates a different
idea from the majority), those with high V—C were more likely to be influenced by the
agent than those with low V-C. By contrast, individuals with high V-C, when exposed to
a low-status minority influence agent, were less likely to be influenced compared to those
with low V-C.

Two other studies provided support to the theory that V-C stresses inequality
while H-C emphasizes equality. In a study of reward allocation preferences in China,
Chen et al. (1997) found that vertical collectivists favored differential reward allocation

(e. g., based on performance, status) but not horizontal collectivists. In another

35

 

JKFHI-LFL'A

 

nib-.1
1

experimental study involving prisoner’s dilemma games, Probst et al. (1999) found that
V-C interacted with the type of prisoner’s dilemma but not H-C. Specifically, vertical
collectivists cooperated more in single prisoner’s dilemma games, and less in intergroup
prisoner’s dilemma games in order to maximize group gains over individual games. On
the other hand, horizontal collectivists did not differ in their cooperation in either type of
social dilemma. Thus, Probst et al.’s (1999) results supported the notion that vertical
collectivists were concerned with sacrificing self-interests for their groups, while
horizontal collectivists were more guided by equality matching, whereby resources
should be divided equally regardless of the type of social dilemma.

What are the implications of H-C and V-C for parallel team members’
commitment to their individual and group goals? Cropanzano et al.’s (1992) goal
hierarchy model suggests that individuals’ higher-level H-C and V-C orientation would
influence their commitment to lower-level task goals. Specifically, since both horizontal
and vertical collectivists emphasize interdependence with other group members (Markus
& Kitayama, 1991; Probst et al., 1999; Triandis, 1995), their higher-order goals concern
the ability to contribute to the pursuit of group goals (Chen et al., 1997). Therefore, I
argue that the attractiveness of achieving the group goal is positively associated with
individuals’ H-C or V-C orientation.

Further, given that collectivists are more likely to believe that other group
members will also contribute to the group task (Barley, 1989), they should also perceive a
stronger link between their effort toward the group task and the attainment of the group
goal. By contrast, because individualists are more skeptical about the amount of effort
other group members will put in, they have less reasons to expect that their own effort

toward the group task will result in the attainment of the group goal.

36

 

Taken together, I propose that because individuals with high H-C or V-C are
likely to have greater valence and expectancy associated with achieving the group goal,
they are more committed to the group goal than individuals with low H-C or V-C. On the
other hand, the reverse is true for individual goal commitment, since individuals high in
H-C or V-C are likely to attach lower values to their individual goal compared to those
with low H-C or V-C.

Hypothesis la: Collectivism (both horizontal and vertical) is positively associated

with group goal commitment.
Hypothesis lb: Collectivism (both horizontal and vertical) is negatively

associated with individual goal commitment.

As Chen et a1. (1997) pointed out, the distinction between H-C and V-C will be
useful when studying specific behavior, or when specific contexts are involved. I defer
my discussion on how H-C and V-C may exert different influences on goal commitment
to a later section when I consider the structural factors of the team.

Personalin

A critical impetus to the resurgence of personality research is the establishment of
the five factor model (F FM) (Mount & Barrick, 1995). There is now widespread
consensus that most personality traits can be described by five dimensions, namely
conscientiousness, agreeableness, extraversion, openness and neuroticism (Costa &
McCrae, 1992). Furthermore, these factors have also been replicated in other cultures.
For instance, McCrae and Costa (1997) administered translated versions of the Revised
NEO Personality Inventory (Costa & McCrae, 1992) in six cultures (German, Portuguese,
Hebrew, Chinese, Korean and Japanese) and found that the five factor structure was

closely reproduced, thus suggesting that the personality trait structure is universal.

37

Another factor that spurs the interest in personality research is meta-analytic
evidence demonstrating that personality is a predictor of individual performance (e.g.,
Barrick & Mount, 1991; Hough, Baton, Dunnette, Kamp, & McCloy, 1990). This finding
significantly heightens the implications personality has for organizational sciences.

In this dissertation, I focus on two of the five personality characteristics. The
omission of the other three personality traits does not imply that they are unimportant.
Rather, as discussed earlier, I chose conscientiousness and agreeableness because they
are theoretically relevant to the resource dilemma induced by the multiple-goal context of
parallel teams. In addition, they also provide an interesting contrast in terms of their
effect on goal commitment, as illustrated by my hypotheses below.

Conscientiousness. According to a literature review by Barrick and Mount (1991),
conscientiousness reflects the qualities of being dependable, careful, responsible,
hardworking, achievement-oriented, and persevering. Costa and McCrae (1992)
described conscientious individuals as likely to adhere to their obligations, possess high
aspiration levels and willingness to work hard to attain goals, and to carry out tasks to
their completion. Hence, because highly conscientious people are hardworking,
achievement-oriented and perseverant, they tend to do what needs to be done in order to
accomplish work

Of the five dimensions of personality, conscientiousness has been identified as the
most consistent predictor of performance, cutting across all occupational groups and all
job-related criteria (Barrick & Mount, 1991; Hough et al., 1990). In addition to
performance, Organ and Ryan’s (1995) meta-analysis on organizational citizenship
behavior (OCB) also revealed a positive relationship between conscientiousness and

OCB, the latter referring to behavior which is discretionary and which goes beyond

38

existing role expectations (Van Dyne, Cummings & Parks, 1995). This finding is also
consistent with the finding that personality is likely to affect the “will-do, ” rather than
the “can-do” aspect of job performance (Borman et al., 1991).

A significant contribution to personality research is the further specification of the
mediating mechanisms between personality traits and the outcome variables. In a study
involving sales representatives, Barrick et al. (1993) found that the positive effects of
conscientiousness on sales volume were mediated through autonomous goal-setting and
goal commitment. Conscientious sales representatives were found to set sales goals on
their own, and were also more committed to these goals. Similarly, Gellatly (1996) found
that the relationship between conscientiousness and performance was mediated by
expectancy and goal level. Specifically, conscientious individuals reported higher levels
of expectancy, which led them to set more difficult goals, which in turn caused higher
task performance.

Both Barrick et al.’s and Gellatly’s findings corroborated with Cropanzano et al.’s
(1992) general fi'amework that higher-order response tendencies (such as personality)
influence specific lower-level goals. For instance, Cropanzano et al. proposed that
individuals high in the response tendency of behavioral regulation focus on the manner in
which their behavior is enacted and therefore, are likely to set goals, monitor them, and
stick closely to their standards. The characteristics of conscientious individuals seem to
correspond to the response tendency that Cropanzano et al.’s described as “high in
behavioral regulation,” thus leading to my next proposition.

Conscientious individuals, due to their dutiful and achievement-oriented nature,
are likely to perceive attaining both individual and group goals to be important, and

hence, should be more committed to both their individual and group goals compared to

39

individuals with low conscientiousness. This is similar to research that has found that
individuals with high need for achievement (Hollenbeck, Klein, O’Leary, & Wright,
1989; Hollenbeck, Williams, & Klein, 1989; Ivancevich & McMahon, 1977) are more
committed to their goals, perhaps due to their inherent desire to excel in tasks assigned to
them.

Besides, conscientious individuals may also posses greater expectancy beliefs
concerning their effort and outcomes, since they generally feel more competent and are
more self-disciplined. Hence, I propose that

Hypothesis 2a: Conscientiousness is positively associated with ggoup goal

commitment.

Hypothesis 2b: Conscientiousness is positively associated with individual goal

commitment.

Agr_eeableness. Agreeableness is a personality dimension that involves
maintaining positive interpersonal relationships, with one end of the continuum
representing individual motives toward selfishness, and the other end representing
concerns for altruism (Digman & Inouye, 1986; Digman & Takemoto-Chock, 1981).
According to Costa and McCrae (1992), agreeableness can be further specified by six
subfacets: the tendency to trust others, to be frank and straightforward, to be altruistic and
selfless, the willingness to cooperate, the lack of arrogance, and the tendency to express
Sympathy (tender-mindedness).

Findings on the relationship between agreeableness and job performance are
mixed. In the meta-analysis by Barrick and Mount (1991), agreeableness was not found
to relate with job performance, even in jobs containing a large social component (e. g.,

sales). However, in a separate meta-analysis, Tett, Jackson, & Rothstein (1991) found

40

that agreeableness had the strongest relationship with job performance. Ones, Mount,
Barrick and Hunter (1994) provided several explanations for the discrepant findings,
including the fewer studies and smaller sample size in Tett et al.’s meta-analysis,
differences in the nature of the jobs and criterion measures examined, and differences in
the process of assigning preexisting scales to personality dimensions.

While the impact of agreeableness on individual job performance may be
equivocal, there is stronger theoretical basis to expect agreeableness to affect
performance in the context of teams. Graziano et al. (1996) proposed that agreeable
people are less competitive by disposition, since their tendency toward interpersonal
harmony induces them to value cooperation. This in turn can bias the way in which
individuals collect and interpret information, such that agreeable people are more likely
to generate positive attributions for others’ behavior. Graziano et al. (1996) conducted an
CXperiment with three-person groups involved in an interdependent task, and found that
the relationship between group agreeableness and group performance was partially
mediated by competitiveness. Members in more agreeable groups perceived less
competition and hence, had better group performance as a result of better coordination.
Likewise, Neuman and Wright (1999) found that individuals’ agreeableness was
positively associated with task performance (e.g., problem solving, planning) and
interpersonal skills (e.g., conflict resolution, communication), as rated by their peers. In
another study, LePine and Van Dyne (2001) hypothesized and found that agreeableness
was positively related to cooperation in a group task context.

These studies suggest that agreeable people, by their cooperative, amiable and
helpful nature, are likely to attach greater values to the achievement of the group goal

than those low in agreeableness. Moreover, since agreeable individuals are more trusting

41

of others’ intentions, they should be less skeptical about their group members’
contribution, which can otherwise reduce the expectancy that their effort will lead to the
achievement of the group goal.

Since agreeableness relates to interpersonal interaction, I propose that it is
relevant in social contexts involving interdependence with others, but not in situations
involving only individual striving. Thus, I expect agreeableness to exert an impact on
group goal commitment but not on individual goal commitment.

Hypothesis 3: Agreeableness is positively associated with group goal

commitment.

To summarize, I have proposed that parallel team members differ in their
individual and group goal commitment as a function of their cultural values and
personality. Specifically, parallel team members with high collectivism (horizontal and
vertical) are likely to be more committed to the group goal than those low in collectivism.
For individual goal commitment, the reverse is true. Conscientious individuals are likely
to be more committed to both individual and group goals than those low in
conscientiousness. Finally, agreeable individuals are likely to be more committed to the
group goal, compared to those low in agreeableness.

Structural Factors

Another category of antecedents proposed in Hollenbeck and Klein’s (1987)
model of goal commitment is structural factors. In this dissertation, I examine two
Structural factors that relate to the design of the parallel team: task interdependence and
team leader status. As explained at the outset of my dissertation, I am restricting my
scope of interest to team-related structural factors and excluding factors that focus more

on the individual task. As such, my next few hypotheses deal with the effects of these

42

team-related factors on group goal commitment and not individual goal commitment,
since these structural factors pertain to the group, rather than to the individual task.

ESLmterdependence. Ki ggundu (1981) defined task interdependence as the
connectedness between jobs such that the performance of one is dependent on the
successful performance of the other. In other words, task interdependence requires group
members to exchange information, resources, coordinate roles, or otherwise perform their
work such that the outcomes of one individual are influenced by the actions of another
(Kiggundu, 1983; Saavedra, Barley & Van Dyne, 1993).

Different work groups face varying degree of task interdependence. Thompson’s
(1967) hierarchy of pooled, sequential and reciprocal methods of coordination describes
an increasing level of dependence among group members. Pooled interdependence refers
to a situation where “each part renders a discrete contribution to the whole,” and hence,
has the least amount of interdependence (p. 54). Sequential interdependence requires one
member to act before another member can act, and hence, has a greater degree of
interdependence. Members with sequential interdependent tasks assume different roles
and perform different parts of the task in a prescribed order. Reciprocal interdependence
refers to the situation in which “the outputs of each become inputs for the others”
(Thompson, 1967 : 55), and has the highest degree of interdependence in Thompson’s
typology. Generally, group members are specialists with different expertise, and group
Performance requires the careful coordination of group members.

Task interdependence is generally thought to enhance cooperation among group
members because of its motivating potential (Campion et al., 1993; Kiggundu, 1983;
Shea & Guzzo, 1987; Steiner, 1972). Since members in an interdependent task perform a

unique role (Saavedra et al., 1993), they are more likely to experience greater

43

responsibility (Kiggundu, 1983; Pearce & Gregersen, 1991; Wageman, 1995) and
indispensability (Kerr & Bruun, 1983; Sheppard, 1993). Using expectancy theory
(Vroom, 1964), Sheppard (1993) argued that increasing task uniqueness in groups can
increase individuals’ perception of the contingency between their personal contributions
and the performance of the collective. This strengthens the expectancy component (i.e.,
the effort-performance relationship), thereby reducing the social loafing phenomenon.

Accordingly, I propose that the degree of interdependence in the group task
should influence group goal commitment positively.

Hypothesis 4: Task interdependence is positively associated with gzoup goal

commitment.

Legn Leader Stm The status system of a group reflects the distribution of
power among its members (Levine & Moreland, 1998) - some members within a group
have higher status than others. Similarly, Barley (1999) referred to status as the relative
power of a person. In this dissertation, I am interested in the status of the team leader vis-
a-vis that of group members, and its impact on group members’ group goal commitment.
I conceptualize status as the leader’s hierarchical position in the organization.

How may the formal status of parallel team leaders affect team members? It is
likely that parallel team members who have a high status leader will perceive their group
as possessing greater organizational legitimacy, compared to those who are led by a low
Status leader. This is because appointing a high status leader for the team suggests that the
team’s mission is important enough to warrant a leader with considerable power and
clout in the organization. Besides, group members under high-status leaders may perceive
that their leaders possess greater reward power, and thus, attach greater values to

attaining the group goal, compared to those led by low-status leaders. Thus, assigning

44

high status leaders to head parallel teams may alleviate motivational problems arising
from members’ perceived lack of organizational support and legitimacy (Cohen &
Lawler, 1992). Accordingly, I propose that
Hypothesis 5: Team leader status is positively associated with ggoup goal
commitment.
Person bv Situation Interactions

In addition to the main effects advanced earlier, I further propose that the
individual differences and structural factors interact with each other to influence goal
comnritrnent. I adopt Mischel’s (1977) situational strength argument to explain the
majority of these interactions. One exception is the hypothesis involving V-C and team
leader status, which I will discuss in greater detail later.

Mischel (1977) proposed that situations can be characterized by their relative
“strength.” Strong situations are situations when there are well recognized and strongly
accepted rules of conduct which constrain and direct behavior. Weak situations, on the
other hand, are ambiguously structured. Mischel argued that situational strength will
moderate the trait-behavior relationship, such that under strong situations, inter-
individual variability in behavior is low, and individual traits have low predictive power.
Conversely, ambiguously structured situations will allow for greater variation in the
interpretation of behavior, and hence, are more conducive for examining the effects of
individual characteristics such as personality.

Following Mischel’s arguments, a general pattern of relationship between
individual differences, situational factors, and the outcome of interest is depicted in
Figure 2. The figure shows that the relationship between individual differences (i.e.,

cultural values and personality) and group goal commitment should be weaker under

45

strong situations than weak situations. In other words, the influence of individual
differences on group goal commitment is constrained by strong situational cues.
However, when the situation is relatively ambiguous, individual differences should play a
more important role in influencing individuals’ group goal commitment.

Figure 2: A figeneral framework depicting Mischel’s situational strengt_h argu_ment

Group Goal
Commitment

f

 

Situational Strength

IIIIIIII Weak

_ Strong

 

 

 

 

Individual
Difference

 

 

The first set of my interactive hypotheses concerns task interdependence. Here, I
am interested in how task interdependence as a situational factor will have different
effects on parallel team members’ group goal commitment, depending on members’
values and personality. Using Mischel’s (1977) situational strength thesis, I argue that
high task interdependence creates a strong situation for group members because it signals
to them that their contribution to the group task is critical and indispensable. As such,
individuals’ cultural values and personality should have less influence on their group goal
Commitment. Conversely, low task interdependence creates a weak situation in which

expectations for contribution to the group are more ambiguous. In this instance, values

46

and personality should play a greater role in influencing individuals’ group goal
commitment.

Applied to collectivism, I expect that group goal commitment should vary less
between individualists and collectivists under high task interdependence, than under low
task interdependence situations. Thus, I propose that in the absence of strong situational
cues to promote contribution to the group task, both horizontal and vertical collectivists
are likely to be more committed to the group goal compared to those low in H-C or V-C.
When strong situational cues are present, group members should be less affected by their
collectivism orientation.

Hypothesis 6a: Task interdependence moderates the relationship between
collectivism (horizontal and vertical) and group goal
commitment, such that the relationship is stronger for
individuals in groups with low task interdependence than for
those with high task interdependence.

Applying this argument to conscientiousness, I propose that high task
interdependence should restrict the effect of conscientiousness on group goal
commitment. This logic is similar to the one adopted by Barrick and Mount (1993), who
found that the positive relationship between conscientiousness and job performance was
greatest for those who had high autonomy (i.e., weak situation) and least for those who
had low autonomy in their jobs (i.e., strong situation). However, instead of job autonomy,
I propose that low task interdependence serves as a weak situation that allows the effects
of conscientiousness to be observed. Since no study has empirically verified this
interaction, the next hypothesis can potentially add insight into existing personality

research.

47

Hypothesis 6b: Task interdependence moderates the relationship between
conscientiouaness and group goal commitment, such that the
relationship is stronger for individuals in groups with low task
interdependence than for those with high task interdependence.

I extend the same reasoning to agreeableness and propose that agreeableness of
parallel team members should have less influence on their group goal commitment in the
high task interdependence, rather than in the low task interdependence condition. This
also mirrors the argument underlying Barrick and Mount’s (1993) finding that the
relationship between agreeableness and job performance was strongest for jobs with high
autonomy and weakest for jobs with low autonomy. Thus, the next hypothesis states that

Hypothesis 6c: Task interdependence moderates the relationship between
agreeableness and group goal commitment, such that the
relationship is stronger for individuals in groups with low task
interdependence than for those with high task interdependence.

The next set of hypotheses focuses on the interactive effects between individual
differences and team leader status. This is premised on the notion that since not everyone
will respond the same way to authority figures, it is important to consider the role of
cultural values and personality in conjunction with the status of the leader.

With regards to collectivism, I propose that team leader status interacts with both
H-C and V-C, albeit in a different manner. For H-C, I apply the same logic used in
Hypotheses 6a-6c. Given that legitimate authority evokes conformity pressure (Cialdini
& Trost, 1998; Michener & Burt, 1975), teams led by a high status leaders are likely to
create a “strong” situation for their members such that individual differences in members’

H-C should have less influence on their behavior. Conversely, teams led by low status

48

leaders provide more discretion for members’ behavior, thus allowing members’ H-C
orientation to play a greater role. Hence,

Hypothesis 7a: Team leader status moderates the relationship between horizontal-
collectivism and group goal comrrritment, such that the
relationship is stronger for individuals in groups with low team
leader status than for those with high team leader status.

For V-C, I propose a different form of interaction. As discussed earlier, the
distinction between the two dimensions of collectivism is the emphasis on inequality
among group members. Unlike horizontal collectivists, vertical collectivists are sensitive
to inequality of status (particularly with regards to social status), and tend to view one
another as different, some having more status than others. An inherent assumption is that
high status individuals possess authority and enjoy benefits that individuals with low
status are not entitled to (Singelis et al. 1995).

Since vertical collectivists are sensitive to status differences, I predict that the
positive relationship between V-C and group goal commitment is influenced by the status
of the team leader. Due to their sensitivity to differences in social status and their
deference to authority (e.g., Ng & Van Dyne, 2001; Singelis et al., 1995), vertical
collectivists are likely to perceive the group goal as considerably more important when
they are led by a leader who has high status than when led by one with little status. On
the contrary, individuals low in V-C are likely to be oblivious to the leader’s status, since
they pay little attention to inequalities in social relationships. Hence, unlike the
situational strength logic employed for earlier hypotheses, I propose in my next
hypothesis that the relationship between V-C and group goal commitment is stronger

when team leader status is high than when team leader status is low (see Figure 3).

49

figure 3: Proposed intergtion between V-C and leader status on group goal commitment

Group Goal
Commitment

4

 

Leader Status

. . . . . Low Leader Status

_ High Leader Status

 

 

 

 

Vertical
’ Collectivism

 

 

The departure from the situational strength argument is premised on the rationale
that individuals with different V-C orientation do not necessarily view high team leader
status as a “strong” situation, since they have divergent opinions regarding power and
authority. Accordingly, the next hypotheses states that

Hypothesis 7b: Team leader status moderates the relationship between m

collectivism and group goal commitment, such that the
relationship is stronger for individuals in groups with high status
team leaders than for those with low status team leaders.

The status of the team leader also has implications for conscientiousness and
agreeableness. As with the situational strength rationale adopted for H-C, I propose that
the effects of conscientiousness and agreeableness on group goal commitment are
Stronger in teams led by low status leaders, than in teams headed by high status leaders.

Hypothesis 7c: Team leader status moderates the relationship between

conscientiouaness and group goal commitment, such that the

50

relationship is stronger for individuals in groups with low
status team leaders than for those with high status team leaders.

