/

University of Southern California Dissertations and Theses

/

Understanding employees' efforts to seek change or quit: a longitudinal study of human and social service workers' cognitive and behavioral responses to their work experiences

(USC Thesis Other)

Understanding employees' efforts to seek change or quit: a longitudinal study of human and social service workers' cognitive and behavioral responses to their work experiences

PDF

Transcript (if available)

Content

UNDERSTANDING EMPLOYEES’ EFFORTS TO SEEK CHANGE OR QUIT:
A LONGITUDINAL STUDY OF HUMAN AND SOCIAL SERVICE WORKERS’
COGNITIVE AND BEHAVIORAL RESPONSES TO THEIR WORK EXPERIENCES

by
Dnika Jones Travis

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(SOCIAL WORK)

December 2006

Copyright 2006 Dnika Jones Travis

ii
Acknowledgements
My academic tenure has been filled with unwavering support from my
dissertation committee members; faculty and staff at USC; my parents, family, and
friends; and, of course, my dear husband.
I first extend my heartfelt gratitude to Dr. Michàlle Mor Barak, my faculty
advisor, mentor, dissertation chair, qualifying exam chair, and research director.
She has continuously provided an atmosphere for excellence in research and
teaching. I value her guidance, admire her grace, and respect her as a leader and
innovator.
I am very grateful to my dissertation committee members: Dr. Robert
Nishimoto and Dr. Sarah Banet‐Weiser, both of whom served on my qualifying
exam committee. I would also like to thank the two additional members of my
qualifying exam committee: Dr. Bruce Jansson and Dr. Madeline Stoner. These four
scholars provided instructive feedback and support that were invaluable
components of my educational experience at USC.
I extend my gratitude to the USC School of Social Work Doctoral Program.
Dr. Jansson, Director of the Doctoral Program, has provided continuous
encouragement and quality leadership. I would also like to thank Malinda
Sampson, Manager of the Doctoral Program; through her support and guidance, I
was never left with a question unanswered.

iii
I express my sincere appreciation to the staff at the USC Center on Child
Welfare (CCW): Dr. Paul Carlo, Donna Toulmin, Sarah Novak, Harold Pyun,
Alberto Reynoso, Lorena Vega, Maria Chavez, Ana Sufuentes, Cheryl Jackson, and
Nancy Raman. Each has made every provision to ensure that the research project, in
which this dissertation was drawn from, was successful in meeting the study goals.
I am proud to be a part of the CCW team and thankful to all for your unremitting
support, valuable feedback, and words of encouragement that helped make
completion of this dissertation possible.
I also thank the faculty and staff at the USC School of Social Work. Every
time that I walked the halls of the School or sought out assistance, I was always
greeted with smiles and encouragement. Additionally, I would like to extend
gratitude to Bin Xie for his statistical consultation.
When I entered the doctoral program, three others joined me as part of my
cohort. I thank Susan, Sang Mi, and Heather for the laughs and camaraderie.
I would like to thank the Council of Social Work Education Minority
Fellowship Program for providing support and the opportunity to focus on my
dissertation and develop connections with other researchers. I am also grateful to
the Title IV‐E Child Welfare Training grant for support of this dissertation research.
As the saying goes, “it takes a village”… To my parents in Houston and
Chicago and to my in‐laws in Knoxville, Raleigh, and Kinston: each of you has

iv
shown me unconditional love and tremendous generosity for which I am eternally
grateful. I also am appreciative of the long‐distance cheers I received from my
grandparents, aunts, uncles, cousins, family members, and friends from all over the
country (from Houston, Richmond, San Francisco, Raleigh, Kinston, Knoxville, DC,
Baltimore, Orlando to Atlanta). Thanks to my daughter, Morgan, for gifting me with
being your mother. In finding and using your voice, I wish for you the will and
courage to follow your heart and honor your integrity.
Finally, I am eternally appreciative of my loving husband, Raphael, whose
fortitude and generosity have been a daily source of support. I am thrilled to have
shared this journey with you by my side. This dissertation is dedicated to you.
I conclude by expressing my gratitude to the child welfare workers that
participated in this research study. I am honored to share a piece of your work
experiences and voice as a part of this manuscript. Thank you for participating in the
study and for putting yourselves on the firing line on behalf of children and their
families.

v
Table of Contents

Acknowledgements..............................................................................................................ii
List of Tables........................................................................................................................vii
List of Figures.......................................................................................................................ix
Abstract..................................................................................................................................x
Chapter 1: Introduction........................................................................................................1
1.1. Overview and Structure of Dissertation........................................................1
1.2. Study Purpose...................................................................................................6
1.3. Study Significance & Innovation....................................................................7
1.4. Relevance to Social Work & Human Services..............................................9
Chapter 2: Review of the Exit/Neglect/Voice Literature..................................................14
2.1. How Do People Respond to Their Experiences at Work?.........................14
2.2. Unfavorable or Favorable Exit/Neglect/Voice Responses...........................18
2.3. Identifying Exit/Neglect/Voice in the Literature...........................................26
2.4. Summary..........................................................................................................35
Chapter 3: Conceptual Model and Research Hypotheses.............................................38
3.1. Introduction to the Conceptual Model........................................................38
3.2. Theoretical Framework..................................................................................42
3.3. Factors Influencing Exit/Neglect/Voice Responses......................................53
3.4. Relationships Among Exit, Neglect, and Voice............................................82
3.5. Indirect Effects.................................................................................................87
3.6. Conceptual Model..........................................................................................89
Chapter 4: Methodology....................................................................................................94
4.1. Overview of Study..........................................................................................94
4.2. Study Site.........................................................................................................94
4.3. Study Participants, Enrollment, & Data Collection...................................95
4.4. Description of Study Sample.........................................................................98
4.5. Measures........................................................................................................104
4.6. Screening of Longitudinal, Quantitative Data..........................................115
4.7. Analytic Procedures....................................................................................118

vi
Chapter 5: Research Findings..........................................................................................123
5.1. Introduction...................................................................................................123
5.2. Descriptive Statistics and Correlations......................................................123
5.3. Assessment of Model Fit..............................................................................126
5.4. Study Hypotheses: Findings.......................................................................133
5.5. Indirect Effects...............................................................................................146
5.6. Summary of Results......................................................................................152
Chapter 6: Discussion.......................................................................................................157
6.1. Overview........................................................................................................157
6.2. Key Findings..................................................................................................158
6.3. Strengths and Limitations...........................................................................175
6.4. Theoretical Implications..............................................................................180
6.5. Practical Implications...................................................................................184
6.6. Suggestions for Future Research................................................................190
6.7. Conclusions...................................................................................................191
Glossary..............................................................................................................................193
References..........................................................................................................................195
Appendix A: Scales from Study Questionnaire............................................................218
Inclusion in Decision‐making............................................................................218
Perceived Supervisory Support.........................................................................218
Role Conflict.........................................................................................................219
Role Ambiguity....................................................................................................220
General Health Questionnaire...........................................................................220
Exit/Neglect/Voice...............................................................................................221
Appendix B: Supplemental Information—Variable Transformation........................222
Appendix C: Supplemental Path Analysis Results......................................................224

vii
List of Tables
Table 1. Categorizations of Exit/Neglect/Voice Responses 20
Table 2. Voice & Related Constructs 37
Table 3. The Role of Diversity Characteristics (Hypotheses Set A) 91
Table 4. The Role Perceptions of Work Climate (Hypotheses Set B) 91
Table 5. The Role Job Stress & Psychological Wellbeing (Hypotheses Set C) 92
Table 6. Relationship Among Exit, Neglect, and Voice (Hypotheses Set D) 92
Table 7. Diversity Characteristics: Indirect Effects (Hypothesis Set E) 93
Table 8. Personal Diversity Characteristics of Study Participants 99
Table 9. Professional Diversity Characteristics of Study Participants 100
Table 10. Recap of Significant Findings From Hypothesis set A 153
Table 11. Recap of Significant Findings From Hypothesis set B 154
Table 12. Recap of Significant Findings From Hypothesis set C 155
Table 13. Recap of Significant Findings from Hypothesis set D 155
Table 14. Recap of Significant Findings from Hypothesis set E 156
Table 15. Skewness Transformation Results 223
Table 16. Path Analysis Results (Phase I Refined Model) 225
Table 17. Path Analysis Results (Phase II Refined Model) 226
Table 18. Path Analysis Results (Phase III Refined Model) 227

viii
Table 19. Path Analysis Results (Phase IV Refined, Cross‐lagged Model) 228
Table 20. Post‐hoc Path Analysis Results: Neglect & Voice Cross‐lagged Model 229

ix
List of Figures
Figure 1. Complaining, Aggressive Voice, & Constructive Voice..................................31
Figure 2. Overview of Relationships Among Study Variables....................................38
Figure 3. What Keeps People From Exercising Voice?...................................................53
Figure 4. Hypothesized Conceptual Model....................................................................90
Figure 5. Final Phase III Path Model (Baseline)............................................................130
Figure 6. Final Phase IV Path Model (Longitudinal, Cross‐lagged Model).............132

x
Abstract
Well‐trained and highly qualified workers are invaluable assets to human
service organizations; yet, these workers face a host of challenges in doing their
jobs. Prior studies have gauged organizational commitment, job satisfaction,
intention to leave, and turnover (Mor Barak, Nissly, & Levin, 2001). This study
forges new ground in its focus on the relationships among organizational factors
and employees’ responses to undesirable work experiences in a human service
setting. Specifically, this study examined the following responses: (1) efforts
toward of quitting (exit), (2) efforts that reflect one’s disengagement from work‐
related tasks or organizational activities (neglect), and (3) efforts to improve
dissatisfying work conditions or promote organizational change (voice) (Farrell,
1983; Rusbult, Farrell, Rogers, & Mainous, 1988). To encourage well‐functioning
organizations and positive client outcomes, greater understanding is warranted in
this regard. The studyʹs purpose is three‐fold: (1) to test a theory‐driven conceptual
model that depicts the relationships among employee diversity characteristics,
perceptions of work climate, job stress, psychological wellbeing, and
exit/neglect/voice responses as outcome variables; (2) to examine the relationship
among exit, neglect, and voice; (3) to test the model longitudinally with respect to
the outcome variables.

xi
The model was developed from social work, organizational psychology,
management, and communication theories to offer a comprehensive
multidisciplinary perspective. Through a longitudinal study of 359 child welfare
workers, the hypothesized model was tested in four phases using path analysis
(Amos 6.0). Both the cross‐sectional and longitudinal path analytic models did fit
the data well and provides evidence that diversity characteristics, perceptions of
work climate, job stress, and psychological wellbeing are important determinants
of employees’ exit/neglect/voice responses to undesirable work experiences.
Theoretically, this study’s findings suggest that inclusion should be conceptually
distinguished from supervisory and organizational support and that role conflict is
distinguishably role ambiguity. Additionally, voice and neglect should be refined
and reconceptualized in accordance with the challenges and demands of a human
service setting. At the practice‐level, these findings provide guidance to
organizational leaders on how to target interventions to retain employees and keep
them fully engaged.

1
Chapter 1: Introduction
1.1. OVERVIEW AND STRUCTURE OF DISSERTATION
Human and social service workers (also referred to as human service
workers) provide invaluable assistance to the individuals, children, families,
organizations, and communities that they serve. These workers are often charged
with delivering and managing quality services, as well as maintaining
accountability to their supervisors, organizations, and clients. Yet, researchers have
demonstrated human service workers, such as child welfare, juvenile and criminal
justice, and employment/training services workers, face common challenges (The
Annie E. Casey Foundation, 2003). Human service workers may feel stifled in
providing optimal care for their clients due to such stressors: high caseloads;
reporting responsibilities to oversight agencies; conflicting demands between
government policies (such as in the court system) and client desires; and limited
flexibility in making decisions (in part, due to hefty regulations) (Lait & Wallace,
2002; The Annie E. Casey Foundation, 2003). In some cases, these workers have
added challenges based on the extent to which they feel supported by their
supervisor or work organization (K. M. Hopkins & Hyde, 2002; W. E. Hopkins,
Hopkins, & Mallette, 2001) and included in critical formal and informal
organizational processes (Mor Barak & Cherin, 1998). As researchers have well
documented, these types of challenges often result in elevated levels of stress (Lait

2
& Wallace, 2002) and the desire to leave their jobs; thus leading to rapid turnover
rates (Mor Barak, Nissly et al., 2001).
When faced with undesirable experiences at work (such as large caseloads or
lack of perceived supervisory support) employees typically may engage in one, a
combination, or all of the following responses (i.e., exit/neglect/voice response)
(Farrell, 1983; Hagedoorn, Van Yperen, Van de Vliert, & Buunk, 1999; Rusbult et al.,
1988):
♦ First, employees may engage in efforts toward quitting one’s job
(referred to as ‘exit’) (Hirschman, 1970). In these efforts, employees may
contemplate, intend, or take steps toward leaving their job.
♦ Second, employees may become neglectful in their work‐related or
organizational efforts (referred to as ‘neglect’) (Hanisch & Hulin, 1990,
1991; Hanisch, Hulin, & Roznowski, 1998). Neglect, or employees’ efforts
that reflect one’s disengagement from work‐related tasks and
organizational activities (e.g., missing a meeting, showing up late for
work) is another way that employees can respond to undesirable
experiences at work (Kidwell & Robie, 2003).
♦ As a third response, employees may make an effort to improve
dissatisfying work conditions or to promote organizational change
(referred to as ‘voice’) (F. Bowen & Blackmon, 2003; Kowtha, Landau, &

3
Beng, n.d.; Van Dyne, Cummings, & Parks, 1995). Their voice efforts may
involve talking with a supervisor about a problem or suggesting ways of
improving organizational processes (Saunders, Sheppard, Knight, &
Roth, 1992; Van Dyne, Ang, & Botero, 2003).
Employees’ exit/neglect/voice responses not only affect (1) the employee, but
also (2) the functioning of human service organizations; as well as (3) the client’s
outcomes. These responses involve both cognitive and behavioral elements.
Cognitions represent individuals’ thought processes that are linked to behaviors,
while behaviors involve individual actions (Bergin & Garfield, 1994). While
cognitive responses are important to gauge because of the focus on intentionality
and thought processes (e.g., thoughts about leaving the job), behavioral responses
(e.g., going on job interviews) provide an opportunity to capture individuals’
constructive or unconstructive actions in a specific area (see Chapter 2, Section 2.1
for a more detailed discussion of exit/neglect/voice responses).
Accordingly, exit/neglect/voice responses can be categorized as unfavorable or
favorable to work organizations. Unfavorable responses are those that can adversely
affect the work organization, which include employee exit and neglect responses. In
the short‐term, the costs associated with exit and neglect can result in dwindling
service quality due to the possibility of employees not fully engaging in task or
organizational activities. In the long‐term, efforts toward quitting one’s job as well

4
as those efforts that reflect one’s disengagement from work‐related tasks and
organizational activities are correlates of employee turnover (Alexander,
Lichtenstein, Oh, & Ullman, 1998; Griffeth, Hom, & Gaertner, 2000; Mor Barak,
Nissly et al., 2001). The costs associated with the turnover can be high and
detrimental to the organization. Human service organizations are often faced with
high costs associated with replacing and training new workers (United States
General Accounting Office [GAO], 2003). Further, gaps in services may be
experienced at the client level and new employees may be faced with rebuilding
trust with the client (American Public Human Services Association [APHSA], 2005;
The Annie E. Casey Foundation, 2003).
One the other hand, voice responses have the potential to foster favorable
outcomes for the employee and for the organization through the promotion of
organizational growth, development, and change (Morrison & Milliken, 2003).
Hence, voice is of special interest to the current study due to its improvement‐
oriented focused. Through voice (efforts to improve work conditions or promote
organizational change), organizations may receive employee feedback, suggestions,
and innovation. Employees are also afforded a greater opportunity to feel engaged
and to contribute to the organizations’ functioning.
Through better understanding of organizational and individual factors
affecting employees’ exit/neglect/voice responses, human service organizations may

5
be better equipped to retain engaged staff that are committed to delivering high
quality services to clients. Hence, the central question of this current study is as
follows:
What factors support or hinder human service workers’ efforts toward quitting their
job (exit); efforts that reflect their disengagement from work‐related tasks or
organizational activities (neglect); or efforts to improve dissatisfying work
conditions or promote organizational change (voice)?
The current research examines this central question in a cross‐sectional and
longitudinal study of 359 human and social service workers. In this study, a theory‐
driven model of the organizational and individual antecedents of human service
workers’ exit/neglect/voice responses to their work experiences is tested using path
analysis (Byrne, 2001; Loehlin, 2004).
To this end, this study is organized as follows: this chapter (Chapter 1)
focuses on the study’s purpose and significance as well as its relevance to social
work and the field of human services. The review of the literature related to
exit/neglect/voice responses is offered in Chapter 2. Chapter 3 presents the study’s
theory‐driven conceptual model as well as the specific research hypotheses. The
detailed methodology is laid out in Chapter 4, which includes a description of the
sample, methods of data collection, measures used, and analytic procedures. The
results of the study are presented in Chapter 5. Last, Chapter 6 concludes with a

6
discussion of key findings. Here the focus is on theoretical and empirical evidence
to support the relationship among study variables, benefits and limitations of the
study, as well as the practical implications and recommendations for future
research.
1.2. STUDY PURPOSE
Previous research has made significant contributions in understanding the
role of organizational and individual factors in contributing to human service
workers’ job satisfaction, stress and burnout, as well as intention to leave and
turnover (see for example Abu‐Bader, 2000; Coyle, Edwards, Hannigan, Fothergill,
& Burnard, 2005; Lait & Wallace, 2002; Mor Barak, Levin, Nissly, & Lane, 2006;
Mor Barak, Nissly et al., 2001; Zlotnik, DePanfillis, Daining, & Lane, 2005). The
current study builds on this body of knowledge by examining a theory‐driven
conceptual model depicting the relationships among employees’ diversity
characteristics, perceptions of their work climate, job stress and psychological
wellbeing, and their exit/neglect/voice responses.
Briefly, as stated in the previous section, employees’ exit/neglect/voice
responses reflect individual cognitions or behaviors in reaction to undesirable
experiences at work. An employee may engage in efforts toward quitting one’s job
(exit), disengage from work‐related tasks or organizational activities (neglect), and
seek to improve dissatisfying working conditions or promote organizational

7
change (voice) (Hirschman, 1970; Rusbult et al., 1988; Withey & Cooper, 1989). In
the case of voice, employees may also engage in change‐based efforts that promote
organizational change in an interest to advance workplace functioning for its own
sake, not just because the working conditions are undesirable (Creed, 2003;
Meyerson & Scully, 1995; Morrison & Phelps, 1999; Van Dyne et al., 1995; Van
Dyne & LePine, 1998). Understanding the impact of organizational as well as
individual factors in contributing to exit/neglect/voice responses is the focus of this
study. Voice, in particular, is of special interest to the current study because of its
potential to increase organizational effectiveness. Therefore, the purpose of the
study is threefold:
1. To test a theory‐driven conceptual model that depicts the relationships among
employee diversity characteristics, perceptions of work climate, job stress,
psychological wellbeing, and exit/neglect/voice responses.
2. To specifically examine the relationship among employees’ exit/neglect/voice
responses.
3. To test the conceptual model longitudinally, at two time points over a six‐
month period, with respect to exit/neglect/voice responses as outcome variables.
1.3. STUDY SIGNIFICANCE & INNOVATION
This study forges new ground in several respects:

8
♦ Unique combinations of organizational and individual variables are
examined in this study for their effects on human service workers’
exit/neglect/voice responses to undesirable work experiences.
♦ Specifically, by focusing on voice, the study seeks to identify factors that
enhance the likelihood of human service workers engaging in voice rather
than exit or neglect. In addition, the study seeks to determine the relationship
among exit/neglect/voice over time.
♦ The study is longitudinal, enabling the antecedents to exit/neglect/voice to be
identified.
Prior studies, documenting organizational and individual factors as
antecedents to employee outcomes, have focused considerable attention on
organizational commitment, job satisfaction, intention to leave, and actual turnover
(for example, see Mor Barak, Nissly, & Levin’s [2001] meta‐analysis). Yet, based on
an extensive review of the literature, no studies in social work and the human
service field have also examined the influence of these factors on the combination of
cognitions and (more uniquely) behaviors relating to exit/neglect/voice responses.
Further, none have looked at the combination of these over time. More uniquely,
voice has not been investigated in human and social services.
By identifying factors that support or hinder employee voice as well as exit
and neglect, the results of this study may provide researchers and practitioners with

9
additional evidence‐based information on how to create the best possible worker
outcomes—decreased turnover of talented workers, reduced job neglect, and
increased worker‐initiated voice and feedback. The results should be particularly
instructive to scholars interested in capturing how employees’ responses to their
work experiences are expressed. Specifically, in the human service arena and at the
practice‐level, this research will provide guidance to organizational leaders seeking
to both retain employees and keep them fully engaged. Most notably, knowing
which factors may actually enhance employeesʹ choice of voice as distinct from exit
or neglect may help guide or augment current approaches to supervision, ultimately
resulting in the best‐improved outcomes for clients and their communities.
1.4. RELEVANCE TO SOCIAL WORK & HUMAN SERVICES
The National Association of Social Workers (National Association of Social
Workers, 2002) declares that the social work profession’s key function is to enrich
human wellbeing. Thus the profession focuses on providing support and guidance
to the individuals, families, and communities served as well as promoting social
justice and human rights (National Association of Social Workers, 2002). In a
parallel path, the field of human services is focused on addressing human needs
based on an interdisciplinary approach that serves the optimal goal to improve
quality of life (National Organization for Human Services & Council for Standards
in Human Services Education, 2006).

10
To meet its service‐focused missions, both the social work profession and the
field of human services are dependent on well‐trained and highly qualified
workers. These workers fulfill critical roles connecting individuals, communities,
and organizations with needed social services. In child welfare, for example,
workers are on the frontlines—counseling clients, investigating reports of abuse or
neglect, responding to emergency situations, testifying in courts, and keeping
agencies abreast of community and individual needs. Those not on the frontlines
are often fulfilling critical managerial and supervisory roles. All in all, child welfare
workers and other human service workers have similar interests and share
comparable challenges (The Annie E. Casey Foundation, 2003). As such, The
National Human Service Organization (National Organization for Human Services
& Council for Standards in Human Services Education, 2006) characterized a human
and social service worker as follows:
… a generic term for people who hold professional and
paraprofessional jobs in such diverse settings as group homes and
halfway houses; correctional, mental retardation, and community
mental health centers; family, child, and youth service agencies, and
programs concerned with alcoholism, drug abuse, family violence,
and aging. Depending on the employment setting and the kinds of
clients served there, job titles and duties vary a great deal.