Hypothesis 7d: Team leader status moderates the relationship between

agreeableness and group goal commitment, such that the
relationship is stronger for individuals in groups with low
status team leaders than those with high status team leaders.

To recapitulate, I have argued that parallel team members’ commitment to their
individual goal and group goal is determined by (1) individual differences such as values
and personality; (2) the team structure such as task interdependence and team leader
status; and (3) the interaction between the individual differences and structural factors.
C_on_sequences of Goal Commitment

The next central question in this thesis is the impact of goal commitment on
performance outcomes. As described in the literature review of goal commitment in
Chapter 1, goal commitment has been demonstrated to have important implications on
task performance. Here, I argue that parallel team members’ performance in their group
task depends on how committed they are to their respective individual and group goals.
Specifically, I propose that individual and group goal commitment will jointly affect
individuals’ performance in the group task. The interaction is presented in Figure 4 to aid
the ensuing discussion.

Figure 4 shows that performance in the group task is a function of both individual
and group goal commitment. First, the graph depicts a positive relationship between
group goal commitment and performance in the group task. This is consistent with
existing research that has shown that goal commitment has a direct impact on the

performance of the relevant task (e.g, Klein et al., 1999).

51

figure 4: Proposed interaction between individual and group goal commitment on

individual performance in the group task

Individual performance
in group task

A

 

Individual Goal Commitment

IIIIIIII LOW

_ High

 

 

 

 

Group Goal
Commitment

 

 

But more importantly, Figure 4 suggests that the positive relationship between
group goal commitment and performance in the group task is contingent on the degree of
individual goal commitment. When there is low individual goal commitment, the
relationship between group goal commitment and performance in the group task is
expected to be stronger. On the other hand, if individual goal commitment is high, the
relationship between group goal comrrritrnent and performance in the group task will be
attenuated.

This argument is based on the fact that individuals have a finite amount of
resources (Naylor et al., 1980) and hence, are restricted in the total amount of time and
effort they can supply. Therefore, even if individuals are motivated to perform a task, the
actual amount of effort they can put into that task is constrained by the amount of effort
they need to contribute to other tasks. To the extent that the individual is equally

committed to several goals, s/he is constrained by how much s/he can contribute to any

52

one task. On the other hand, if the individual is highly committed to one goal and not to
the other, s/he will be willing to work hard on one task at the expense of the other.

Figure 4 reflects the above argument. When individuals are committed to both the
individual and group goals, performance in the group task is suppressed by the fact that
resources also need to be channeled to the individual task. Hence, group goal
commitment has a weaker relationship with performance in the group task because
resources are also being channeled to the individual task. By contrast, the link between
group goal comrrritrrrent and performance in the group task will be stronger when
individuals are less committed to the individual goal because in this case, there are more
resources available for individuals with high group goal commitment to dedicate these
resources to the group task.

Hypothesis 8: The positive relationship between group goal comrrritrnent and
individual performance in the group task is moderated by
individual goal commitment, such that the relationship is stronger
under low individual goal commitment than under high individual
goal commitment.

Thus far, I have developed theoretical arguments for the effects of both distal
(cultural values and personality) and proximal factors (goal commitment) on individuals’
performance in the group task in the presence of competing demands from their
individual tasks. In the next chapter, I present the methodology used to test the

hypotheses advanced in this chapter.

53

CHAPTER 3

METHOD

Setting

I simulated a parallel team environment in a laboratory setting to test my
hypotheses. As Kerlinger (1986: 369) pointed out, the primary aim of laboratory
experiments is to “test hypotheses derived fi'om theory, to study the precise interrelations
of variables and their operation, and to control variances under research conditions that
are uncontaminated by the operation of extraneous variables.” Hence, I chose an
experimental design to test my theory because of two reasons. First, laboratory
experiments can eliminate extraneous influences that may contaminate or confound the
independent and/or dependent variables through the random assignment of subjects.
Second, laboratory experiments provide a means to assess for evidence of causal, and not
merely correlational, relationships among the variables that the study examines.

Power Analysis

The power of a statistical test is the probability that it will yield statistically
significant results (Cohen, 1987). For multiple regression analyses, Cohen (1987; Cohen
& Cohen, 1983) recommended using the f 2 statistic as an index of effect size to
determine the sample size required by a particular study. The f 2 statistic is defined as the
proportion of systematic variance explained by a set of predictors, over and above what is
accounted for by another set of predictors (see Equation 4.5.1. in Cohen & Cohen, 1983:
155)

I based the computation of my expected f 2 statistic on studies with similar

research questions and design. Two examples are Barley’s studies on social loafing

54

(1989, 1993). Both studies examined the main and interaction effects of collectivism and
structural features of the group (e.g., shared responsibility), used a pancultural sampling
strategy involving Chinese and US. subjects, adopted an experimental design, and did
not involve actual interaction among group members during performance of the group
task.

According to the R2 reported in Barley’s studies, the f 2 that can be expected from
my study is .04 — an effect size that Cohen (1987) classified as small. Setting my
significance criterion to be .05, and desired power at .80, the sample size required for my
study was 310 (see Equation 4.5.2 in Cohen & Cohen, 1983: 155).

Participants

I employed a pan-cultural sampling strategy and recruited subjects from an
individualistic and a collectivistic country to avoid the statistical problem of restrictive
ranges on each cultural dimension within each nation (e. g., Barley, 1989; 1993; Leung &
Bond, 1989). According to Hofstede’s (1984) study, the US. is an individualistic society,
while Singapore, being an Asian culture, is more collectivistic. Participants were
undergraduate students enrolled in a management course in two large universities in the
US. and in Singapore. In both countries, the experiment was incorporated into the course
curriculum, and eight cash awards were awarded to top performers in each country.

Based on the power analysis reported above, I recruited a total of 324 participants,
but used only 318 in my data analyses. The responses of four participants from the U.S.,
and 2 from Singapore were excluded because they appeared not to have understood the
instructions of the experiment. For instance, some of these students thought they had 25
minutes to work on each set of task; others thought they needed to choose to work on one

set of task only.

55

38% of the 318 participants were males (U.S. = 73; Singapore = 48), and 62%
were females (U.S. = 85, Singapore = 112). There were significantly more female
participants from Singapore than from the US. (x2 mm) = 8.85, p < .01). However, this
should not bias my analyses given that (1) I use individual level measures of cultural
values (hence, country is not a substantive variable of interest), and (2) gender is a
control variable in all my analyses.

Exppriment

Participants were told to assume the role of a junior manager in a large
manufacturing firm. Participants were told that in addition to their regular job duties as a
manager, they also belong to a team that manages college recruitment of entry-level staff.
Participants’ overall performance was determined by their performance in their individual
task as well as their group’s performance (i.e., everyone in the group will receive a same
score for the group’s performance).

Ias_k_s. Participants were given twenty-five minutes to work on two sets of tasks:
an individual task (i.e., regular job duties as a junior manager) and a group task (i.e.,
duties as a member of the recruitment team). Both tasks were comparable in terms of
complexity. I describe them in greater detail below.

The individual task was adapted from Saavedra et al.’s (1993) merit bonus task.
Participants were told that one of their regular job functions was to conduct annual
performance appraisal of their subordinates. The steps for evaluating a subordinate
involved (1) reading a short description of the subordinate; (2) rating him or her on four
dimensions (effort, ability, performance and team-orientation) using a 3-point scheme
(low, average, high); (3) assigning weights to each dimension based on the division of the

subordinate (a weighting scheme was provided); (4) computing a merit increase for the

56

subordinate based on the weighted scores; (5) making a merit increase recommendation
(low, moderate or high). Thus, there was a correct and unique numerical solution for
every subordinate, which served as the performance index for participants in their
individual tasks.

The group task required participants to make hiring recommendations for job
applicants and involved similar steps as the individual task in arriving at the correct
solution. Participants were asked to read descriptions of job applicants and rate them on
four dimensions (academic performance, confidence, understanding of job nature, and
team-orientation) using a 3-point scheme (low, average, high). As with the individual
task, each division had different weighting schemes for job applicants. Participants had to
go through the 5 steps outlined above to arrive at a decision of whether to hire, reject, or
keep the applicant temporarily on hold.

Manipulations. The experiment was a 2 (low vs high group task interdependence)
X 2 (low vs high team leader status) fully crossed factorial design. Participants in all
conditions were assigned goals for their individual and group tasks.

Task interdependence was manipulated by leading participants to either believe
that they could process an entire job application without input from other group members,
or that they needed input fiom others in order to complete a job application. In actuality,
the nature of the task was the same in both task interdependence conditions. In the high
t_as_l_<_iaterdependence condition. participants were told that there were three sequential
parts to the job applicant evaluation task, and that each group member could complete
only one part of the task. Specifically, the three components of the task were: (1)
preparing job applicant’s evaluation form for rating, (2) rating job applicant, (3)

preparing an official letter of recommendation to top management. In actual fact, all

57

participants in this condition were given only the second step of the task to complete (i.e.,
rating job applicant). To enhance the reality of the interdependent nature of the task,
participants received evaluation forms that contained handwritten information of job
applicants so that it appeared as though another group member had already completed the
first part of the group task. In addition, the research assistant routinely collected
participants’ completed forms in the guise of passing these forms to the “next group
member” for “him/her” to complete the third step of the task.

In the low task interdependence condition. participants were told that they could
complete the entire evaluation of the job applicant by themselves and that no input from
other group members was needed.

Team leader status was manipulated by informing participants of their
hierarchical position in the firm vis-a-vis the team leader’s both verbally and through an
organizational chart (see Appendix A). In the ugh team leader statas condition,
participants were told that the idea of a recruiting team staffed by junior managers from
the various divisions originated from the vice-president. As a result, the vice-president
was given the mandate by top management to head the team and to handle the current
campus recruitment drive. In the low team leager stalua condition, participants were told
that the idea of the recruiting team came fi'om a junior manager in the HR division.
Subsequently, he was asked by top management to head the team and handle the campus
recruitment drive.

Procedure. There were two parts to the study. First, participants took part in a 75-
minute experiment on campus. Second, about one week later, they completed two
questionnaires measuring cultural values and personality.

The experiment was conducted in a classroom on campus. Every participant

58

received a package containing the following materials as s/he arrived.

1. Consent form (see Appendix A).

2. A one-page write-up on the role to be assumed in a hypothetical company in this
experiment and an organizational chart (see Appendix A).

3. A package containing reference and practice materials for the individual and group
tasks.

Every session was conducted with three groups, each comprising 6 participants.
Each person in the group was randomly assigned the role of a junior manager of a
division (there were 6 divisions altogether). After signing the consent form, participants
read a one-page summary describing the background of the hypothetical company and
their roles as a junior manager and a member of a campus recruitment team. An
organizational chart was appended to the background information and participants were
told to familiarize themselves with the hierarchy of the organization. After participants
read their roles, I described to them the two tasks that they needed to complete during the
experiment. One was the subordinate evaluation task assigned by their immediate
supervisor in the division (i.e., individual task), and the other was the job applicant
evaluation task assigned by the recruitment team leader (i.e., the group task).

To ensure that participants understood their tasks, I trained them with several
practice samples of subordinate and applicant evaluations. The training and practice
lasted for about 15 minutes. This was followed by a short “ice-breaking” session for
participants belonging to the same group to get to know one another.

M Before timing the actual performance, two sets of goals were assigned.
Participants were told that the goal for the individual task was to complete all 20

subordinate evaluations within the 25 minutes.

59

For the group task, a total group goal (to be achieved by the entire group) and an
individual goal (to be achieved by individual group members) were assigned. While the
goal for individual group members was held constant at 20 for the entire sample, the total
group goal varied according to the task interdependence condition. Specifically, the total
group goal was 40 for the high task interdependence condition and 120 for the low task
interdependence condition. This discrepancy was necessary so that the total group goal
(at the group level), when translated to the group goal for the individual member, was 20.
These figures were derived using the following logic. For the high task interdependence
condition, since participants were told that a complete evaluation required the input of
three persons, and given that there were six members in the group (each person to
complete 20 evaluations), the total group goal was 2X20 = 40. In the low task
interdependence condition, since members could complete one evaluation by him/herself,
the total group task was 6X20 = 120. Thus, in actuality, the goals at the individual level
were identical across all the conditions.

When the goals were assigned, participants filled up a questionnaire that asked for
their individual goal commitment, group goal commitment, and group goal priority. Once
everyone had completed the questionnaire, I distributed the actual tasks and started
timing for 25 minutes. In the high task interdependence condition, the RA and I went
round to collect completed applicant evaluation forms in order to make participants
believe that they were involved in an interdependent group task.

When the 25 minutes were up, participants were told to stop working on the tasks
and to fill up a short questionnaire that included manipulation checks for task
interdependence and team leader status.

Measures

60

Individualism-Collectivism. Given the proliferation of I-C scales in the field, a
secondary objective of this dissertation is to compare these various measures for their
content and empirical properties. To keep the length of the I-C survey manageable, I
used three existing scales that have received considerable attention in organizational
research. The first was the original 32-item horizontal-vertical I-C scale developed by
Singelis et al. (1995). This was the first scale that attempted to make the distinction
between the horizontal and vertical dimensions of l-C. Specifically, H-C was measured
with 8 items (e.g., “If a co-worker gets a prize, I would feel proud;” “The well-being of
my co-workers is important to me”), V-C with 8 items (e.g., “Parents and children must
stay together as much as possible”), H-I with 8 items (“I often do my own thing;” “One
should live one’s life independently of others”), and V-I with 8 items (“Winning is
everything;” “It annoys me when other people perform better than I do”).

The second scale (Triandis & Gelfand, 1998) was a shorter version of Singelis et
al.’s horizontal-vertical I-C scale in which four out of the 16 items were not found in the
original scale. One item tapped at V-C (“It is important to me that I respect the decisions
made by my groups”) and the other 3 items tapped at H-I (“I’d rather depend on myself
than others;” “I rely on myself most of the time; I rarely rely on others;” “My personal
identity, independent of others, is very important to me”).

The third scale was a 20-item scale reported in Wagner’s (1995) study, which was
in turn constructed from existing measures from Wagner and Moch (1986; 9 items), Erez
and Barley (1987; 2 items), and Triandis and colleagues (Triandis, Bontempo, Villareal,
Asai, & Lucca, 1988; Hui, 1988; 9 items). The factor analysis reported in Wagner’s
(1995) study indicated that a 5-factor structure was appropriate. The five factors

described (1) personal independence and self-reliance; (2) importance accorded to

61

competitive success; (3) value attached to working alone; (4) norms about subordination
of personal needs to group interests; and (5) beliefs about the effects of personal pursuits
on group productivity.

In comparing the content of these various instruments, it appears as though there
is some convergence between Wagner’s (1995) 5 factors of I-C and the horizontal-
vertical I-C conceptualization. For instance, the first factor describing personal
independence, and the third factor describing preference to work alone, are similar to the
concept of H-I — the dimension of LC that characterizes people as self-reliant, but who
are not particularly interested to distinguish themselves from others (Triandis & Gelfand,
1998). Wagner’s (1995) second factor, competitive success, captures the definition of V-I
- the dimension that describes people who want to be distinguished and to acquire status
(Triandis & Gelfand, 1998). The fourth and the fifth factors which described norms and
beliefs about subordinating one’s personal needs to the group’s interest (Wagner &
Moch, 1986) are similar to V-C, the dimension that describes people who emphasize the
integrity of the group, and are willing to sacrifice their personal goals for the sake of the
group.

Since this dissertation purports to examine H-C and V-C, I used all the H-C and
V-C items in Singelis et al.’s (1995) and Triandis and Gelfand’s (1998) scales. 1 also
included items from the two factors reported in Wagner’s (1995; Wagner & Moch, 1986)
study because they appeared to capture elements of V-C. My objective is to assess if they
overlap with existing items by Singelis et al. (1995) and Triandis and Gelfand (1998) to
tap at the latent construct of V-C.

To recapitulate, H-C was measured with 8 items from the original Singelis et al.

(1995) scale, while V-C was measured with items from three different scales: 8 items

62

fiom Singelis et al. (1995), 1 item from Triandis & Gelfand (1998), 4 items on norms
concerning subordinating personal interests to group interests from Wagner & Moch
(1986), and 3 items on beliefs about the effects of pursuing personal interests on group
productivity from Wagner & Moch (1986). Hence, a total of 24 items were examined for
H-C and V-C.

A complete list of the items for I-C is presented in Appendix B.

ConacientiousnesL and Aggeeableness. The Personal Characteristics Inventory
(PCI) developed by Barrick and Mount (1995) was used to assess participants’
conscientiousness and agreeableness. The PCI has been administered to over 2,000
individuals, including students, managers, sales representatives, retail clerks, and
production workers. The factor structure of the PCI replicates the big five dimensions
established by other instruments, such as the NBO P-I-R (Costa & McCrae, 1992). I
chose to administer the PCI since it comprehensively measures the five personality
constructs with considerably fewer items, while maintaining reasonable alpha and test-
retest reliabilities (alphas for the Big Five primary scales ranged from .82 to .87).

Conscientiousness in the PCI comprises three subscales: dependability,
achievement striving and efficiency, while agreeableness comprises two subscales:
cooperation and consideration. Each subscale consists of 10 items. Since my hypotheses
involving personality were made with the broad dimensions of conscientiousness and
agreeableness, my analyses were conducted with the general personality constructs rather
than with their sub—dimensions.

Goal Commitment. The measures of group goal commitment and individual goal
commitment were based on the 9-item goal commitment scale developed by Hollenbeck

et al. (1989). Although other operationalizations of goal commitment can be found in

63

existing literature [e.g., single-item self-reported measures (e.g., Erez & Zidon, 1984;
Yukl & Latham, 1978) and the discrepancy measure of self-set versus assigned-goal (e. g.,
Tubbs, 1993; Tubbs & Dahl, 1991)], Hollenbeck et al.’s (1989) scale widely-used and
well-validated. For instance, Klein, Wesson, Hollenbeck, Wright and Deshon (in press)
tested the measurement properties of the 9-item scale using a combination of meta-
analytic and structural equation modeling techniques. This novel method combined the
individual strengths of meta-analysis and structural equation modeling to provide a robust
assessment of the goal commitment measure. Results of the study suggested that a subset
of the full scale, consisting of five items, best reflected a unidimensional goal
commitment construct. In addition, this sub-scale was found to hold up well under
different situations, such as measurement timing (i.e., whether goal commitment was
measured before, during/after the task), goal origin (self-set versus assigned goal), and
task complexity.

In my study, consistent with prior research, I used the full 9-item scale and
modified items to tailor them to the group goal and the individual goal respectively.
Hence, group goal commitment consisted of 9 items such as “I am strongly committed to
pursuing the goal for my group task;” “It’s hard to take the goal for my group task
seriously (R);” “Quite frankly, I don’t care if I achieve the goal for my group task or not
(R).” Likewise, individual goal commitment consisted of 9 items that referred to the
individual task, such as “I am strongly committed to pursuing the goal for my individual
task;” “It’s hard to take the goal for my individual task seriously (R);” “Quite frankly, I
don’t care ifI achieve the goal for my individual task or not (R).” Appendix B presents a
complete list of the 18 items measuring both group and individual goal comnritrnent.

Group Goal Priorig. In addition to the goal commitment scales described above, I

64

also included a four-item group goal priority scale that asked subjects how committed
they were to achieving the group goal vis-a-vis the individual goal (e.g., “achieving the
goal for my individual task is a greater priority than achieving the goal for my group task
(R)”; “it means more to me that I achieve the goal for my individual task than to achieve
the goal for my group task”). Thus, this scale requires respondents to prioritize the
individual and group goals.

Including the group priority scale in my study serves two purposes. First, given
that no prior study has examined two parallel set of goal commitment items for an
individual and group task, little is known about whether the two goal commitment
constructs can be discriminated when used concurrently. Hence, the group priority scale
is another approach for assessing commitment when there are multiple tasks and multiple
goals. Second, including the group goal priority scale allows me to assess the
relationships of individual and group goal commitment with another relevant construct,
thereby providing additional evidence for their construct validity. In this case, I expect
group goal commitment to be positively, while individual goal commitment to be
negatively, related to group goal priority.

Pedomce Outcomes. The primary dependent outcome is individuals’
performance in their group task. This is the total number of correct job applicant
evaluations made by each participant. A correct job applicant evaluation includes having
accurate scores for all dimensions of the evaluation as well as making the right hiring
decision. An individual’s performance in the group task ranged from 0 to 20.

The dual-task nature of parallel teams suggests that another important outcome,
even though it is not explicitly proposed in my hypotheses, is individuals’ performance in

their individual task. This is the total number of correct employee evaluations made by

65

each participant. A correct employee evaluation includes having accurate scores for all
dimensions of the evaluation as well as making the right merit increase recommendation.
As with individuals’ performance in the group task, individuals’ performance in the
individual task ranged from 0 to 20.

Control variable. I control for gender in all my analyses, since it is possible that
females, being traditionally more nurturing and less competitive, may view group goal
differently from their male counterparts (Bagly, 1987; Bagly and Wood, 1991). Male is
coded 1, and female is coded 2.

Manipulation checks. Manipulation check for task interdependence was assessed
with four items adapted from Pearce and Gregerson’s (1991) task interdependence scale
(“I need to obtain information/materials from other group members to complete my
portion of the group task;” “I can complete my portion of the group task without any
contribution from my group members (R);” “My own performance in the group task is
dependent on receiving accurate input from other group members;” “The way I perform
my group task has a significant impact on other group members”). Results of a pilot
(n=111) demonstrated that perceived task interdependence was significantly higher in the
high task interdependence condition (mean = 4.03) than in the low task interdependence
condition (mean = 2.07; p< .01). The coefficient alpha for the scale was .82.

Manipulation check for team leader status was assessed with six items such as
“My team leader has higher status in the organization than me;” “My team leader has
greater organizational influence over others than me;” “My team leader has greater power
in the organization than me.” Pilot subjects in the high leader status condition reported
higher levels of perceived leader status (mean = 3.87) than those in the low leader status

condition (mean = 2.35; p< .01 ). The coefficient alpha for the scale was .94.

66

Data Analyses

Before testing my hypotheses, I used a combination of exploratory and
confirmatory factor analysis techniques to assess the validity and cultural equivalence of
the constructs in my model. My general approach to creating variables for testing my
substantive relationships is as follows. First, I performed a principal component factor
analysis (with varimax rotation) of items measuring the personality, values, and goal
commitment constructs. Items that did not load on to the anticipated factor, or those
which demonstrated high cross-loadings (i.e., loadings of .40 or more, and where the
difference with items from other factors is less than .30), were discarded. I then
conducted confirmatory factor analyses on the surviving items to assess the goodness-of-
fit, as well as the measurement equivalence of the construct between the US. and
Singapore samples.

The test of measurement invariance is to establish “whether or not, under different
conditions of observing and studying phenomena, measurement operations yield
measures of the same attribute” (Horn & McArdle, 1992). Hence, establishing
measurement equivalence is a prerequisite for any substantive conclusions to be drawn
from cross-cultural studies. In this dissertation, perhaps the most fundamental and critical
form of measurement equivalence to establish is configural invariance — whether the
pattern of salient (nonzero) and nonsalient (zero or near zero) factor loadings is similar
across the US. and Singapore samples (Steenkamp & Baumgartner, 1998). A lack of
configural invariance would suggest that a construct is operationalized differently in the
two countries, rendering it impossible to create a common scale for the entire pool of
subj ects. To test for configural invariance, I conducted separate confirmatory factor

analyses for the US. and Singapore samples, and compared their significant versus non-

67

significant paths. Only items that had significant factor loadings in both countries were
used to create scales for my subsequent data analyses.

To test the substantive relationships proposed in my dissertation, I performed
hierarchical ordinary least squares regressions in all my statistical analyses. The general
analytical strategy involved entering the control variable (gender) in the first block,
followed by main effects, followed by product terms when interactions were tested. The
statistical significance of the block of variables entered in each step was assessed with an
F-test of the change in R-square. Within significant block of variables, I examined the
statistical significance of each variable with a t-statistic.

To test whether I-C values and personality influenced group and individual goal
commitment (Hl — H3), I conducted two separate sets of regressions, one for group goal
commitment, and the other for individual goal commitment. In each set, I entered gender
in the first step, and the values and personality in the second step. To test whether
structural factors influenced group goal commitment (H4 — H5), I regressed group goal
commitment onto the dummy codes representing task interdependence and leader status.
To test whether structural factors moderated the relationships between individual
differences and group goal commitment (H6 -H7), I conducted two separate sets of
moderated regression analyses: one for task interdependence (H6a —- H6c) and one for
team leader status (Hypotheses 7a-7d). In each of these two regressions, gender was
entered in the first step, individual differences in the second step, structural variable in
the third step, and the corresponding product terms (individual difference variable X
structural variable) in the last step.

To test for the joint effects of individual and group goal commitment on

performance in the group task (H8), I entered gender in the first step, individual and

68

group goal commitment in the second step, and the product term of individual and group

goal commitment in the final step.

69

CHAPTER 4

RESULTS

Measures

H-C and V-C. A principal components factor analysis (with varimax rotation) of
the 24 items revealed 6 factors with eigen values greater than 1 which explained 57% of
the total variance. Six items exhibited high cross loadings and were discarded (4 V-C
items from Singelis et al., 1 V-C item fi'om Triandis & Gelfand; l H-C item from Singelis
et al.). Another principal components analysis with the remaining 18 items demonstrated
4 factors which explained 57% of the total variance. The first factor extracted consisted
of 7 H-C items from Singelis et al. (1995) [eigenfactor value = 4.72]; the second factor
consisted of the 4 norm items from Wagner and Moch (1986) [eigenfactor value = 2.18];
the third factor consisted of 4 V-C items from Singelis et al. (1995) [eigenfactor value =
1.88]; and the fourth factor consisted of the 3 belief items from Wagner & Moch
[eigenfactor value = 1.43].

A CF A of these 18 items revealed a reasonably good fit for the four-factor
structure (RMSBA = .05; CF I = .93; GFI = .91) and all factor loadings were significant at
the p=.05 level. In addition, compared to this four-factor structure, an alternative one-
factor structure (I-C only; RMSBA = .13; CFI = .57; GFI = .72; mam.) = 542, p< .01) or
an alternative two-factor structure (H-C and V-C only; RMSBA = .10; CFI = .77; GF I =
.82; Afar.” = 247.58, p< .01) demonstrated significantly poorer fit. I also tested for a
two-factor structure (H-C and V-C) where V-C is a second-order construct represented by
three lower-level constructs: the Singelis et al.’s V-C scale, and Wagner and Moch’s 2

factors. This model was not significantly better than the initial four-factor model (Ax2(df=2)

70

= 2.58, p> .05).

Taken together, these results seem to refute my apriori expectation that the two
factors in Wagner and Moch’s (1986) scale dealing with submitting personal interests to
the group interest would collapse with Singelis et al.’s (1995) V-C items to form one V-C
factor. One explanation for this could be attributed to the context in which the items were
constructed - Singelis et al’s items involved family relationships and obligations, while
Wagner and Moch’s items were concerned with work group relationships and
obligations. Further, results also demonstrated that the values-belief distinction of the two
subdimensions in Wagner and Moch’s scale was robust in the current sample.

For my substantive analyses, 1 adhered to the original V-C items developed by
Singelis et al. (1995). Since results demonstrated that Wagner and Moch’s (1986) items
could not be combined with the existing V-C measures, and given that these items were
not written apriori to measure V-C, I exclude them from the rest of the analyses. Hence,
two composites were formed for my analyses: H-C consisting of 7 items (or = .83), and
V-C consisting of 4 items from the Singelis scale (or = .65). A within-culture CFA also
showed that all the 11 items had significant loadings with their respective factors, thus
establishing configural invariance (i.e., the pattern of salient factor loadings is the same)
for the H-C and V-C measures across the US. and Singapore subjects.

The specific items are presented in Table l.

mcientiousness and Agreeableness. While the PC] is an established personality
instrument in the U.S., little is known about its generalizability to Singapore subjects.
Hence, I conducted a within-culture principal components analysis (with varimax
rotation) to assess whether the 50 items load correctly onto conscientiousness and

agreeableness in the two countries. I specified two factors since I am interested in the

71

 

 

 

a «Mm — Hnsux

N0. EU

.3. EU<

m0. Emu

m5. <mm2~H

ca. 8 H0 26.8% 008 20 EH88 >8 .HH £008 b0.» >080 H 85 53300 8 00mt00m 2:03 H .2
we. $8600 85 080800 H HH 80>0 $288 H8 0808 2:03 8:3 0H0 2:03 H .2
00. 00.8.00“ 0.3 0.08808 :83 00808 00 805080 x000 200%. 800808 HHH80HH .0
mm. 0.0808 mm £098 00 805080 H80 088 82220 05 08083 .w
R. .000: 8 08 >05 80:3 800808 8.00» H8 20: 2 0000 H8 .3 H .H.

on. .8008 H8 85H? 80: 2003 H .bHsomcHH. 8 0003 000808 80% 0 .HH .0

me. .0008 0103 H8 855 808.8: 88808 00 08 80H 85.38 mm H .m

me. 8080 830 880800 H 8085 000w H00H H .v

mm. .8050 at? 085 8.580% 8 0.8803 .08 0H. .m

mm. .08 00 8.888 mm 80010300 H8 H0 8:09:03 05. .N

mm. .0008 H00H 2003 H .088 a 30w 08:03-00 0 H .H

U-> 0-: 0.80:

 

 

a8.8832 U-> 084“ 0-3 SH 8038?. 88mm CSMEESU .H 2an

72

general dimensions of conscientiousness and agreeableness, rather than with their sub-
facets. Results showed that 2 conscientiousness items and 5 agreeableness items loaded
differently onto the factors across the two cultures. After removing the 7 items, I
conducted a CFA on the remaining 43 items using the entire sample. Given the large
number of items, I averaged items with similar factor loadings (based on EFA results) to
form composite variables before analyzing them in the CFA (Yuan, Bentler, & Kano,
1997). Specifically, 7 composite variables for conscientiousness (with 4
conscientiousness items each), and 5 composite variables for agreeableness (with 3
agreeableness items each) were formed.

Results of the CF A demonstrated an acceptable fit for a 2-factor structure
(RMSBA = .08; CFI = .93; GP] = .92). Moreover, an alternative one-factor structure
yielded a significantly worse fit (RMSBA = .16; CFI = .73; GFI = .74; AX2(df=l) = 263.08,
p< .01). Finally, a within-culture CFA demonstrated that all items had significant
loadings with their respective factors, thus demonstrating configural invariance of the two
personality measures.

Hence, conscientiousness was made up of 28 items (on = .89) and agreeableness

was made up of 15 items (a = .79). Scores for these range from 1 (low) to 3(high).

Goal commitment. A principal components factor analysis showed that the 9
individual goal commitment items and their 9 counterpart group goal commitment items
exhibited high cross-loadings and did not always load correctly onto their respective
factors. A CF A specifying a two-factor structure confirmed that the error terms of the
individual goal commitment items were highly correlated with their counterpart group
goal commitment items. For instance, a two-factor CF A that did not allow for correlated

errors of similarly worded individual and group goal commitment items yielded a poor fit

73

(RMSEA = .14; CF I = .61; GF I = .76). When the errors for individual goal commitment
items were allowed to correlate with the errors of their counterpart group goal
commitment items, the fit indices improved significantly (RMSBA = .06; CFI = .92; GFI
= .91; Ax2(df=9) = 620.57, p< .01).

In order to minimize the problem of correlated errors, I selected 4 pairs of
individual goal and group goal commitment items that had the least correlated errors
between them. A CFA of these 8 items, specifying 2 factors, suggested that one pair be
dropped to improve the fit. The resulting 3 pairs of goal commitment items were (1) I am
strongly committed to pursuing the goal for my individual (group) task; (2) Quite fiankly,
I don’t care if I achieve the goal for my individual (group) task or not w); (3) It wouldn’t
take much for me to abandon the goal for my individual (group) task (R). The CFA result
of these 6 items demonstrated a weak fit for a 2-factor structure (i.e., individual goal
commitment and group goal commitment) (RMSBA = .11; CFI == .90; GFI = .96). The
alternative l-factor structure demonstrated a significantly worse fit (RMSBA = .12, CFI =
.89; GFI = .95; made” = 4.52, p< .05).

However, a within-culture CFA of the 6 items revealed that the 2-factor structure
provided a significantly better fit than the l-factor structure in the Singapore sample
(612%,): 6.97 p< .01), but not in the us. sample (AX2(df=l) = 13, 12> .05). A multigroup
SEM analysis confirmed that the covariance between individual and group goal
commitment differed for the two countries, since releasing the constraint of covariance
invariance improved the fit significantly (Ax2(df=1) = 12.42, p_< .01). Taken together, these
results imply that the US. subjects were unable to discriminate between individual goal
commitment and group goal commitment. All factor loadings were significant in both
samples.

74

 

"‘Illl

\

At

I‘QIQIH. -.

.3832 888 HooNHHVBHESm 8m $588 2an a

 

2.3.3. lawman
ca. HmU
ea. EU<
ea. 50
HH. <mmH>H~H

mo. .SHV VHmS .3on >8 88 Row 05 885% 9 08 .88 50:8 8.3 r8283 «H .o
Nm. .VHmS 3on >8 .88 How 05 $8.86 8 38888 >Hwaobm 8a H .m
vs. .9: 8: 8 VH3 9.8m >8 8H How 05 0538 H .HH 8.3 :86 H 588% 880 .v
S. .9: “as 3323 as 5 new 05 83% 9 ea .3 5:8 83 8.5303 a .m
cm. .va8 H8232: >8 .88 How 05 $8.88 8 38888 >Hwaobm 8m H .N

on. .90 Ho: Ho on8 H8232: >8 8H How 05 2628 H .3 28 «.826 H >393 880 .H

 

”H8890 388:0
H80 5.9.0 H30 8: an:

 

 

 

“8.8862 80828800 H80 9:80 H88 :32ng 88 mag 83on 388888.800 .m 033.

75

.mquHoon 58am 3833887. can $8088 033. a

 

 

 

2%.: fix
ca. EU
ea. EU<
3. any
ac. <mmH>H~H

gas Eases E 5 8
co. 3 26: H 85 x83 «outoam :3, H JHmS anew >8 8H How 05 2638 8 .580 8H .3
.onS 98w >8 8H Row 08 2628
on. 8 55 x93 35368 >8 8H Hmow 05 26.83 H 35 08 9 208 838 HH 6
.vaS H3238:
on. >8 8H 82: V7.8 anew >8 8H How 2: 0528 H 38 83.588 0.88 8 H .w
.va8 ~5on >8 8H Hmow
am. 05 war/038 .85 >880th 539% m E onS 33388 >8 5H Row 08 m8>o£o< .H.
G. .QHV vaS 8.on >8 8H Bow 2: 88838 8 08 SH 558 8H8 #82503 HH .0
mm. .vaS 8on >8 8H How 05 8885 8 38888 >Hmaobm 85 H .m
E. .QHV Ho: 8 onS 9.on >8 8H How 05 0523 H HH 28 rag H >23 880 .v
mm. .8 one @6365 E E in ca conga 9 2: é :88 83 races» a .m
mm. .on8 H3329: >8 5H Hwow 05 9.888 3 33888 >Hw8obm :8 H .N
S. .9; an s as 32%.: E he wow 05 0528 :8 28 CE _ .9253 380 .8
$.85 .835 .855
H30 H30 H30
macaw 9.26 E: 28.:

 

 

 

«krona HmoG 38.5 93 808H88o0 H80 :80 8288800 H80 326385 SH mHm>Hmaq< 88mm 30388980 .m 038.

76

For the purpose of adhering to my hypothesized model, I created a 3-item
individual goal commitment scale and a 3-item group goal commitment scale using the 6
items described above. The internal consistencies of individual goal commitment and
group goal commitment were .55 and .66 respectively. Table 2 presents the CF A results
for the 6 items.

Group goal priorig. A principal components factor analysis revealed that all the 4
items loaded onto one factor which explained 51% of the total variance. A CFA with
group goal priority, individual goal commitment and group goal commitment items
specified with a 3-factor structure yielded an acceptable fit (RMSBA = .08; CFI = .90;
GFI = .94). An alternative l-factor structure yielded a significantly worse fit (Ax2(df=3) =
229.15, p< .01). Table 3 presents the CFA results.

Within-culture CF As demonstrated that the 3-factor structure was robust in both
the US. and Singapore samples, and all item-factor loadings were significant. The
internal consistency for the four-item group goal priority scale was .68.

Further, group goal priority appeared to have some evidence of convergent and
discriminant validity. As expected, it was positively related to group goal commitment (r
= .29, p< .01) and negatively related to individual goal commitment (r = -.14, p< .05).
Likewise, it was positively related to individual performance in the group task (r= .27, p<
.01), and negatively related to performance in the individual task (r = -.29, p< .01).
These relationships were all consistent with the definition of group goal priority, thus
providing evidence supporting construct validity.

Finally, a principal components factor analysis (with varimax rotation) of items
measuring H-C, V-C, the composite variables of conscientiousness and agreeableness,

group and individual goal commitment, and group goal priority, reveal 6 factors (with

77

 

eigen factor values greater than 1) which explain 56% of the total variance. Specifically,
all the items measuring H-C, V-C, conscientiousness, agreeableness and group goal
priority loaded correctly to their respective factors, thus demonstrating discriminant
validity for these constructs. Group goal commitment and individual goal commitment
items, however, loaded onto the same factor, once again raising caution for subsequent
results and interpretations involving these constructs.

Descriptive Statistics

Table 4 presents the descriptives, Cronbach’s alphas and inter-item correlations of
all the variables examined in this dissertation. Variables that are not in my formal
hypotheses, but examined in my post-hoc analyses, are also included. They are individual
task performance, group goal priority, country in which the experiment was conducted
(1=U.S., 2=Singapore), and the two manipulation checks for task interdependence and
leader status (perceived task interdependence and perceived leader status).

Several relationships are worth highlighting. First, as anticipated, individual task
performance is negatively correlated with individual performance in the group task (r = -
.56, p< .00). Nonetheless, the magnitude of the trade-off between individual and group
performance is not as big as I initially expected, suggesting that other third measures
(such as ability) may have influenced performance outcomes.

Second, the correlation between individual goal commitment and group goal
commitment in the total sample is .52 (p< .00). The positive direction indicates that
psychologically, subjects did not make a trade-off between striving to achieve their
individual goal and their group goal. Indeed, several findings pointed to construct validity
problems with the two goal commitment constructs. For instance, as noted earlier, these

two measures did not have discriminant validity in the US. sample. Further, the

78

 

 

79

2.1%»... «tQN... aawN. «¢%~.u m6? «NH?
8. :8.- ..8.- .2. :8.- 8. 8.- 1.3.- :2. :2.- 8. of .5588
3. 2.- 2. 2. 8.- 8.-
8.- 8. 3.- 8. 2. 8. 8.- 8.- 8.- 8. 2.. 83 .3593
:8.- .2. 2.. 2.- 2.. 2.-
.2. :8.- .2- 3.- .2. 3. S.- 8. s.- S. a. m; Emacs-H 8283.. .2
.2.- 2.. 2.. 2.. 2.- 2.-
.E. .3. 2.- 3.- 8.- 8. 8.- 8. .2.- 8.- o... 2. .53333 58.8 .2
2. 2. 2. 2. a. .2.
:2. 8. 8. 8. 2. :2. 8.- .:. :2.- 2. a. 28 gasxazuzoeom:
2.. 2.. 8. 2.. :2.- 2.
-- 3.- 2.- 8. 8.- 8. :2.- 8. ..8.- 8. on. 8. .88203225 .3 .2
a...» :2. S.- ..Sz :2. S. :2. .2. :2. 2. 828
at :8. 8. :2. :2. 8. .2. .3; .2. 2.. 8.8 3.5382»... .a
a: 2. :8. 2.. :2. .2. 3.. :8. 2. 3.2
33 3. :8. 3.- :2. 8. 3.- :8. 2.. 2.8 $83,582.88 .w
«2.» :8. 2.- 8.- 2.- 3.- 2. 2. 22
G3 :8. 3.- S.- 8. 8.- 8. S. 2....- o-> .8
3..» 2. .2. :2. 8.. .2. 2. 22
83 8. E. .3. 2.- .S. a... 8.4 0-: a
takN... «aRN.
93 .3. :8. :8.- ..8. 8. «3 £6838 955 .m
at :2. .2. 2. 2. 8%
$3 :2. 8. .2. S. 28 855.888....» 2: .4
$3 2.. :8. 2. 22
63 .2. :8. 2. 8..” 2.8.58.8» 9.20 .m
- :2.- 86 on. xmsgazusstoa .8
-- 82 a; :5 82w 5er ._
S a m 8 e m 4 m N _ .3 5.: .o_§a>

 

 

 

 

beam 5 838$ .6 .8328 @5482 £28.20 .3585 38:88 .4 03mm

o—.Vm. mo.Vm* S.Vm..:..

Bommwfim MN rmd u_ o

0380.: MN .22: H: 2.

253m 332 :88 32 No .253 5:3— :88 :wE n_ o

8:859:52: 223 32 No 605230285 223 :32 .1 a

$88 REE-8 :o :83 m:o_3_oboo\mo>umtomow
08 938285: 20:26 8:8 88.8: :2 $53: 323 958 mom—Sm 20: E 859-: 2:0me :0 0:: 2222: 9:03:80 .

 

 

.2. :3. .2. .2. .888 .2
- :2. .3. 8. .58.: .3
28.8 :3. .2. .88 .88.: 8383: .2

- 8. .33 .88.: 58-8 .2

85 3.5 28.3 828.8 ._ 2

3 2 2 : 83.2.;

 

 

 

APEOUV v 03::-

8O

correlations of individual and group goal commitment with performance outcomes are
not always in the expected direction. While group goal commitment demonstrated sound
relationships with individual performance in the group task (r = .23, p< .00) and
individual task performance (r = -.10, p_< .10), individual goal commitment did not, as
evidenced by its weak positive correlation with individual performance in the group task
(r= .11, p< .10) and non-significant relationship with individual task performance (r= .03,
p_> .05). These counter-intuitive relationships provided additional evidence that individual
goal commitment may not be distinctively different from group goal commitment. Hence,
subsequent results involving group goal commitment and individual goal commitment
should be interpreted with caution.

As expected, group goal priority is positively correlated with group goal
commitment (r = .29, p< .00) and negatively correlated with individual goal commitment
(r = -.14, p< .05).

Although not formally hypothesized, several relationships with the performance
outcome variables are notable. Specifically, three out of the four individual differences
have a positive, albeit weak effect on individual performance in the group task: H-C (r =
.12, p< .05), conscientiousness (r = .17, p< .00), and agreeableness (r = .15, p< .05). For
individual task performance, only agreeableness is significantly negatively correlated (r =
-.14, p< .05). All these relationships are in the expected direction, and suggest that
besides influencing performance outcomes via the goal commitment constructs,
individual differences may also directly affect performance outcomes.