In their comprehensive report on “The Unsolved Challenge of System
Reform: The Condition of the Frontline Human Services Workforce”, The Annie E.
Casey Foundation (2003) conservatively estimated that over 7.5 million people

11
encompass the human service workforce—not including supervisory or
administrative positions or those childcare workers. These include: 870,000 child
welfare workers, two million youth services worker (4 million part‐time), 300,00
juvenile justice workers, 500,000 employment and training. Many find working in
social work or human services rewarding yet stressful. Often workers have heavy
workloads and limited rewards for extra work‐related effort (The Annie E. Casey
Foundation, 2003). Managers and administrators, although not on the frontlines,
also experience stressors. As David S. Liederman, the Director of the Child Welfare
League of America portrayed (State of Child Welfare Services, 1993):
Running a national human services organization is a tough job. You
have to keep one eye on the mission and the other on the bottom
line. You have to provide high‐quality services to your members,
work collaboratively with your board, and keep staff morale high.
You have to criss‐cross the country, keynoting conferences,
addressing leadership gatherings, and attending annual meetings.
You have to meet the press, inspire the troops, and work with
colleagues in other national organizations. And you have to
remember everybodyʹs name! You have responsibilities to yourself,
to your organization, to your members, to the human services field,
and to the nation at large.

All of these types of organizational challenges can result in high turnover
among human service workers. The American Public Human Services Association
(2005) cited stress as the most common problem related to preventable turnover.
Preventable turnover is characterized as leaving the organization due to factors that
do not include death, retirement, family considerations (i.e., marriage, spouse job

12
move, or parenting), or returning to school. Workers have reported such stressors as
high and demanding workloads, lack of work/family balance, limited time with
clients due to paperwork and travel time (APHSA, 2005) as well as low pay, risk of
violence, staff shortages, administrative burdens, inadequate supervision,
inadequate training (United States General Accounting Office [GAO], 2003). These
all contribute to preventable worker turnover.
As the challenges of working in human services generate unfavorable
outcomes for work organization and its employees, detriments to clients exist as
well. An emergent body of literature has found support for the connection between
organizational context and client outcomes (Glisson & Durick, 1988; Glisson &
Hemmelgarn, 1998; Glisson & James, 2002; Morris, Bloom, & Kang, 2002; Yoo, 2002;
Yoo & Brooks, 2005). As examples, Glisson and Hemmelgarn (1998) found that an
organizational climate with increased role clarity and cooperation and low conflict
linked to the successful outcomes for child welfare services. Yoo and Brooks (2005)
have had related findings in their study of approximately 4000 children and 13
networks organizations. They found that workers’ perceptions of greater
habituation of work, strong

leadership qualities, and supervisor and coworker
support were associated with positive client outcomes. Last, Morris, Bloom, and
Kang (2002) found organizational culture and climate were predictors of consumers’
mental and physical wellbeing over time.

13
The current study is relevant to the social work profession and the human
service field because workers are germane to helping their organizations
effectively meet client needs. Particularly, this study focuses on modeling
organizational and individual antecedents to employees’ exit (efforts toward
quitting one’s job), neglect (efforts that reflect one’s disengagement from work‐
related tasks or organizational activities) and voice (efforts to improve dissatisfying
work conditions or promote organizational change). As stated, working human
and social service organizations offers many unique challenges for workers and
administrators. This study builds on an existing knowledge base to gain insight
into the factors that encourage well‐qualified workers to remain in their jobs,
engage in their work, and promote organizational change.

14
Chapter 2: Review of the Exit/Neglect/Voice Literature
This chapter reviews the exit/voice/neglect constructs as discussed in the
social work, human service, organizational psychology, and business and
management literatures. The first part of the review provides a general description
of exit/neglect/voice as responses to undesirable experiences at work. As discussed,
an employee may engage in the following: (1) efforts toward quitting one’s job
(exit), (2) efforts that reflect one’s disengagement from work‐related tasks and
organizational activities (neglect), and (3) efforts to improve dissatisfying work
conditions or promote organizational change (voice). Next, a discussion of the
responses as unfavorable versus favorable to work organizations is provided. In
this, exit and neglect are considered unfavorable, while voice is considered to have
favorable outcomes for organizational effectiveness. Last, a detailed review of exit,
neglect, voice responses as well as related constructs is provided to aid determining
their antecedents (see Chapter 3 for a discussion of the antecedents to
exit/neglect/voice as well as the hypothesized conceptual model).
2.1. HOW DO PEOPLE RESPOND TO THEIR EXPERIENCES AT WORK?
Faced with work challenges (i.e., perceived lack of support, feelings of
exclusion, job stress, and job dissatisfaction), employees develop cognitions and
behaviors in reaction to undesirable experiences at work. Cognitions represent
individuals’ thought processes, while behaviors involve individual actions.

15
According to Bergin (1994), individuals’ thought processes play a role and even
guide their responses or behaviors with regard to a given circumstance. Further,
cognitions may affect specific behaviors but also are influenced by individual
mood, attitudes, and prior experiences. On the other hand, behaviors are the
specific actions, reactions, and responses that individuals engage in based on their
environment. While cognitive responses are important to gauge because of the
focus on intentionality and thought processes that eventually affect behaviors (e.g.,
thinking about leaving the job), behavioral responses (e.g., going on job interviews)
provide an opportunity to capture employee favorable or unfavorable actions.
For this study, cognitive and behavioral responses are categorized as
exit/neglect/voice responses that are based on the work of Hirshcman (1970), Rusbult
et al. (1982) and Farrell (1983). Hirschman’s seminal work on exit, voice, and loyalty
(individuals’ feelings of allegiance toward their organization) responses to
dissatisfying conditions laid the foundation for Rusbult’s et al. (1982) and Farrell’s
(1983) exit, voice, loyalty, neglect model. Neglect was added to the model as a fourth
response in a study of interpersonal romantic relationships (Rusbult et al., 1982) and
dissatisfaction in organizational settings (Farrell, 1983). Over the last two decades,
these constructs have been comprehensively investigated in organizational and
social psychology, management, and political science studies (see Dowding, John,
Mergoupis, & Vugt, 2000).

16
Exit involves employees’ efforts toward quitting their job (Hirschman, 1970).
This response has both cognitive (e.g., intending to leave one’s job) and behavioral
(e.g., searching for a new job) elements (Farrell & Rusbult, 1992; Rusbult et al., 1988)
that can create physical or psychological distance between employees and their
work organization (Carmeli, 2005; Rosse & Hulin, 1985). Exit is an important
variable to gauge because it is well documented as a predictor of employee turnover
(e.g. Firth, Mellor, Moore, & Loquet, 2004; Mayes & Ganster, 1988). Specific to this
study, employee retention is a critical challenge for human service organizations
(Light, 2003; The Annie E. Casey Foundation, 2003; United States General
Accounting Office [GAO], 2003).
Neglect involves efforts that reflect one’s disengagement from work‐related
tasks and organizational activities (Hanisch & Hulin, 1990, 1991; Hanisch et al.,
1998; Kidwell & Robie, 2003). By engaging in neglect, employees may to some extent
psychologically remove themselves from their work activities or the organization
(Withey & Cooper, 1989). As such, neglect efforts simply may involve any element of
not working including spending less time at work, coming to work late, making
errors, and working slowly. Sometimes, neglect is an alternative response when an
employee feels no hope or that a specific situation cannot be resolved (Rusbult et al.,
1988).

17
Voice, a voluntary (F. Bowen & Blackmon, 2003) response involves employee
efforts to improve dissatisfying work conditions or promote organizational change
(F. Bowen & Blackmon, 2003; Kowtha et al., n.d.; Rusbult et al., 1988; Saunders et al.,
1992; Van Dyne et al., 2003; Van Dyne et al., 1995). Hirschman (1970) defined voice
as “any attempt at all to change, rather than to escape from, an objectionable state of
affairs” (p.30). This can involve discussing problems with a supervisor, providing
solution focused suggestions to human resources departments or consulting with a
union or organizational expert (Rusbult et al., 1998). However, the construct has
been extended to reflect employee behaviors that are geared toward promoting
organizational change (F. Bowen & Blackmon, 2003; Dowding et al., 2000; Kowtha et
al., n.d.; Van Dyne et al., 1995; Zhou & George, 2001). Because of the vast options
available, voice is not as straightforward as exit. As Hirschman explained, voice can
take different forms and meanings depending on the context.
The current study focuses on exit/neglect/voice as cognitive and behavioral
responses to undesirable experiences at work. Although, employee loyalty was
originally included in Rusbult’s et al. and Farrell’s framework, loyalty is not
examined in this dissertation for several reasons, as discussed by Withey and
Cooper (1989):
1. The conept itself lacks clarity in the literature.
2. It is more attitudinal than behavioral.

18
3. Loyalty has differing implications for organizational practice. Loyalty is
positive if it results in defense and support of the organizational. Blind
loyalty, on the other hand, is the absence of a desire to change things that are
detrimental to the organization or clients.
Therefore, this study focuses on cognitions and behaviors that are considered either
favorable or unfavorable to work organizations. Again, voice is viewed as having the
potential to foster favorable outcomes resulting in enhanced organizational
effectiveness while neglect and exit are responses that can have unfavorable results.
2.2. UNFAVORABLE OR FAVORABLE EXIT/NEGLECT/VOICE RESPONSES
Farrell (1983) and Rusbult et al. (1988) have classified exit/neglect/voice
responses along a continuum of destructive versus constructive responses. In this,
neglect and exit are forms of destructive responses that are viewed as costly (refer to
the following section 2.3) to the organization or the individual (Hagedoorn et al.,
1999). On the other hand, voice is considered a constructive response. Through voice,
employees may obtain an outlet for their grievances and contribute to the
organization’s effectiveness. As may be evident, voice is generally preferred over exit
(M. G. Keeley, Jill W., 1991). Organizations would rather have qualified, possibility
innovative, workers to help address workplace challenges. Although voice may
benefit a work organization, voice is not always a welcomed response. For example,
employees who engage in whistleblowing may be viewed as disloyal and as

19
individual who intend to cause harm to their work organizations (Bather & Kelly,
2005; Greene & Latting, 2004; Near & Miceli, 1996).
This study adopts Farrell’s and Result’s destructive/constructive continuum
in categorizing the cognitive and behavioral responses as either unfavorable or
favorable (see Table 1). Consistent with destructive response, unfavorable responses
(e.g., exit and neglect) are those that have limited benefits and grave implications to
organizations as well as its workers and clients. In contrast, favorable and
constructive voice responses are those that organizations generally may want to
encourage although they may not initially welcome these types of responses. In all,
regardless of the employee response, exit/neglect/voice efforts can impact the
employees’ relationship with their organization as well as impact organizational
effectiveness (Hagedoorn et al., 1999). For those reasons, the way that human
services workers deal with challenges at work has implications for organizational,
employee, and service‐related outcomes.

20
TABLE 1. CATEGORIZATIONS OF EXIT/NEGLECT/VOICE RESPONSES
Type of
Response
Response Definition Example
Unfavorable Exit

♦ Efforts toward
quitting one’s job
♦ Thinking about quitting
♦ Searching for a new job

Neglect

♦ Efforts that reflect
one’s disengagement
from work‐related
tasks and
organizational
activities

♦ Avoiding a supervisor
♦ Taking extended breaks
♦ Working slower than
usual on job tasks
Favorable Voice

♦ Efforts to improve
dissatisfying work
conditions or promote
organizational change
♦ Talking to a supervisor
♦ Taking initiative to
implement a new idea

Exit: Organizational, Employee, & Service‐related Outcomes
Exit involves employees’ efforts toward quitting their job (Hirschman, 1970)
and is a well‐documented predecessor to actual employee turnover (Griffeth et al.,
2000). Accordingly, the outcomes of exit are comparable to those associated with
employee turnover. Turnover rates in human services are elevated (The Annie E.
Casey Foundation, 2003). In child welfare, 2003 estimates ranged from 12% for
front‐line supervisors to 22% for child protective service workers with most of the
turnover being preventable (American Public Human Services Association
[APHSA], 2005). Additionally, The Annie E. Casey Foundation (2003) reviewed
annual estimates and found that the average turnover rates differed based on type
of organization: 20% in public agencies and 40% in private agencies. Other areas of
human service share similarly high turnover rates. For example, in the field of

21
juvenile justice, annual rates ranged from 40‐80% (The Annie E. Casey Foundation,
2003). Challenges with turnover are not only problematic for front line workers, but
also for supervisors, managers, and administrators as well. David Liederman, the
Director of the Child Welfare League of America, declared, “There is a lack of stable
leadership in child welfare. In the last 2 years there has been a 50 percent turnover
among State directors of child welfare programs” (State of Child Welfare Services,
1993).
Unfortunately, elevated rates of turnover cost human service organizations
at the client, employee, and organization levels. Client level outcomes involve the
loss in consistency and continuity of care (American Public Human Services
Association [APHSA], 2005; United States General Accounting Office [GAO], 2003)
and, in the case of child welfare, can impact the achievement of one of the most
precious child welfare goals—permanence (Hess, 1992). Unfavorable outcomes at
the employee level can be financial (loss of income and benefits), emotional (Withey
& Cooper, 1989), as well as create stress for co‐workers who stay on the job. At the
organizational level, unfavorable outcomes include the financial costs of losing
(separation costs) and hiring employees (American Public Human Services
Association [APHSA], 2005). Separation costs include severance pay and possible
legal expenses associated with employees leaving their work organization. New hire
costs involve the monies and time spent hiring new employees (recruiting,

22
screening, and assessing candidates; time spent interviewing candidates; orientation
and training of frontline workers and supervisors; and in some cases travel and
relocation expenses). APHSA’s (2005) estimate of turnover costs was $108 million
dollars for 8 of the 18 states reporting turnover in their national survey of human
service providers. Organizations can also be burdened with indirect costs such as a
loss of productivity and general disruption of the team dynamics (Mor Barak,
Nissly et al., 2001). Balfour and Neff (1993) refer to this as the cost associated with a
loss of human capital. This is especially relevant to human services where the
employees’ knowledge and skills are in a sense the product. In this respect, the
service provided by the human service worker is produced and consumed at the
same time and the customer acts as a co‐producer of the service (Woodside, Frey, &
Daly, 1989). Balfour and Neff further asserted that turnover rates greater than 20
percent are distressing to human capital and the effectiveness of a human service
organization.
It is noteworthy that not all turnover is problematic for organizations. On
some occasions, organizational turnover may actually benefit an organization
(Powell & York, 1992; Todd & Deery‐Schmitt, 1996). For instance, new employees
may bring innovative ideas and diverse experiences and toxic employees may be
filtered out. In these cases, organizations should welcome the change. However, too
much turnover often can negatively impact the quality of services provided,

23
especially when well qualified and well trained employees vacate their jobs (Mor
Barak, Nissly et al., 2001).
Neglect: Organizational, Employee, & Service‐related Outcomes
Counterproductive or deviant work behaviors have been classified as those
that harm individuals (e.g., gossiping) or those that negatively impact the
organization (e.g., tardiness) (Robinson and Bennet, 1995). Neglect is a form of a
counterproductive work behavior with regard to employees’ efforts that reflect their
disengagement from work‐related tasks and organizational activities. The
consequences of neglect to the organization as well as to employee engagement
warrant continued investigation within the human and social services. At the
organizational level, engagement in counterproductive behaviors often involves a
violation of expectations, values, and practices that can ultimately disturb service
quality (Duffy, Ganster, & Pagon, 2002 In Press; Robinson & Bennett, 1995). Withey
& Cooper (1989, p.523) discussed the possible indirect effects of counterproductive
behaviors at the individual level. Employees may adversely (and unknowingly) be
affected due to the potential loss of opportunities or personal and professional
reputation. Further, by engaging in neglect over time employees may feel a lack of
success at completing one’s job. Finally, employees may be subjected to disciplinary
action.

24
What are the benefits of having an employee who is not fully participating
or putting forth sufficient effort in their jobs? In a review of the peer‐reviewed
organizational literature, no direct benefits to the employee or organization for
workers’ counterproductive behaviors were found. Nonetheless, at times taking on
the job personal slack time may ultimately aid in employee productivity. For
example, socialization above and beyond designated work breaks can provide
employees opportunities to vent and find solutions to work‐related stressors.
Voice: Organizational, Employee, & Service‐related Outcomes
With a focus on optimal organizational effectiveness, the exercise of voice to
improve work conditions or promote organizational change does not inevitably
suggest that workplace conditions are good or bad (Morrison & Phelps, 1999).
Rather, employees’ involvement can be a critical part of an organization’s growth
and development, serving as a vehicle for feedback, learning, and innovation that
contributes to organizational performance (M. Keeley & Graham, 1991; Morrison &
Phelps, 1999). Organizational leaders recognize this and even may anticipate
employee contribution or create more empowerment‐based practices (Piderit, 2000).
Despite the prospective benefits of voice, many employees may feel that if
they offer ideas and input these may be overlooked or dismissed (Milliken,
Morrison, & Hewlin, 2003; Morrison & Phelps), especially if they hold minority
views (Noelle‐Neumann, 1984, 1991). Further, because some employees view

25
engaging in voice as risky, they may be fearful to seek opportunities to change
dissatisfying working conditions or promote organizational change. Engaging in
voice can result in real or perceived consequences for employees.
Sometimes employees face isolation from coworkers, supervisors, or the
organization (Cortina & Magley, 2003) or fear that their views will be dismissed or
will not make a difference (Edmondson & Detert, 2005). The “mum effect” (Conlee
& Tesser, 1973) also helps explain employee reluctance to exercise voice. Employees
may not want to express dissatisfaction or discuss a problem due to their uneasiness
with being the conveyer of bad news, especially to those in higher organizational
ranks (Conlee & Tesser, 1973).
Further, if employee voice efforts are mistaken for venting or complaining,
then voice and complaining can share similar detrimental consequences (see the
upcoming Chapter 3, Section 3.3 for a description of the difference between
complaining or venting and voice). Kowalski (2002) posited that the negative costs to
complaining include an experience of rejection or being ostracized from others. In
addition, employees may be viewed as the fabled “boy who cried wolf”. As such,
when a co‐worker, supervisor, or manager experiences the complaining as habitual
(similar to the boy that cried wolf inciting panic in the community when there was
no threat only to be ignored when the actual danger lurked), the employee

26
ultimately risks not being heard when an actual undesirable situation occurs (i.e.,
the wolf actually appears).
Finally, costs associated with voice may occur at the organizational level as
employees feel that their voice efforts are not heard. Individuals who may go above
their job responsibilities by attempting to enact change or express themselves, but
do not feel heard or validated, may become discontent and critical of the
organization. This, in turn, may negatively impact the functioning of the
organization, such as, increased employee turnover (e.g., Olson‐Buchanan, 1996).
2.3. IDENTIFYING EXIT/NEGLECT/VOICE IN THE LITERATURE
To determine the antecedents to exit/neglect/voice an extensive review of the
literature was conducted. However, a widespread search of the social work as well
as the human and social services literature failed to identify any research related to
Rusbult’s (1988) and Farrell’s (1983) exit/neglect/voice constructs in these settings.
Therefore, this study leans on related constructs of exit/neglect/voice in human
services where available. Otherwise, the study draws extensively from business and
management and organizational psychology literature. Each of these is discussed
below.
Related Constructs: Exit, Intention to Leave, & Turnover
This study utilized the current research on turnover and intention to leave to
understand the antecedents of exit. Although similar to exit, intention to leave

27
reflects individuals’ thoughts about leaving (Mor Barak, Nissly et al., 2001). Exit, on
the other hand, involves individuals’ efforts toward quitting one’s job (Farrell, 1983;
Rusbult et al., 1982). For example, an intention to leave scale item asks: “Do you
intend to look for another job within the next year?” (Abrams, Ando, & Hinkle,
1998); while an exit item asks: “Have you recently spent time looking for another
job?” (Rusbult et al., 1988). Intention to leave and exit are assessed at the individual
level while workers are still employed, while turnover is normally assessed at the
organizational level and involves examining the rate at which employees actually
leave the organization. In two separate meta‐analyses of determinants of turnover,
researchers consistently found that intention to leave and withdrawal behaviors are
strong predictors of actual turnover (Griffeth et al., 2000; Mor Barak, Nissly et al.,
2001). Accordingly, each has similar correlates; therefore this study borrows from
human and social services (including child welfare), and the organizational and
psychology literatures in determining the antecedents to exit.
Neglect and Other Forms of Counterproductive Workplace Behaviors
Neglect, as conceptualized and measured in this study, is one form of
counterproductive and deviant work behavior that involves efforts that limit ones’
involvement in work‐related tasks and organizational activities (e.g., showing up
late for work or meetings and not putting much effort into work) (Bennett &
Robinson, 2000). Therefore, researchers have argued that neglect comprises an

28
employee’s withdrawal or psychological removal from their work environment
(Carmeli, 2005; Rosse & Hulin, 1985). To fully understand neglect and its
antecedents, this study leans on related constructs related to counterproductive
work behaviors in organizational psychology, sociology, and management
literatures. The broad interdisciplinary approach reflects the absence of related
research in human and social services.
Counterproductive work behaviors, in general, are employee behaviors that
have the potential to harm the welfare of the organization, its members, or its
services and clients (P. E. Spector & Fox, 2005). Similar in scope and definition is the
concept of employee deviance (Robinson & Bennet, 1995). Greenbelt and Scott
(1996) dichotomized employee deviance based on production deviance versus
property deviance and D.K. Peterson (D. K. Peterson, 2002) also discussed political
deviance. Production deviance infringes on the work organization’s standards,
norms, and practices that can affect service quality and delivery. Examples of
production deviance closely mirror neglect (as conceptualized in this study) and
include misuse of time (working on personal matters while on the clock, taking a
longer break than acceptable, or working slower than usual) (Robinson & Bennett,
1995). Other forms of production deviance may occur that fall under the umbrella of
withholding effort (employee lack of effort related to the execution of their job)

29
(Bennett and Naumann, 2004; Kidwell & Robie, 2003). According to Kidwell &
Robie, (2003) and Kidwell (1993) these include:
• Free riding is a lack of participation in team‐based activities with the
anticipation of reaping benefits without much effort.
• Social loafing involves reduced effort when working in groups or teams
• Shirking is employees’ propensity to give less than full effort by engaging
in leisure activities while at work often without consequences.
• Job neglect (specifically studied here) involves employees’ lack of
participation or full effort in work task or organizational related
activities.
In contrast, property deviance involves employee behaviors that are harmful
to the organizational dynamics or others assets (e.g., aggressive behavior, sexual
harassment, verbal or physical intimidation) (Robinson & Bennett, 1995). Political
deviance is another variation of employee deviance that includes but is not limited
to employing favoritism, blaming others for mistakes, and gossiping (D. K.
Peterson, 2002).
Despite the differences in magnitude between neglect and an obviously
harmful counterproductive behaviors such as theft, these can be positively related
(Gruys & Sackett, 2003). Based on this notion, Marcus and Schuler (2004) took a
generalist (or broad) perspective in their research on counterproductive and deviant

30
work behaviors to gain insight into its antecedents. This study also adopts a
generalist perspective focusing on production deviance and withholding effort to
understand the antecedents to neglect.
A Closer Look at Workers’ Voice and Related Constructs
Voice versus complaining
For the antecedents to voice, this study relies on the voice literature as well as
its related concepts (see Table 2 for a detailed description of the related concepts).
As previously stated, voice involves efforts aimed at improving dissatisfying work
conditions or promoting organizational change (Creed, 2003; Hagedoorn et al., 1999;
Rusbult et al., 1988). For conceptual clarification, when voice responses are
discussed, I am referring to constructive voice responses as opposed to aggressive
voice or complaining. Constructive voice responses are responses geared to shifting
ones relationship with the organization or solving problems in the workplace, as
opposed to complaining or aggressive voice (Hagedoorn et al., 1999). Complaining is
the expression of subjective dissatisfaction or feelings of discontent that has both
negative consequences and personal benefits (Kowalski, 2002). Kowalski argued
that dissatisfaction is a sufficient but not required part of complaining and in some
cases individuals view complaining as a means to an end. This has some overlap
with aggressive voice. Aggressive voice involves attempts at change for the purpose
of winning without concern for the organization or others (Hagedoorn et al., 1999).