Interestingly, while the manipulations of group task interdependence and group
leader status have no impact on individual performance in the group task, they exert a

negative impact on individual task performance (task interdependence: r = -.27, p< .00;

81

 

leader status: r = -.11, p< .10).

Finally, country is significantly correlated with many of the substantive variables.
With respect to task performance outcomes, participants from Singapore performed
worse in the group task (r = -.20, p< .00) and better in the individual task (r = .26, p<
.00), compared to participants from the US. Consistent with these findings, Singapore
subjects reported lower group goal commitment (r = -.16, p< .01) compared to their US.
counterparts, although the other related measures of individual goal commitment and
group goal priority are not significantly different between the two countries.

With respect to individual differences, Singapore subjects reported lower H-C (r =
-.24, p< .00), higher V-C (r= .12, p< .05), lower conscientiousness (r = -.27, p< .00), and
lower agreeableness (r = -.37, p< .00). Of interest is the different pattern of relationships
between H-C and V-C with the country variable. The correlations suggest that while US.
subjects had greater concern for harmony and cooperation with co-workers than
Singapore subjects, they were less willing to sacrifice their personal interests for the sake
of an ingroup such as their family. Thus, these findings reinforce the importance of
making the horizontal-vertical distinction of collectivism (Singelis et al., 1995; Triandis,
1995)
Manipulation Checks

A one-way AN OVA indicated that subjects in the high task interdependence
condition (coded 1) perceived greater group task interdependence than those in the low
task interdependence condition (F = 54.18, p< .00). Similarly, subjects in the high leader
status condition (coded 1) reported having a team leader who had greater power in the
organization than those in the low leader status condition (F = 131.97, p< .00). Hence,

both manipulations worked in the expected direction.

82

 

Tests of Hypotheses

£ain Effects of Individual Differences (H1 — H3). Hypotheses 1 to 3 dealt with
the effects of values and personality on individuals’ group goal commitment and
individual goal commitment. Specifically, I proposed that collectivism (both H-C and V-
C; Hla), conscientiousness (H2a) and agreeableness (H3) would be positively related to
group goal commitment. Results in Table 5 show that H-C has a significant positive
relationship with group goal commitment (B=.l6, p< .05) but VC, unexpectedly, is
negatively related with group goal commitment ([3= -.14, p_< .05). Conscientiousness
(B=.Ol, p> .05) has no effect on group goal commitment, while agreeableness is
marginally positively related to group goal commitment (B=.l 1, p< .10). Hence, results
provide partial support for Hla and H3 but not for H2.

Next, Hlb proposed that collectivism (both H-C and V-C) would have a negative
impact on individual goal commitment while H2b proposed that conscientiousness would
have a positive impact on individual goal commitment. Results in the second column of
Table 5 approach traditional levels of significance for V-C ([3= -.12, p< .10) and
conscientiousness ([3= .19, p< .01), but not H-C ([3: .10, p_> .10). Hence, Hlb is partially
supported and H2b is fully supported.

Main EffectL of Structural Factors (H4 — HSL H4 and H5 predicted that task
interdependence and leader status would be positively related to group goal commitment.
Results in Table 6 (task interdependence) and Table 7 (leader status) did not support
either hypothesis (task interdependence: B= -.03, 13> .10; leader status: B= .05, p> .10).

Interactions between Individual_ and Structurafiactors (H6 — H7). H6a to H6c

proposed that task interdependence would interact with collectivism (both H-C and V-C;

83

 

H6a), conscientiousness (H6b) and agreeableness (H6c) to affect group goal
commitment. Table 6 shows the results of these interaction terms entered simultaneously
into the regression.

While none of the interaction terms in the block is statistically significant, the

interaction terms involving H-C and conscientiousness appear to approach significance.

Given that statistical tests of moderators typically have very low power (Aquinis, 1995;
Evans, 1985), I conducted two separate regression analyses in order to alleviate the
problem: one involving H-C X task interdependence, and the other involving
conscientiousness X task interdependence (see Table 6a). Results support both interaction
terms (H-C X TI : B= -1.44, p< .01; Consc X TI : B: -.85, p< .05).

Figures 5 and 6 present the interaction plots using a median split of the H-C and
conscientiousness measures. Figure 5 shows that the effect of H-C on group goal
commitment is stronger under a low task interdependence condition than under a high
task interdependence condition. However, while I had proposed that H-C will have less

of an impact under high task interdependence, Figure 5 suggests the opposite. Hence,
results only partially support the prediction advanced in H6a.
Figure 6, however, shows an unexpected pattern of interaction for
conscientiousness and task interdependence. While I had predicted that conscientiousness
would have a stronger positive effect on group goal commitment under low task
interdependence compared to high task interdependence, the graph suggests that
conscientiousness exerted opposite effects on group goal commitment under different
task interdependence conditions. Specifically, conscientiousness has a positive impact on
group goal commitment under low task interdependence, but a negative impact under a

high task interdependence condition. Hence, H6b is not supported.

84

 

T_able 5. Hierarchical IggreSSion Results for Individual Differences and Goal

Commitment (H l -H3)ll

 

 

Group Goal Individual Goal

Step Predictor Variables Commitment Commitment
1. Gender” -.08 (.07) -.04 (.07)
R2 .01 (.01) .00 (.00)

2. H-C .16* (.18'”) .10 (15*)
V-C -.l4* {-.12’) -.12' (-.13*)
Conscientiousness .01 6 06) .19“ * (.20**)
Agreeableness .1 1' C 1 0') -.01 (-. 02)
AR2 .05** (.07**) .06** (.07 **)
Total R2 .06 (.08) -06 607)

Overall F(5, 264)

3.37** (4.12**)

 

3.43* (4.42**)

 

' Table contents are beta-weights. Figures in (italics) are beta weights based on

original scales.
" Male coded 1, Female coded 2.

**p<.01 *p<.05 'p<.10

85

_!?

_T_able 6. Hierarchigl Moderated Regression Results for Individual Differences aa_d Task

Interdependence on Group Goal Commitment (H4, H6a-H6c) ’

 

Step Predictor Variables

Step 1
Genderb
AR2

Step 2
Horizontal Collectivism
Vertical Collectivism
Conscientiousness
Agreeableness

ARI

Step 3
Task Interdependence (TI)°
AR;

Step 4
TI X Horizontal Collectivism
TI X Vertical Collectivism
TI X Conscientiousness
TI X Agreeableness

ARI

Overall R2
Overall F (10, 259)

Group Goal Commitment

-.13* (—.10')
.02* (.01')
.17* (.19")

-.14* (.1319
.01 (06)
.11' (.10')
.05** (.07**)

-.03 (-. 06)

.00 (.00)

-.83 (p=.l6) (.100) @=.11)

-.34 (.23)
-.57 (p= .17) (—.43)
.15 (1.04)
.02 (.02)
.09 (.10)
2.63“ (2.88")

 

" Table contents are beta-weights. Figures in (italics) are beta weights based on original

scales.

b Male coded 1; Female coded 2;

c High task interdependence coded 1; Low task interdependence coded 0.

** p_< .01 *p<.05 'p<.10

 

 

 

20m. nova. Eva...

o e38 35303585 55 >65 5 pogo 85230535 5mg nwmm o .N @550 £55m ._ @550 552 a

82.8..- 355.8 no woman $5503 83 2a $5ng 5 8.3me 55503.83 93 $558 033. .

 

 

«.38 .88 $3... .155 «:36 :8... ass-a :85
as.» me. am 38,—. Re.» 5. «~— 38,—.
S: .8. .5. «.8.» :8. .54

«G s .3- 88280235 #3- x 3.80 «.32 :8. _- 85280225 0.3- x 0-:
v new 8 88m
«8.» 8. «.2 :3 8. «.2
«8 r» 3.- 085880285 6.35 «S r» 3.- 085880235 0.8-8
m 58m m 85
9N§¢ 2.. «m4 «gins.» «N6. «“4
a»: 8. 333282.28 «.8: .2. 0-:
m new N mam
SS .8. ”.2 93 .8. «:4
81 .8. .880 as: .2. .528
_ new 5 new
Evangoo moEmta> 853.5 gnaw 50:55:80 moi—$5» “353.5 moam
.30 5.80 .80 380

 

 

 

 

-2 .8 338m :28 oz .8 035-

87

Table 7. Hierarchical Moderated Regression Results for Individual DifferencesaLeader

Status and Group Goal Commitment (H5, H7a-H7d) ‘

 

 

 

Step Predictor Variables Group Goal Commitment

Step 1

Genderb -.13 * (-.10')
AR2 .02* (.01')
Step 2

Horizontal Collectivism .17* (. 19* *)

Vertical Collectivism -. 14* {-.13 *)

Conscientiousness .01 (06)

Agreeableness .11' (.10’)
AR: .05** ( 07“.)
Step 3

Leader Status (LS)c ~05 ('- 04)
AR1 .00 (.00)
Step 4

LS X Horizontal Collectivism -.02 (-. 64)

LS X Vertical Collectivism -.48 (-.30)

LS X Conscientiousness .11 {-.15)

LS X Agreeableness -.15 (-. 07)
An2 .01 (.01)
Overall R2 .08 (.09)
Overall F (10, 259) 212* (2.61 **)

 

“ Table contents are beta-weights. Figures in (italics) are beta weights based on

original scales.

" Male coded 1; Female coded 2.

c High leader status coded 1; Low leader status coded 0

**p<.01 *p<.05 *p<.10

88

F_igure 5. Interaction between HL and Ta_s_k Interdependence on Group Goal

Commitment

Group Goal
Commitment

‘ Low Task Interdependence

High Task Interdependence

 

> Horizontal Collectivism

 

Low High

F igge 6. Interaction between Conscientiousness and Task Interdependence on Group

 

 

Goal Commitment
Group Goal
Commitment
‘ Low Task Interdependence
High Task Interdependence
, Conscientiousness
Low High

89

H7a —- H7d predicted that leader status would interact with H-C (H7a), V-C
(H7b), conscientiousness (H7c), and agreeableness (H7d) to influence group goal
commitment. Results in Table 7 show that none of the interaction terms is significant.

Effect; of Goal Commitment. The last hypothesis predicted that group goal
commitment would interact with individual goal commitment to affect individual
performance in the group task, such that the relationship between group goal commitment
and performance would be stronger when individual goal commitment was low. Results
in Table 8 did not support the interaction hypothesis — only a main effect of group goal
commitment on individual performance in the group task is found ([3= .25, p< .01).
Table 8. Hiegrchical Moderemed Regression Results for Individualan Group Goal

Commitment on Performance in Group Task (H8)a

 

Performance in

 

Step Predictor Variables Group Task
Step 1
Genderb .09 (.09)
AR2 .01 (.01)
Step 2
Individual Goal Commitment (IGC) -.02 {-.12')
Group Goal Commitment (GGC) .25M (30")
ARZ .06** (. 06"”)
Step 3
IGCXGGC ~13 (“-15)
AR: .00 (.00)
Overall R2 ~07 (.0 7)
Overall F“, 313) 5.28“ (5. 90")

 

' Table contents are beta-weights. Figures in (italics) are beta weights based on original
scales; b Male coded 1; Female coded 2

**p<.01 *p<.05 'p<.10

9O

Post-Hoc Analyses

Consistent with the data analytic strategy outlined in Chapter 3, the results
reported above employed variables that were created based on results of factor analyses.
While the objectives of this approach were to establish discriminant validity and cultural
equivalence of the measures, there are limitations arising from the sensitivity of factor
analyses to sample size. Specifically, given the relatively small sample size in my
dissertation, using factor analyses to create variables may capitalize on sampling error —
the deviation between a sample statistic and population parameter that results because the
sample size is smaller than the population size. Since the measures examined in this
dissertation (i.e., conscientiousness, agreeableness, H-C, V-C, individual and group goal
commitment) have been empirically validated by larger samples in previous studies, an
alternative approach to creating variables to test my hypotheses is to use the original
scales without eliminating items based on factor analyses results. This approach also has
the advantage of yielding higher reliabilities because of the greater number of items
underlying each variable. Thus, I conducted another set of analyses to test Hypotheses 1
to 8 using the original scales for conscientiousness, agreeableness, horizontal and vertical
collectivism, and individual and group goal commitment. The descriptives, Cronbach’s
alphas and inter-item correlations of the variables using the original scales are reported in
bold italics in Table 4 on page 79.

Limitations of the measures also led me to examine four other sets of post-hoc
relationships not formally hypothesized in my original model in an attempt to better
understand the data and to suggest possible areas for future research. Here, I outline the
rationale and report the results of these post-hoe analyses. The first set of analyses was

targeted to address the issue of the lack of discriminant validity between individual goal

91

commitment and group goal commitment demonstrated by the US. subjects. Specifically,
I replaced the two goal commitment measures with the group goal priority scale - a
relative measure that captures one’s commitment to the group goal vis-a-vis one’s
commitment to the individual goal. I then reanalyzed the main and interaction effects of
individual differences and structural factors (i.e., H1 — H7) on group goal priority.

The second set of analyses aimed to ascertain if subjects’ psychological
perceptions of the experimental manipulations, rather than the actual manipulations,
would affect goal commitment. Indeed, several scholars have proposed the idea that
environments are created through individual and social processes rather than defined by
objective realities (e.g., Berger & Luckmann, 1967; Salancik & Pfeffer, 1978; Schutz,
1967; Weick, 1977). For instance, Weick (1977) argues that objects and events cannot
become part of a person’s environment without the person actually participating in the
creation of that environment. This is one explanation why people in the same situation
may perceive things differently and consequently, react differently to it. Similarly,
Salancik and Pfeffer’s (1978) social information theory highlights the importance of
understanding how social context, rather than objective task dimensions, can influence
employees’ attitudes toward their jobs. Hence, taking this social constructionist approach,
I replaced the dummy codes representing the experimental manipulations with the
manipulation checks that assessed subjects’ perceptions of the situation, and examined
their impact on group goal priority.

The third set of analyses aimed to ascertain if the effects of values, personality
and structural factors on individual performance in the group task were fully or partially
mediated by group goal commitment and group goal priority. According to the

procedures outlined by Baron and Kenny (1986), I tested for mediation by conducting

92

three separate regression analyses for each relationship. For instance, to assess if the
effect of H-C on individual performance in the group task is fully mediated by group goal
commitment, I estimated three regressions (all controlling for gender): (1) regressing
group goal commitment on H-C, (2) regressing individual performance in the group task
on H-C; and (3) regressing individual performance in the group task on H-C and group
goal commitment. A full mediation is supported when the effects of H-C are significant
in the first two regressions, but become non-significant when entered simultaneously with
group goal commitment in the third regression. I conducted similar tests for all the other
individual differences and structural factors (both objective and perceived).

The fourth set of analyses explored whether the country in which the experiment
was conducted influenced the pattern of results. In all my prior analyses, country was
excluded since it was not a variable of interest on its own. Rather, its function was to
prevent the problem of restricted ranges in the cultural measures, which may potentially
occur if these values were collected from one homogeneous sample. Nonetheless, it is
interesting and important to determine if the data suggest different patterns of
relationships in the two cultures. Results of these within-culture analyses can also provide
some insight as to whether the theory articulated in this dissertation might apply
differentially based on national culture. Here, I examined the bivariate correlations of
variables examined in this dissertation within the US. and Singapore subsamples. In
particular, my focus is on identifying significant correlations that have opposite signs in
the two subsamples.

In all the post-hoe analyses outlined above, I reported two sets of results: one
based on variables created from factor analyses results, and one based on variables using

original scales (presented in (bold italics) in Tables 9-13).

93

Post-Hoc Results

H1-H8 using original scales. In general, results based on variables created with
the original scales did not differ considerably from those based on variables created using
factor analyses outcomes. Specifically, two differences were observed. First, the
interaction between conscientiousness and task interdependence on group goal
commitment (H6b) became non-significant in the current analyses using original scales.
This is not surprising given that the significant interaction reported earlier was found only
when conscientiousness and task interdependence were entered separately into a
regression equation (see Tables 6 and 6a), implying a weak interactive effect which may
not be robust. Second, unlike previous analyses, individual goal commitment was
negatively related to individual performance in the group task at the traditional level of
significance when original scales were used ([3: -.12, p< .10). This relationship is logical
since individuals more committed to achieving their individual goal should focus more on
their individual task at the expense of their group task. However, given that this result
was found only under one of the two approaches to creating variables, this relationship
should be interpreted with caution.

Group goal priority. Subjects who scored high in group goal priority indicated
that they were more committed to achieving the group goal than to the individual goal,
and vice versa.

Drawing from the theoretical rationale outlined for H1 and H3, I expected that H-
C, V-C, and agreeableness would be positively related to group goal priority. Following
the logic that conscientiousness is positively related to both individual goal commitment
and group goal commitment (H2a and H2b), I did not expect conscientiousness to

influence group goal priority, since conscientious people are likely to view both goals as

94

equally important. Next, I also expected group task interdependence and team leader
status to influence group goal priority positively. Finally, using Mischel’s (1977)
situational strength argument, I expected the same interactions between individual
differences and structural factors on group goal priority. Specifically, under weak
situations such as low task interdependence and low leader status, values and personality
should have a greater influence on group goal priority than under strong situations.

Results in Table 9 show that agreeableness is positively related to group goal
priority ([3= .21, p< .01), but not H-C or V-C. Contrary to expectation, conscientiousness
had a marginal negative relationship with group goal priority (B= -.12, p< .10). Results
also indicated that neither task interdependence (Table 9) nor leader status (Table 10) is
significantly related to group goal priority. Next, analyses involving interactions between
task interdependence and individual differences (Table 9) show that none of the four
interaction terms is statistically significant when entered simultaneously into the
equation. However, conscientiousness did interact with task interdependence to influence
group goal priority in the absence of the other three interactions ([3= -.65, p< .10). Figure
7 shows a completely crossed-over interaction, demonstrating that conscientiousness has
a positive impact on group goal priority for individuals under low task interdependence,
but a negative impact for those under high task interdependence. This pattern contradicts
the anticipated interaction predicted by the situational strength argument, but is consistent
with the observed interaction when group goal commitment is used as the dependent
variable (see Figure 6).

Analyses involving interactions between leader status and individual differences

did not yield any significant results (see Table 10).

95

 

Table 9. Post-Hoc: Regrassion Results for Effects of Individual Differences and Task

Lniejdependence on Group Goal Priority ‘

 

 

Step Predictor Variables Group Goal Priority
Step 1
Gender b .06 (.06)
AR2 .00 (.00)
Step 2
Horizontal Collectivism .05 (.04)
Vertical Collectivism -.06 (-. 04)
Conscientiousness -.12’ (-. 13 *)
Agreeableness .21 ** (21 "9
AR’ .06** (. 05*)
Step 3
Task Interdependence (TI)c -09 (09)
AR’ .01 (.01)
Step 4
TI X Horizontal Collectivism -.36 (-. 32)
TI X Vertical Collectivism -.15 (.04)
TI X Conscientiousness -.65 (p= .14) (-.83')
TI X Agreeableness .17 (41)
AR2 .01 (.01)
Overall R2 .08 (.07)
Overall F (10, 259) 109* (2. 03 *)

 

 

" Table contents are beta-weights. Figures in (italics) are beta weights based on original

scales.

" Male coded 1; Female coded 2;

c High task interdependence coded 1; Low task interdependence coded 0

** Q < .01 *p<.05 'p<.10

J

 

 

 

_T_able 10. Past-Hoe: Regression Results for Effects of Individual Differences and Leader

 

 

 

Status on Group Goal Priority 3
Step Predictor Variables Group Goal Priority
Step 1
Genderb .06 (.06)
AR2 .00 (.00)
Step 2
Horizontal Collectivism .05 (.04)
Vertical Collectivism -.06 {-.04)
Conscientiousness -.12’ (-. 13 *)
Agreeableness .21 ** (21 **)
AR2 .05* * (. 05*)
Step 3
Leader Status (LS)c -03 (03)
AR2 .00 (.00)
Step 4
LS X Horizontal Collectivism -.45 (—. 70)
LS X Vertical Collectivism -.58 (-. 38)
LS X Conscientiousness .36 (.43)
LS X Agreeableness -.18 (-.1 7)
AR: .01 (.02)
Overall R2 .06 (.07)
Overall F00. 259) 2.02* (1.82?)

 

“ Table contents are beta-weights. Figures in (italics) are beta weights based on original

scales.

" Male coded 1; Female coded 2;

° High leader status coded 1; Low leader status coded 0;

**p<.01 *p<.05 'p<.10

97

figure 7. Post-Hoczlnterarction between Coascientiousness aad Tas_k Interdependence on

 

Group Goal Priority
Group Goal
Priority
+ Low Task Interdependence
High Task Interdependence

 

, Conscientiousness

 

Low High

Analyses based on variables created with original scales yielded similar results for
this set of post-hoc regressions involving group goal priority as the outcome.

Perceptual measures of structural getors. Before testing the effects of perceived
task interdependence and perceived leader status on goal commitment, I assessed whether
perceptions of the situation were significantly influenced by gender, values and
personality after partialling out the effects due to actual manipulations. This test was
predicated on the notion that different individuals may construe the same situation
differently (e.g., Weick, 1977). For instance, it is plausible that collectivists, given their
interdependent self-construal, tend to view a situation as more interdependent than
individualists. Or, females or individuals with high V-C may tend to view the group
leader as having more power because of their sensitivity to status differences, compared
to males or individuals lower in V-C.

Results in Table 11 show very weak support for the above speculations. Actual

98

 

task and leader manipulations explain 18% of the variance in perceived task

interdependence and 31% in perceived leader status respectively. Female subjects

reported greater task interdependence (B= 0.10, p< .10) and greater leader status ([3: .15,

p_< .01), while agreeable individuals tended to report lower levels of leader status (B= -
.12, p< . 05). No other individual differences measured in the study are statistically
significant in predicting perceptions of the experimental conditions.

Results based on variables created with original scales yielded only one slight
difference — H-C was positively related to perceived task interdependence at a marginal
level of significance (p= .ll, p< .10; see Table 11). This could be that horizontal

collectivists, consistent with their emphases on fostering common goal and social
harmony, tend to view all group tasks as being more interdependent, regardless of the
actual nature of the tasks.

The next set of analyses ascertained whether perceived measures of task
interdependence and group leader status, as well as their interactions with individual
differences, influenced group goal priority. Specifically, I would expect greater perceived
task interdependence and perceived greater leader status to be positively associated with

group goal priority. Further, subjects who perceived stronger situations (i.e., greater task

interdependence or higher leader status) should be driven more by their perceptions of the

situations rather than by their values and personality, compared to subjects who perceived
weaker situations.

Results show that perceived task interdependence had a positive impact on group

goal priority ([3= .19, p< .01) (Table 12) but not perceived leader status (B: .10, p> .10)

(Table 13). Further, while none of the interaction terms between perceived task

99

 

 

ofivm. mag-Va... ~o.Vm..:..

o 02000 5.33 520.62 305 5 02000 558m .232 :93 e

.N 668 6666-.“ x 668 6.62 .

o 0008 85330525 033 305 5 0208 080053525 2.6-3 55m 6

.838 355.5 :0 women 5550? Son 03 3535 5 monE 5333-33 0.3 558:8 23-.- e

 

 

6.:- g $8.2 a. e... :65 9.8.5 :8...— as 6,.— .665

«:63 on "m =65 :3 3. am .665

«:8.» :8. 24 a. .3 .8. ”.2
«:2 a .2; 6663666,. :3 8. 6663666...
«3 ¢ 8.- 80555520500 as .\ ao- 80535520500
«mg no. 5558230 350.65 Go r» S... 55>u8=oo 3555
«we r» 3.- 5558230 3505.85 a ~ 3 no. 5558230 350583
p.20 ..m _. o 6260 S: .2. .6260

a new N new

«at ~M.» «4mm. "“4 «1&3» aemu. «g
«:63 :2. 33% 6663 a... N: :8. .8628865 6.68

2 92m 2 9on

553m “2305 352083525
35023 m2n53> “3500.5 3% 25d- 005023 m2n53> 58500-5 98m

 

 

 

 

 

Maw-8W Han-005 0053qu 53 3:25 0525 030.“- 0o>moouom no 30:80th 355305 How 533% ”com-30m .: 23,—.

100

_T_able 12. Post-Hoe: Regression Results for Effects of Individug Differences and

Perceived Ta_slc Interdependence on Group Goal Priority 3

 

 

Step Predictor Variables Group Goal Priority
Step 1
Genderb .05 (.05)
AR2 .00 (.00)
Step 2
Horizontal Collectivism .05 (04)
Vertical Collectivism -.08 (-.03)
Conscientiousness -.1 1t (-. 13')
Agreeableness .21" (21 **)
AR2 .06** * (.05 *)
Step 3
Perceived Task Interdependence .19" * (19")
AR: .04“: (. 04“)
Step 4
PTI X Horizontal Collectivism -.55 (-. 32)
PT] X Vertical Collectivism -.34 (-. 33)
PT] X Conscientiousness .42 (21)
PTI X Agreeableness -.31 (-.19)
AR2 .01 (.00)
Overall R2 .11 (.09)
Overall Fawn 2.93** (2. 52 **)

 

 

“ Table contents are beta-weights. Figures in (italics) are beta weights based on original
scales.

b Male coded 1; Female coded 2.

**p<.01 *p<.05 'p<.10

101

 

flble 13. Post-Hoe: Regression Results for Effects of Individual Differences and

Perceived Leader Status on Group Goal Prioritg

 

 

Step Predictor Variables Group Goal Priority

Step 1

Gendera .05 (.05)
AR: .00 (.00)
Step 2

Horizontal Collectivism .05 (.04)

Vertical Collectivism -.O6 {-.03)

Conscientiousness -.l 1' (-. 12’)

Agreeableness .21* * (.21 **)
AR2 .05** (.05*)
Step 3

Perceived Leader Status (LS)b -10 (10)
AR’ .01 (.01)
Step 4

PLS X Horizontal Collectivism

PLS X Vertical Collectivism
PLS X Conscientiousness

-1.36* (425*)
-.35 (-.18)
1J7* (97D

PLS X Agreeableness 1.30M (1-57")
AR2 .07** (.07**)
Overall R2 .13 (.13)
gavmu pm 2.56) 3.76“ (3.82“)

 

 

a Table contents are beta-weights. Figures in (italics) are beta weights based on original

scales.
b Male coded 1; Female coded 2.

**p<.01 *p<.05 'p<.10

102

interdependence and individual differences is significant in predicting group goal priority
(Table 12), three out of the four interaction terms involving perceived leader status are
significant (Table 13). Specifically, perceived leader status interacted with H-C ([3: -
1.36, p< .05), conscientiousness ([3: 1.17, p< .05), and agreeableness (B: 1.30, p< .01) in
predicting group goal priority. Figures 8 — 10 illustrate these interactions respectively.

Contrary to expectations, the graphs suggest that H-C (Figure 8) and
agreeableness (Figure 9) exert a greater impact on group goal priority under high
perceived leader status than under low perceived leader status. In other words, when
subjects perceived their group leader to have greater power in the organization, those who
were more horizontally collectivistic and more agreeable reported higher levels of group
goal priority, compared to those who were less collectivistic and less agreeable.

For conscientiousness, the interaction plot (Figure 10) suggests that the effects of
conscientiousness on group goal priority differ under high versus low perceived leader
status. Specifically, conscientiousness has a positive relationship with group goal priority
under high perceived leader status, but a negative relationship with group goal priority
under low perceived status. This pattern also suggests that the perception of leader status
matters only for individuals high in conscientiousness.

All analyses based on variables created with original scales yielded similar pattern
of results for this set of analyses involving individual differences, perceived measures of
structural factors and group goal priority (see Tables 12 and 13).

fists of Mediation. In this set of analyses, I tested if group goal commitment and
group goal priority mediated the effects of the distal individual differences and situational
factors on individual performance in the group task.

Following Baron and Kenny’s (1986) criteria for mediation, results using

103

Figare 8. Post-Hoe: Intergtion between H-C and Perceived Leader Status on Group Goal

 

 

Priority
Group Goal
Priority
‘ High Perceived Leader Status
/ Low Perceived Leader Status
, Horizontal Collectivism
Low High

figure 9. Post-Hoe: Interaction between Agreeableness and Perceived Leader Status on

Group Goal Priority

Group Goal
Priority

A High Perceived Leader Status

Low Perceived Leader Status

/

 

+ Agreeableness
Low High

104

Figure 10. Post-Hoe: Interaction between Conscientiousnewd Perceived Leader Status

on Group Goal Prioritv

Group Goal
Priority

\\ Low Perceived Leader Status

, Conscientiousness

High Perceived Leader Status

 

 

Low High

variables created based upon factor analyses outcomes demonstrated that group goal
commitment fully mediated the effects of H-C and agreeableness on individual
performance in the group task. Specifically, both agreeableness and H-C were
significantly related to group goal commitment (H-C: B: .16, p< .05; agreeableness: [3:
.16, p< .01), and both were significantly related to individual performance in the group
task (H-C: B: .11, p< .10; agreeableness: [3: .14, p< .05). When group goal commitment
was added with H-C to predict performance, H-C became insignificant (B: .07, 12> .05).
Likewise, for agreeableness, its effect on performance was not significant when group
goal commitment was entered simultaneously (agreeableness: B: .11, p< .10). No other
factors were mediated by group goal commitment to influence individual performance in
the group task.

Results using variables created with original scales yielded a different picture.

Group goal commitment only partially mediated the effects of agreeableness on

105

individual performance in the group task. Specifically, agreeableness was significantly
related to group goal commitment (agreeableness: B: .18, p< .01), and to individual
performance in the group task (agreeableness: B: .16, p< .05). When group goal
commitment was added in the regression equation, the effect of agreeableness became
marginally significant (B= .11, p< .10), thus suggesting partial mediation. Further, group
goal commitment did not mediate the relationships of both conscientiousness and H-C
with individual performance in the group task.

When group goal priority was examined as the mediator, results using variables
based on factor analyses outcomes demonstrated that only agreeableness was mediated
by group goal priority to influence individual performance in the group task. Specifically,
agreeableness was significantly related to group goal priority in the first equation ([3: .19,
p< .01), but was non-significantly related to task performance when entered
simultaneously with group goal priority in the third equation (B= .09, 12> .05), thus
demonstrating a full mediation. No other factors were mediated by group goal priority to
influence individual performance in the group task.

Results based on variables created with original scales yielded only a slight
difference. The relationship between agreeableness and individual performance in the
group task was being partially, instead of fully mediated by group goal priority.
Specifically, agreeableness was significantly related to group goal priority in the first
equation (B= .18, p< .01), but was marginally significantly related to task performance
when entered simultaneously with group goal priority in the third equation ([3: .11, p>
.10), thus demonstrating a partial mediation.

Coungy differences. A comparison of the bivariate correlations between the US.

106

and Singapore subsamples revealed one contrasting relationship between the leader status
manipulation and individual performance in the group task: the correlation was negative
in the US. sample (r= -.19, p< .05) but positive in the Singapore sample (r= .22, p< .01).
In other words, having a high status leader in the parallel team had an unexpected
negative impact on members’ performance in the group task in the U.S., but a positive
impact on members’ performance in Singapore. The opposite effects also explain why
leader status in the entire sample did not have a significant main effect on individual
performance in the group task.

Results based on variables created with original scales yielded a similar pattern of
results. In the next chapter, I discuss these findings as well as their implications on theory

and practice.

107

CHAPTER 5

DISCUSSION

In light of the growing popularity of parallel teams in organizations, there is a real
need for research to understand the unique issues faced by individuals belonging to such
teams. I contend that one such issue is the dilemma (between pursuing personal goals
versus group goals) created by the concurrent existence of two tasks (i.e., an individual
task and a group task). Specifically, the general research question I address is given the
limited amount of time, what factors would influence parallel team members’
prioritization of goals, and consequently, performance in the group task?

To recapitulate briefly, Kanfer’s (1990) distal-proximal framework of
motivational theories, as well as Hollenbeck and Klein’s (1987) goal commitment theory,
provided the rationale for this dissertation to address the effects of three sets of variables:
the proximal construct of goal commitment, and distal constructs such as individual
differences and situational characteristics. Specifically, I theorized that horizontal and
vertical collectivism, conscientiousness, and agreeableness would influence both
individual goal commitment and group goal commitment. Task interdependence and
team leader status were posited to have main, as well as interactive effects with the
individual differences, on group goal commitment. Finally, I proposed that individual and
group goal commitment would interact with each other to influence individuals’
performance in their group task.

Results based on an experiment conducted with students in the US. and
Singapore highlighted three main findings. First, construing individual and group goal

commitment as separate distinct constructs to reflect the dual-task nature of parallel

108

teams may or may not be viable. This equivocal conclusion is based on the intriguing
phenomenon observed in my data that while the Singapore subjects distinguished
between the two goal commitment constructs relatively well, the US. subjects did not.
Thus, it could be that the viability of having two separate goal commitment constructs in
a multiple-goal environment is subject to cultural nuances. Nonetheless, group goal
commitment and the alternative construct -- group goal priority (examined in post-hoe
analyses), had a significant positive impact on individual performance in the group task.
This is a critical link in this study since group goal commitment was posited as the
proximal mechanism that links the distal factors with the performance outcome.

The second finding is that individuals with different collectivism orientation and
personality traits are likely to react differently to resource dilemma situations posed by
competing individual and group tasks in a parallel team environment. Specifically, a
robust finding was that agreeable individuals were more committed to their group goal,
which in turn led to them to perform better in the group task compared to their
counterparts with lower agreeableness. Individuals high in H-C and conscientiousness
also performed better in the group task. However, whether these relationships were
mediated by group goal commitment depended on how the variables were created (i.e.,
based on factor analyses outcomes or original scales). This inconsistency suggests
caution in interpreting these relationships, and also implies that more research is required
in future in order to replicate the results.

The third finding concerns the impact of task interdependence and leader status on
team members’ group goal commitment/priority. The actual manipulations of task
interdependence and leader status had no significant effects, although post-hoc results

demonstrated that perceived task interdependence had a positive impact on parallel team

109

members’ group goal priority. It appears that a more fruitful way of theorizing about the
effects of task interdependence and leader status in parallel teams is to consider their
interactions with individual differences (rather than focusing on main effects) — implying
that these structural characteristics invoke different reactions from different people.
Specifically, a robust finding is that high task interdependence seemed to “backfire” for
individuals with high H-C— they reported less group goal commitment than their high H-
C counterparts under low task interdependence. On the other hand, high task
interdependence seemed to have no effect on individuals with low H-C.

The remaining of this chapter is organized into four parts to provide a more
detailed discussion of my results, as well as to highlight some of the implications,
limitations, and future directions of this research. Specifically, the first section discusses
the empirical results relating to my formal hypotheses as well as post-hoc speculations,
and develops plausible rationale for unsupported or unexpected relationships. The second
section highlights the contributions of this study to research and practice. Finally, in the
last section, I discuss limitations of the current research, and conclude by proposing
several future directions that may further enhance our understanding of parallel teams.
Discussion of Empirical Results

The sequence of presentation in this section is as follows. I first discuss the
empirical results relating to goal commitment and group goal priority, since these goal
constructs are theorized to be the link between individual differences/ structural factors
and individual performance in the group task. I then discuss the effects of values,
personality, and structural factors (task interdependence and leader status) on the
hypothesized outcome of goal commitment, as well as post-hoe outcomes of group goal

priority and individual performance in the group task. Finally, I discuss the interactions

110

between individual differences and structural factors on group goal commitment and
group goal priority.

Goal Commitment. Consistent with the dual-task nature of parallel teams, I
proposed that parallel team members would develop two sets of goal commitment, one
for their individual task and one for their group task. Results from this study, however,
demonstrated an intriguing phenomenon: while Singapore subjects discriminated between
individual goal commitment and group goal commitment reasonably well, U.S. subjects
did not. Given that both groups of subjects were given the same background information,
instructions and materials, and went through the same steps during the experiment, I
ruled out explanations due to potential confounds or contaminants with the experimental
procedures.

One tentative speculation is that because subjects essentially worked
independently on both their individual and group tasks, U.S. subjects may have viewed
both tasks as equally reflective of their personal competence, and hence, made no
distinction between the two goals. This is consistent with Markus and Kitayarna’s (1991)
thesis that individuals in the West, such as the U.S., are typically motivated by
individually rooted needs or motives — the motive to enhance self-esteem, the motive to
achieve, and the motive to self-actualize. As such, U.S. subjects might have viewed the
achievement of both individual and group goals to be equally important to their self-
esteem and consequently, treated the performance in both tasks as an overall indicator of
their effectiveness.

It may be that Singapore subjects were better able to discriminate between
individual and group goal commitment because of their clearer distinction between the

self and the surrounding context, or what Markus and Kitayama (1991) termed “self-in-

111

relation-to-other” construal that is more typical of Eastern or Asian cultures. As such,
instead of viewing the achievement of the two tasks as reflecting one’s personal
competence, they may have viewed the achievement of each task as having different
implications on themselves and on the group. This delineation between the self and others
could have been made more acute by the possibility that Singapore subjects, being more
vertically collectivistic than their US. counterparts, might have viewed themselves as
distinct from their groups due to the temporary nature of the groups formed in this study.

Another reason could be due to the fact that the goal commitment scales were
measured prior to the beginning of the actual tasks. Even though subjects were told that it
was impossible to achieve both goals within the limited time, it could be that US.
subjects, for some reason (e.g., social desirability bias), did not make a psychological
trade-off between pursuing the two goals before the commencement of the tasks. One
way of ascertaining if this was indeed the case is for future research to measure goal
commitment at the start, as well as in the middle, of the performance trials.

Because of the lack of discriminant validity for the goal commitment constructs
with the US. data, it would be misleading for me to make any inferences based on the
results of my analyses for H1 to H7. Thus, I discuss in the next section the results of a set
of post-hoe analyses using group goal priority — an alternative measure that taps group
goal commitment in relation to individual goal commitment.

Moving beyond construct validity issues, I examined the implications of
individual and group goal commitment on individual performance in the group task.
Specifically, hypothesis 8 predicted that the relationship between group goal commitment
and individual performance in the group task would be moderated by individual goal

commitment. Results did not support this interaction. However, group goal commitment

112

 

did have a positive impact on individual performance in the group task and contributed
6% of variance explained. Further, results using the original 9-item goal commitment
scales also demonstrated that individual goal commitment was negatively related to
individual performance in the group task, thus suggesting additive rather than interactive
effects. Given the relatively a-theoretical nature of parallel team research to date, these
findings suggest that group and individual goal commitment are viable mechanisms
underlying members’ performance in the group task, and represent an important step in
theory development with respect to parallel teams.

The interaction between individual goal commitment and group goal commitment
may have failed to produce the predicted results because of two statistical reasons. First,
given the high correlation between individual goal commitment and group goal
commitment, multicollinearity may have prevented a significant interaction. Second, the
relatively restricted ranges in both individual and group goal commitment (standard
deviations = .58 and .62 respectively) could have also contributed to the non-significant
result.

Group Goal Prioritv. In addition to addressing the construct validity problem
mentioned earlier, analyzing group goal priority also has the benefit of providing another
look at the results involving goal commitment. Besides, the group goal priority construct
has demonstrated relatively sound properties, as evidenced by the moderate correlations
(in the expected direction) with related constructs such as individual and group goal
commitment, as well as the performance outcomes.

Nonetheless, there are some caveats to using the group goal priority measure.
First, I included it as post-hoc analyses and it obviously contradicts my earlier theoretical

formulation that individual goal commitment and group goal commitment can exist

113

independently (however, this assertion was refuted by the US. data). Second, combining
individual goal commitment and group goal commitment into one relative scale results in
some loss of information — a case similar to the use of difference scores. For example, in
examining the correlations for the individual goal commitment and group goal
commitment constructs with other individual differences in the Singapore sample (I
omitted the US. sample since the two goal commitment constructs lacked validity), I
observe that conscientiousness had a positive impact on individual goal commitment, but
a negative impact on group goal commitment. These individual relationships, however,
were masked by the group priority scale since the latter showed no significant
relationship with conscientiousness.

The post-hoe analysis on group goal priority and individual performance in the
group task yielded a similar positive relationship as that of group goal commitment (7%
of variance explained), thus suggesting group goal priority as a potential mechanism that
mediates the relationships between the distal factors and individual performance in the
group task.

Qdividug Differences (H1 — H3). This set of hypotheses proposed effects of
individual differences on group and individual goal commitment. Overall, three
individual differences consistently predicted group goal commitment regardless of how
they were created (i.e., using factor analyses outcomes vs. using original scales): as
predicted in Hla and H3, H-C and agreeableness were positively related to group goal
commitment. Unexpectedly, however, V-C was negatively related to group goal
commitment. With respect to individual goal commitment, two hypotheses were
supported consistently across the two approaches of creating variables: V-C (Hlb) was

negatively, while conscientiousness was positively (H2b) related to individual goal

114

commitment.

However, in the post-hoe analyses where I regressed these individual differences
on group goal priority, only agreeableness and conscientiousness demonstrated
significant relationships. Specifically, agreeableness was positively, while
conscientiousness was negatively related to group goal priority. The failure to replicate
the effects of H-C and V-C using group goal priority should raise caution with the
findings involving their relationships with group goal commitment (given the ambiguity
in the nature of the goal commitment construct). Nonetheless, I offer a tentative
speculation to the unexpected negative relationship between V-C and group goal
commitment here. It could be due to the nature of the group involved in this study. Recall
that V-C tapped at obligations to family members -— a very strong “ingroup.” On the other
hand, the groups simulated in this study were temporary groups with no past history, nor
anticipation for future interaction. This stark contrast in the nature of the group may
explain why individuals who view the family ingroup as very important may view the
outgroup as peripheral and hence, reacted negatively to the goal pursued by the outgroup.
Barley (1993) for instance, found that collectivists working in outgroups performed worse
than individualists working in outgroups. This is consistent with past research that
demonstrates that the ingroup-outgroup distinction is more critical to collectivists than
individualists. However, it should be noted that Earley’s study employed a generic I-C
scale that did not make distinctions between the horizontal and vertical dimensions.
Following my interpretation of the current finding, future research interested in the
influence of ingroups versus outgroups may consider using a scale such as V-C that
targets a strong ingroup so that any effects would be more pronounced.