31
On the contrary, voice and related constructs share a common thread—employees’
constructive behaviors aimed at making a positive difference in their workplace or
work lives. In the following section, I discuss the different constructive voice
mechanisms through a review of related constructs.
FIGURE 1. COMPLAINING, AGGRESSIVE VOICE, & CONSTRUCTIVE VOICE

Aggressive
Voice
Seek to change for
purpose of winning
(limited concern for
organization or others)
Express feelings of
pain or
dissatisfaction
Seek to:
Change dissatisfying conditions
Promote organizational change
Complaining or
Venting
Constructive Voice &
Related Constructs

What are the different voice mechanisms?
Based on Hirshcham’s (1970) and Farrell’s (1983) original conceptualization
of voice, the construct was traditionally framed as an employee response to
dissatisfying workplace conditions. Yet, researchers and theorists expanded this
concept to more specifically highlight constructive efforts to improve organizational
or workplace functioning. This refined perspective stand in contrast to critical or
complaint‐based behaviors that lack an orientation toward change. Therefore,
constructs related to constructive voice share a similar goal of improving undesirable
work conditions or promoting organizational change. These can include: speaking

32
up (Premeaux & Bedeian, 2003), taking charge (Morrison & Phelps, 1999), tempered
radicalism (Meyerson & Scully, 1995), issue selling (Dutton, Ashford, OʹNeill, &
Lawrence, 2001), whisteblowing (Miceli & Near, 1992), organizational citizenship
behavior (Borman, 2004; Dennis W. Organ & Greene, 1974), and extra‐role behavior
or personal initiative (Benkhoff, 1996; Frese, Kring, Soose, & Zempel, 1996; Van
Dyne et al., 1995; Van Dyne & LePine, 1998). Silence (Morrison & Milliken, 2000) is
an opposing construct; yet informative due to its shared antecedents to voice.
Because this study takes into account the traditional and refined conceptualizations
of voice in considering its antecedents and outcomes, below I further define and
outline the scope and focus of each voice mechanism.
Speaking up involves “openly state one’s views or opinions about workplace
issues” (Premeaux, 2001, p.vi); whereas, taking charge (Morrison & Phelps, 1999)
and tempered radicalism (Meyerson, 2001; Meyerson & Scully, 1995) reflect
behaviors geared toward changing the status quo. The focus of speaking up is on an
employee’s willingness to discuss workplace concerns with limited apprehension of
retaliation. Premeaux (2001; Premeaux & Bedeian, 2003) further argued that
speaking up does not necessarily involve challenging the status quo but rather is a
product of an employee’s interest in discussing a workplace issue. Alternatively,
taking charge is contextualized as “voluntary and constructive efforts, by individual
employees, to effect organizationally functional change with respect to how work is

33
executed within the contexts of their jobs, work units, or organizations” (Morrison
& Phelps, 1999, p. 403). In this respect, the focus of taking charge extends to change‐
oriented responses that require effort beyond, for example, speaking up about a
concern. These behaviors focus on creating productive change in work
organizations.
Tempered radicalism is similar to taking charge; however, the focus is not just
on the behavior but also on an individual’s state of being. Tempered radicalism
involves the processes by which those committed to not only their organization, but
also their personal and community identity affect change (Meyerson & Scully, 1995,
p. 586). Tempered radicals, as organizational change agents, utilize small wins (i.e.,
selective approaches to create change, driven by unanticipated opportunities) and
“local, spontaneous, action” (expression of feelings or beliefs geared to creating
change) (Meyerson & Scully, 1995).
Issue selling (Dutton et al., 2001) is geared toward enhanced organizational
effectiveness by focusing on a strategic issue as opposed to changing organizational
processes and procedures (Morrison & Phelps, 1999). Issue selling, specifically,
comprise the behaviors taken by individuals in a quest to bring attention to higher‐
ups regarding the activities that performance outcomes within the organization
(Dutton et al., 2001).

34
Principled organizational dissent and whistleblowing share similar foci;
however, the former is primarily internal to the organization and latter is external.
Principled organizational dissent involves employee efforts geared to protest or
change the status quo (Graham, 1986). Employees’ drive to engage in principled
organizational dissent is based on their personal objection to organizational
practices or polices and their assessment of the seriousness of issue. Similarly, when
an employee uncovers objectionable or illegal workplace practices and policies, they
employee may engage in whistleblowing (Near & Miceli, 1996). Whistleblowing is the
terminology used characterize employees disclosure of information to an
administrator, media, or source outside the organization (Near & Miceli, 1996).
Organizational citizenship behavior (Borman, 2004; Dennis W. Organ & Greene,
1974), personal initiative (Frese et al., 1996 & Zempel, 1996), and extra‐role behaviors
(Van Dyne et al., 1995; Van Dyne & LePine, 1998) are broad‐based constructs that
may or may not include voice. These constructs reflect employees’ voluntary
behaviors that extend above‐and‐beyond their daily work duties or formal
requirements of the job. Examples include devoting extra hours to a work task as
needed or helping out a co‐worker in the successful completion of a task (Borman,
2004). These types of efforts also can involve voice related behaviors such as offering
a innovation suggestion that would enhance service quality (K. M. Hopkins, 2002;

35
Van Dyne & LePine, 1998) or making a suggestion to a supervisor to improve the
efficiency of a work processes (Rusbult et al., 1988).
Last, silence (Bell, Edmondson, Meyerson, Nkomo, & Scully, 2003; Milliken &
Morrison, 2003; Noelle‐Neumann, 1984) is also a construct that is used in the
organizational and psychology literature to describe the absence of voice, however it
is not diametrically opposed to voice. Silence involves holding back input that may
be beneficial to the organization (Morrison & Milliken, 2003). Silence, however, is
not necessarily negative, in fact, it may be strategic tool for action among those who
lack socially‐defined power (Bell et al., 2003)
This study relies extensively on voice and related constructs in the theoretical
development of the hypothesized model and in explaining employees’ reluctance to
engage in voice.
2.4. SUMMARY
This chapter provided a description of the exit/voice/neglect constructs as
responses to undesirable experiences at work. As discussed, employees may engage
the following: (1) efforts toward quitting one’s job (exit), (2) efforts that reflect one’s
disengagement from work‐related tasks and organizational activities (neglect), and
(3) efforts to improve dissatisfying work conditions or promote organizational
change (voice). These responses can be categorized as unfavorable versus favorable
to work organizations, employees, or clients. In this, exit and neglect are considered

36
unfavorable, while voice is considered to have favorable outcomes for organizational
effectiveness. Each of the constructs has been more extensively explored in the
organizational psychology, business and management literature than in social work
or human services. Therefore, this study drew from the social work and human
service literature where available as well as related constructs in other fields where
this body of literature is more comprehensively developed. The following chapter
(Chapter 3) presents the conceptual model detailing the relationships among study
variables, building on these literatures to determine the antecedents to employees’
exit/neglect/voice responses.

37
TABLE 2. VOICE & RELATED CONSTRUCTS
Construct Driver (scope) Definition
Voice (Farrell, 1983; Hirschman,
1970; Rusbult et al., 1988)
Efforts to improve dissatisfying work conditions or promote
organizational change
Speaking up (Premeaux &
Bedeian, 2003)
Improving organizational
functioning (internal)
Attempts to discuss a workplace
issues within the organization
(e.g., supervisor, administrator)
Taking charge (Morrison &
Phelps, 1999)
Improving organizational
functioning (internal)
Voluntary and productive efforts
to affect organizational change
Tempered radicalism (Meyerson,
2001; Meyerson & Scully, 1995)
Changing and improving
organizational functioning;
challenging the status quo
(internal)
Processes by which those loyal to
their work and personal identity
seek organizational change
through constructive activities
Issue selling (Dutton et al., 2001) Creating strategic
organizational change
(internal)
Efforts that bring attention to the
organization about trends or
developments (i.e., issues) that
influence performance
Principled organizational dissent
(Graham, 1986)
Creating a buzz or change about
an objectionable policy or
practice
Attempts to protest and/or change
the organizational status quo out
of an objection to a policy or
practice (internal)
Whistleblowing (Near & Miceli,
1996)
Disclosing unethical practices
(internal or external)
Disclosure of illegal, immoral, or
illegitimate organizational
practices
Broadly‐related constructs
Org. citizenship behavior; Extra‐
role behavior; Individual initiative
(Moorman & Blakely, 1995; D.
W. Organ, 1997; Van Dyne et al.,
1995)
Going beyond ones job
responsibilities (internal)
Voluntary behaviors that extend
over and beyond ones job
description or responsibility
Opposing voice construct
Silence (Milliken & Morrison,
2003; Morrison & Phelps, 1999)
Withholding of expression that
may be change‐oriented
Not speaking up about an
organizational idea or concern
that may benefit the organization

38
Chapter 3: Conceptual Model and Research Hypotheses
3.1. INTRODUCTION TO THE CONCEPTUAL MODEL
Figure 2 depicts a simplified version of the study’s conceptual model (refer
to Figure 4 for the comprehensive model). This model illustrates the general
relationships among employee: (1) diversity characteristics, (2) perceptions of work
climate, (3) job stress and psychological wellbeing, and (4) exit/neglect/voice
responses. To provide context, detailed descriptions of each of the major variables
follows:
FIGURE 2. OVERVIEW OF RELATIONSHIPS AMONG STUDY VARIABLES

39

Individualsʹ personal and professional diversity characteristics may be
categorized based on their dominant and non‐dominant group status according to
societal norms (Calvert & Ramsey, 1996). In the United States, men, Whites,
Christians, and heterosexuals are all normatively dominant group members based
on personal diversity characteristics (Calvert & Ramsey, 1996). Similarly with
regard to work‐related characteristics, those with higher educational levels
(Graduate versus Bachelor’s degree), managerial rank (as opposed to direct service
workers), and longer agency tenure (seasoned employees) are considered dominant
group members because these characteristics increase people’s positions of power
or influence. Thus, one’s social group membership offers advantages over other
social groups based on the dominant status of that group (Mor Barak, 2005).
Keltner, Gruenfeld, and Anderson (2003) illustrated this power differential in the
following statement:
Elevated power, we propose, involves reward‐rich environments
and freedom and, as a consequence, trigger approach‐related
positive affect, attention to rewards, automatic cognition, and
disinhibited behavior. In contrast, reduced power is associated with
increased threat, punishment, and social constraint, and thereby
cognition, and situationally constrained behavior (p.265).

This power differential has led researchers to examine how dominant and non‐
dominant group status relates to worker outcomes (Calvert & Ramsey, 1996).

40
Perceptions of work climate comprise employees’ felt experiences of inclusion
in decision‐making as well as supervisor and organizational support. Perceived
inclusion in decision‐making reflects the extent to which employees perceive that
they experienced opportunities to participate in and influence decision‐making
within the organization and their work groups (Mor Barak & Cherin, 1998).
Inclusion in decision‐making is a subcomponent of Mor Barak & Cherin’s (1998)
inclusion‐exclusion model which reflects the continuum of an individual’s
perceptions regarding the extent to which he or she feels part of vital organizational
processes (i.e., access to information and resources, involvement in work group
activities, and inclusion in decision‐making). Perceived supervisor and
organizational support involves employees’ global views about the extent to which
the supervisor or the organization, respectively, values their contributions and cares
about their wellbeing (Eisenberger, Fasolo, & Davis‐LaMastro, 1990; Kottke &
Sharafinski, 1988; Rhoades & Eisenberger, 2002).
Job stress and psychological wellbeing focus on employee perceptions of their
wellbeing based on their role conflict and role ambiguity as well as their mental
health status. For this study, job stress is a function of role conflict and role
ambiguity, both well‐known work stressors. Role conflict is the extent to which
expectations of one or more roles mirror the reality of others’ perceptions of that
role, whereas role ambiguity reflects the extent to which an employee is unclear

41
about work‐related expectations (House & Rizzo, 1972; Rizzo, House, & Lirtzman,
1970). Individuals’ general mental health status or psychological standing reflects
their psychological wellbeing (Goldberg, McDowell, & Newell, 1996). This
specifically relates to employees’ levels of happiness, anxiety, stress, or depression
in general.
Employees’ responses to undesirable work experiences are captured using
Rusbult et al. (1988) and Withey and Cooper’s (1989) exit/neglect/voice constructs.
Both exit and neglect are considered unfavorable responses that may adversely
affect work organizations (Hagedoorn et al., 1999). Exit involves efforts toward
quitting one’s job (e.g., thinking about quitting and sending out resumes). Neglect
involves employees’ efforts that limit ones’ work‐related tasks and organizational
activities (Bennett & Robinson, 2000; Farrell, 1983; Kidwell & Bennett, 1993;
Kidwell & Robie, 2003). The type of employee neglect examined in this study does
not directly relate to an employee’s ineptitude in meeting clients needs, but rather
it involves efforts such as showing up late to work, missing mandatory meetings,
or working slower than usual. Voice (a favorable response) includes employees’
efforts to improve dissatisfying work conditions or promote organizational change
(Farrell, 1983; Hagedoorn et al., 1999; Mayes & Ganster, 1988; Rusbult et al., 1988;
Van Dyne & LePine, 1998). Two examples include: talking to a supervisor about a

42
work issue or suggesting a new innovation for improved work group
communications.
3.2. THEORETICAL FRAMEWORK
Figure 2, the basic research model, depicts the hypothesized relationships
among the study variables based on ecological (person‐in‐environment fit), social
exchange, role stress, empowerment, and communication (spiral of silence) theories.
Person‐in‐Environment Fit
Person‐in‐environment fit supports the overall structure of the proposed
model. Based on an ecological perspective, individuals are best understood in the
milieu of their fit with their environment (Feldman, 2003). Fit is determined by
cultural influences on individuals’ attitudes and behaviors (Feldman, 2003). A
ʺgoodness‐of‐fit” occurs as balance between individuals and their environment is
achieved (Germain, 1979). According to Germain, a ʺmisfitʺ may erode individuals’
physical, psychological or social needs; thus, resulting in stress or decreased
satisfaction. Others scholars have agreed that a lack of person‐in‐environment fit
can negatively impact psychological wellbeing, satisfaction, and behaviors
(Bronfenbrenner, 1979; Lerner, 2002). Therefore, the more a person’s work
environment is meets one’s needs or interests, the greater the proponents to be
satisfied with the job (Ellickson & Logsdon, 2001). Consistent with this theory, the
research model proposes several variables that will assess perceptions of fit:

43
perceptions of inclusion, perceived organizational support, and perceived
supervisory support. More positive experiences in each of these variables, suggest
greater person‐in‐environment fit.
Social Exchange Theory
Social exchange theory (Blau, 1964) lends support to the hypothesized
relationships among perceptions of work climate (inclusion in decision‐making,
supervisor support, and organizational support) and employees’ responses
(exit/neglect/voice). According to social exchange theory, how an individual feels
about a relationship is based on the balance between their efforts in the relationship
and the anticipated or actual rewards (Blau, 1964; Cook, 1977). Based on the norm of
reciprocity, employees reciprocate their perceptions of inclusion and support with
increased commitment to their organization, greater work effort in usual job tasks,
and extra‐role performance. As an example, employees may be more inclined to
attend mandatory meetings or support an organizational initiative if they and their
work team receive an email about their good work.
These theoretical propositions are particularly relevant to the current study
since Farrell and colleagues based their original conceptualization of
exit/voice/neglect on the premise of social exchange (Farrell, 1983; Rusbult et al.,
1988). Similarly, Eisenberger and colleagues (1986) used social exchange theory as
the foundation for perceived organizational support theory. According to these

44
researchers, as employees feel included in decisions affecting their work group or
organization and feel supported by the organization, they may reciprocate by
intending to stay in the job (e.g., not thinking about quitting or not initiating steps to
look for a new job). Also, they may reciprocate by being fully engaged in their work
(e.g., showing up to meetings, putting forth full effort in work tasks, or speaking up
about a concern with the intent on creating positive change).
The social exchange framework also extends to models of psychological
contract breach (Coyle‐Shapiro, 2002; Morrison & Robinson, 1997), which further
helps to explain the relationship between (a) inclusion and support, and (b)
exit/neglect/voice responses. “Perceived contract breach results in a sense of
discrepancy between what is promised and what is fulfilled” (Coyle‐Shapiro, 2002,
p. 927). In a sense, a psychological contract breach symbolizes a sense of betrayal
from the perspective of the employee (Morrison & Robinson, 1997). Based on the
norm of reciprocity, the notion of psychological contract breach reflects employees’
perceptions that their organization or supervisor has not lived up to expectations or
fulfilled its obligations. This, in turn, may influence employee behavior.
Accordingly, as employee experience a psychological breach—that is, they perceive
that their organization or supervisor does not involve them in decision‐making or
support—the employees may reciprocate attitudinally by decreased job satisfaction
or organizational commitment; or behaviorally by quitting or disengaging from

45
work‐related tasks or organizational activities (as assessed in this study) (Coyle‐
Shapiro, 2002). The difference between organizational or supervisory support in a
social exchange relationship and a psychological contract is that the former
addresses reciprocity between the supervisor/organization and the employee, while
the latter addresses employees’ responses when expectations and anticipated
rewards are not fulfilled (Coyle‐Shapiro, 2002).
In summary, social exchange theory offers a straightforward perspective of
how perceptions of work climate are likely to affect employees’ exit/neglect/voice
responses. Based on this theory, employees who feel included and supported by their
organization or supervisors will feel valued and cared about. As a result, the
employee is likely to reciprocate this positive feeling by remaining with the
organization, participating and putting forth sufficient effort in their work, and
seeking opportunities to affect workplace change.
Empowerment Theory
The development and application of empowerment theory has blossomed
since the 1980s within academic disciplines such as management (Herrenkohl,
Judson, & Heffner, 1999; Lee & Koh, 2001), organizational psychology (Block, 1991;
Spreitzer, 1996, 1997), nursing and health promotion (Laschinger, 1996; Laschinger
& Havens, 1996; Laschinger, Sabiston, & Kutszcher, 1997; Rissel, 1994), community
psychology (Perkins & Zimmerman, 1995; N. A. Peterson & Zimmerman, 2004;

46
Zimmerman, 1995), and social work (Carr, 2003; Gutierrez, 1994; Pearlmutter &
Bartle, 2000). At the same time, work organizations have integrated empowerment‐
focused practices into popular management or leadership technologies such as
participative design (Cabana, 1995; Emery, 1993), organizational learning (Ellinger,
Watkins, & Bostrom, 1999), self‐managing work groups (Cummings, 1978; Robbins
& Fredendall, 1995), and organizational development (Latting & Blanchard, 1997).
With the explosion of empowerment‐related theory, research, and practice‐level
interventions, definitions of empowerment vary considerably according to the
discipline. This study uses an organizational‐level understanding of empowering
management practices that aim to increase employees’ felt sense of engagement in
their jobs.
Organizationally‐based empowerment is at its best when the leadership is
shared and its members are afforded opportunities to develop skills and influence
organizational processes (Zimmerman, 1990). However, empowerment at the
organizational‐level has broadly been considered synonymous with employee
involvement (Wilkinson, 1998) and participation in decision‐making (Nykodym,
Simonett, Nielsen, Welling, 1994). These approaches reflect a type of structural
empowerment that involves the delegation of decision‐making authority while
fostering employee autonomy (Dewettnick, Singh, & Buyens, 2003). Wilkinson
(1998), however, critiqued the use of empowerment when employees’ involvement

47
is limited to the purpose of aiding work organizations in achieving goals, yet
employees have a lack of direct power or authority.
Hence, the current study draws from both academic‐ (Spreitzer & Doneson,
forthcoming) and practitioner‐ based (Carroll, 1994) definitions of empowerment as
foundations for understanding employees’ responses to undesirable work
experiences. In their conceptualization of empowerment, Spreitzer and Doneson
focused on processes and outcomes for the employee:

Empowerment enables employees to participate in decision
making, helping them to break out of stagnant mindsets to take a
risk and try something new. Empowering practices allow
employees to decide on their own how they will recover from a
service problem and surprise‐and‐delight customers by exceeding
their expectations rather than waiting for approval from a
supervisor. And perhaps most importantly, empowerment is
viewed as critical in the process of organizational change. Rather
than forcing or pushing people to change, empowerment provides
a way of attracting them to want to change because they have
ownership in the change process. (p.2)

Carroll’s (1994) definition of empowerment focused on interaction between the
worker and organization:
A shared understanding between managers and employees
that employees are trusted and valued partners. Because they
are in the best position to assume responsibility for individual
and team results, employees take action and make decisions
that support the organizationʹs goals. Managers acting as
coaches align goals, measures, and resources in a climate of
trust and open feedback. (p.3)

48
Therefore, an empowerment perspective is used to support the relationship among
employee perceived inclusion in decision‐making, their feelings of job stress, their
sense of psychological wellbeing, and their exit/neglect/voice responses. Accordingly,
empowerment is believed to foster a positive connection between employees and
the work organization (Foster‐Fishman & Keys, 1997) based on employees’ felt
sense of self‐efficacy, control, job autonomy (Honold, 1997; Zimmerman, 1990).
Based on this perspective, as work organizations and supervisors engage in
empowering management practices—the formal and informal organizational
policies and procedures aimed at encouraging employees to take on more
responsibility in their jobs (Wilkinson, 1998)—employees are expected to be more
proactive in affecting workplace change through voice and less inclined to engage in
exit or neglect. Employees’ feelings of empowerment may reduce job stress because
of the supports provided as well as positively heighten their sense of psychological
wellbeing (Gutierrez, 1994). Zimmerman (1990) further explained an outcome of
empowerment on workers “empowered individuals may not always make the best
(or correct) choices, but they may know that they can choose whether to fight or
retreat, to be dependent or independent, and to organize or wait” (p.174).
Therefore, as depicted in the proposed model, empowering management
practices are proposed to influence stress, psychological wellbeing, exit, voice, and
neglect.