Besides the main effects found for these individual differences on group goal

115

commitment, post-hoe analyses demonstrated that group goal commitment mediated
some of the relationships between individual differences and individual performance in
the group task. Perhaps the most robust finding is that the effects of agreeableness on
group task performance were mediated by group goal commitment and the related group
goal priority construct, since these mediating relationships were demonstrated under both
methods of creating variables (using factor analyses outcomes vs. using original scales).
Hence, these findings are consistent with the distal-proximal motivational framework
underlying this dissertation. In addition, they represent a contribution to current research
because existing work on agreeableness has been limited primarily to traditional work
teams. Besides, the impact of agreeableness on task performance has typically been
equivocal. For instance, the meta-analysis by Barrick and Mount (1991) concluded that
agreeableness has no impact on performance. One reason offered is that since
agreeableness is associated more with social interaction rather than achievement-striving,
it should theoretically have more impact on social outcomes (e. g., team viability; Barrick
et al., 1998) rather than performance outcomes. However, this study suggests that
agreeableness may be a more meaningful and predictive personality trait when group
tasks are involved (see also Neuman & Wright, 1999). Perhaps agreeableness is an even
more predictive personality trait in parallel teams when competing group and individual
tasks are involved, given that the helpful and altruistic nature of agreeableness is
particularly relevant to the resource dilemma posed by the competing individual and
group interests.

A less consistent result was found for H-C. When variables were created based on
factor analyses outcomes, H-C was fully mediated by group goal commitment in

influencing individual performance in the group task. However, this mediating

116

relationship was not supported when variables were created using the original scales.
Despite these inconsistent results, the findings provide some preliminary insight to the
current state of parallel team research. First, they suggest that group goal commitment is
a potential mediator for H-C with regards to performance in the group task. Second, they
suggest the presence of other mechanisms not examined here that may have stronger and
more consistent explanatory power concerning the H-C — group task performance link.
Future attempts to replicate the existing relationships and to explore other potential
mediators will help advance our current understanding of performance in parallel teams.
An interesting finding that was consistent across the two approaches to creating
variables is the lack of mediation for conscientiousness and individual performance in the
group task. Instead, results showed a direct positive relationship, suggesting the
possibility of a different mechanism for conscientiousness other than group goal
commitment or priority. One reason why the effect of conscientiousness on performance
in the group task is not mediated by the same mechanism as agreeableness (and plausibly
H-C) could be due to the fact that unlike H-C and agreeableness, conscientiousness does
not explicitly involve concern for others. In fact, results showed that prior to the
commencement of the tasks, conscientious individuals were more likely to report greater
individual goal commitment, and lower group goal priority than those with low
conscientiousness. However, it appears that despite the greater concern to get their own
job done, conscientious parallel team members still performed better in the group tasks
than their less conscientious counterparts. It could be that the dependable nature of
conscientious individuals induces a sense of responsibility toward the group, which
subsequently over-rides the initial commitment to achieving the personal goal.

Task Interdependence and Team Leader Status (H4-H5). Hypotheses 4 and 5

117

respectively predicted that group task interdependence and group leader status would
have a positive impact on group goal commitment. Both hypotheses were not supported.
Neither were they supported when group goal priority was used as the dependent
outcome in my post-hoc analyses. This lack of results may be due to several reasons.

The first is a generalizability issue. Almost all the existing studies conducted on
group task interdependence and leader authority/power involved traditional teams, where
participants were full-time group members. As such, little is known whether the effects of
such group structural interventions generalize to parallel teams where members have
fewer interactions within the teams, and have their own responsibilities outside the teams.
It could be that the presence of an individual task might have overwhelmed the
motivating potential of such team structural interventions. Or, it could be that because
different individuals in parallel teams reacted differently to task interdependence and
leader status, the simple main effects of these interventions on individual performance in
the group task were masked. I discuss the interaction effects in greater detail in the next
subsection.

The second reason why task interdependence and leader status did not influence
outcome variables of interest is a methodological explanation. An issue is the strength of
the task interdependence and leader status manipulations. I noted earlier that although
both manipulation checks were significant in the predicted direction, the variance
explained by the actual manipulations were 15% and 33% for task interdependence and
leader status. Even though these levels of variance explained are not uncommon in
experimental studies, it could be insufficient to invoke the expected reactions from
subjects in this particular setting.

Another explanation for the lack of results related to the manipulations is that it is

118

individuals’ perceptions, rather than objective realities, that matter (Weick, 1977). First, I
tested the plausibility that demographic, values and personality would affect individuals’
perceptions of task interdependence and team leader status. Results, however, provided
only very weak support: female subjects tended to report higher leader status, while
agreeable subjects tended to report lower leader status. The former finding could be due
to the fact that because women generally occupy less dominant and powerful roles
compared to men (Eagly, 1987; Eagly and Wood, 1991), they tend to perceive greater
hierarchical distance between themselves and their leaders. The negative impact of
agreeableness on perceived leader status is less clear. One speculation is that perhaps
because less agreeable people are generally more hostile toward others, they put a greater
hierarchical distance between themselves and their leaders.

Post-hoc results also show that perceived task interdependence had a positive
impact on group goal commitment (and group goal priority), thus providing some support
for H4. However, it should also be cautioned that since perceived task interdependence
and group goal commitment (goal priority) were self-reported measures, their
relationships could be artifacts of a response-response bias. Perceived leader status did
not have a significant relationship with either group goal commitment or group goal
priority.

Notwithstanding the lack of support for H4 and H5, it is interesting and important
to note that task interdependence and team leader status (as objective realities) are
negatively correlated with individual task performance (for task interdependence: r = -
.27, p< .01; for leader status: r = -.11, p< .10). These relationships imply that while
having a highly interdependent team task and a high status team leader generally failed to

motivate parallel team members to contribute more to the group, they actually detracted

119

from their individual task performance. Hence, it seems that the dilemmas created by the
dual- task nature of parallel teams are not restricted to parallel team members alone;
organizations, in structuring parallel teams to foster member contribution, may risk
detracting parallel team members from their personal job duties. This may in turn incur
huge costs to the organizations.

InteractionsLbetween structural and individual factors (H6-H7). H6 predicted that
high task interdependence would dampen the effects of collectivism (H6a),
conscientiousness (H6b) and agreeableness (H6c) on group goal commitment. Under the
first method of creating variables (i.e., based on factor analyses outcome), results show
that only the interactions involving H-C and conscientiousness were significant. Further,
only the H-C X task interdependence interaction conforms to the situational strength
argument in that the slope between H-C and group goal commitment is weaker under
high task interdependence than under low task interdependence (see Figure 5). For the
interaction involving conscientiousness, a completely cross-over graph is obtained (see
Figure 6), demonstrating that high task interdependence has a positive impact on the
group goal commitment of individuals with low conscientiousness, but a negative impact
for those with high conscientiousness.

When variables were created based on their original scales, the interaction
between task interdependence and conscientiousness was not significant. However, the
post-hoe analysis that replaced group goal commitment with group goal priority yielded a
significant interaction for task interdependence and conscientiousness regardless of the
method of creating variables, thus suggesting a rather robust finding. The form of the
interaction (see Figure 7) — a completely cross-over graph, was similar to the earlier

interaction involving group goal commitment. (It should be noted again that since the

120

interaction between H-C and task interdependence was not significant when group goal
priority was replaced as the outcome variable, the interaction should be interpreted with
caution).

According to these interactions, the unintended consequence of high task
interdependence for high conscientious individuals is potentially insightful. In a recent
study of college students by Shaw, Duffy and Stark (2000), the authors found a similar
negative impact of task interdependence for group members who had low preference for
group work. Interestingly, the authors also found that ability was negatively related to
preference for group work. Taken together, Shaw et al.’s findings may help interpret the
unintended consequence of task interdependence on high conscientiousness individuals
reported here. Perhaps because individuals with high conscientiousness had higher levels
of ability, they had lower preference for group work, and thus, when placed in an
incongruous situation of high task interdependence, reported lower levels of group goal
priority (or commitment).

The next set of hypotheses predicted that team leader status would moderate the
effects of H-C, V-C, conscientiousness and agreeableness on group goal commitment.
Results were consistent across the two approaches to creating variables, and showed that
none of the interactions was significant. However, when leader status was replaced with
perceived leader status, three significant interactions emerged under both methods of
creating variables. As highlighted in the earlier section, this discrepancy could be
attributed to a weak manipulation of leader status, a social constructionist explanation,
common method bias, or a combination of the three. Nonetheless, a discussion of these
post-hoe interactions can potentially provide some important insight to parallel teams.

Two interactions involving H-C and agreeableness are similar in form, as

121

demonstrated in Figures 8 and 9 respectively. The third interaction involving
conscientiousness (Figure 10) yields a slightly different picture. Across all the three
interactions, perceived leader status has a positive impact on the group goal priority of
individuals with high H-C, high agreeableness, and high conscientiousness. The
difference between conscientiousness and the other two interactions (i.e., H-C and
agreeableness) lies with the effect of perceived low leader status on group goal priority:
while low perceived leader status has a weak positive impact on individuals with high H-
C and agreeableness, it has a negative impact on those with high conscientiousness. All
these interactions also demonstrate that Mischel’s (1977) situational argument did not
apply to the current context, thus refuting the general logic underlying Hypothesis 7.

Interestingly, the form of the interaction demonstrated by H-C and agreeableness
is similar to the form of interaction predicted for V—C and team leader status in H7b (see
Figure 3). I argued in H7b that because vertical collectivists are deferent to authority,
they are likely to perceive group goal as considerably more important when they are led
by a leader who has high status, than when led by a leader who has little status. Hence, it
appears that even though individuals high in H-C, conscientiousness and agreeable
people are not, theoretically, more sensitive to status differences than those low in these
dimensions, they responded more to authority in terms of prioritizing their goals in this
study.

The interaction involving conscientiousness is especially intriguing when
juxtaposed with the interaction between conscientiousness and task interdependence:
(perceived) leader status and task interdependence had opposite effects on individuals
with high conscientiousness — they respond positively to leader status, but negatively to

task interdependence. Perhaps the achievement-oriented nature of conscientious

122

individuals predisposes them to view the group task assigned by a high status leader to be
more critical in demonstrating their competence, compared to a group task assigned by a
low status leader (since high status leaders are ofien construed as having greater reward
power). On the contrary, as discussed earlier, high task interdependence may have
lowered their motivation because of their preference for independent work, thus reducing
their group goal priority.

In the next section, I discuss the implications of these findings on research and
practice.
Implications

On theory. This study uses parallel teams as a context to investigate a multiple-
goal environment in which two goals (an individual goal and a group goal) compete for
individuals’ limited resources. Given the scarcity of research on multiple goals (e.g.,
Austin & Vancouver, 1996; Keman & Lord, 1990), there is a need for a greater
understanding of what factors and mechanisms drive individuals’ reactions to multiple-
goal striving. Besides highlighting the importance of studying multiple-goals, this
research extends previous research in several ways. First, I extend the goal commitment
construct to a multiple-goal setting. Specifically, I argued that parallel team members will
develop two distinct sets of goal commitment: one for their individual goal, and one for
their group. Second, I proposed, and demonstrated to some extent, that goal commitment
and its related group goal priority construct can be construed as proximal constructs that
mediate the effects of individual differences on task performance. Third, I expanded the
range of the antecedents of goal commitment to include individual differences such as
collectivism, conscientiousness, agreeableness, and team design factors such as task

interdependence and team leader status.

123

As discussed in the earlier section, results provided support to some of my
predictions, but raised many important questions as well. Perhaps the most critical
question is whether my construal of two distinct sets of goal commitment is viable.
Results based on the US. subjects seemed to refute that, although methodological flaws
may have influenced these. As mentioned before, it could be the timing of the questions,
or the strong social desirability cues invoked by an experimental context, that prevented
discrimination between commitment to the individual goal and to the group goal. To
address the first issue, future research can measure goal commitment halfway in an
experiment, when subjects actually feel the tension between achieving the individual and
group goals. To address the second issue, field studies involving real parallel teams may
alleviate the strong social desirability cues evoked by experimental settings, and thus,
allow another look at the viability of having distinguishable sets of goal commitment.

Nonetheless, results of this study provided encouraging support for the group goal
priority construct — a relative measure that taps at individuals’ group goal commitment
vis-a-vis their individual goal commitment. Even though the current measure is tailored
to the context of parallel teams (individual and group goal), future research interested in
multiple goals can modify the items to suit their specific context of interest.

In addition, the findings that group goal commitment mediated the effects of
agreeableness on individual performance in the group task also inform existing research
in several ways. First, existing studies on multiple goals (Keman & Lord, 1990; Schmidt
et al., 1984) have typically focused on the effects of structural factors (e.g., reward,
specificity of goal, feedback), rather than individual differences. Thus, the current
dissertation extends previous multiple goal studies by demonstrating that personality (and

potentially cultural variables) can be important factors to consider in multiple-goal

124

striving situations.

Second, these findings also inform personality research in two ways. First, the
absence of a measured mediating variable in most personality research may explain why
very few studies have found a direct relationship between agreeableness and task
performance (e.g., Barrick & Mount, 1991). Second, it also suggests that while
conscientiousness may seem to have a similar positive effect on performance in the group
task as agreeableness, it does so through a different mechanism. Thus, future research
should explore what are the different mechanisms driving the effects of agreeableness
and conscientiousness on group members’ performance. This is potentially important
because the different motivational bases of agreeable and conscientious group members
suggest that they are likely to react differently to different situations and/or interventions.

Results of this dissertation also suggest a less positive, more qualified evaluation
of task interdependence as a team structural intervention to elicit cooperation from team
members. First, task interdependence as an objective reality, did not have a direct impact
on group goal commitment or individual performance in the group task. Although
perceived task interdependence was found to influence group goal commitment (and
group goal priority) positively, the potential inflation of results due to common method
bias suggests that fixture research should attempt to replicate the finding. Nonetheless,
Magjuka and Baldwin’s (1991) study may provide some supporting evidence that
perceived task interdependence can be beneficial to parallel teams. In a study of
employee involvement programs (a form of parallel team), the authors found that
heterogeneous teams (i.e., teams where members come from diverse jobs) evaluated their
team performance more positively than homogeneous teams, presumably because they

were more likely to perceive that they had greater informational variety and richness as a

125

result of the diverse backgrounds of their members. Indirectly, such diversity could be
construed as a form of “resource interdependence” (Johnson, Johnson, & Stanne, 1989)
whereby members perceive that individuals in the group possess unique resources that are
required for successful execution of the group task.

Second, given that task interdependence is shown to affect the group goal
commitment of individuals with high H-C (and possibly high conscientiousness)
negatively, this study extends the existing literature by suggesting some boundary
conditions to the motivating potential of task interdependence (e. g., Kerr & Bruun, 1983;
Kiggundu, 1983; Pearce & Gregersen, 1991; Saavedra et al., 1993; Wageman, 1995).
One boundary condition is the type of teams examined. Most existing studies on task
interdependence were conducted with traditional teams engaged full-time in the group
task, rather than a parallel team where members had their own tasks to manage as well.
Another boundary condition is individual differences, since the majority of the existing
studies did not examine how different individuals may react differently to task
interdependence. Thus, future research should continue to examine the generalizability of
the motivating potential of task interdependence to other types of teams, and to different
types of individuals.

Even though the results involving team leader status are generally disappointing,
some preliminary insights gained from this study should receive further attention in
future research, given the prevalence of hierarchical relationships in organizations. For
instance, the findings that individuals with high H-C, high conscientiousness, and high
agreeableness reported greater group goal priority when they perceived having a high
status leader, compared to individuals with low scores on this dimensions, suggest that

the personality of subordinates may be an important consideration for future leadership

126

research (Phillips & Bedeian, 1994). This is consistent with Hollander’s (1992)
recommendation that leadership research should develop a more active conception of the
follower role in reacting to leader qualities.

Finally, this study also raises questions concerning Hofstede’s (1984) dated
findings on I-C, and suggests that the conventional wisdom that US. is relatively more
individualistic than countries in Asia should be reexamined for three reasons. First, as
demonstrated in this study, there are at least two types of collectivism (horizontal and
vertical) that can distinguish two countries in opposite directions. Given that Hofstede’s
study used a broad-based measure of I-C, the finer distinctions in collectivism among
these countries are lost. Second, Hofstede’s study also did not differentiate between the
different states fiom which the US. responses were collected (understandably so, since
their his focus was to compare across countries). Yet, there can be substantial variation in
I-C even within the US. alone (Vandello and Cohen, 1999), with the Deep South
reported as being the most collectivistic. Michigan, according to the study, was ranked in
the middle. Hence, to maximize the range on I-C, future research involving cross-cultural
samples between US. and other countries could consider the US. state in which the
study will be conducted, or could collect data from multiple geographic areas in the US.

The third reason why Hofstede’s findings could be dated is due to the rapid
economic development of Asia in the past decade. As a result of the economic affluence
and the influx of modernity, the societal and personal values of Asians may have changed
gradually over the years to be more individualistic in nature.

In light of these reasons, and consistent with the advice of many scholars (e.g.,
Lytle et al., 1995), cross-cultural research should be theoretically grounded on specific

dimensions of cultural values, rather than based on the broad and fuzzy distinction

127

provided by national boundaries.

On Practice. What are some of the lessons learned fi'om this research and how can
they inform practice?

First, organizations adopting parallel teams need to be aware of the inevitable
trade-off in the performance outcomes of team members in their individual and group
tasks. This is an obvious, but critical point, since the trade-off can incur serious
consequences for the organization. Often times, employees may lack the bigger
perspective to realize which of the two tasks is more critical to the organization. When
left on their own, they may inappropriately invest more time and energy in one preferred
task at the expense of the other. This also suggests that communication between
supervisors of parallel teams and supervisors of the functional areas from which parallel
team members are drawn is critical. Such communication can potentially enhance
supervisors’ understanding of their employees’ workload and reduce the level of role
conflict experienced by parallel team members (Colquitt, 1999), which can otherwise be
detrimental to members’ affect and even physical health (Kahn, 1981). This
communication can also take the form of setting appropriate priorities for parallel team
members, so that the larger organizational interests can be made known to them. In this
way, organizations are more likely to reap the benefits intended from parallel teams.

This study also sheds light on two forms of interventions that managers may
consider in implementing parallel teams. The first relates to the selection of parallel team
members — which type of employees is more suitable for parallel teams? According to
this study, the most unequivocal personality found to predict group goal commitment,
and consequently, individual performance in the group task, is agreeableness. Hence,

agreeable employees are potentially good candidates, to the extent that they have the

128

requisite ability for the job, and that the group task is of considerable significance to the
organization. Choosing agreeable employees to be part of a team that serves only a minor
function, however, runs the risk of detracting them from their own regular jobs, which
may be costly to the organizations.

Conscientious employees may be the most ideal in terms of managing a balance
between both individual and group tasks. On the one hand, they are unlikely to neglect
their individual tasks since they generally tend to view their personal goal achievement to
be more important than their group goal achievement. On the other hand, their
dependable nature is likely to induce them to contribute more to the group compared to
the less conscientious individuals.

Choosing employees based on their collectivism values may be a trickier issue.
Results of this study demonstrate that it is important for managers to distinguish between
two types of collectivism — horizontal (a concern for harmony and cooperation) and
vertical (a concern for group solidarity and authority). Such a distinction is important
because given the “part-time” nature of parallel teams, vertical collectivists may view
such teams as peripheral and an interference to their personal jobs. Horizontal
collectivists, on the other hand, seem to embrace parallel teams better as they are more
likely to demonstrate greater commitment to the group goal compared to those with low
H-C.

The second form of intervention is the structural design of parallel teams. Results
of this study imply that a blanket adoption of highly interdependent tasks in the team, or
appointing a senior company official to head the team, may not be wise. For instance,
these team interventions may actually detract parallel team members from their personal

duties. Besides, not everyone responds favorably to them. For instance, task

129

interdependence may backfire for parallel team members are who are high in
conscientiousness. Similarly, the motivating potential of a high status team leader may
differ across countries, thus requiring the manager to consider carefully the local cultural
norms and values toward authority and power. One finding that emerged from this study
is that high team leader status elicited opposite reactions from subjects in the US. and
Singapore, thus highlighting the importance of tailoring the leadership intervention to the
local context.
Limitations

There are several issues and limitations in this study that deserve some discussion.
First, the effect sizes reported in this study are generally small, thereby raising the
question of how important and relevant these findings are. However, using Abelson’s
(1985) batting example, minuscule values of variance explained in one single episode can
cumulate over time to provide a substantial amount of explanation for a phenomenon in
the long run. I contend that this cumulation of small effect sizes can indeed occur in
parallel teams where members work on a series of short-term projects. Besides, the
practical significance of these findings can be substantial if the parallel team is formed
for a critical purpose where poor performance can result in huge costs to the organization.

The second issue deals with the experimental nature of the study. While the
laboratory setting provided a tighter control over potential confounds or contaminants, it
also resulted in a contrived parallel team setting. Specifically, I had used two similar
tasks for the individual and group work. In real life, however, many parallel teams are
formed for projects that may differ substantially from the regular work duties that the
group members have. Nonetheless, having two similar tasks was imperative, given that

my primary interest lies with team members’ values/ personality and the team’s structural

130

features, not the nature of the tasks. Also, the temporary nature of the teams (as with
other research studies conducted with teams in a laboratory setting) meant that the study
could not model the complex dynamics of real-life teams. However, this criticism may be
mitigated by the fact that parallel teams in organizations, unlike full-time teams, typically
have shorter life span characterized by relatively limited interactions among group
members.

Third, it should be noted that Singapore and the US. differed on a variety of
cultural and contextual dimensions other than the cultural values examined in this model.
Even though the predictions of this study were focused on the variation in specific
cultural dimensions, some country influences were evident, and for which I had no clear
explanation.