49
Role Theory
Role theory (Biddle, 1986; Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964)
offers insight into how job stress affects psychological wellbeing and worker
outcomes. From an organizational perspective, role theory posits that employees
have particular roles in their jobs and perform their work accordingly to those roles.
These roles by and large are molded by the organizational context or workgroup
culture (Dobreva‐Martinova, Villeneuve, Strickland, & Matheson, 2002) as well as
the worker’s social positions (Biddle, 1986). As individuals partake in their various
work‐related roles, they consequently experience occupational role stress (Dobreva‐
Martinova, 2002). Role stress is defined as individual perceptions of the extent to
which they experience conflict, ambiguity, and overload from their personal lives
and work environments (Tetrick, 1992).
Role conflict and role ambiguity are most often the focus of empirical
research with examinations into their antecedents and outcomes (Dobreva‐
Martinova, 2002). Conceptualized by Kahn et al. (1964), role conflict is the extent to
which people’s expectations of their roles mirror the actual reality of that role;
whereas role ambiguity reflects the extent to which an employee is unclear about
their expectations others have of them. From the perspective of role theory, when a
person experiences competing demands, a form of role conflict, he or she will
experience stress and diminished work efforts, more so than if they did not have the

50
expectations imposed (Rizzo et al., 1970). Further Rizzo et al. purported that as
employees experience ambiguity in their jobs, they will cope by taking steps to
solve the problem or they will avoid the sources of stress. As a result, the experience
of ambiguity increases the likelihood of felt anxiety and results in decreased
performance.
In this study, exit, neglect, and voice are three means to respond to stressors
arising from role conflict and role ambiguity. Therefore, as reflected in the
hypothesized model, as employees experience stress associated with role conflict
and role ambiguity, they will experience psychological distress and be more likely
to engage in exit, neglect, and voice.
Spiral of Silence & Employee Voice
Noelle‐Neumann’s (1984; 1991) ‘Spiral of Silence’ theory offers an innovative
perspective on employee voice. The ‘spiral of silence’ model explains why people
may feel a desire to remain silent when faced with the possibility of opposing public
opinion with their views. The model is based on three premises:
(1) People have a general sensitivity to public opinion,
(2) People fear isolation and have an awareness of the behaviors that will result
in exclusion, and
(3) People are weary to express their minority views, primarily out of fear of
being isolated.

51
According to the spiral of silence theory, the more that individuals view their
opinions as similar to prevailing public opinions, they more that may express those
views. If public opinion shifts and individuals note that their views are dissimilar,
they will be less inclined to share them (F. Bowen & Blackmon, 2003).
Specific to the proposed study, as employees feel included in the
organizational process or supported by their supervisor or organization, they may
be less inclined to remain silent and more inclined to make suggestions about
workplace problems or talk with a supervisor (voice). In exercising voice, employees
are considered key sources of creativity and innovation and organizational leaders
may recognize and even may anticipate employee contribution (Piderit, 2000);
however, many employees refrain from constructively expressing their viewpoints
(Huang, Macbeth, Dodge, & Jacobstein, 2004). Why would people be reluctant to
exercise voice under these circumstances? On one hand, employees’ diversity
characteristics or organizational status may lead them to feel out of place in
exercising voice. For example, a relatively new employee may just be beginning to
negotiate his or her way around the organization’s culture or a person in a low job
position may be more reluctant to talk to a supervisor about their ideas focused on
improving an organizational process.
On the other hand, employees’ reluctance to exercise voice may be based on
their fear of negative public opinion (F. Bowen & Blackmon, 2003; Milliken et al.,

52
2003). Initially, some may fear potential negative consequences associated with
speaking up (e.g., a loss of reputation or promotional opportunity). Others may fear
that a relationship with a co‐worker, supervisor, or administrator will be damaged.
Last, an employee may feel that if they offer their ideas and input these may be
overlooked or dismissed, or even deemed risky (Milliken & Morrison, 2003;
Milliken et al., 2003) especially if they hold minority views (Noelle‐Neumann, 1984,
1991)
In addition to the individual dynamics, certain organizational climate
variables may prevent a person from using voice. For example, lack of (a) inclusion
in decision‐making, (b) supervisory support, or (c) organizational support may give
substance to individual fears of negative repercussions for exercising voice. Figure 3
summarizes this argument based on the spiral of silence theory.

53
FIGURE 3. WHAT KEEPS PEOPLE FROM EXERCISING VOICE?

3.3. FACTORS INFLUENCING EXIT/NEGLECT/VOICE RESPONSES
The Role of Diversity Characteristics
Individual factors do not necessarily determine workplace behaviors alone.
Consistent with the person‐in‐environment fit theory, a combination of individual
and organizational factors best predict workplace behaviors (D. K. Peterson, 2002).
Considering this, researchers have demonstrated that certain diversity
characteristics may be predictive of specific factors that directly and indirectly affect
exit/neglect/voice responses. Understanding the influence of employees’ diversity

54
characteristics is important because “there is a fundamental difference between
attributes that make a person unique human beings and those that—based on group
membership rather than individual characteristics—yield negative or positive
consequences” (Mor Barak, 2005, p.122). This study explores the relationship
between diversity characteristics and employees’ choice of exit/voice/neglect
responses based on empirical and theoretical literature. Briefly, as discussed in the
previous Section 3.1., non‐dominant group members include Latinos, African
Americans, Asians, women, bachelor‐degreed employees, direct service workers,
and shorter‐tenured employees, as well as others who hold low socially‐defined
hierarchical status in this society in comparison to their dominant group members.
Dominant group members include Caucasians, men, graduate‐degreed employees,
employees in managerial or supervisory positions, and seasoned employees.
Diversity characteristics   perceptions of work climate
Diversity may affect work climate variables including employee perceived
inclusion, supervisory support, and organizational support. In formative studies of
diversity and inclusion, Mor Barak and colleagues have investigated the
relationships among demographic characteristics and perceptions of inclusion.
Specifically, the researchers found that women and members of racial/ethnic
minorities were more likely to feel excluded from important organizational
processes (Mor Barak & Cherin, 1998; Mor Barak, Cherin, & Berkman, 1998; Mor

55
Barak, Findler, & Wind, 2003; Mor Barak & Levin, 2002). Other researchers have
reported similar findings in their examinations of diversity characteristics and
perceived exclusion to critical work processes (Milliken & Martins, 1996). In all,
those who hold an outsider as compared to insider status (similar to non‐dominant
group status) within a particular organization may feel marginalized and therefore
a lack a sense of belonging or inclusion within their work organization (Stamper &
Masterson, 2002).
Research that explored relationships among diversity characteristics and
perceptions of supervisor and organizational support have used these
characteristics as control variables (Rhoades & Eisenberger, 2002). A few researchers
did report significant findings related to diversity characterizes and perceived
support. Nissly et al. (2005) found that graduate‐degreed workers reported lower
levels of social support from their supervisors than bachelor‐degreed workers
among a sample of child welfare workers. In a study of organizational stress, job
satisfaction, and turnover among social workers, Acker (2004) found that workers
with a higher level of education reported lower levels of support from management.
Scott (n.d.) found that age was not statistically correlated with perceived
supervisory support and Eisenberger et al. (2002) found that organizational tenure
was not associated to supervisor or organizational support in a longitudinal study.

56
Considering that researchers have found differences based on diversity
characteristics, the current study hypothesizes that non‐dominant groups will
perceive that they experience less inclusion in decision‐making and support than
dominant group members. This hypothesis is further supported by the perspective
that one’s social group membership may offer advantages over other social groups
based on the dominant status of that group (Calvert & Ramsey, 1996; Keltner et al.,
2003; Mor Barak, 2005). Therefore, as individuals hold a membership in a less
dominant group, they will have an increased propensity to perceive that they
experience lower levels of inclusion in decision‐making or support: Therefore, the
following is proposed:
HA. As compared to dominant group members, non‐dominant group members
perceive that they experience the following:
1. lower levels of inclusion in decision‐making and
2. lower levels of supervisory and organizational support.
Diversity characteristics   job stress & psychological wellbeing
Based on the influence of power in an organizational context, non‐dominant
and dominant group members may experience differences in their job stress and
psychological wellbeing. Stone‐Romero, Stone, and Salas’s (2003) proposed that
cultural background may influence how individuals’ conceptualize their role within
an organization as well as their behavioral intention regarding that role. Moreover,

57
Brooker and Eakin (2001) posited that power influences job stress through two
components:
(1) The meaning that individual attribute to the events and features of the
work environment. For example, employees may attach symbolic
importance to specific workplace circumstance (e.g., who has the corner
office or receives more rewards).
(2) The distribution of stressors among employees in which those with less
organizationally‐based hierarchical power may more likely be in job
positions that physically and economically taxing.
Therefore, employees who hold membership in a dominant group may
experience diminished stress because of the privileges that accrue to those of a
higher status. For example, women employees (non‐dominant) in human service
settings may have a greater propensity to experience high levels of stress and
pressures than men (dominant), even though these organizations may have
proportionately more women than men (Ratliff, 1988). These may be further
compounded as an individual holds a combination of non‐dominant group
characteristics.
Stone‐Romero’s et al. (2003) and Brooker and Eakin’s (2001) propositions
can be applied to the relationship among diversity characteristics and employees’
psychological wellbeing. As such, Mor Barak, Findler, and Wind (2003) found that

58
African American and women workers reported lower psychological wellbeing as
measured using the General Health Questionnaire (GHQ) than the other ethnic
groups and men. Further, Anderson and Berdhal (In Press) found that high power
individuals (those with more perceived influence) reported experiencing more
positive emotions (e.g, happiness) than less dominant groups. This may be
attributed to perceptions or actual experiences of discrimination for those
historically marginalized groups. Rollock and Gordon (2000) surmised that
racism and other “isms” (e.g., sexism, ageism) have deleterious effects on mental
health. Thus, with regard to the job stress and psychological wellbeing, the
following is hypothesized:
HA. As compared to dominant group members, non‐dominant group members
will experience:
3. greater role conflict and role ambiguity.
4. a lower sense of psychological wellbeing
Diversity characteristics   exit
The relationship among diversity characteristics and employee intention to
leave or turnover has been well documented, but with variable results across
disciplines and particularly in the human services. In a review of the literature
related to gender and turnover, inconsistent findings have emerged. Women have
been found to be more likely to quit their jobs than men (Mor Barak et al., 2006).

59
However, Alexander, Lichtenstein, Oh, and Ulman (1998) had an opposite finding:
women nursing employees in a psychiatric setting were less likely to intend to quit
than men. Furthermore, several researchers have found no relationship between
gender and turnover intentions (Dickinson & Perry, 1998; Mor Barak, Nissly et al.,
2001; Nissly et al., 2005; Valentine, 2001) and actual turnover rates (Griffeth et al.,
2000).
Human services studies assessing the relationship between ethnicity and
intention to leave and turnover have also found mixed results. Valentine (2001)
found that race and intention to look for other work were positively related.
Valentine’s findings indicated that racial minorities were more engaged in job
search behaviors that Caucasians. Milliken and Martins (1996) also found that
members of minority groups were less likely to remain in their jobs. Similarly Abu‐
Bader (Abu‐Bader) found that majority group members had lower turnover than
non‐majority group members in a study of social workers in Israel. On the other
hand, Nissly et al. (2005), Dickinson and Perry (1998) and Griffeth (2000) found a
lack of support for the relationship between ethnicity and turnover or turnover
intentions. Somers and Birnbaum (2001) specifically investigated the difference
between Blacks and Whites in their work attitudes, including job withdrawal
intentions. The authors also found a lack of correlation between race and intention
to leave. The lack of significant relationship between ethnicity and turnover

60
intentions received further support from Mor Barak et al.’s meta‐analysis (2001) on
predictors of turnover in the human services.
Professional factors have been studied more extensively studied in relation
to intention to leave or turnover. Balfour & Neff (1993) and Nissly et al. (2005) found
that graduate‐degreed workers had a greater probability of turnover and greater
reported intention to leave than those with bachelor’s only degrees. Mor Barak et al.
(2001) also found a significant positive correlation between educational level and
actual turnover. Balfour & Neff (1993) attributed this relationship to master’s level
employees having higher expectations and increased perception of job alternatives
(that come with higher educational attainment) that ultimately increases their
chance leaving their job. In contrast, in a longitudinal study, Krausz (1995) found
that education was not a predictor of intention to leave a job or profession in a
hospital setting.
With regard to job position, empirical research supports the notion that job
position is unrelated to intention to leave or turnover (Krausz et al., 1995;
Landsman, 2001; Michaels & Spector, 1982). However, some have had opposing
findings. Nissly et al. (2005) found a positive correlation between child welfare
direct service workers and their propensity to leave. In contrast, Blankertz &
Robinson (1997) found that supervisors were more likely to consider quitting their
jobs than were direct service employees.

61
Last, numerous studies have investigated the influence of organizational
tenure. Several researchers have found a negative relationship between job tenure
and intention to leave, with seasoned employees less likely to intend to leave their
work organization (Balfour & Neff, 1993; Bussing, Bissels, Fuchs, & Perrar, 1999; de
Croon, Sluiter, Blonk, Broersen, & Frings‐Dresen, 2004). However, Michaels and
Spector (Michaels & Spector, 1982) found that job tenure was unrelated to turnover
or turnover intentions. Because existing research has produced mixed results, a
nondirectional hypothesis, rather than directional hypothesis, is proposed regarding
the relationship between diversity characteristics and exit:
HA.5. As compared to dominant group members, non‐dominant group members
will differ in their exit responses.
Diversity characteristics   neglect
Empirical studies specifically examining the relationship between job neglect
and individual’s diversity characteristics are limited. Rather, researchers have
focused on a range of related counterproductive workplace behaviors as outcome
variables. Researchers have found that employees who are young, new to their jobs,
work part‐time, and are in low paying positions are more likely to engage in
production or property deviance (Frank, 1989, Hollinger & Clark, 1983). Lau, Au,
and Ho (2003) found that employees who are young and women were more likely
to be absent from their job in a meta‐analysis of 40 published studies. A separate

62
research study of 3,800 employed and self‐employed individuals in the British
workforce revealed that men had higher reported lateness than women and those
with limited educational qualifications had lower counts of lateness (Clark, Peters,
& Tomlinson, 2005). In contrast, Koslowsky, Sagie, Krausz, and Dolman (1997)
found demographic characteristics were not significantly associated with lateness in
a meta‐analysis of empirical articles that examined lateness as an outcome variable.
Despite these findings, researchers have been critical of examinations into
demographic characteristics as determinants of counterproductive workplace
behaviors. They declared that significant relationships might, in part, be a result of
the job environment (see Robinson & OʹLeary‐Kelly, 1998) and related to social and
interpersonal factors. Hence, examining the literature relating to diversity
characteristics and work withdrawal (job neglect) as a specific form of
counterproductive work behavior reveals scarce and conflicting information for
drawing conclusions about child welfare workers as a study population. Based on
the evidence reviewed, a hypothesis of no relationship between diversity
characteristics and neglect is proposed.
HA.6 As compared to dominant group members, non‐dominant group members
do not differ in their neglect responses.

63
Diversity characteristics   voice
The dominant/non‐dominant group framework and the spiral of silence
theory, lead credence to the proposition that diversity characteristics influence
employees’ engagement in voice responses. Therefore, in the hypothesized model,
certain employee diversity characteristics serve as antecedents of employee efforts
to improve dissatisfying work conditions or promote organizational change (voice).
Specifically, the literature suggests that employees who are different from the
‘mainstream’ may be less inclined to talk about aspects of their identity (F. Bowen &
Blackmon, 2003). This reluctance may then spark a ‘spiral of silence’ in which
employees refrain from exercising their voice and expressing dissatisfaction about
their work group and organization (F. Bowen & Blackmon, 2003). Five studies
illustrate this relationship.
In an exploratory study, Milliken, Morrison, and Hewlin (2003) discussed
job position as a factor contributing to employee lack of voice. The authors stated
that compared to older, more experienced, and higher ranking colleagues, workers
who are of a younger age, inexperienced, or have a low ranking position are more
aware of the potential unfavorable consequences to speaking up. These workers
may then attribute their relative silence to their low power and lack of credibility
within the work organization. Similarly, in a study of personal and situational
factors that contribute to employee voice responses, LePine and Van Dyne (1998)

64
found that women, non‐Whites, and those without a college education have a lower
propensity to engage in voice than men, and their White, and college educated
counterparts. In a related study, Bowen and Blackmon (F. Bowen & Blackmon, 2003)
also argued that differences based on diversity characteristics may impact
employees’ voice. Based on the framework of invisible/visible difference, these
authors’ proposed as non dominant group members are uncomfortable in revealing
their sexual identity, they may be less willing to speak out or voice concerns in other
areas. Although, sexual identity is not examined in this study, this finding provides
perspective on the challenges of non‐dominant groups in exercising voice.
In contrast to the previously discussed studies, Bell, Meyerson, Nkomo, &
Scully (2003) examined ethnic differences between Black and White women’s’
tempered radicalism. They qualitatively captured the life histories and workplace
changes of 120 women. They found that Black women were more likely to express
themselves through voice as a verbal action, while White women were more focused
on trying to understand the problematic situation rather than acting.
A few researchers have specifically looked at the role of power derived from
hierarchical position in relationship to voice. For example, Islam and Zyphur (2005)
conducted an experiential study exploring the relationship between power
(operationalized by a simulation of a CEO versus line worker position), social
dominance (belief in hierarchy in which a person places one social group as

65
dominant or subordinate), and voice. Although the researchers did not find that
power had a direct relationship with voice, they did find that a social dominance
orientation moderated the relationship between hierarchical position and voice.
Anderson and Berdhal (2002) found that high power individuals (those with more
perceived influence) have a greater tendency to express their attitudes and opinions
than people with low power. Therefore, based on the spiral of silence theory and
supporting research the following is hypothesized:
HA.7. As compared to dominant group members, non‐dominant group members
engage less in voice responses.
The Role of Work Climate Variables
Inclusion in decision‐making
Employee perceived inclusion might also influence the extent to which
employees feel supported by their supervisor or work organization. Allen, Shore,
and Griffeth (2003) argued that human resource management practices are the
mechanisms by which work organizations can convey that they care about the
wellbeing of its members as well as value member contributions. Accordingly,
employees judge these practices as determinants of whether they think their
organization values them or cares about their wellbeing (Eisenberger et al., 1986). In
studies of salespersons and business professionals, perception of human resource
practices (including involvement in decision‐making) was a correlate of perceived

66
organizational support (Allen et al., 2003). Empowerment theory further supports
the relationship between these two constructs. Accordingly, when an employee
experiences a sense of empowerment from being involved in organizational or work
group decision‐making, he or she may feel that the organization values their
contribution and cares about their wellbeing. These outcomes are components of
perceived supervisory and organizational support. Hence the following is
hypothesized:
HB.1. The more employees perceive that they are included in decision‐making, the
more they will perceive their supervisor or organization as supportive.
Inclusion also indirectly relates to exit/neglect/voice responses through
employees’ job stress. In a systematic review of 40 articles that examined the work‐
related correlates of psychological wellbeing and sickness, Michie and Williams
(2003) found that limited opportunities for participation in decision‐making was
associated with psychological distress, including depression, anxiety, and an overall
reduced sense of psychological wellbeing. In an earlier study, Schuler (1980) also
found that employee participation was related to perceptions of role‐related conflict
and ambiguity. In this vein, the more that employees were involved in decision‐
making processes, the lower their levels of job stress associated with their
expectations and clarity about their work roles. Additionally, the organizational

67
context and supervisors influenced these perceptions. For that reason, an inverse
relationship is hypothesized between inclusion and decision‐making and job stress:
HB.2. The more employees perceive that they are included in decision‐making, the
less they will experience job stress associated with role conflict or role
ambiguity.
Research indicates that inclusion and related constructs can be predictors of
employees’ psychological wellbeing. Mor Barak and colleagues (Mor Barak et al.,
2003; Mor Barak & Levin, 2002) found that employee perceptions of inclusion
predicted job satisfaction and wellbeing, even when personal and professional
diversity characteristics were controlled. Research also supports this relationship
based on the extent to which employees are afforded opportunities to participate in
decision‐making. Accordingly, participative work environments have been linked
to positive mental health and wellbeing of employees (Mackie, Holahan, & Gottlieb,
2001; Witt, Andrews, & Kacmar, 2000). Last, job stress, which is an outcome of
participative decision‐making, has been linked to employee wellbeing (Sparks,
Faragher, & Cooper, 2001). These findings are consistent with empowerment
models in that as employees perceive that they have opportunities to be involved in
the critical organizational processes, such as decision‐making, they will feel
empowered which will positively affect their mental health and psychological

68
wellbeing. Ultimately, this sense of heightened psychological wellbeing may
influence their workplace behaviors:
HB.3. The more employees perceive that they are included in decision‐making, the
greater their sense of psychological wellbeing.
As noted in the previous section (3.2), social exchange theory and
empowerment theories offer a rationale for the relationships between work climate
variables and exit/neglect/voice. In essence, these theories suggest that as employees
feel included in the decision‐making process and supported by their supervisors
and the organization they will respond by reducing exit and neglect efforts, and
engaging in greater voice as a change‐oriented effort. This theoretical perspective is
supported by empirical research that looks at the influence of inclusion in decision‐
making, supervisory support, and organizational support on employees’
exit/neglect/voice efforts.
Perceived inclusion in decision‐making reflects the extent to which
employees feel involved in decisions that impact their work group, organization,
and job functioning (Mor Barak et al., 1998). Researchers reported that employees
who felt included were less likely to have an intention to leave (Mor Barak et al.,
1998). Similar findings were found between participative decision‐making and
perceived work control in relation to employee turnover (P.E. Spector, 1986). By
extension, these findings may also support the relationship between inclusion in

69
decision‐making and neglect. That is, employees who are involved in the decisions
that affect their jobs may be less inclined to withdraw from work activities (e.g.,
avoiding supervisors, working slowly). On the other hand, as employees experience
feeling empowered due to their involvement in critical organizational processes,
they may be more likely to be invested in the functioning and success of the
organization (Foster‐Fishman & Keys, 1997). This sense of empowerment results in
employees increasing their contributions such as voice responses (Foster‐Fishman &
Keys, 1997).
HB. The more employees perceive that they are included in decision‐making,
4. the less they will engage in exit responses.
5. the less they will engage in neglect responses.
6. the more they will engage in voice responses.
Perceived Supervisory Support
Because of the multifaceted role of supervisors, researchers have found that
supervisory support is vital to generating positive worker outcomes, including
reducing role conflict and role ambiguity as forms of job stress (Kelloway & Day,
2005; Nissly et al., 2005). Further, Michie’s and Williams’ (2003) systematic review of
empirical studies found that a variety of workplace factors, including a lack of
supervisory support, contributed to poor psychological wellbeing. Duffy, Ganster,
and Pagon (In Press) also found that perceptions of social undermining by

70
supervisors is positively correlated with workers’ somatic complaints, an indicator
of employee mental health. Social undermining is the contrary to social support in
that it includes behaviors that impede an employees’ ability to maintain positive
relationships, success in the workplace, or a positive relationship. Supervisors or
managers can unknowingly or intentionally cause distress in an employee, simply
based on their hierarchical positions of power (Sparks et al., 2001). Therefore, the
current study hypothesizes that perceived supervisory support is inversely related
to role conflict and role ambiguity and positively related to psychological wellbeing.

HB. The more employees perceive their supervisor as supportive,
7. the less they will experience job stress associated with role conflict or
role ambiguity.
8. the greater their sense of psychological wellbeing.
Research on exit, intention to leave, and turnover has determined that a host
of organizational antecedents. Most notably, the quality and type of supervision has
been cited as one of the most instrumental factors contributing to intention to leave
or turnover (American Public Human Services Association [APHSA], 2005;
Landsman, 2001; United States General Accounting Office [GAO], 2003 et al., 2001,
Samantrai’s, 1992; Vinokur‐Kaplan, 1991). In the nursing field, Ribelin (2003)
declared, “nurses don’t leave hospitals, they leave managers” (p.18). Based on prior

71
research, this sentiment also appears to apply to human service workers and their
supervisors. Smith (2004) found that perception of supervisory support and
competence were correlates of intention to leave or actual turnover. In a study of
mental health service providers, Acker (2004) found that social support and
assistance from supervisor were significant correlates of intention to leave.
Hagedoorn et al. (1999) found a negative relationship between satisfaction with
supervisor and exit. Last, Nissly et al. (2005) found that social support received from
supervisors was a significant predictor of intention to leave among child welfare
workers.
While the quality and type of supervision has been cited as one of the most
instrumental factors contributing to intention to leave or turnover, few studies
directly examined the role of supervisory support in predicting neglect. For example,
Hagedoorn et al. (1999) found a negative relationship between satisfaction with
supervisor and neglect. In simpler terms, those who were satisfied with their
supervisors were less likely to neglect their work. In a study of a variety of work
settings (corporate, human service, manufacturing, and retail industries), Kidwell
and Bennett (2003) found that job neglect was inversely related to three different
forms of supervisory support: (1) task visibility—the extent to which employees
perceive that their supervisors have an awareness of their effort and performance;
(2) supervisor expertise—the expectation that supervisors provide expertise, and (3)

72
supervisor consideration—the extent to which supervisors offer positive accolades
and encouragement. Finally, Duffy, Ganster, and Pagon (In Press) found that
perceptions of social undermining by supervisors was a significant predictor of
counterproductive work behaviors ranging from taking extended breaks to theft.
Social undermining may be used as an antonym of social support in that it involves
intentional behaviors that impede on an employee’s ability to maintain positive
relationships, success in the workplace, or a positive reputation. Consequently, the
presence of social support has similar effects as the absence of social undermining.
Limited studies were found that directly examined the role of supervisory
support in predicting voice. However the lack of empirical research does not
diminish the likely importance of supervisors as champions for employee voice. In a
study of operating room personnel, Edmondson (2003) found that supervisors with
non defensive responses to workers’ comments fostered psychological safety that
aided workers in speaking up. Hagedoorn et al. (1999) found positive relationships
between supervisor satisfaction and employees’ voice among teachers and nurses. In
other words, those who were satisfied with their supervisors were more likely to
voice their concerns.
Social exchange theory, empowerment models, and the spiral of silence
theory illuminate the hypothesized relationships among supervisory support and
exit, neglect, and voice. Accordingly, as employees feel their supervisors support

73
them through caring and appreciation of their contribution, they may feel more
empowered to take steps to improve dissatisfying conditions or promote
organizational change. In contrast those who do not perceive their supervisor as
supportive may be less likely to engage in such behavior. Thus the following is
hypothesized:
HB. The more employees perceive their supervisor as supportive,
9. the less they will engage in exit responses.
10. the less they will engage in neglect responses.
11. the more they will engage in voice responses.
Perceived organizational support
Despite the profound direct and indirect influence of supervisory support
on employees’ job stress and psychological wellbeing, these relationships do not
occur in a vacuum. Employee‐perceived organizational support can affect
employees in the same manner as perceived supervisory support. Based on social
exchange theory and the norm of reciprocity, high levels of perceived
organizational support could help reduce employee stress and strengthen
psychological wellbeing. Those resultant feelings could create better clarity about
the expectations and clarity about employees’ psychological wellbeing (Eisenberger,
Armeli, Rexwinkel, Lynch, & Rhoades, 2001).
HB The more employees perceive their work organization as supportive,

74
12. the less they will experience job stress associated with role conflict or role
ambiguity.
13. the greater their sense of psychological wellbeing.
Consistent with social exchange theory, perceived organizational support
can also affect employee exit/neglect/voice responses (Eisenberger et al., 1990).
Rhoades and Eisenberger (2002) and Eisenberger et al. (1990) found that strong
perceptions of organizational support contributed to decreasing intention to leave
and actual turnover. Allen, Shore, and Griffeth (2003) also found that perceived
organizational support was correlated with intention to leave as mediated by
commitment and satisfaction. Although no research was found supporting the
relationship between organizational support and neglect, social exchange theory
would predict that this relationship stems from the same factors that affect
supervisory support and intention to leave. That is, when employees perceive that
that their organization does not value their work and care about their wellbeing,
they may be more likely to engage in neglect oriented behaviors, such as working
slowly or taking extended breaks.
Perceived organizational support may also affect employees’ willingness to
exercise voice. Organizations can send verbal and nonverbal messages that sticking
with the status quo is most advantageous for employee success (Perlow & Williams,
2003). Edmondson and Detert (2005) suggested that those organizational leaders in

75
higher ranking positions than direct supervisors may directly and indirectly send
such message because they signify organizational standards, norms, and acceptable
practices. A few studies support this linkage. Morrison and Phelps (1999) found that
top management openness (the degree to which top management is believed to
encourage and support suggestions for change) was significantly related to taking
charge (a construct similar to voice). In a study of 140 child and family service
workers, Hopkins (2002) found that perceptions of organizational support were
correlates of employees’ organizational citizenship behaviors (a broad‐based
construct that includes voice). Accordingly, employees who feel that their
organization is caring and values their contribution may feel more empowered and
therefore engaging in voice as a step toward create organizational change.
Based on the importance of organizational support in fostering or hindering
exit, neglect, and voice, the following is hypothesized.
HB. The more employees perceive their work organization as supportive,
14. the less they will engage in exit responses.
15. the less they will engage in neglect responses.
16. the more they will engage in voice responses.

76
The Role of Job Stress & Psychological Wellbeing
Job Stress (role conflict and role ambiguity)
From a broad‐based perspective, occupational or job stress is a function of
workplace conditions in which the job requirements are incongruent with
employees’ skills and expectations (National Institute for Occupational Safety and
Health (NIOSH), n.d.). Role stress—specifically role conflict and role ambiguity—is
a form of job stress related to the extent to which one’s expectations of their role
mirror the actual reality of that role (role conflict) and the extent to which an
employee is unclear about their role (role ambiguity) (Kahn et al., 1964). Although,
the current study focuses on employees’ job stress associated with role conflict and
role ambiguity, job stress in general may affect human service workers’
psychological wellbeing and propensity to engage exit/neglect/voice responses.
Job stress can affect psychological wellbeing (Lundberg, 2000) as defined by
the absence of such symptoms as depression and anxiety (Debreva‐Martinova,
2002). Johnson and Indik (1997) reported that clinical depression affects more that
17.5 million working age individuals per year, which, can be partially attributed to
work‐related stress. In a systematic review of 40 articles that examined the work‐
related correlates of psychological wellbeing and sickness, Michie and Williams
(2003) found that role conflict and ambiguity were associated with poor
psychological wellbeing. Further, Boles, Johnston, and Hair (1997) found that role

77
conflict not ambiguity was related to emotional exhaustion. Thus, the following is
hypothesized:
HC.1. The more employees experience job stress associated with role conflict or role
ambiguity, the lower their level of psychological wellbeing.
How individuals deal with stress in general is illustrated by the
“fight/flight” response (Eisler & Levine, 2002; Mayes & Ganster, 1988). When
individuals are subjected to threats or stress, they can response with attachment or
withdrawal behaviors. The latter behaviors—flight responses—are considered
withdrawal behaviors (such as exit and neglect) and in some cases voice may be
considered a fight response (Mayes and Ganster, 1988). With exit considered a flight
response, it not surprising that researchers have consistently found that stress is one
of the strongest predictors of intention to leave and turnover (de Croon et al., 2004;
Mor Barak, Nissly et al., 2001; Williams, Konrad, Scheckler, Pathman, & al, 2001)
and particularly preventable turnover (APHSA, 2005). Preventable turnover is
characterized as leaving the organization due to factors that do not include death,
retirement, family considerations (marriage, spouse job move, or parenting), or
returning to school. Specifically, in a meta‐analysis of turnover among human
service workers, Mor Barak et al. (2001) found a positive correlation between job
stress and intention to leave, but a lack of statistically significant correlation with
actual employee turnover. In a study of mental health service providers, Acker

78
(2004) also found that role conflict and role ambiguity were statistically significant
correlates of intention to leave.
The relationship between job stress and neglect is expected to mirror the
relationship between job stress and exit. As such, researchers have found that as job
stress occurs employee have greater propensity to engage in counterproductive
work behaviors (P. E. Spector & Fox, 2005). For example, Jones (1981) and Jones and
Boye (1994) found that burnt‐out nurses extended work breaks more often. Burnout
is a form of chronic stress that is long term and unresolved (Maslach, Schaufeli, &
Leiter, 2001). Further, Kidwell and Robie (2003) found that role ambiguity was
positively related to withholding effort, a construct that includes job neglect. In this
case, those employees who are unclear about their job responsibilities may chose
differing forms of neglectful work behavior. Last, in a review of studies linking
stressors to several different types of counterproductive work behaviors Spector
and Fox (2005) found that counterproductive work behaviors are a response to job
stressors and conditions that create downbeat emotions.
Role theory supports this connection between stress and neglect. Based on
role stress theory. Rizzo et al. (1970) purported that as individuals experience
inconsistency in their lives due to role conflict, the resulting stress will cause less
effective performance than if the expectations were not imposed. Further, Rizzo et
al. (1970) purported that when dealing with role ambiguity, an individual may use a

79
variety of defense mechanisms to deal with the situation as well as perform less
effectively. Thus, it can be inferred that neglect, as a flight response, is a form of a
defense mechanism that helps individuals cope with adverse circumstances. Based
on these findings and role stress theory, it is hypothesized that role conflict and
ambiguity will be positively related to employees’ efforts with regard to quitting
their job (exit) and limiting their work‐related tasks and activities (neglect).
To date, no research was found that depicted the relationship between
organizational‐based stress and voice. However, because voice can be considered a
fight response (Mayes & Ganster, 1988), perhaps the experience of role stress will
contribute to greater voice. However, a lack of empirical evidence serves as a caution
against forming a directional hypothesis.
Therefore, the following is hypothesized with regarding the relationships
between job stress and the exit/neglect/voice responses:
HC.   The more employees experience job stress associated with role conflict or role
ambiguity,
2. the more they will engage in exit responses.
3. the more they with engage in neglect responses.
4. the more they will engage in voice responses.

80
Psychological wellbeing
Many organizational studies have focused on mental health as an outcome
variable and therefore empirical research considering its influence on worker
outcomes are less prevalent (Wright, Bonett, & Sweeney, 1993). Nevertheless, poor
mental health and psychological wellbeing can have unfavorable consequences for
work organizations and their employees (Diener, 2000). Conversely, research
demonstrates that happiness (as a trademark of wellbeing) and positive affect are
linked to positive worker outcomes (Lyubomirsky, King, & Diener, 2005).
Lyubomirsky et al. conducted an extensive review of over 200 empirical cross‐
sectional and longitudinal studies that addressed whether happy people are
successful in many aspects of their life. The researchers found that workers who
were happy and satisfied were more likely to perform well because they had lower
reports of counterproductive withdrawal behaviors such as absenteeism, turnover,
and retaliatory behaviors. Happiness is part of psychological wellbeing because it
reflects those “who experience frequent positive emotions, such as joy, interest, and
pride and infrequent (though not absent) negative emotions, such as sadness,
anxiety, and anger” (Lyubomirsky et al., 2005). Koeske and Kirk (1995) also found
that psychological wellbeing was related to employee retention among social
workers. Thus:

81
HC.5. The more employees experience a heightened sense of psychological
wellbeing, the less they will engage in exit responses.
Several studies suggest that constructs related to psychological wellbeing,
such as negative emotions and affectivity (e.g., fear and anger) are positively
correlated with neglect. Spector and Fox (2005) proposed that personality differences
might help explain counterproductive work behaviors because of the linkage
between an individual’s personality and their propensity to experience negative
emotions, which, in turn, help to predict a variety of counterproductive work
behaviors. Martinko, Gundlach, and Douglass (2002) shared this perspective. These
authors suggested that individuals high in negative affectivity may be more focused
on the negative aspects of their lives and their workplaces, and; hence may be more
inclined to engage in a counterproductive behavior. Finally, Boyd (1997) found that
lower quality of employee‐related health and wellbeing reduced productivity and
increased absenteeism. Person‐in‐environment fit theories are consistent with this
perspective. Martinko (2002) maintained that employee engagement in
counterproductive workplace behaviors is a “result of a complex interaction
between the person and the environment in which the individual’s causal reasoning
about the environment and expected outcomes drive the individual behavior”
(p.41). Therefore, it is hypothesized that:

82
HC.6. The more employees experience a heightened sense of psychological
wellbeing, the less they will engage in neglect responses.
Empowerment theory supports the hypothesized relationship between
psychological wellbeing and voice. As employees feel a sense of confidence and
capability in their lives, they may be more willing to engage in voice‐related
behaviors. Empirical research on related constructs to wellbeing and voice also
support this hypothesized relationship. LePine and Van Dyne found that self‐
esteem (LePine & Van Dyne, 1998) as well as the absence of neuroticism (LePine &
Van Dyne, 2001) were positively related to individuals’ engagement in voice
responses. The authors characterized neurotic individuals as those who are
emotionally unstable and have a lower sense of self‐worth. Graham (1986)
suggested that employees who engaged in principled dissent (a change‐oriented
extra role behavior) had more self‐confidence, which aided them in understanding
the feasibility of embarking on organizational change. Thus:
HC.7. The more employees experience a heightened sense of psychological
wellbeing, the more they will engage in voice responses.

3.4. RELATIONSHIPS AMONG EXIT, NEGLECT, AND VOICE
Investigators have proposed that relationships exist among exit, neglect, and
voice (Dowding et al., 2000; Withey & Cooper, 1989). Withey and Cooper (1989)

83
proposed one possible linear sequence among them: an employee may voice their
concerns to a supervisor and then quit if those attempts are successful. Other times
the behaviors may work in conjunction with one another (M. Keeley & Graham,
1991). For instance, an employee may decide to talk with a supervisor about a
specific problem while looking for another job. However, only a few studies have
examined the relationships among exit, neglect, and voice and these mostly have
been cross‐sectionally. While these studies have made important contributions to
the literature, Withey & Cooper (1989) declared that longitudinal assessment (6
months or more) are advantageous. Others have shared a similar sentiment.
Dowding et al. (2000) have concerns about use of cross‐sectional designs to study
the relationships among these three constructs. The proposed study will fulfill this
need by assessing the influence of these relationships on each other over time.
Therefore, to aid in determining the connections among exit, neglect, and voice, each
of the hypothesized relationships (discussed in the following section) includes the
element of time: a six‐month time lag between baseline and follow‐up responses
Exit   Neglect & Voice
Because exit involves employees’ efforts toward quitting their job (and not
actual employee turnover), this study sought to determine its influences over time.
Essentially, this study hypothesizes the following: As employees who engage in
efforts toward quitting their job but remain in the organization, for whatever

84
reason, will be more likely to engage in neglect over time. Additionally, these
employees would be less likely to subsequently engage in change‐related efforts
such as voice. While, no empirical research was found that support these
relationships, the hypothesis is supported based on psychological contract breach
(Coyle‐Shapiro, 2002; Morrison & Robinson, 1997). Based on this notion, employees
who have unfilled expectations about their job may then consider quitting. Effort
toward quitting can increase psychological distance between the employee and the
work organization (Carmeli, 2005; Rosse & Hulin, 1985), and in turn increases their
propensity to disengage from work‐related tasks and organizational activities (e.g.
working slowly, avoiding supervisors) as well as their desire to engage in efforts
that may improve dissatisfying working conditions or promote organizational
change (i.e., talking to supervisor about a specific organizational problem).
Therefore, this study hypothesizes (1) a positive relationship between exit and
neglect and (2) a negative relationship between exit and voice over time:
HD. The more employees’ engage in exit at baseline,
1. the more they will engage in neglect responses at the six‐month follow‐
up.
2. the less they will engage in voice responses at the six‐month follow‐up.

85
Neglect   Exit & Voice
As previously discussed, neglect involves efforts that reflect one’s
disengagement from work‐related tasks and organizational activities (Bennett &
Robinson, 2000; Hanisch et al., 1998; Rusbult et al., 1988). From the perspective of
person‐in‐environment fit (Feldman, 2003; Germain, 1978, 1979), as employees
engage in efforts that limit their involvement and participation in organizational
functions, activities, or tasks; they may experience a disconnect (or lack fit) with
their work organization. Therefore, these employees may be more likely to engage
in efforts toward quitting and less likely to engage in voice efforts over time. As
such, Withey and Cooper (1989) found that exit and neglect were related because
both are forms of withdrawal—the former from work and the latter from the job. In
this, neglect is considered a antecedent to exit or a replacement when other
alternatives are not available (Withey & Cooper, 1989). Further, in the case of voice,
employee lack of fit over time may create a lower inclination to engage in efforts to
improve dissatisfying conditions or promote organizational change.
HD. The more employees’ engage in neglect at baseline,
3. the more they will engage in exit responses at baseline and at the six‐
month follow‐up.
4. the less they will engage in voice responses at the six‐month follow‐up.