Future direction;

Given the paucity of research conducted with parallel teams, there are still many
questions that future research can and should address. Obviously, one important need is
to replicate the findings reported here in order to test the robustness of the experimental
results. For instance, a field survey conducted with employees who are involved in
parallel team initiatives can provide critical evidence concerning the discriminability of
individual goal commitment and group goal commitment. The effects of task
interdependence and team leader status should also receive further investigation, perhaps
with more levels of manipulation, and/or with more fine-grained dimensions. For
instance, future research can manipulate more contrasting levels of task interdependence
(e. g., pooled versus team interdependence -- Saavedra etal., 1993) and team leader status
(e.g., supervisor versus peer vs subordinate) in order to increase the variance of the

independent variables. Research can also investigate other forms of interdependence

131

(e.g., task versus goal interdependence; Wageman, 1995) and various operationalizations
of status that may be relevant to the culture that is being sampled (e. g., gender, education,
age; Earley, 1999). For instance, according to Earley’s (1999) study, gender,
race/ethnicity and age are three demographic characteristics that managers in the US.
cited as important markers that are used to judge the amount of status (or relative power)
a stranger has. Given that the team leader in the current dissertation is unknown to
participants, perhaps fiiture study can either replace or combine the current status
manipulation with these demographic variables to strengthen the manipulation of leader
status.

Based on this dissertation, several extensions can also be made to enrich our
understanding of parallel teams. To begin with, future research should establish
empirically one assertion that I have made in this dissertation: that parallel team members
may have viewed their teams more as an “outgroup” than an “ingroup” due to the
temporary and part-time nature of the team (compared to members in traditional full-time
teams). Although this may seem superfluous given the obvious differences in the nature
of parallel teams versus full-time traditional teams, this assertion remains a theoretical
speculation until research has empirically proven it to be true. Besides, the degree to
which parallel team members perceive their team to be an ingroup/outgroup may itself be
a substantive variable of interest. In particular, parallel team members who score high in
vertical collectivism may be especially affected by their perception of ingroup-outgroup
membership (assuming there will be sufficient variance in the ingroup-outgroup variable
even within parallel teams). Thus, future research should consider incorporating
perceptions of ingroup-outgroup membership not only to establish the difference between

parallel teams and full-time teams, but also to test for its moderating potential with

132

regards to the values and personality of parallel team members.

One boundary condition imposed in the current dissertation is the fixed amount of
time given to participants to work on the individual and group tasks. In real life, however,
employees often have a greater flexibility in expanding the amount of time for their work.
Thus, future research may relax this boundary condition and design studies that allow
participants to work longer on their tasks if they choose to. Such studies can be
potentially insightful for ascertaining what personality or value characteristics may
predict parallel team members’ willingness to work longer hours in order to complete
both their individual and group assignments.

Another avenue for future research is to adopt a different theoretical angle to
studying parallel teams. Role theory (Katz & Kahn, 1978), for instance, may provide a
potentially interesting and relevant perspective. Unlike the current view of parallel team
members as relatively passive role-takers (who try to manage between their individual
and group tasks), future studies may examine parallel team members as active role-
makers who may manage their multiple roles by attempting to influence various role-
senders (e. g., direct supervisor and team leader) (e.g., Graen, 1976; Graen & Scandura,
1987). The presence of multiple roles also suggests that role constructs such as role
conflict or role overload maybe important mediating mechanisms to parallel team
members’ stress and other negative reactions.

Following from the above, a relevant framework for examining parallel team
members’ potential dissatisfaction is the exit-voice-loyalty-neglect framework
(Hirschman, 1970; Rusbult, Farrell, Rogers & Mainous, 1988), which offers a
comprehensive typology of employee reactions to workplace dissatisfaction. Specifically,

exit refers to leaving the organization (or in this case, the team). Voice refers to active

133

efforts undertaken to improve conditions, such as discussing problems with supervisors.
Loyalty means passively but optimistically waiting for conditions to improve. Neglect
refers to passively allowing conditions to deteriorate through reduced interest or effort.
The performance in the group task variable examined in this dissertation is perhaps
closest to “neglect” in the framework, since poor performance in the group task is
symptomatic of reduced interest and/or effort in the group task. Future research may
consider looking into behaviors that are opposite of neglect, such as helping and related
extra-role behaviors in parallel teams (Van Dyne et al., 1995). Or, future research can
also look into other outcome variables associated with exit, voice, and loyalty. This is
important because exit has been found to be a problem plaguing the quality circles
examined by Adam (1991); voice may be crucial for parallel team members especially if
they are to seek help or solutions when faced with role overload or conflict; and loyalty is
essential if it is desirable that members stick with the teams regardless of their individual
workload.

Future efforts should also be directed to focus on the process in which parallel
team members allocate their time and effort to their individual and group tasks. This is an
interesting question that can yield critical insight as to how parallel team members cope
with their multiple-goal striving. For instance, the criterion that members use to decide
when to stop working on one task and move to the other may differ from one person to
another. Some may use the degree of completion on one task as a marker, such as moving
on to the group task only when the individual task has been completed. Others may use
time as a marker, such as moving on to the group task when half the time has been used
up. These different processes of managing the two tasks not only reflect the motivational

bases of parallel team members, but can also have potential implications on their task

134

performance in dynamic real-life situations. For instance, parallel team members
described in the first instance may never get down to working on their group task if their
individual work is continually being piled on them. Parallel team members described in
the second instance may have less than the expected amount of time to work on their
group task, if the total amount of time allotted is suddenly shortened.

Future research should also examine more structural interventions, both at the
team and organizational level, for their effects on parallel team members’ commitment to
their group task. At the team level, one potential factor to consider is accountability to
team members or to the team leader for one’s portion of the work. Since accountability
has been found to reduce social loafing in traditional teams (e. g., Earley, 1989), future
research may examine whether such benefits extend to parallel teams. Also, having
parallel team members be accountable to the group leader can also have the benefit of
enhancing the relevance of the status of the leader.

In addition, future research should also consider the broader organizational
context, which can have a substantial impact on the amount of legitimacy and resources
accorded to parallel teams. Potentially relevant factors include organizational culture
(group-oriented versus individual-oriented), reward system (independent versus
interdependent versus mixed), the extent to which parallel teams are used in the
organizations, and training programs in time and stress management.

Finally, in order to arrive at a robust set of findings and recommendations for
parallel teams, future research need to engage in various methodologies (e.g., survey,
laboratory experiment, experimental simulation, field study with primary and secondary
data, field experiment, judgment task, etc.). Such triangulation of results can enhance the

internal and external validity of our findings (Campbell & Fiske, 1959; Scandura &

135

Williams, 2000), and thus, increases the confidence and clarity of our recommendations
to managers currently involved or interested in parallel teams.
Conclusion

Results of this dissertation demonstrated that individual differences and goal
commitment/priority are important factors that can influence parallel team members’
performance in their group task. Nonetheless, the unsupported and unexpected findings
of task interdependence and leader status (and their interactions with individual
differences) also raise many interesting research questions that future research can
examine. Given the increasing popularity of parallel teams in organizations, and the
paucity of studies conducted on them, there is a real need for academic research to gain a
greater understanding of how parallel team members cope with their individual and group

tasks.

136

APPENDICES

137

APPENDIX A

EXPERIMENTAL MATERIALS

138

CONSENT FORM

This study simulates a day at work as a manager.

If you choose to participate, you will be asked to take part in a 90-minute study
conducted on campus, and to complete a personality questionnaire in class. In
exchange for your participation, you will receive the FULL miscellaneous points
for your MGT 325 class. You also stand a chance to win a small cash award if
you do well. If you choose not to participate in this study, you can approach your
instructor for alternative miscellaneous credit exercises. Hence, your participation
in this research is STRICTLY voluntary. You are free to decline to answer any
questions or terminate your participation at any time. YOUR RESULTS WILL BE
KEPT PRIVATE AND CONFIDENTIAL, to the maximum extent allowable by law.

If you have any questions or concerns concerning this study, please contact Linn
Van Dyne in the Management Department at 353-5415, or at
vandyne@msu.edu.

 

CONSENT STATEMENT

My signature below indicates that the procedures of the study have been
explained to me and that my participation is strictly voluntary.

Participant’s Name (please print):

 

Participant’s Signature:

 

Date: Group #:

 

Gender (M or F):

No. of years of working experience (if any) :

139

SAMPLE BACKGROUND INFORMATION
(High Team Leader Status - Support Department)

Instruction: In the next 90 minutes, please imagine yourself assuming the role of a junior
manager in a large manufacturing firm. Please read the following information carefully in
order to better understand your job scope in the company. Please also refer to the
organizational chart provided here so that you know where you are located in the
hierarchies of the firm.

Company background. HomeTech Manufacturing Inc. is a large public listed
company that manufactures a wide range of household products. HomeTech has
approximately 20,000 employees located in various parts of the US. You were
hired by the Michigan office four years ago, and you are now a junior manager
in the SUPPORT department. You like the corporate culture, and have found
your job as a manager challenging and fulfilling (please turn over to see the
organizational chart).

As a junior manager, your primary job responsibilities include the following:

1. Oversee subordinates under your charge. You need to plan ahead and
delegate work to the appropriate employees, and to monitor their
performance. Currently, you have 20 subordinates under you.

2. Ensure efficient and effective services to other departments. You occasionally
initiate meetings with the heads of other departments for feedback and new
ideas.

3. Train newcomers. You need to make sure that newcomers fit well into the
department.

4. Evaluate your subordinates and provide merit bonus recommendations at
year-end.

You report primarily to Pat Neuwen, the senior manager in-charge of the
support department. You are known to him as a capable employee with
tremendous potential for advancement up the corporate ladder.

In addition to your regular job responsibilities described above, you are also a
member of the campus recruitment team. Your team leader, Chris
SandersonI is the vice-president of the firm. Chris was the one who
championed the idea of a recruitment team that is staffed by managers from
different divisions. Currently, the recruitment team consists of junior managers
from six different divisions of the firm, including yourself. The scope of the
recruitment team includes three major phases: planning campus visits, screening
applicants for site visits and interviews, and finally, recommending hiring
decisions to top management. By this time of the year, you have finished the first
two phases of the recruitment process, and are expected to proceed to
evaluating applicants soon.

140

0.5: 9:“ 3%

  

$2 $2 $2 E2 .92
L a :. :. :. :,

 

.92 $2
:. :

cw: a2 cw: $2 $2 $2
:. :. .: .:. S S

 

 

9::

Lawn—BE LoEom

Atomaswv
commas—2 SEom

 

 

 

305::— a. 88<v
commas—2 Siam

Amines—m .Smbm a. m:
Lowe—32 5.5m

Awaiazv
Swag—«2 LoEom

 

 

 

Ease

Cowman—2 coEom

 

 

Acouozuoiv
Lamaze—z noEom

 

 

_ a

E

 

_

£83393 85:58:: 25>

 

 

_
_

H5285 92>

 

 

 

 

 

 

 

 

$55 59:25
EoEmEm

 

 

$.50

.332 anu 25.53.82 25>

.07: QZEDEUac—DZE EUHHH—ZO:
mDh<Hm “HA—(dd =0:— | EEO A<ZO~H<N~Z<U¢O ”aw—gm

141

SAMPLE BACKGROUND INFORMATION
(Low Team Leader Status — Support Department)

Instruction: In the next 90 minutes, please imagine yourself assuming the role of a junior
manager in a large manufacturing firm. Please read the following information carefully in
order to better understand your job scope in the company. Please also refer to the
organizational chart provided here so that you know where you are located in the
hierarchies of the firm.

Company background. HomeTech Manufacturing Inc. is a large public listed
company that manufactures a wide range of household products. HomeTech has
approximately 20,000 employees located in various parts of the US. You were
hired by the Michigan office four years ago, and you are now a junior manager
in the SUPPORT department. You like the corporate culture, and have found
your job as a manager challenging and fulfilling (please turn this page over to see
the organizational chart).

As a junior manager, your primary job responsibilities include the following:

5. Oversee subordinates under your charge. You need to plan ahead and
delegate work to the appropriate employees, and to monitor their
performance. Currently, you have 20 subordinates under you.

6. Ensure efficient and effective services to other departments. You occasionally
initiate meetings with the heads of other departments for feedback and new
ideas.

7. Train newcomers. You need to make sure that newcomers fit well into the
department.

8. Evaluate your subordinates and provide merit bonus recommendations at
year-end.

You report primarily to Pat Neuwen, the senior manager in-charge of the
support department. You are known to him as a capable employee with
tremendous potential for advancement up the corporate ladder.

In addition to your regular job responsibilities described above, you are also a
member of the campus recruitment team. Your team leader, Chris
SandersonI is the junior mamer in the Hgman Resource department Chris
was the one who had to coordinate with all team members on the activities of the
campus recruitment team. Currently, the recruitment team consists ofjunior
managers from six different divisions of the firm, including yourself. The scope of
the recruitment team includes three major phases: planning campus visits,
screening applicants for site visits and interviews, and finally, recommending
hiring decisions to top management. By this time of the year, you have finished
the first two phases of the recruitment process, and are expected to proceed to
evaluating applicants soon.

142

$.50 £252 :38
0:25:28.— 58>

95: 9E =e>

  

um: em: a: $2 aw:
:. L _. ._ _. .— _. .1.

 

T

cw: cw: hm: hm:
:. b. :. .:.

LwE E2 E2 cw:
: La :. :

 

 

3::

Lemma“: 8_:om

€93sz
Lemmas: LoEom

 

 

 

 

325:5 ow 38$
Emacs—2 LoEom

 

 

Essa: seem a m:
Swag—2 LoEow

 

 

REESE
Downs“: LoEom

 

 

Esme
comm—BE LoEum

 

 

$532625
Lowe—52 LoEom

 

 

S

Em

£83.59; 03:58:: .:S>

mDh<Hm Amman": BOA I .5350 A<ZOE<N~Z<UKO mus—gm

_

 

“528$ 92>

 

 

 

 

 

$050 322%
E0285

 

 

_

.07: UZEDS<§ =0HH§OI

143

APPENDIX B

MEASURES

144

 

Individualism-Collectivism Scales

ingelis eLal.’s (1995) H-C items

If a co-worker gets a prize, I would feel proud.

The well-being of my co-workers is important to me.

To me, pleasure is spending time with others.

I feel good when I cooperate with others.

It is important for me to maintain harmony within my work group.

My happiness depends very much on the happiness of those around me.
If a group member were in difficulty, I would help within my means.

I like sharing little things with my group members.

P3999939!"

S_in_gelis et al.’s (1995) V—C Items

9. Parents and children must stay together as much as possible.

10. It is my duty to take care of my family, even when I have to sacrifice what I want.
11. Family members should stick together, no matter what sacrifices are required.

12. I would do what would please my family, even if I detested that activity.

13. I would sacrifice an activity that I enjoy very much if my family did not approve it.
14. Children should be taught to place duty before pleasure.

15. I hate to disagree with others in my group.

16. I usually sacrifice my self interests for the benefit of my group.

Singelis et al.’s (1995) PM Items

17. I often “do my own thing.”

18. I prefer to be direct and forthright when I talk with people.
19. One should live one’s life independently of others.

20. What happens to me is my own doing.

21. I enjoy being unique and different from others in many ways.
22. I am a unique individual.

23. I like my privacy.

24. When I succeed, it is usually because of my abilities.

flgelis et al.’s (1995) V-I Items

25. Winning is everything.

26. It annoys me when other people perform better than I do.

27. It is important to me that I do my job better than others.

28. I enjoy working in situations involving competition with others.
29. Competition is the law of nature.

30. When another person does better than I do, I get tense and aroused.
31. Without competition, it is not possible to have a good society.

32. Some people emphasize winning. I am ngt one of them.

Triandis and Gelfgd’s (1998) Items

33. It is important to me that I respect the decisions made by my groups (V-C).
34. I’d rather depend on myself rather than others. (H-1)

35. I rely on myself most of the time; I rarely rely on others. (H-1)

36. My personal identity, independent of others, is very important to me. (H-I)

145

Other I-C Item_s Reported in Wagner’s (1995) Study

37. Only those who depend on themselves get ahead in life.

38. To be superior, a person must stand alone.

39. If you want something done right, you’ve got to do it yourself.

40. What happens to me is my own doing.

41. In the long run, the only person you can count on is yourself.

42. I prefer to work with others in a group rather than working alone.

43. Given the choice, I would rather do a job where I can work alone rather than
doing a job where I have to work with others in a group.

44. Working with a group is better than working alone.

45. People should be made aware that if they are going to be a part of a group
then they are sometimes going to have to do things they don’t want to do.

46. People who belong to a group should realize that they’re not always going to get
what they personally want.

47. People in a group should realize that they sometimes are going to have to make
sacrifices for the sake of the group as a whole.

48. People in a group should be willing to make sacrifices for the sake of the group’s
well-being.

49. A group is more productive when its members do what they want to do rather than
what the group wants them to do.

50. A group is most efficient when its members do what they think is best rather than
doing what the group wants them to do.

51. A group is more productive when its members follow their own interests and
concerns.

Individual (Group) Goal Commitment (Adapted from Hollenbeck et al., 1989)

1. It’s hard to take the goal for my individual on task seriously.

2. It’s unrealistic for me to expect to reach the goal for my individual (group) task.
3. It is quite likely that the goal for my individual (group) task may need to be
revised, depending on how things go.

Quite frankly, I don’t care if I achieve the goal for my individual (goup) task or
not.

I am strongly committed to pursuing the goal for my individual (group) task.

It wouldn’t take much for me to abandon the goal for my individual rou task.
I think the goal for my individual (group) task is a good goal to shoot for.

I am willing to put forth a great deal of effort beyond what I’d normally do to
achieve the goal for my individual rou task.

There is not much to be gained by trying to achieve the goal for my individual

(gimp) task.

:5

”>195”

.0

146

REFERENCES

Abelson, R. P. (1985). A variance explanation paradox: When a little is a lot.
Psychological Bulletin. 97. 129-133.

Adam, B. E. Jr. (1991). Quality circle performance. Journal of Management, 17, 25-
39.

Aguinis, H. (1995). Statistical power problems with moderated multiple regression
in management research. J ourrlal of Management, 21, 1141-1158.

Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure,
process, and content. Psychological Bulletin. 120, 33 8-375.

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

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

Barrick, M. R., & Mount, M. L. (1993). Autonomy as a moderator of the
relationships between the Big Five personality dimensions and job performance. Journal
of Applied Psychology, 78, 111-118.

Barrick, M. R., & Mount, M. L. (1995). The Personal Chflcteristics Inventory
manual. Unpublished manuscript, University of Iowa, Iowa City.

Barrick, M. R., Mount, M. K., & Strauss, J. P. (1993). Conscientiousness and
performance of sales representatives: Test of the mediating effects of goal setting. Journal
of Applied Psycholggy, 78, 715-722.

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

Berger, P. L., & Luckmann, T. (1967). The socigl construction of reality. London:
Penguin.

Bettenhausen, KL. (1991). 5 years of groups research: What we have learned and
what needs to be addressed. Joumgl of Management, 17, 345-381.

Bochner, S., & Hesketh, B. (1994). Power distance, individualism/collectivism, and
job-related attitudes in a culturally diverse work group. Journ_al of Cross-Cultural

Psychology, 25, 233—257.

147

 

Borman, W. C., White, L. A., Pulakos, E. D., & Oppler, S. H. (1991). Models of
supervisor job performance ratings. Journ_al of Applied Psychology, 76, 863-872.

Boyacigiller, N. A., & Adler, N. J. (1991). The parochial dinosaur: Organizational
science in a global context. Academy of Management Review. IQ 262-290.

Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validity by
the multitrait-multimethod matrix. Psychological Bulletin, 56. 81-105.

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.

Campion, M. A., Medsker, G. J ., & Higgs, A. C. (1993). Relations between work
group characteristics and effectiveness: Implications for designing effective work groups.
Personnel Psychology, 46, 823-850.

Carver, C. S., & Scheier, M. F. (1981). Action and self-regrfltion: A control-theory
approach to human behavior. New York: Springer-Verlag.

Chan, D. K.S., Gelfand, M. J., Triandis, H. C., & Tzeng, O. (1996). Tightness-
looseness revisited: Some preliminary analyses in Japan and the United States.
International Journjal of Psychology. 31. 1-12.

Chen, C. C., Meindl, J. R., & Hunt, R. G. (1997). Testing the effects of vertical and
horizontal collectivism: A study of reward allocation preferences in China. Journal of
Cross Cultural Psychology, 28, 44-69.

Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity,
and compliance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds), The handbook of
social psychology (4th ed., vol. 2, pp. 151-192). Boston, MA: McGraw Hill.

Cohen, J. (1987). Statisticalfi power analysis for the behavioral sciences (2"d ed.).
Hillsdale, New Jersey: Lawrence Erlbaum Associates.

Cohen, J., & Cohen, P. 1983. Applied multiple reggssion/correlation anmsis for
the behavioral sciences (2"d ed.). Hillsdale, New Jersey: Lawrence Erlbaum Associates.

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

Colquitt, J. A. (1999). The impact of procedural justice in teams: An analysis of
individual task and team moderators. Unpublished doctoral dissertation, Michigan State
University.

 

148

Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-
PI-R) and NBC Five-Factor Inventory (NEO-FFI) professional manual. Odessa, FL:
Psychological Assessment Resources.

Cotton, J. L. (1993). Employee involvement: Methods for improving performance
and worl_(_attitudes. Newbury Park, CA: Sage Publications.