86
Voice   Neglect & Exit
None of the existent literature on voice or neglect, examined the relationship
among the two constructs. However, several researchers have examined the
relationship between organizational citizenship behavior (OCB) and
counterproductive work behaviors. OCB is broader construct reflective of
employees’ extra‐role behavior (efforts that extend beyond one’s job). Voice is
considered one form of OCB. In a meta‐analysis of the relationship between OCB
and counterproductive workplace behaviors, Dalal (2005) found a modest but
negative relationship between OCB and counterproductive work behaviors. In a
separate study of Nigerian employees, Ladebo (2005) found that as employees did
not engage in organizational activities, they exhibited a greater tendency toward
work withdrawal behaviors such as being late or absent. Based on these findings,
the following is hypothesized:
HD.6 The more employees’ engage in voice at baseline, the less they will engage in
neglect responses at baseline and at the six‐month follow‐up.
In a study of hospital administrators and non‐supervisory nurses, Spencer
(1986) examined the relationship between employee voice and turnover. The
researchers found that when employees feel that they have opportunities to voice
their dissatisfaction, with the goal of promoting change, they are more likely to stay
in the organization. Similarly, Paré and Tremblay (2004) found that organizational

87
citizenship behaviors were negative correlates of intention to withdraw from one’s
job. As the researchers explained: “Frequent helping behaviors motivate individuals
to reciprocate such inducements by staying with the organization” (p. 22). Thus:
HD.5. The more employees’ engage in voice at baseline, the less they will engage in
exit responses at baseline and at the six‐month follow‐up.
3.5. INDIRECT EFFECTS
Figure 4 illustrates the extensive combinations of hypothesized indirect
effects in the conceptual model. As discussed in the preceding sections, employees’
diversity characteristics, perceptions of work climate, job stress and psychological
wellbeing and exit/neglect/voice responses all may influence one another.
Accordingly, each of the variables may have both direct and indirect relationships
among each other. This section details the general hypothesized relationship,
specifically as they pertain to the indirect effects.
Work Climate Variables as Mediators
HE.1. Inclusion in decision‐making will mediate the relationships among
employees’ diversity characteristics and their perceptions of supervisory or
organizational support.
HE. Work climate variables (perceived inclusion in decision‐making, supervisory
and organizational support) will mediate the relationships among:

88
2. diversity characteristics and job stress (role conflict, role ambiguity) and
psychological wellbeing.
3. diversity characteristics and exit/neglect/voice responses.
Wellbeing Variables as Mediators
HE. Job stress—role conflict and role ambiguity—will mediate the relationships
among:
4. diversity characteristics and exit/neglect/voice responses.
5. work climate variables and exit/neglect/voice responses.
HE. Psychological wellbeing will mediate the relationships among:
6. diversity characteristics and exit/neglect/voice responses.
7. work climate variables and exit/neglect/voice responses.
Exit/Neglect/Voice as Mediators
HE. Exit will mediate the relationship among the model variables at baseline (i.e.,
diversity characteristics, work climate variables, job stress and psychological
wellbeing, neglect, and voice) and the following:
8. neglect responses at the six‐month follow‐up.
9. voice responses at the six‐month follow‐up.
H E. Neglect at baseline will mediate the relationship among the model variables
at baseline (i.e., diversity characteristics, work climate variables, job stress
and psychological wellbeing, and voice and the following:

89
10. exit responses at baseline and at the six‐month follow‐up.
11. voice responses at the six‐month follow‐up.
H E. Voice at baseline will mediate the relationship among the model variables at
baseline (i.e., diversity characteristics, work climate variables, job stress and
psychological wellbeing, and the following:
12. neglect responses at baseline and at the six‐month follow‐up.
13. exit responses at baseline and at the six‐month follow‐up.
3.6. CONCEPTUAL MODEL
As reflected by the preceding hypotheses, this study proposed a theory‐
driven conceptual model focused on both organizational and individual factors that
may be contributing to employees’ exit/neglect/voice responses. A unique aspect of
this study is the focus on the relationship among exit, neglect, and voice responses
over two‐time points (baseline and six‐month follow‐up). Specifically, Figure 4
depicts the hypothesized relationships among:
• personal and professional diversity characteristics;
• perceptions of work climate (inclusion in decision‐making,
supervisory support, and organizational support;
• job stress and psychological wellbeing; and
• exit/neglect/voice responses over two points in time.

90
FIGURE 4. HYPOTHESIZED CONCEPTUAL MODEL

All model hypotheses are restated in Tables 3 to 7. Based on a
multidisciplinary theoretical framework, the hypothesized model posits that as
employees perceive their work climate as inclusive and supportive, levels of job
stress may be reduced, personal wellbeing may be increased which, in turn, can
ease exit and neglect responses as well as boost voice over two time points.

91
TABLE 3. THE ROLE OF DIVERSITY CHARACTERISTICS (HYPOTHESES SET A)
Hypothesis Set A
As compared to dominant group members, non‐dominant groups perceive that they:
A.1 will experience lower levels of inclusion in decision‐making.
A.2 perceive that they experience lower levels of supervisory and organizational support.
A.3 will experience greater role conflict and role ambiguity.
A.4 will experience a lower sense of psychological wellbeing.
A.5 will differ in their exit responses.
A.6 do not differ in their neglect responses.
A.7 will engage less in voice responses.

TABLE 4. THE ROLE PERCEPTIONS OF WORK CLIMATE (HYPOTHESES SET B)
Hypothesis Set B
The more employees perceive that they are included in decision‐making,
B.1 the more they will perceive their supervisor and organization as supportive.
B.2 the less they will experience job stress associated with role conflict or role ambiguity.
B.3 the greater their sense of psychological wellbeing.
B.4 the less they will engage in exit responses.
B.5 the less they will engage in neglect responses.
B.6 the more they will engage in voice responses.
The more employees perceive their supervisor as supportive,
B.7 the less they will experience job stress associated with role conflict or role ambiguity.
B.8 the greater their sense of psychological wellbeing.
B.9 the less they will engage in exit responses.
B.10 the less they will engage in neglect responses.
B.11 more they will engage in voice responses.
The more employees perceive their work organization as supportive,
B.12 the less they will experience job stress associated with role conflict or role ambiguity.
B.13 the greater their sense of psychological wellbeing.
B.14 the less they will engage in exit responses.
B.15 the less they will engage in neglect responses.
B.16 more they will engage in voice responses.

92
TABLE 5. THE ROLE JOB STRESS & PSYCHOLOGICAL WELLBEING (HYPOTHESES SET C)
Hypothesis Set C
The more employees experience job stress associated with role conflict or role ambiguity,
C.1 the lower their level of psychological wellbeing.
C.2 the less they will engage in exit responses.
C.3 the less they will engage in neglect responses.
C.4 the more they will engage in voice responses.
The more employees experience a heightened sense of psychological wellbeing,
C.5 the less they will engage in exit responses.
C.6 the less they will engage in neglect responses.
C.7 more they will engage in voice responses.

TABLE 6. RELATIONSHIP AMONG EXIT, NEGLECT, AND VOICE (HYPOTHESES SET D)
Hypothesis Set D
The more employees’ engage in exit at baseline,
D.1 the less they will engage in neglect responses at the six‐month follow‐up.
D.2 the more they will engage in voice responses at the six‐month follow‐up.
The more employees’ engage in neglect at baseline,
D.3 the more they will engage in exit responses at baseline and at the six‐month follow‐
up.
D.4 the less they will engage in voice responses at the six‐month follow‐up.
The more employees’ engage in voice at baseline,
D.5 the less they will engage in neglect responses at baseline and at the six‐month follow‐
up.
D.6 the less they will engage in exit responses at baseline and at the six‐month follow‐
up.

93
TABLE 7. DIVERSITY CHARACTERISTICS: INDIRECT EFFECTS (HYPOTHESIS SET E)
Hypothesis Set 5
HE.1 Inclusion in decision‐making will mediate the relationships among employees’
diversity characteristics and their perceptions of supervisory or organizational
support.
Work climate variables (perceived inclusion in decision‐making, supervisory and
organizational support) will mediate the relationships among:
HE.2 diversity characteristics and job stress (role conflict, role ambiguity) and psychological
wellbeing.
HE.3 diversity characteristics and exit/neglect/voice responses.
Job stress—role conflict and role ambiguity—will mediate the relationships among:
HE.4 diversity characteristics and exit/neglect/voice responses.
HE.5 work climate variables and exit/neglect/voice responses.
Psychological wellbeing will mediate the relationships among:
HE.6 diversity characteristics and exit/neglect/voice responses.
HE.7 work climate variables and exit/neglect/voice responses.
Exit will mediate the relationship among the model variables at baseline (i.e., diversity
characteristics, work climate variables, job stress and psychological wellbeing, neglect, and
voice) and the following:
HE.8 neglect responses at the six‐month follow‐up.
HE.9 voice responses at the six‐month follow‐up.
Neglect at baseline will mediate the relationship among the model variables at baseline (i.e.,
diversity characteristics, work climate variables, job stress and psychological wellbeing, and
voice and the following:
HE.10 exit responses at baseline and at the six‐month follow‐up.
HE.11 voice responses at the six‐month follow‐up.
Voice at baseline will mediate the relationship among the model variables at baseline (i.e.,
diversity characteristics, work climate variables, job stress and psychological wellbeing, and
the following:
HE.11 neglect responses at baseline and at the six‐month follow‐up.
HE.11 exit responses at baseline and at the six‐month follow‐up.

94
Chapter 4: Methodology
4.1. OVERVIEW OF STUDY
This study employed both cross‐sectional and longitudinal panel designs to
investigate the relationships among human service workers’ diversity
characteristics, perceptions of work climate, job stress and psychological wellbeing,
and their exit/neglect/voice responses to undesirable work experiences. For the study,
surveys were distributed to human service workers at baseline and at the six‐month
follow‐up.
This research was conducted as a part of a large‐scale, longitudinal (three
time periods), mixed‐methods (surveys and interviews) study entitled, “Supervision
and Workforce Retention in Child Welfare.” The large‐scale study was conducted
under the auspices of a university‐based training center for employees working in
children and families services. Additionally, the large‐scale study built upon
previous research that investigated employee intention to leave and actual turnover
(Mor Barak et al., 2006; Mor Barak, Nissly et al., 2001; Nissly et al., 2005).
4.2. STUDY SITE
A large child welfare agency in the western region of the United States
served as the study site for the current study. Comparable to organizations of
similar size, the organization has over 3000 employees in direct children’s services

95
functions. The organization employs over 5000 individuals and maintains
approximately 12,000 children in custody.
4.3. STUDY PARTICIPANTS, ENROLLMENT, & DATA COLLECTION
This study used an availability sample of child welfare workers that were
recruited for participation while attending one of 30 targeted required or voluntary
training sessions offered at a university‐based training center. In total, 364
individuals agreed to participate in the longitudinal study. Three individuals
agreed to participate but did not complete the baseline questionnaire. Two
questionnaires were excluded because the participants did not turn in their signed
informed consent forms. The elimination of these five questionnaires resulted in
three hundred and fifty‐nine (n = 359) participants at baseline.
Potential study participants were provided general information about the
study during the morning of the required or voluntary training sessions. At that
time, research team members presented general and sufficient information to enable
participants to make informed decisions about the potential of enrolling in the
study. During the lunch break, participants enrolled in the study after researchers
provided further detail regarding the purpose of the study. All participants were
reassured that their participation would be voluntary and their responses would be
kept confidential. Study participants enrollment comprised the completion of the
questionnaire, and, most importantly, the signed IRB (Institutional Review Board)

96
approved informed consent form. As a token of appreciation, participants were
served pizza and soda for lunch while completing questionnaires.
Participants’ contact information was captured for tracking and informed
consent purposes; however, actual questionnaire responses were kept anonymous
through the creation of unique identifiers. The unique identifier consisted of a pre‐
established administrative code set by the researchers that linked respondents’
contact information and their actual questionnaires. This code was not, however,
listed on the informed consent forms. The informed consent form only holds
participants’ name and signatures. All information is kept secured and confidential.
Further, questionnaires, contact information, and informed consent forms are kept
in separate and secured locations. Finally, only aggregate data will be presented in
this dissertation and subsequent reports of findings.
To obtain six‐month follow‐up data, participants were asked to provide their
name and contact information on a cover sheet at the time of baseline data
collection. This cover sheet had a pre‐established administrative code that linked
survey responses to participants’ contact information. To ensure confidentiality, any
identifying information was stored and locked separately from the questionnaire
responses. In addition, the name and contact information of the participants was not
indicated on the actual follow‐up questionnaire, only the code.

97
Six‐month follow‐up data were collected from July 2004 until September 2005.
All six‐month follow‐up questionnaires were distributed by mail. Study
participants’ administrative codes were inserted on the top of their questionnaires
so that the latter questionnaires could be matched to baseline surveys. In all, 341 of
the initial 364 participants were mailed six‐month follow‐up surveys. Twenty‐three
surveys were not mailed because:
a. 21 employees completed the baseline survey but did not provide follow‐up
contact information and
b. two participants did not complete the informed consent; thus, their baseline
surveys were not included in the study.
Of those six‐month follow‐up surveys mailed, six were returned from the
post office. In these instances, attempts were made to contact participants via
telephone or email to obtain the correct address. However, the participants were not
able to be located using this mechanism.
Six individuals quit the organization during the study time period. Among
those who quit, three individuals also returned the follow‐up survey based on their
last date of employment and these surveys were included in the analyses.
To facilitate as high a response rate as possible, 158 first attempt contacts
(97% phone calls; 3% emails or mailing) were made to those who had not returned
their questionnaires within two weeks after the questionnaires had been mailed.

98
Furthermore, second attempt contacts included a mailing of 65 repeat
questionnaires with a reminder letters to applicable employees at the one‐month
mark as recommended by Rubin and Babbie (2001). We received an additional 79
completed questionnaires after the first initial follow‐up contact (regardless of
method) was made with the participant.
Thus, based on the initial mailing as well as first and second follow‐up
contacts, 187 six‐month follow‐up surveys were received resulting in a 55.8%
response rate of 341 mailed surveys mailed and 52.1% of the 359 eligible study
participants. Both these response rates slightly exceed Rubin’s and Babbie’s (2001)
suggestion that response rates greater than 50.0% are adequate for analysis and
reporting.
4.4. DESCRIPTION OF STUDY SAMPLE
Personal and Professional Diversity Characteristics
Descriptions of employee diversity characteristics help provide an
understanding of the characteristics of the study sample. For this study, employees
were surveyed on their personal (gender, ethnicity, age) and professional
(educational level, job position, and job tenure) characteristics. As Table 8 shows the
majority of sample at baseline were women (84%) and under 40 years of age (62%).
The average age was 37 (SD = 11.48). Ethnically, the sample population was rather
diverse: Caucasian (31%), Latino/a (30%), African American (22%), Asian (12%).

99
TABLE 8. PERSONAL DIVERSITY CHARACTERISTICS OF STUDY PARTICIPANTS
Baseline
(N = 359)
Six‐month follow‐up
(N = 187)
Frequency
(n)
Valid
%
Frequency
(n)
Valid
%
Gender
Women 300 83.6 155 82.9
Men 59 16.4 32 17.1

Age mean = 36.8; median = 33; sd = 11.5 mean = 38.5; median = 34; sd = 12.1
<30 117 32.6 48 25.7
30‐39 106 29.5 61 32.6
40‐49 56 15.6 31 16.6
50‐59 44 12.3 25 13.4
60+ 36 10.0 22 11.8

Ethnicity

Caucasian 111 31.1 71 38.5
Latino/a 107 30.0 47 25.1
African
American/Black
77 21.6
36 19.8
Asian 44 12.3 25 13.4
Other 18 5.0 6 3.2

With regard to the study sample’s professional characteristics (see Table 9),
most have a graduate degree (76%) with a more than half of those having Master in
Social Work (MSW) degrees. The overwhelming majority of the participants were
direct service workers (82%). The remaining workers were either in supervisory
(14%) or managerial (3%) positions. Participants’ length of tenure in their work
organization (at the time of baseline data collection) ranged from less than 1 month
to 37 years with 30% employed for less than 1 year; 27% between 1 and 5 years, 19%
between 5 and 10 years, and the remaining were employed over 10 years.

100

TABLE 9. PROFESSIONAL DIVERSITY CHARACTERISTICS OF STUDY PARTICIPANTS
Baseline
1

(N = 359)
6‐month follow up
(N = 187)

Frequency
(n)
Valid Percent
(%)
Frequency
(n)
Valid Percent
(%)
Highest Degree
Obtained

BA 96 26.9 34 18.2
BSW 26 7.3 8 4.3
MA/MS/MFT 73 20.4 38 20.3
MSW 155 43.4 104 55.6
PHD or Other 7 1.9 3 1.6

Job Position

Direct service
worker
282 81.7
146 79.3
Supervisor or
Manager
63 18.3
38 20.7

Job tenure


Less than 3
months
50 14.0 ‐‐ ‐‐
3 months to 1yr 57 16.0 46 25.3
1+ to 5 yrs 96 27.0 47 24.7
5+ to 10 67 18.8 42 22.6
10+ to 15 46 12.9 25 13.4
15+ yrs 40 11.2 26 14.0

mean = 6.9; median = 3.7; sd = 6.7 mean = 7.0; median = 5.0; sd = 6.6

1

Analyses of Six‐month Follow‐up Non‐response (Study Attrition)
Of the 359 study participants, 186 participants completed both baseline and
six‐month follow‐up surveys (six‐month follow‐up respondents), 173 completed
baseline only (six‐month follow‐up non respondents), and one participant

101
completed six‐month follow‐up only. Although over 50% of the participants
completed the six‐month follow‐up survey, a consequence of non‐response effects
are bias and sampling variance of the estimates (Fowler, 2002). Therefore,
differences between six‐month follow‐up respondents and non‐respondents were
analyzed on key diversity characteristics and for all scale variables.
First, chi square (χ
2
) tests were conducted to examine differences between
baseline and six‐month follow‐up data on gender, ethnicity, educational level, and
job position as well as a t‐test for job tenure. Results revealed no statistically
significant differences between six‐month follow‐up respondents and non‐
respondents, on the following variables: gender (χ
2
= .17; p = .68), ethnicity (χ
2
= 9.02;
p = .11), job position (χ
2
= 2.28; p = .35), and job tenure (t = ‐1.18; df = 354; p = .24).
The only significant difference between six‐month follow‐up respondents
and non respondents was based on level of education (χ
2
= 25.85; p <= .000):
Graduate‐degreed workers were 65.7% of baseline respondents, while 77.5% of six‐
month follow‐up respondents. Also, there was a difference in age (t = ‐1.97; df = 336;
p < .05), with those participants who completed both baseline and six‐month follow‐
up questionnaires being older (37.96; sd = 12.04) than those who completed baseline
questionnaires only (mean = 35.49; sd = 10.73).
Similarly, t‐tests were conducted to assess differences between six‐month
follow‐up respondents and non‐respondents on baseline study variables. This helps

102
in understanding patterns of attrition from baseline to six‐month follow‐up. With
respect to the scale variables, perceived organizational support was the only
variable in which a statistically significant association was found between six‐
month follow‐up respondents and non respondents (t = 2.70; df = 351; p = .01). Those
who completed baseline only questionnaires (six‐month follow‐up non
respondents) reported higher scores of perceived organizational support (mean =
3.31; sd = 1.15) than those who completed both baseline and six‐month follow‐up
surveys (mean = 2.99; sd = 1.08). No statistically significant differences existed on all
other scale variables (inclusion in decision‐making, perceived supervisory support,
role conflict, role ambiguity, psychological wellbeing, exit, neglect, and voice).
Comparisons Between the Sample and the Agency Population
Single‐sample t‐test and binominal tests were conducted to assess the extent
to which the study sample was representative of the population. These comparisons
were based on the work organization’s data employee data from 2000‐2001 of a
subset of workers (n = 3724 of the total agency population of 6726) that held the
same job titles as the study sample. Comparisons were made based on employees’
diversity characteristics: age, gender, race/ethnicity, educational level, agency
position, and job tenure.
The sample was generally similar to the agency’s population, but it over‐
represented or under‐represented certain characteristics compared to the

103
organization’s employees overall. Specifically, the participants were younger, there
were more women represented, less African American and more Asian/Pacific
Islanders, higher proportion of line‐workers, more highly educated and with less
tenure compared to the agency’s employees.
• Age: the study sample (mean = 37 years; sd = 11.5) was on an average five
years younger than the population mean of 42 (t = ‐8.3, df = 338, p < .001).
• Gender: the study sample had about 10% more women (84%) than the
agency population (74%) (binomial test yielded p < .001).
• Race/Ethnicity: the study sample had less African Americans and more
Asian/Pacific Islanders and Other ethnic groups than the population
(binomials for African Americans, Asian/Pacific Islanders, and Others =
.000). The sample was not statistically different with regard to Caucasians
and Latino/a’s.
• Level of Education: the study sample was more highly educated (65%
graduate degreed) than the population (51%; binomial test yielded p < .001).
• Agency Position: the study sample included a higher proportion of line
workers (82%) as compared to the population (78%) (binomial yielded p =
.046).
• Job tenure: our sample (mean = 6 years) had shorter job tenure than the
population mean of 9 (t = ‐7.85, df = 355, p < .001).

104
4.5. MEASURES
Data were obtained via a 9‐page questionnaire that assessed employees’
diversity characteristics, perceptions of their work climate, job stress, and
psychological wellbeing as well as exit/neglect/voice responses. All study variables had
well‐established psychometric properties. Additionally, Cronbach’s alpha, which
indicates a scale’s internal consistency reliability or the extent to which one scale’s
items measures the same construct, was computed for each measure. According to
Devillis (1991) coefficient greater or equal to .70 provides a respectable degree of
reliability and coefficients between .65 and .70 demonstrate minimally acceptable
reliability.
Composite scores were computed for each scale by first reverse coding items
as needed. With reverse coding, negatively or positively worded items were
rescored in the opposite direction to match the other items in the scale. (See
Appendix A for the specific scale items.) Then, a mean composite score was
calculated for cases with 50% or more of the scale items completed.
Diversity Characteristics
Respondents were asked to provide information on their personal and
professional diversity characteristics. Personal diversity characteristics included
gender, race/ethnicity, and age. Gender was precoded as 1 ‘women’ and 0 ‘men’.
The race/ethnicity categories drew from current census methods providing

105
opportunities for employees to check more than one category; the categories
included: Caucasian, African‐American, Latino/a decent, Asian decent (Asian
Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, or other Asian), and Other
(including American Indian or Alaska Native, Native Hawaiian, Guamanian or
Chamorro, Samoan and other Pacific Islander and other non specified ethnic
groups). The respondents’ age was captured as a continuous variable and
transformed into six categories (<30, 30‐39, 40‐49, 50‐59, 60+) for the presentation of
descriptive statistics in Table 8.
Professional diversity characteristics included employees’ (1) level of
education (i.e., highest degree obtained: BA, BSW, MA/MS/MFT, MSW, PHD,
Other), (2) current job tenure (in years), and (3) current agency position (manager,
supervisor, or line worker).
Perceptions of Work Climate
Perceptions of work climate reflect employees’ felt experiences of inclusion
in decision‐making as well as supervisor and organizational support. Inclusion in
decision‐making is conceptualized as employees’ perceptions of their experiences
related to participating in and influencing decision‐making within the organization
and their work groups. The construct was measured using the 5‐item subscale of
Mor Barak’s and Cherin’s Perception of Inclusion Scale (Mor Barak, 2005; Mor Barak
& Cherin, 1998). The subscale captured employees’ perceptions of their contribution

106
to their job and organizational functioning. Sample items include “I have a say in the
way my work group performs its tasks”, “I am often invited to voice my opinion in meetings
with management higher than my immediate supervisor”, and “I am able to influence
decisions that affect my organization.” Responses were rated on 6‐point scale
(1=strongly disagree to 6 = strongly agree) and higher scores indicated greater
perceived inclusion. The Perception of Inclusion scale as well as the subscales have
demonstrated strong internal consistency and convergent validity in its original
application (Mor Barak & Cherin, 1998) and in a series of cross‐national studies
(Mor Barak et al., 2003). Reliability coefficients have ranged from .72 to .75 for
influence in decision‐making subscale and .81 to .90 for the entire inclusion scale
(Mor Barak & Cherin, 1998; Mor Barak, Findler, & Wind, 2001; Mor Barak et al.,
2003; Mor Barak et al., 2006). For the current study, the alpha coefficient was .71 at
baseline and .74 at the six‐month follow‐up.
Perceived supervisory and organizational support involves employees’
global views about the extent to which supervisors or the work organization,
respectively, values their contributions and cares about their wellbeing. Both
perceived supervisory and organizational support were measured using an 8‐item
scale, originally developed by Eisenberger, et al. (1986). The original Survey of
Perceived Organizational (SPOS) scale was 24‐items and reduced to eight items per
a recommendation by Rhoades and Eisenberger (2002). As Rhoades and Eisenberger

107
stated, “Because the original scale is unidimensional and has high internal
reliability, the use of shorter versions does not appear problematic. Prudence
nevertheless dictates that both facets of the definition of POS (valuation of
employees’ contribution and care about employees’ well‐being) be represented in
short versions of the questionnaire.” (p. 699) The items for the SPSS and SPOS scales
are exactly the same except the word “supervisor” replaced “organization”, as
previously carried out by Eisenberger, Stinghamber, Vandenberghe (2002) and
Rhoades et al. (2001). Sample items include: “My supervisor/the organization values my
contribution to [the organization’s] well‐being”, “My supervisor/the organization really
cares about my well‐being”, and “My supervisor/the organization takes pride in my
accomplishments at work.” Responses were rated on 7‐point scale (1 = strongly
disagree to 7 = strongly agree). Four items within each scale were reverse scored so
that higher scores indicated greater perceived support. Both the POS and PSS have
been used extensively in a variety of settings and have demonstrated strong validity
as well as reliability with scores ranging from .81 to .94 for PSS and .74 to .88 for
POS (Aselage & Eisenberger, 2003; Eisenberger et al., 1986; Griffin, Patterson, &
West, 2001; Rhoades & Eisenberger, 2002). For the current study, the alpha
coefficients were acceptable for both measures (baseline: PSS = .88; POS = .83; six‐
month follow‐up: PSS = .90; POS = .86).

108
Inclusion in decision‐making, perceived supervisory support, and perceived
organization support are conceptually distinct constructs that reflect aspects of
employees’ view of the work climate within their organization. An exploratory
factor analysis (principal component extraction with varimax rotation) was
conducted on baseline data to statistically support the discrimination among the
work climate constructs using SPSS. In this analysis, all the self‐reported work
climate items were evaluated for their loadings on respective predetermined factors.
Results indicated the emergence of five factors (explaining 64.41% of the total
variance) based on the criterion of eigenvalues greater than 1 and an evaluation of
scree plots. Although five factors were found, all items loaded strongly on their
respective factors (minimum loading was .43 with most loadings >.62). However,
the items on both the supervisory support and organizational support generated
two factors per construct. This appears to be a measurement artifact because the
items from each scale were split respectively based on positively versus negatively
worded items. All inclusion in decision‐making items loaded respectively onto one
factor.
Job Stress and Psychological Wellbeing
Job stress and psychological wellbeing focus on employees’ perceptions of
their wellbeing based on organizational circumstances (i.e., role conflict and role
ambiguity) and their current mental health status. For the current study, job stress

109
was conceptualized based on employees’ role conflict and role ambiguity. Role
conflict is the extent to which employees’ expectations of one or more roles does not
mirror the relations of others’ perceptions of that role, where role ambiguity reflects
the extent to which an employee is unclear about work‐related expectations (House
& Rizzo, ; Rizzo et al., 1970). Originally developed by Rizzo et al. (1970), the widely
used 8‐item role conflict and 6‐item role ambiguity scales measure the extent to
which individuals feel stress related to the expectations and clarity of their work.
Role conflict items include: “I have to work on things that should be done differently”; “I
seldom receive an assignment without the manpower to complete it”; and “I do things that
are apt to be accepted by one person but not accepted by others.” Examples of role
ambiguity items include: “I feel uncertain about how much authority I have”, “Iʹm not
sure what my responsibilities are”, and “Explanation is clear of what has to be done (on the
job).” Responses were rated on a 6‐point scale, 1 = strongly disagree to 6 = strongly
agree. Higher scores indicated greater stress for both scales. As such, all of the items
on the role ambiguity scale were reverse scored for consistency with the direction of
the role conflict scale. These scales are widely used and have well‐established
psychometric properties (Acker, 2004; Netemeyer, Johnston, & Burton, 1990; M. F.
Peterson, Smith, Akande, Ayestaran, & et al., 1995). Alpha coefficients for previous
studies have generally ranged from .56 to .85 for role conflict .58 to .87 for role
ambiguity. However, lower than desirable reliabilities was based on a cross‐cultural

110
comparison (21 countries) of the construct. For the most part, the reliability
coefficients exceed the .70 criteria (M. F. Peterson et al., 1995).
For the current study sample, the Cronbach’s alpha coefficients for the role
conflict scale was .77 for both time periods. The alpha coefficient for role ambiguity
was .83 at baseline and .86 at the six‐month follow‐up.
Individuals’ general mental health status or psychological standing reflects
their psychological wellbeing (Goldberg et al., 1996). This specifically relates to
employees’ levels of happiness, anxiety, stress, or depression. Goldberg’s (1972)
General Health Questionnaire, a well‐validated global measure of current mental
health status, was used to assess psychological wellbeing. Originally the instrument
was 60 items but shortened versions (i.e., GHQ‐30, GHQ‐28, GHQ‐20, GHQ‐12) are
available. For this study the GHQ‐12 was used. The GHQ‐12 is brief and easy to
complete and has been tested for reliability and validity in a variety of different
settings and cultures (Coffey, Dugdill, & Tattersall, 2004; Daradkeh, Ghubash, & El‐
Rufaie, 2001; Jacob, Bhugra, & Mann, 1997; Mor Barak et al., 2003; Schmitz, Kruse, &
Tress, 1999; Tait, French, & Hulse, 2003). Prior reliability estimates have ranged
from .65 to .91 depending on the setting and cultural context. The GHQ gauges the
extent to which respondents have recently experienced a particular symptom or
behavior. For example: “Have you recently felt you couldn’t overcome your difficulties?”
and “Have you recently been feeling unhappy or depressed?” Items were rated on a 4‐

111
point scale from 1 = much worse than usual, 2 = worse than usual, 3 = same as usual,
and 4 = better than usual. A lower score indicated psychological distress. The
Cronbach’s alpha for the current study was at .89 at baseline and .90 at the six‐
month follow‐up indicating strong internal consistency.
An exploratory factor analysis (principal component extraction with varimax
rotation) was conducted on baseline data to statistically support the discrimination
among the job stress and psychological wellbeing. In this analysis, all the self‐
reported role conflict, role ambiguity, and psychological wellbeing items were
evaluated for their loadings on respective predetermined factors. Results indicated
the emergence of five factors (explaining 64.41% of the total variance) based on the
criterion of eigenvalues greater than 1 and an evaluation of the scree plot. Although
five factors were found, all items loaded strongly on their respective factors
(minimum loading was .45 with most loadings >.60). Specifically, the 8 items in the
role conflict scale and 6 items in the role ambiguity scale loaded on their respective
factor. However, the items on the GHQ scale generated three factors. In an effort to
evaluate the individual loadings of the GHQ, a secondary factor analysis was
conducted forcing 3 factors onto all the items. In this all the minimum GHQ item
loading was .49, indicating that all the GHQ items reflect one construct for the
purposes of this study.

112
Exit/Neglect/Voice
Exit, neglect, and voice responses respectively are conceptualized as
employees’ efforts toward quitting their job; efforts that reflect disengagement from
work‐related tasks or organizational activities; and efforts to improve dissatisfying
work conditions or promote organizational change (Farrell, 1983; Hagedoorn et al.,
1999; Rusbult et al., 1988). The exit/neglect/voice scales were adopted from Rusbult,
Farrell, Rogers, Mainous III (1998) and measured generalized tendencies toward
each response. All items were rated on a 6‐point Likert‐type scale, from 1 (strongly
disagree) to 6 (strongly agree). The exit scale was 4‐items with higher scores
indicating a greater propensity to engage in efforts related to quitting one’s job.
Sample items comprised: “I have recently spent time looking for another job”, and “I
often think about quitting.”
In its original formation the neglect scale was 6‐items; however, one item, “I
care very little what happens to this organization as long as I get a paycheck”, was
removed due to two reasons. First, the item differed conceptually from the other
five items (examples listed below), reflecting aspects of loyalty, which was not
measured in the current study (see Chapter 1, section 2.1 for an explanation of the
exclusion of employee loyalty). Also, after reliability analysis was conducted, results
revealed the item’s low contribution to the Cronbach’s alpha. Sample items from the
resultant neglect scale included: “Sometimes when I don’t feel like working I will work

113
slowly or make errors” and “I try to keep out of sight of my supervisor so I can talk to co‐
workers, take breaks, or do other personal business”. Higher scores are reflective of a
greater tendency to disengage from work‐related tasks and organizational activities.
Last, the voice scale was four items with higher scores indicating greater
efforts toward improving dissatisfying workplace conditions or promoting
organizational change. Sample items included: “When I think of an idea that will
benefit my organization I make a determined effort to implement it”, and “I sometimes
discuss problems at work with my employer.”
In general, Rusbult’s original formation of the exit/neglect/voice scales has
been widely adapted and has demonstrated varied psychometric properties.
Rusbult (1988) tested the psychometric properties of the exit, voice, and neglect scales
in different variations across three studies. In all variations, the reliability
coefficients for exit (.76 to .97) and neglect (.69 to .82) were acceptable, but varied for
voice (.45 to .77). Rusbult’s measures have been adapted in other settings and shown
fairly high internal consistency. As an example, Hagedoorn et al. (1999) found the
Cronbach’s alphas for their versions of exit/voice/neglect scales were high (>= .80).
Turnley and Feldman (1999) adaptations of Rusbult’s exit/voice/neglect measures also
had acceptable reliability coefficients (> = .77). In contrast, other researchers have
found low internal consistency for voice and neglect measures. Withey’s and
Cooper’s (1989) found reliability estimates that were acceptable for exit (> .70);

114
however, the estimates for voice and neglect were lower than desired (< .50). Rusbult
et al. (1988) attributed the lower than desired reliability estimates for voice to the
heterogeneity of the construct or the differing forms of voice responses (see Chapter
2, Section 2.3 for a review). The researchers were not surprised by the low reliability
estimates, but rather suggested that specific voice responses do not necessarily occur
in conjunction with other voice efforts. For example, an individual may talk to a
supervisor about a problem but may not engage in any other change‐oriented
behavior. The same reasoning can be applied to neglect; an employee may regularly
miss meetings, yet still put forth extra‐effort in their work‐related tasks. Rusbult et
al. (1988) stated that these lower than desirable reliability coefficients do not trump
the usefulness of such measures because the goal is to identify similarities within a
specific construct that differ from another.
For the current study, the Cronbach’s alpha for the exit scale (baseline = .87;
six‐month follow‐up = .89), neglect (baseline = .70; six‐month follow‐up = .74 ) were
acceptable. The alpha coefficients for the voice scale (baseline = .69; six‐month
follow‐up = .68) was lower but also acceptable, considering the complexity of the
contrast (Withey & Cooper, 1989).
An exploratory factor analysis (principal component extraction with varimax
rotation) was conducted on baseline data to statistically support the discrimination
among the exit/neglect/voice constructs. Results indicated the emergence of four

115
factors explaining 53.79% of the total variance based on the criterion of eigenvalues
greater than 1 and an evaluation of the scree plot. Although four factors were found,
all items loaded strongly on their respective factor (minimum loading was .45 with
most loadings >.60). Specifically, the 4‐items in the exit and 4‐ items in the voice scale
loaded on their respective factor. The items on the neglect scale generated two
factors; however, considering the complexity of the response (Rusbult et al., 1988),
this was not alarming. Most importantly, there were no cross‐loadings among the
measures.
4.6. SCREENING OF LONGITUDINAL, QUANTITATIVE DATA
Evaluation of Normality
All interval/ratio level variables were analyzed for normality. Job tenure,
inclusion in decision‐making, perceived supervisory support, perceived
organizational support, role conflict, role ambiguity, psychological wellbeing were
examined using baseline data. Exit/neglect/voice responses were evaluated using
baseline and six‐month follow‐up data.
Most statistical tests carry the underlying assumption that all variables as
well as combinations of variables are normally distributed. Graphically, the
distribution of each variable was assessed based on looking at the imposed normal
curve. For the statistical test of normality, both skewness and kurtosis were
examined. A normal distribution occurs when both the skewness and kurtosis are

116
zero. Skewness is related to the symmetry or tilt of the distribution with a normal
distribution having a skewness of 0. Kurtosis reflects the peakedness of the
distribution, either too tall or too flat. To test for the skewness and kurtosis, the
statistics were divided by their standard errors. Researchers suggested that
skewness and kurtosis values range between +2 and –2 (Wuensch, 2005).
Of the ten variables, three (inclusion in decision‐making, perceived
organizational support and role conflict) fell within acceptable ranges. Perceived
supervisory support and psychological wellbeing were negatively skewed; whereas
role ambiguity, exit, neglect, and voice were positive skewed. To alleviate skewness
of the distribution, square root transformations were conducted as recommended
by Tabachnick and Fidell (2000). (The formula differed for positively and negatively
skewed variables, refer to Appendix E).
All variables except voice and exit, showed improvement in skewness and
kurtosis after the square root transformation. Voice at baseline and exit at the six‐
month follow‐up, however, was moderately positively skewed prior to the
transformation, then became moderately negatively skewed after the
transformation. Thus, as recommended by Tabachnick and Fidell (2000), the
variables remained in their original composition. Further, neglect remained in its
original formation for consistency and interpretability in the longitudinal analysis to

117
exit and voice. Appendix E offers more detail regarding the transformation of the
skewed variables.
Assessment of Collinearity
The calculated bivariate correlations were assessed to determine whether
collinearity‐related problems existed among all permutations of the study variables.
These included all interval variables at baseline (inclusion in decision‐making,
supervisory support, and organizational support, role conflict, role ambiguity,
psychological wellbeing, and exit(T1), neglect(T1), and voice(T1)), and the outcome scale
variables at the six‐month follow‐up (exit(T2), neglect(T2), and voice(T2)). Collinearity
reflects a near perfect linear relationship between two variables; thus resulting in a
high correlation (Pedhazur, 1997). Collinearity is problematic because the high
correlation create difficulty in differentiating between the variables. Thus,
collinearity can negatively affect statistical analyses, resulting in inaccurate
regression coefficients estimates (Pedhazur, 1997). In general, multicollinearity is
considered problematic when the absolute value of the correlations are greater than
.85 (Kline, 1998). Based on this criterion, no pairs in this study were multicollinear.
Despite this, the high correlation between job tenure and age (r=.71) resulted in only
using job tenure in the path analysis.

118
4.7. ANALYTIC PROCEDURES
Descriptive Statistics & Bivariate Correlations
Descriptive statistics were computed for all demographics and scale
variables. For the demographic variables (gender, ethnicity, age, educational level,
job position, and job tenure), frequency and percentage statistics were computed.
Measures of central tendency and dispersion were also computed for the interval‐
level demographic variable— age and job tenure. Last, bi‐variate correlations were
examined for all the scale variables as identified in Tables 11 and 12.
Assessment of Theory‐driven Model
Path analysis with Amos version 6.0 (Arbunkle, 2005) using the maximum
likelihood estimation method was utilized to test the theory‐driven hypothesized
model. Further, all hypotheses were examined using the path‐analytic approach.
Path analysis is a statistical technique used to test relationships among variables by
defining the predictive relationships based on theoretical and empirical
researchOne assumption related to path‐analytic modeling is that all variables are
continuous or interval level. Similar to other forms of regression modeling, a
common practice is to create dichotomous variables by recoding categorical
variables. Therefore, for the analyses, categorical and ordinal variables were
recoded in to dummy variables as follows: gender (1 = women, 0 = men); African
American (1,0); Latino/a (1,0); Asian (1,0); educational level (1 = graduate degree, 0 =

119
bachelor’s degree); and job position (1 = direct service worker, 0 = manager or
supervisor). All other model variables were at the interval/ratio levels. These
included job tenure, perception of inclusion in decision‐making, perceived
supervisory support, perceived organizational support, role conflict, role ambiguity,
psychological wellbeing, exit/neglect/voice.
To achieve optimal model parsimony, the hypothesized model was tested in
four parts using a stepwise approach (based on the method employed by Mor Barak
et al, (2005). The description and sequence of testing each phase is detailed below.
Phases I, II, and III involved analyses of baseline variables only (n = 359), while
Phase IV was a longitudinal analysis of baseline and six‐month follow‐up responses
(n = 186)
Phase I modeled the relationship among diversity characteristics and work
climate variables. All hypothesized paths (see Figure 4) were included. Non‐
significant paths (p >= .05; C.R. < 1.96) were removed and the refined model
was retested for model fit. In addition to an assessment of the goodness‐of‐
fit statistics, Likelihood ratio tests (chi‐square difference tests) were used to
determine if the removal of non‐statistically significant paths significantly
changed the overall fit at the p <.05 levels After this assessment, the refined
model formed the basis for the next phase and was incorporated into the
subsequent analyses.

120
Phase II tested the refined Phase I model (diversity characteristics   work
climate) with the addition of the job stress and psychological wellbeing as
outcome variables. As described above, the paths were evaluated for model
misfit and non‐significant paths removed; then the refined Phase II model
was assessed again for fit. The refined model was used as a component of
Phase III analyses.
Phase III built on the earlier step and incorporated exit, neglect, voice as
outcome variables. This model was tested and refined in accord with the
technique used in the earlier steps. The final results from Phase III analyses
created the refined, most parsimonious cross‐sectional model.
Phase IV examined the cross‐lagged relationship among exit, neglect, and
voice utilizing the refined model of Phase III as a base. This model was also
tested and refined by removing non‐significant paths in accordance with the
technique used in the earlier phases of the analysis. The final results from
Phase IV analyses created the refined, most parsimonious longitudinal
model.
Several goodness‐of‐fit indices were used to determine the consistency of the
conceptual model’s fit with the data as recommended by Byrne (2001). Indicators of
acceptable model fit include the following: a non statistically significant chi‐square
( χ
2
); NFI and CFI values greater than 0.90 (values range between 0 and 1) with

121
values close to .95 indicating superior fit; a root mean square error approximation
(RMSEA) value of less than .10 (with <.05 demonstrating a good fit, .07 representing
reasonable errors of approximation in the population, and .08‐.10 indicating
mediocre fit, and values greater than >.10 equivalent of a poor fit), Hoelter’s critical
N (CN) was used to assess the adequacy of sample size: values greater than 200
indicated that the model adequately represents the same data. Finally, the Expected
Cross‐validation Index (ECVI) values were used to compare the full initial models
with refined, more parsimonious models. The smaller the ECVI value as compared
among 2 or more models, the better the fit.
Examination of Research Hypotheses
All hypotheses were examined using path analysis testing both cross‐
sectional and longitudinal data. Further, findings from the path analysis of the
longitudinal results generated in Phase IV of the analyses were used to evaluate the
outcome variables—exit/neglect/voice—over two points in time (baseline and six‐
month follow‐up). Assessment of supported hypotheses was based on the level of
significance of each pathway (p values less than .05 were considered significant)
(Diamantopoulos & Siguaw, 2000).
Indirect effects were tested using a causal steps technique (Baron and Kenny
(1986). For an indirect effect to occur these conditions must be met: (1) the
relationship between the independent variable and dependent variable must be

122
statistically significant; (2) the path from the independent variable to the mediator
must be significant; (3) the path from the mediator to the dependent variable must
be significant; and (4) for complete mediation, the significant relationship between
the independent variable and the dependent variable must be dropped when
including the mediator. Sobel’s test of indirect effects was used to test for significant
mediation effects (Preacher & Hayes, 2004).

123
Chapter 5: Research Findings
5.1. INTRODUCTION
This chapter presents the current study’s research findings related to the
organizational and individual predictors of exit/neglect/voice. The chapter begins
with a discussion of the descriptive statistics and correlation analyses. Then, a
discussion of the assessment of model fit for the hypothesized, theory‐driven model
(see Chapter 3, Figure 4) is provided. Finally, the results from testing the study
hypotheses through path‐analytic modeling are provided.
5.2. DESCRIPTIVE STATISTICS AND CORRELATIONS
Descriptive statistics for the perceptions of work climate, job stress and
psychological wellbeing variables are shown in Table 10. Higher scores represent
stronger perceptions of inclusion in decision‐making, supervisory support, or
organizational support; greater stress associated with role conflict or role ambiguity;
and a heightened sense of psychological wellbeing. Table 11 details the descriptive
statistics for the exit/neglect/voice responses at baseline and at the six‐month follow‐
up. Higher scores for exit and neglect respectively represent more efforts toward
quitting one’s job and disengagement from work‐related tasks and organizational
activities, while higher voice scores reflect more efforts in improving work
conditions or promoting organizational change. As shown, exit scores increased

124
over the two time points with participants reporting more efforts toward quitting
their jobs.
TABLE 10. DESCRIPTIVE STATISTICS OF WORK CLIMATE, STRESS, & WELLBEING VARIABLES
Scale
1
Baseline

Mean SD
Perception of inclusion in decision‐making 3.36 1.13
Perceived supervisory support 4.69 1.05
Perceived organizational support 3.14 1.12
Role conflict 3.75 1.07
Role ambiguity 2.35 .84
Psychological wellbeing 2.92 .43
1
All scales had minimum and maximum values of 1 to 6 with the
exception of psychological wellbeing, scaled from 1 to 4.