Cropanzano, R., James, K., & Citera, M. (1992). A goal hierarchy model of
personality, motivation and leadership. In L. L. Cummings and B. Staw (Eds), Research
in organizational behavior (vol. 15, pp. 267-322). Greenwich, CT: JAI Press.

Crown, D. F., & Rosse, J. G. (1995). Yours, mine, and ours: Facilitating group
productivity through the integration of individual and group goals. Organizational
Behavior and Human Decision Processes. 64. 138-150.

Dachler, H. P., & Mobley, W. H. (1973). Construct validity of an instrumentality-
expectancy-task-goal model of work motivation. Journ_an Applied Psychology. 58. 397-
418.

Digman, J. M., & Inouye, J. (1986). Further specification of the five robust factors
of personality. Journan Personality and Social Psychology, 50, 116-123.

Digman, J. M., & Takemoto-Chock, N. K. (1981). Factors in the natural language
of personality: Reanalysis, comparison, and interpretation of six major studies.
Multivariate Behaviopal Research. 16. 149-170.

Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation.
Hillsdale, NJ: Lawrence Erlbamn Associates.

Eagly, A. H., & Wood, W. (1991). Explaining sex differences in social behavior: A
meta-analytic perspective. Personalm and Social Psycholw Bulletin. 17. 306-315.

Earley, P. C. (1989). Social loafing and collectivism: A comparison of the United
States and the People’s Republic of China. Administrative Science Quarterly. 34. 565-
581.

Barley, P. C. (1993). East meets West meets Mideast: Further explorations of
collectivistic and individualistic work groups. Academy of Management Journal. 36. 319-
348.

Earley, P. C. (1999). Playing follow the leader: Status-determining traits in relation
to collective efficacy across cultures. Organizational Behavior and Human Decision
Processes. 80. 192-212.

Barley, P. C., & Gibson, C. B. (1998). Taking stock in our progress on

individualism-collectivism: 100 years of solidarity and community. Journal of
Management, 24, 265-304.

149

Erez, M., & Earley, P. C. (1987). Comparative analysis of goal setting strategies
across cultures. Jouraal of Applied Psycholgy. 72. 658-665.

Erez, M., & Earley, P. C. (1993). Culture. self-identig. and work. New York:
Oxford University Press.

Erez, M., Gopher, D., & Arzi, N. (1990). Effects of goal difficulty, self-set goals,
and monetary rewards on dual task performance. Organizational Behavior and Human
Decision Processes. 47. 247-269.

Erez, M., & Somech, A. (1996). Is group productivity loss the rule or the

exception? Effects of culture and group-based motivation. Academy of Management
Journal 39 1513-1537.

 

Erez, M., & Zidon, I. (1984). Effect of goal acceptance on the relationship of goal
difficulty to performance. Journal of Applied Psychology. 69. 69-78.

Farh, J. L., Earley, P. C., & Lin, S. C. (1997). Impetus for action: A cultural
analysis of justice and organizational citizenship behavior in Chinese Society.
Administrative Science Quarterly. 42. 421-444.

Frost, P. J., & Mahoney, T. A. (1976). Goal setting and the task process: An
interactive influence on individual performance. Organizatioaal Behavior and Humap
Performance 17 328-350.

 

Gagne, R. M., & Fleishman, E. A. (1959). Psychology and human performance.
New York: Holt.

Gellatly, I. R. (1996). Conscientiousness and task performance: Test of a cognitive
process model. Journal of Applied Psychology, 81, 474-482.

Gladstein, D. L. (1984). Groups in context: a model of task group effectiveness.
Administrative Science Quarterly, 29, 499-517.

Graen, G. B. (1976). Role making processes within complex organizations in M. D.

Dunnette (Ed.), Handbook of industrial and organizational psychology (pp. 1201-1245).
Chicago: Rand McNally.

Graen, G. B., & Scandura, T. A. (1987). Toward a psychology of dyadic
organizing. In L. L. Cummings and B. A. Staw (Eds), Remch in Organizational
Behavior (vol. 9, pp. 175-208). Greenwich, CT: JAI Press.

Graziano, W. G., Hair, E. C., & Finch, J. F. (1997). Competitiveness mediates the

link between personality and group performance. Journal of Personality Ad Social
Psychology, 73, 1394-1408.

150

Guzzo, R. A., & Shea, G. P. (1992). Group performance and intergroup relations in
organizations. In M. D. Dunnette & L. M. Hough (Eds.), H_andbool_<&f industrial and
organizatiogl psychology (vol. 3, pp. 269-313). Palo Alto, CA: Consulting Psychologists
Press.

Hackman, J. R. (1987). The design of work teams. In J. W. Lorsch (Ed.), Handbook
of organizational behavior (pp. 315-342). Englewood Cliffs, NJ: Prentice-Hall.

Hirschman, A. O. (1970). Exit. voice. and loyalty: Respon_ses to decline in firms,
organizations and statea. Cambridge, Mass: Harvard University Press.

Hofstede, G. (1984). Culture’s consequences: International differences in work-
related values. Beverly Hills: Sage.

Hollander, E. P. (1992). Leadership, followership, self, and others. Leadership
Quarterly, 3(1), 43-54.

Hollenbeck, J. R., & Klein, H. J. (1987). Goal commitment and the goal-setting
process: Problems, prospects, and proposals for future research. Journal of Applied
Pswholgy. 72. 212-220.

Hollenbeck, J. R., Klein, H. J., O’Leary, A. M., & 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. Joumal of Applied Psychology, 74,
18-23.

Horn, J. L., & McArdle, J. J. (1992). A practical and theoretical guide to
measurement invariance in aging research. Expgrimental Aging Research. 18. 117-144.

Hough, L. M., Eaton, N. K., Dunnette, M. D., Kamp, J. D., & McCloy, R. A.
(1990). Criterion-related validities of personality constructs and the effect of response
distortion on those validities. Journal of Applied Psychology. 75. 581-595.

Hui, C. H. (1988). Measurement of individualism-collectivism. Journal for
Research in Personality, 22, 17-36.

Ilgen, D. R., & Hollenbeck, J. R. (1991). The structure of work: Job design and
roles. In M. D. Dunnette & L. M. Hough (Eds), Handboola of industrial and
organizational psychology_(2“‘l ed., vol. 2, pp. 165-207). Palo Alto, CA: Consulting
Psychologists Press.

Ilgen, D. R., & Pulakos, E. D. (1999). The changm’ g nature of performance:
Implications for staffing, motivation, and development. San Francisco, CA: Jossey-Bass.

151

Ivancevich, J. M., & McMahon, J. T. (1977). Black-white differences in a goal
setting program. Organizational Behavior and HumaaPerformance. 20. 287-300.

Johnson, D. W., Johnson, R., & Stanne, M. B. (1989). Impact of goal and resource
interdependence on problem-solving success. Journal of ScLial Psychology. 129. 621-
629.

Kahn, R. L. (1981). Work_and heaLh, New York: Wiley.

Kanfer, R. (1990). Motivation theory and industrial and organizational psychology.
In M. D. Dunnette and L. Hough (Eds), Handbook of industrial and organizational
psychology (2"d ed., vol. 1., pp. 75-170). Palo Alto, CA: Consulting Psychologists Press.

Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An
integrative/aptitude-treatment interaction approach to skill acquisition. Journal of Applied
Journal 74 657-690.

 

Karau, S. J ., & Williams, K. D. (1993). Social loafing: A meta-analytic review and
theoretical integration. Journal of Personality and Social Psychology, 65, 681-706.

Katz, D., & Kahn, R. L. (1978). The sociaipsycholngy of organizations (2nd ed.).
New York: Wiley.

Kerlinger, F. N. (1986). Foundations of behavioral reseaLh. New York: CBS
College Publishing.

Keman, M. C., & Lord, R. G. (1990). Effects of valence, expectancies, and goal-
performance discrepancies in single and multiple goal environments. Journal of Applied
Psychology. 7; 194-203.

Kerr, N. L., & Bruun, S. E. (1983). Dispensability of member effort and group
motivation losses: F rec-rider effects. Journal of Personality and SociaILPsychology. 44.
78-94.

 

Kiggundu, M. N. (1981). Task interdependence and the theory of job design.
Academy of Management Review. 6. 499-508.

Kiggundu, M. N. (1983). Task interdependence and job design: Test of a theory.
Organizational Bmvior and Human Performame. M. 145-172.

Kirkman, B. L., & Shapiro, D. L. (1997). The impact of cultural values on
employee resistance to teams: Toward a model of globalized self-managing work team
effectiveness. Academy of Management Review. 22. 730-757.

Klein, H. J. (1991). Further evidence of the relationship between goal setting and

expectancy theories. Organizational Behavior and Human Decision Processes. 49, 230-
257.

152

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 ., Wesson, M. J., Hollenbeck, J. R., Wright, P. M., & DeShon, R. P. (In
press). The assessment of goal commitment: A measurement model meta-analysis.
Organizational Behavior and Human Decision Processes.

Latane, B., Williams, K. D., & Harkins, S. (1979). Many hands make light the
work: The causes and consequences of social loafing. Journal of Personflg and Social

Psychology, 37, 822-832.

Latham, G. P., & Locke, E. A. (1991). Self-regulation through goal setting.
Organizational Behavior and Human Decision Processes, 50. 212-247.

Lawler, E. E. III. (1999). Employee involvement makes a difference. Journal for

Mia; and Participation, 22(5), 18-20.

Lawler, E. E. III, & Cohen, S. G. (1992). Desigm’ g pay systems for tea_m_s. CEO
Publication.

Lawrence, P. R., & Lorsch, J. W. (1969). Developing organizations: Diagnosis and
action. Reading, MA: Addison-Wesley.

LePine, J. A., & Van Dyne, L. (2001). Voice and Cooperative Behavior as
Contrasting Forms of Contextual Performance: Evidence of Differential Relationships
with Big-5 Personality Characteristics and Cognitive Ability. Journal of Applied

Psychology, 86, 326-336.

Leung, K., & Bond, J. W. (1989). On the empirical identification of dimensions
for cross-cultural comparisons. Journal of Cross-CulturJal Psychom. 20. 133-151.

Levine, J. M., & Moreland, R. L. (1998). Small groups. In G. T. Gilbert, S. T. Fiske,
& G. Lindzey (Eds). Handbook of Social Psychology (4th ed., vol. 2, pp. 415-469).
Boston: McGraw Hill.

Locke, E. A. (1968). Toward a theory of task motivation and incentives.
Organizational Behavior and Humg Performance. 3. 157-189.

Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task
mrformance. Englewood Cliffs, NJ: Prentice-Hall.

Locke, E. A., Latham, G. P., & Erez, M. (1988). The determinants of goal
commitment. Apademy of Managment Review. 13. 23-39.

Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. (1981). Goal setting and
task performance: 1968-1980. Psychological Bulletin. 90. 125-152.

153

Locke, E. A., Smith, K. G., Erez, M., Chah, D. 0., & Schaffer, A. (1994). The
effects of intra-individual goal conflict on performance. Journal of Management, 20, 67-
91.

Lytle, A. L., Brett, J. M., Barsness, Z. I., Tinsley, C. H., & Janssens, M. (1995). A
paradigm for confirmatory cross-cultural research in organizational behavior. In B. M.
Staw (Ed.), Research in organizational behavior (vol. 17, pp. 167-214). Greenwich, CT:
JAI Press.

Magjuka, R. J ., & Baldwin, T. T. (1991). Team-based employee involvement
programs: Effects of design and administration. Personnel Psychology. 44. 793-812.

Maier, N. R. F. (1955). Psychology in industry. Boston: Houghton-Mifflin.

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for
cognitions, emotion, and motivation. Psychological Review, 98, 224-253.

Matsui, T., Kakuyama, T., Lou, M., & Onglatco, L. U. (1987). Effects of goals and
feedback on performance in groups. Journal of Applied Psychology, 72, 407-415.

May, D. R., & Schwoerer, C. E. (1994). Employee health by design: Using
employee involvement teams in ergonomic job redesign. Personnel Psychology. 47. 861-
876.

Mche, R. R., & Costa, P. T. Jr. (1985).Updating Norman's "adequacy taxonomy":
Intelligence and personality dimensions in natural language and in questionnaires.

Journal of Personality and Social Psychology. 49. 710-721.

McCrae, R. R., & Costa, P. T. Jr. (1997). Personality trait structures as human
universal. American Psychologist. 5; 509-516.

Mead, M. (1967). CoowaLion and competition among primitive people. Boston:
Beacon Press.

Michener, H. A., & Burt, M. R. (1975). Components of “authority” as determinants
of compliance. Journg of Personality and Social Psychology. 3L 606-614.

Mischel, W. (1977). The interaction of person and situation. In D. Magnusson & N.
S. Endler (Eds). Personality at the crossroads: Current issues in interactional psychology
(pp. 333-352). Hillsdale, NJ: Erlbaum.

Mitchell, T. R., & Silver, W. S. (1990). Individual and group goals when workers

are interdependent: Effects on task strategies and performance. Journal of Applied
Psycholpgy, 75. 185-193.

154

Mount, M. K., & Barrick, M. R. (1995). The big five personality dimensions:
Implications for research and practice in human resources management. Research in
Personnel and Human Resources Mangement. 13. 153-200.

Naylor, J. C., Pritchard, R. D., & Ilgen, D. R. (1980). A theory of behayior in
organizations. New York: Academic Press.

Neuman, G. A., Wagner, S. H., & Christiansen, N. D. (1999). The relationship
between work-team personality composition and the job performance of teams. Group

and Organization Management, 24, 28-45.

Neuman, G. A., & Wright, J. (1999). Team effectiveness: Beyond skills and
cognitive ability. Journal of Applied Psychology. 84. 376-689.

Ng, K. Y., & Van Dyne, L. (2001). Individualism-collectivism as a boundary
condition for effectiveness of minority influence in decision making. Organizational
Behavior and Human Decision Processes. 84. 198-225.

Oldham, G. R. (1975). The impact of supervisory characteristics on goal
acceptance. Academy of Management Jouraal. 18. 461 -475.

Ones, D. 8., Mount, M. K., Barrick, M. R. & Hunter, J. E. (1994). Personality and
job performance: A critique of the Tett, Jackson, and Rothstein (1991) meta-analysis.

Personnel Psychology, 47, 147-156.

Orbell, J ., & Dawes, R. (1981). Social dilemmas. In G. Stephenson & J. H. Davis
(Eds), Progess in applied social psycholrgy (vol. 1, pp. 37-65). Chichester, England:
Wiley.

Organ, D. W., & Ryan, K. (1995). A meta-analytic review of attitudinal and
dispositional predictors of organizational citizenship behavior. Personnel Psychology, 48,
775-802.

Parsons, T., & Shils, E. A. (1951). Toward a general theory of action. Cambridge,
MA: Harvard University Press.

Pearce, J. L., & Gregersen, H. L. (1991). Task interdependence and extrarole
behavior: A test of the mediating effects of felt responsibility. Journal of Applied

Psychology, 76, 838-844.

Pervin, L. A. (1980). Personalm theory and assessment. New York: Wiley.
Phillips, A. S., & Bedeian, A. G. (1994). Leader-follower exchange quality: The

role of personal and interpersonal attributes. Academy of Management Journal-37. 990-
1001.

155

Probst, T. M., Camevale, P. J., & Triandis, H. C. (1999). Cultural values in
intergroup and single-group social dilemmas. Organizational Behavior and Humaa
Decision Processes, 77, 171-191.

Rusbult, C. E., Farrell, D., Rogers, 6., & Mainous, A. G. 111. (1988). Impact of
exchange variables on exit, voice, loyalty, and neglect: An integrative model of responses
to declining job satisfaction. Afldemy of Management Journal. 31. 599-627.

Saavedra, R., Earley, P. C., & Van Dyne, L. (1993). Complex interdependence in
task-performing groups. Journal of Applied Psycholcgy. 78. 61-72.

Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to
job attitudes and task design. Administrative Science Qparterly, 23, 224-253.

Scandura, T. A., & Williams, E. A. (2000). Research methodology in management:
Current practices, trends, and implications for future research. Academy of Management
Journal 43 1248-1264.

 

Schmidt, K. H., Kleinbeck, U., & Brockmann, W. (1984). Motivational control of
motor performance by goal setting in a dual-task situation. Psychological Research. 46,
129-141.

Schneider, B. (1983). Interactional psychology and organizational behavior. In B.
M. Staw & L. L. Cummings (Eds), Research in organizational behavior (vol. 5, pp. 1-
31). Greenwich, CT: JAI Press.

Schutz, A. (1967). The phenomenology of the social world. Evanston, IL:
Northwestern University Press.

Shaw, J. D., Duffy, M. K., & Stark, E. M. (2000). Interdependence and preference
for group work: Main and congruence effects on the satisfaction and performance of
group members. Journal of Management, 26, 259-279.

Shea, G. P., & Guzzo, R. A. (1987). Groups as human resources. In K. M. Rowland
and G. R. Ferris (Eds), Research in Human Resources and Personnel Management (vol.
5, pp. 323-356). Greenwich, CT: JAI Press.

Sheppard, J. A. (1993). Productivity loss in performance groups: A motivation
analysis. Psychological Bulletin. 113. 67-81.

Singelis, T. M., Triandis, H. C. Bhawuk, D. P. S., & Gelfand, M. J. (1995).
Horizontal and vertical dimensions of individualism and collectivism: A theoretical and
measurement refinement. Cross-Cultural Reseaih, 29, 240-275.

Sniezek, J. A., May, D. R., & Sawyer, J. E. (1990). Social uncertainty and

interdependence: A study of resource allocation decisions in groups. Organizational
Behavior and Hurnan Decision Processes. 46. 155-180.

156

Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing measurement invariance
in cross-national consumer research. Journal of Consumer Resea_rc_h, 25, 78-90.

Steers, R. M. (1975). Task-goal attributes, achievement, and supervisory
performance. Organizational Behavior and Human Performance. 13. 292-303.

Steiner, I. D. (1972). Group process and productivity. New York: Academic Press.

Sundstrom, E., De Meuse, K. P., & Futrell, D. (1990). Work teams: Applications
and effectiveness. American Psycholajgist. 45(2). 120-133.

Terborg, J. R. (1981). Interactional psychology and research on human behavior in
organizations. Academy of Managament Review. 6. 569-576.

Tett, R. R, Jackson, D. N., & Rothstein, M. (1991). Personality measures as
predictors of job performance: A meta-analytic review. Personnel Psychology. 44. 703-
742.

Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill.

Triandis, H. C. (1995). Individualism and collectivism. Boulder, Colorado:
Westview Press.

Triandis, H. C., Bontempo, R., Villareal, M. J., Asai, M., & Lucca, N. (1988).
Individualism and collectivism: Cross-cultural perspectives on self-ingroup relationships.
Journal of PersonAity and Social Psychology, 54, 323-338.

Triandis, H. C., & Gelfand, M. J. (1998). Converging measurement of horizontal
and vertical individualism and collectivism. Journalf of PersonaliL and Socifl
Psychology. 74. 118-128.

Tubbs, M. E. (1993). Commitment as a moderator of the goal-performance relation:
A case for clearer construct definition. JouLnaI of Applied Psychology, 78, 86—97.

Tubbs, M. E., & Dahl, J. G. (1991). An empirical comparison of self-reported and
discrepancy measures of goal commitment. Journal of Applied Psychology, 76, 708-716.

Vandello, J. A., & Cohen, D. (1999). Patterns of individualism and collectivism
across the United States. Journal of Personality and Social Psychology, 77. 279-292.

Van Dyne, L., Cummings, L. L., & Parks, J. M. (1995). Extra-role behaviors: In
pursuit of construct and definitional clarity (A bridge over muddied waters). In B. M.
Staw & L. L. Cummings (Eds), Resaarch in Organizational Behavior (vol. 17, pp. 215-
285). Greenwich, CT: JAI Press.

Vroom, V. H. (1964). chk_ and motivation. New York: Wiley.

157

Wageman, R. (1995). Interdependence and group effectiveness. Administrative
Science Quarterly. 40. 145-180.

Wageman, R., & Baker, G. (1997). Incentives and cooperation: The joint effects of

task and reward interdependence on group performance. Journal of Organizational
Behavior 18 139-158.

 

Wagner, J. A. III. (1995). Studies of individualism-collectivism: Effects on
cooperation in groups. Academy of Management Journal. 38, 152-172.

Wagner, J. A. III, & Moch, M. K. (1986). Individualism-Collectivism: Concept and
measure. Group and Organization Studies, 11, 280-304.

Weick, K. E. (1977). Enactrnent processes in organizations. In B. M. Staw and G.
R. Salancik (Eds), New direction in organizational bfl/ior (pp. 267-300). Chicago: St.
Clair Press.

Weingart, L. R., & Weldon, E. (1991). Processes that mediate the relationship
between a group goal and group member performance. Human Performance. 4. 33-54.

Wofford, J. C., Goodwin, V. L., & Premack, S. (1992). Meta-analysis of the
antecedents of personal goal level and of the antecedents and consequences of goal
commitment. Journ_al of Management, 18, 595-615.

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.

Yuan, K. H., Bentler, P. M., & Kano, Y. (1997). On averaging variables in a
confirmatory factor analysis model. Behaviormetrikaj4. 71 -83.

Yukl, G. A., & Latham, G. P. (1978). Interrelationships among employee

participation, individual differences, goal difficulty, goal acceptance, goal
instrumentality, and performance. Personnel Psychology, 31, 305-323.

158

IIIIIIIIIIIIIIIIIIIII

llljlllzljjljljljljjjjljll