TABLE 11. DESCRIPTIVE STATISTICS OF EXIT/NEGLECT/VOICE RESPONSES AT TWO TIME POINTS
Scale
1
Baseline

Six‐month follow‐up

Mean SD Mean SD
Exit(T1) 2.70 1.44 3.12 1.62
Neglect(T1) 2.31 0.94 2.30 .95
Voice(T1) 3.33 1.03 3.30 .99
1
All scales had minimum and maximum values of 1 to 6.

Bivariate correlations for the scale variables at baseline are presented in
Table 11. As shown, most of the model variables were statistically significantly
correlated at baseline. However, the following bivariate pairs did not have a
statistically significant correlation: (1) inclusion in decision‐making and role conflict

125
(2) perceived supervisory support and voice; and (3) psychological wellbeing and
voice.
TABLE 11. BI‐VARIATE CORRELATIONS OF SCALE VARIABLES
Scale
1

1 2 3 4 5 6 7 8 9
1: Perception of
inclusion in decision‐
making
1
2: Perceived
supervisory support
.37** 1
3: Perceived
organizational
support
.37** .22** 1
4: Role conflict ‐.04 ‐.22** ‐.39** 1
5: Role ambiguity ‐.23** ‐.21** ‐.14** .17** 1
6: Psychological
wellbeing
.20** .15** .23** ‐.27** ‐.22* 1
7: Exit(T1) ‐.21** ‐.24** ‐.35** .33** .27* ‐.34** 1
8: Neglect(T1) ‐.22** ‐.16** ‐.25** .27** .19** ‐.28** .31* 1
9: Voice(T1) .17** ‐.06 ‐.17** .34** ‐.13* ‐.02 .12 14** 1
Bivariate correlations with respect to outcome variables at the six‐month follow‐up (N =
186)
Exit(T2) ‐.15* ‐.05 ‐.10 ‐.03 .13 ‐.25** .59* .14 .01
Neglect(T2) ‐.17 ‐.20** ‐.16* .12 .13 ‐.11 .31 .56* .05
Voice(T2) .24** ‐.19* ‐.18 .28** ‐.05 ‐.12 .18** .19 .67**
1
Parirwise deletion. Actual number of cases not reflected.

The bivariate correlations among exit, neglect, and voice at the six‐month
follow‐up are illustrated in Table 12. As shown, (1) exit(T2) had a positive statistically
significant correlation with neglect(T2) and (2) voice(T2) had a statistically significant
positive correlation with neglect(T2). However, exit(T2) and voice(T2) did not have a
statistically significant correlation.

126
TABLE 12. BI‐VARIATE CORRELATIONS AMONG EXIT, NEGLECT, AND VOICE AT 6‐MONTH
FOLLOW‐UP
Scale
1
Exit (T2) Neglect (T2) Voice (T2)
Exit (T2) 1
Neglect (T2) .236** 1
Voice (T2) .075 .171* 1
1
Parirwise deletion. Actual number of cases not reflected.
5.3. ASSESSMENT OF MODEL FIT
Selected goodness‐of‐fit statistics were evaluated in accord with the criteria
specified in Chapter 4, Section 4.7. The conceptual model (see Chapter 4, Figure 4)
represents the hypothesized relationships among diversity characteristics,
perceptions of work climate, job stress and psychological wellbeing, and
exit/neglect/voice response over two points in time. As discussed, the model was
tested in four phases to achieve optimal model parsimony (based on the method
employed by Mor Barak et al, 2005). Phases I, II, and III involved analyses of data
from all respondents at baseline (n = 359). Phase IV is a longitudinal analysis of data
from baseline and six‐month follow‐up respondents (n = 186). Detailed statistics
(unstandardized and standardized estimates as well as levels of significance) for
each of the paths for each Phase of the analyses is offered in Appendix C.
Phase I: Diversity Characteristics   Work Climate Variables
In Phase I, relationships among diversity characteristics and work climate
variables were modeled. Personal diversity characteristics included gender and

127
ethnicity; professional characteristics comprised educational level (bachelors or
graduate degree), job position (direct service worker or supervisor/manager), and
job tenure. The work climate variables consisted of perception of inclusion in
decision‐making, perceived supervisory support, and perceived organizational
support. An examination of the fit indices for Phase I demonstrated that the full
model fit the data well (χ
2
= .210; df = 1; p = .647; NFI = 1; CFI = 1; RMSEA = .000; CN
= 6556). The ECVI value was .86; this statistic is useful in comparing two or more
models as discussed below.
To create the refined model, the statistical significance of all hypothesized
paths were examined for the extent to which they added misfit to the model. Then,
the non‐significant paths were removed, creating a more parsimonious model. The
refined model fit the data well (χ
2
= 18.397; df = 16; p = .301; NFI = .971; CFI = .996;
RMSEA = .020; CN = 512). The refined model was also improved from the initial,
full model without a statistically significant difference, thereby not decreasing
model fit (∆χ
2
= 18.187, ∆df = 15). Further, the ECVI value of .325 was lower than in
the initial, full model. As a result, the final, most parsimonious model signifies the
best fit to the data overall. This refined model accounted for 5.6% of the variability
perception of inclusion in decision‐making, an 18.5% in perceived supervisory
support, and 34.8% in perceived organizational support.

128
Phase II: Diversity   Work Climate   Wellbeing
Phase II tested the refined Phase I (diversity characteristics to work climate)
model with the addition of the following variables: role conflict, role ambiguity, and
psychological wellbeing. Examination of the fit indices for Phase II yielded an
overall χ
2
value of 35.005 (df = 18, p = .009). Given that the chi‐square statistics is
sensitive to sample size (Byrne, 2001), the following fit statistics helped demonstrate
a well‐fitting model: NFI = .958; CFI = .977; RMSEA = .051; CN =296. As discussed in
Phase I, the ECVI value (.578) was used to compare this model with the refined one.
As conducted in Phase I, the statistical significance of all hypothesized paths
were examined for the extent to which they added misfit to the model. Then, the
non‐significant paths were removed, creating a more parsimonious model.
Accordingly, the refined Phase II model was indicative of a well‐fitting model (χ2 =
52.043; df = 38; p = .064; NFI = .938; CFI = .981; RMSEA = .032; CN = 368). The refined
model, improved from the initial, full model with out a decrease in fit (∆χ2 = 17.038 ,
∆df = 18) yielded an ECVI value of .514 (lower than in the initial Phase II model).
This refined model accounted for 2.8% of the variability in inclusion in decision‐
making, 18.1% in perceived supervisory support, and 33.8% in perceived
organizational support, 22.7% in role conflict, 16.5% in role ambiguity, and 1.14% in
psychological wellbeing.

129
Phase III: Diversity   Work Climate   Wellbeing   Exit/Neglect/Voice (E/N/V)
Phase III of the path analysis built upon the earlier steps and incorporated
exit/neglect/voice variables based on the study design. As detailed through this study,
exit/neglect/voice respectively reflect efforts toward quitting one’s job, disengaging
from work‐related tasks and organizational activities, and improving dissatisfying
work conditions or promoting organizational change. This comprehensive model
was tested and refined in accordance with the technique used in the earlier steps.
Goodness‐of‐fit indices revealed a well‐fitting model (χ
2
= 54.771, df = 39, p = .048,
NFI = .953; CFI = .985; RMSEA = .034; CN = .784, ECVI = .784).
Estimation of the refined, final cross‐sectional model (in which all non‐
significant paths were removed) also demonstrated a good fit to the data (χ2 =
75.311; df = 61; p = .103; NFI = .936; CFI = .986; RMSEA = .026; CN = 382). The refined
model was improved from the initial, full model (∆χ
2
= 20.54, ∆df = 22) and yielded
an ECVI value of .719 (lower than in the initial Phase III model). The refined Phase
III model accounted for 2.9% of the variance in inclusion in decision‐making, 17.9%
in perceived supervisory support, and 33.3% in perceived organizational support,
22.9% in role conflict, 15.5% in role ambiguity, 1.15% in psychological wellbeing,
27.6% in exit, 18.8% in neglect, and 29.7% in voice. Figure 5 below depicts the final
refined, most parsimonious cross‐sectional model.

130
FIGURE 5. FINAL PHASE III PATH MODEL (BASELINE)
Phase III: Model fit statistics:
χ
2
= 58.402, df = 61, p = .571; NFI = .934; CFI = 1.00 RMSEA = .000; CN = 255; ECVI = 1.916
R
2
Inclusion = .179 R
2
Sup. support = .179 R
2
Org. Support = .333
R
2
Role conflict = .229 R
2
Role ambiguity = .155 R
2
Psychological wellbeing = .115
R
2
Exit = .276 R
2
Neglect = .188 R
2
Voice = .297

Notes:
• Only significant paths (p<.05) are shown.
• The following significant paths were suppressed for clarity of presentation.
Predictor Outcome
Job tenure (‐) Supervisory support
Job tenure (‐) Role ambiguity
Job tenure (+) voice
Latino (+) neglect
African American (‐) Role conflict
African American (‐) Voice
Inclusion (+) Role conflict
Inclusion (+) Psychological wellbeing
Inclusion (+) Voice
Inclusion (‐) Neglect
Supervisory support (‐) Exit
Role conflict (+) Wellbeing
Role conflict (+) Exit

131
Phase IV: Diversity   Work Climate   Wellbeing   E/NV over time
Phase IV examined the cross‐lagged relationships among exit, neglect, and
voice (over two points in time) utilizing the refined model of Phase III as a base.
Because this analysis included data from baseline and six‐month follow‐up
respondents, the number of cases (n = 186) is lower than the earlier phases of the
analyses. Examination of fit statistics for the Phase IV model revealed that the
model fit the data well (χ
2
= 58.402, df = 61, p = .571; RMSEA = .000; NFI = .934; CFI =
1.00) with adequate sample size (CN = 255) and an ECVI value of 1.916. Although
the full model fit the data well, refinements were made to improve model
parsimony through the removal of non‐significant paths. The refined Phase IV
model demonstrated improved model fit without decrease in fit (∆χ
2
= 127.829; ∆df
= 52), with an ECVI value of 1.720. The fit indices from the refined model yielded a
χ
2
value of 126.231 (df = 113, p =.186; RMSEA = .025; NFI = .857, CFI = .981; CN =
202). This model accounted for 3.6% of the variance in inclusion in decision‐making,
9.0% in perceived supervisory support, and 28.6% in perceived organizational
support, 22.6% in role conflict, 8.9% in role ambiguity, and 7.7% in psychological
wellbeing. The model respectively explained 23.5%, 12.9%, 32.7% in exit, neglect and
voice at baseline as well as 39.7%, 31.5%, 47.3% in exit, neglect and voice at the six‐
month follow‐up. Figure 6 below depicts the refined version of the refined, most
parsimonious longitudinal model.

132
FIGURE 6. FINAL PHASE IV PATH MODEL (LONGITUDINAL, CROSS‐LAGGED MODEL)

Phase IV: Model fit statistics:
χ
2
= 58.402, df = 61, p = .571; NFI = .934; CFI = 1.00 RMSEA = .000; CN = 255; ECVI = 1.916
R
2
Inclusion = .036 R
2
Sup.support = .095 R
2
Org. Support = .286
R
2
Role conflict = .227 R
2
Role ambiguity = .099 R
2
Psychological wellbeing = .090
R
2
Exit(T1) = .272 R
2
Neglect(T1) = .142 R
2
Voice(T1) = .333
R
2
Exit(T2 = .453 R
2
Neglect(T2) = .370 R
2
Voice(T2 = .518

Notes:
• Only significant paths (p<.05) are shown.
• The following significant paths were suppressed for clarity of presentation.
Predictor Outcome
Graduate degree (‐) Organizational support
Job tenure (‐) Supervisory support
Job tenure (‐) Role ambiguity
Job tenure (+) Voice(T1)
Latino (+) Neglect
African American (‐) Role conflict
African American (‐) Voice(T1)
Inclusion (+) Role conflict
Inclusion (+) Psychological wellbeing
Inclusion (‐) Neglect(T1)
Inclusion (+) Voice(T1)
Role conflict (‐) Psychological wellbeing
Role conflict (‐) Exit(T2)
Supervisory support (‐) Exit
Supervisory support (+) Neglect(T2)

133
5.4. STUDY HYPOTHESES: FINDINGS
To restate, the conceptual model assessed the direct and indirect effects of
diversity characteristics, perceptions of work climate, job stress, and psychological
wellbeing on exit/neglect/voice over two points in time. In this section, the hypotheses
represented by each model path were examined for significance and direction. All
hypotheses were examined based on the findings generated from the cross‐sectional
(Phase III) and the longitudinal, cross‐lagged (Phase IV) analyses. Sobel’s test was
used to test for significant indirect relationships in these models (Preacher & Hayes,
2004). Tables that present the unstandardized estimates (and its standard error),
standardized estimates, and probability levels for all phases are available in
Appendix E. For findings directly related to Phase III (cross‐sectional analysis), and
Phase IV (longitudinal analyses) refer to Tables 18 and 19, respectively; these tables
depict findings related to the hypothesized relationships.
Hypotheses Set A: The Role of Diversity Characteristics
Diversity characteristics   perceptions of work climate
This hypotheses set investigated the direct effects of diversity characteristics
on perceptions of work climate at two points in time. Each hypothesis included on
non‐dominant or dominant group diversity characteristics. Women, racial/ethnic
minorities, bachelor‐degreed workers, direct service workers, and inexperienced
employees were classified as non‐dominant group members. Caucasians, Whites,

134
graduate‐degreed workers, managers/supervisors, and seasoned employees were
classified as dominant group status.
HA.1 As compared to dominant group members, non‐dominant groups perceive that they
experience lower levels of inclusion in decision‐making. This hypothesis was
partially supported in the Phase III cross‐sectional analysis due to the
significant pathway between job position and inclusion in decision‐making.
Results revealed that direct service workers had lower perceived inclusion in
decision‐making than managers/supervisors (β = ‐.169, p = .002). No other
significant direct relationships were found in support of this hypothesis.
HA.2. As compared to dominant group members, non‐dominant groups perceive that they
experience lower levels of supervisory and organizational support. Results revealed
that employees of Latino/a decent had significantly lower reported levels of
perceived supervisory support (β = ‐.102, p = .041) than other ethnic groups.
No other significant paths were found in the hypothesized direction.
However, two significant paths were contrary to the hypothesized
relationship as follows: (1) Graduate‐degreed workers reported lower levels
of perceived organizational support than did those with Bachelor’s degrees
(β = ‐.390, p < .001) and (2) seasoned employees reported lower levels of
perceived supervisory (β = ‐.199, p < .001) and organizational support (β= ‐
.390, p < .001) than shorter‐tenured employees.

135
Diversity characteristics   job stress and psychological wellbeing
HA.3. As compared to dominant group members, non‐dominant group members will
experience greater role conflict and role ambiguity. Two significant paths were
found; however the relationships were contradictory to the predicted
direction; hence, this hypothesis was not supported. Women experienced
less reported role ambiguity than men (β = ‐.107, p = .031) and African
Americans experienced lower role conflict than Caucasians (β= ‐159, p <
.001). Seasoned employees reported less job stress associated with role
ambiguity at baseline (β = ‐.349, p < .001) than shorter‐term employees
lending some support to this hypothesis. No other statistically significant
pathways were found.
HA.4 As compared to dominant group members, non‐dominant group members experience
a lower sense of psychological wellbeing None of the pathways were statistically
significant, demonstrating a lack of support for this hypothesis.
Diversity characteristics   exit/neglect/voice
HA.5. As compared to dominant group members, non‐dominant group members will differ
in their exit responses. Hypothesis A.5 posits that non‐dominant and dominant
group members will differ in their efforts toward quitting their job.
However, due to mixed findings in the literature with regard the
relationship between specific diversity characteristics and the outcomes

136
variables, directionality was not predicted. Findings from the path analyses
supported this hypothesis only with regard to job tenure and job position
(and not gender, ethnicity, or educational level). Employees with longer job
tenure (i.e., seasoned employees) had less exit(T1) responses than those with a
shorter job tenure in the Phase III cross‐sectional analysis (β = ‐.145, p = .006)
and in the Phase IV longitudinal analysis (β = ‐.242, p = .001). Direct service
workers (β = ‐.153, p = .013) also had less exit(T2) efforts than managers or
supervisors in the Phase IV longitudinal analysis.
HA.6. As compared to dominant group members, non‐dominant group members do not
differ in their neglect responses. As hypothesized, most of the non‐dominant
groups members (women, bachelor degree holders, direct service workers,
inexperienced employees) did not differ from their dominant groups
counterparts in their neglect responses. However, differences based on
ethnicity were found due to the statistically significant path between Latinos
and neglect, which was not as anticipated. Latinos respondents reported
higher scores on the neglect scale indicating a greater propensity to engage in
efforts that reflect disengagement from work‐related tasks or organizational
activities (β = .146, p = .003) than other ethnic groups. Hence, this hypothesis
was not supported.

137
HA.7. As compared to dominant group members, non‐dominant group members will
engage in less voice responses. This hypothesis was partially supported. Two of
the five non‐dominant group categories held statistically significant
pathways to voice as predicted: ethnicity (African American: β = ‐.098, p =
.043; Latinos: β = ‐.124, p = .011) and job tenure (β = .299, p = .006). While
women were less likely to engage in voice in earlier phases of the analyses
(Phase I & II), no statistically significant paths existed in the final cross‐
sectional model. However, in the Phase IV longitudinal analysis, women
reported engaging in more voice responses more than men (β = .073, p = .039),
which is not as predicted. No other statistically significant paths existed
from the diversity characteristics to voice at either time point.
Hypotheses Set B: The Role of Work Climate Variables,
Inclusion in Decision‐making
HB. The more employees perceive that they are included in decision‐making,
1. the more they will perceive their supervisor or organization as supportive. The
paths from perception of inclusion in decision‐making to perceived
supervisory support (β = .387, p < .001) and perceived organizational
support (β = .369, p < .001) were statistically significant in the anticipated
direction. Hence, this hypothesis was fully supported.

138
2. the less they will experience job stress associated with role conflict or role
ambiguity. Although the pathway between inclusion in decision‐making
and role conflict was statistically significant at baseline (β = .273, p <
.001), the relationship was not in the hypothesized direction. No other
direct effects (through role conflict and role ambiguity) were found in
any phase of the analyses. Based on these results, this hypothesis was not
supported.
3. the greater their sense of psychological wellbeing. Perception of inclusion in
decision‐making had statistically significant direct effects on
psychological wellbeing (β = .144, p = .006). As such, this hypothesis was
fully supported.
4. the less they will engage in exit responses. This hypothesis was not
supported due to the lack of statistically significant paths from
perception of inclusion in decision‐making and exit in all phases of the
analyses.
5. the less they will engage in neglect responses. The statistically significant
direct pathway between perception of inclusion in decision‐making and
neglect(T1) in the Phase III cross‐sectional analysis (but not the
longitudinal analysis) provides partial support for hypothesis (β = ‐.183,
p = .006).

139
6. the more they will engage in voice responses. This hypothesis was partially
supported based on the statistically significant direct path from
perception of inclusion in decision‐making to voice(T1) (β = .172 ‐., p < .001)
in the Phase III cross‐sectional, but not the longitudinal analysis.
Perceived Supervisory Support
HB. The more employees perceived their supervisor as supportive at baseline,
7. the less they will experience job stress associated with role conflict and role
ambiguity. Support for this hypothesis was provided due to the direct
statistically significant relationships from perceived supervisory support
to (1) role conflict (β= ‐.205, p < .001) and (2) role ambiguity (β = ‐.207, p <
p < .001.) in the Phase III cross‐sectional analysis.
8. the greater their sense of psychological wellbeing. This hypothesis was not
supported due to a lack of a statistically significant pathway between
perceived supervisory support and psychological wellbeing.
9. the less likely they are to engage in exit responses. This hypothesis was
supported due to the statistically significant path between perceived
supervisory support and exit(T1) (β = ‐.110, p = .026) in the Phase III cross‐
sectional analysis. However, this relationship was not found in the Phase
IV longitudinal analysis with exit(T2) as the outcome variable.

140
10. the less they will engage in neglect responses. In the Phase III cross‐sectional
analysis, no statistically significant pathway between perceived
supervisory support and neglect(T1) was present. Yet, (as hypothesized)
perceived supervisory support at baseline predicted neglect(T2) (β = ‐.151,
p = .016) at the six‐month follow‐up in the Phase IV longitudinal
analysis. Thus, this hypothesis was partially supported.
11. the more they will engage in voice responses. This hypothesis was not
supported due to a lack of statistically significant pathways between
perceived supervisory support and voice in all phases of the analyses.
Organizational Support
HB. The more employees perceived their work organizations as supportive,
12. the less they will experience job stress associated with role conflict or role
ambiguity. Statistically significant direct paths from perceived
organizational support to role conflict (β = ‐.411, p < .001) and role
ambiguity (β = ‐.237, p < .001) in the Phase III cross‐sectional analysis
lend full support for this hypothesis.
13. the greater their sense of psychological wellbeing. This hypothesis was not
supported to due a lack of a statistically significant pathway from
perceived organizational support to psychological wellbeing in all
phases of the analysis.

141
14. the less they will engage in exit responses. This hypothesis was partially
supported due to the statistically significant direct path between
perceived organizational support and exit(T1) (β = ‐.191, p < .001) in the
Phase III cross‐sectional analysis. A statistically significant pathway from
organizational support to exit(T2) was not found in the Phase IV
longitudinal analysis.
15. the less they will engage in neglect responses. No statistically significant
direct path was found to support this hypothesis.
16. the more they will engage in voice responses. This hypothesis was not
supported due to a lack of a statistically significant direct pathway in all
phases of the analysis.
Hypotheses Set C: The Role of Job stress and Psychological Wellbeing
Job stress (role conflict and role ambiguity)
HC. The more employees experience job stress associated with role conflict and role
ambiguity,
1. the lower their level of psychological wellbeing. The hypothesized direct and
significant pathways from role conflict (β = ‐.242, p < .001) and role
ambiguity (β = ‐.146, p = .004) to psychological wellbeing provide full
support for this hypothesis.

142
2. the more they will engage in exit responses. At baseline, statistically
significant pathways were found from (1) role conflict (β = .159, p = .002)
and (2) role ambiguity (β = .104, p = .037) to exit(T1). However, this did not
hold at the six‐month follow‐up; thereby providing partial support for
this hypothesis.
3. the more they with engage in neglect responses. Role conflict (β = .157, p =
.004) and role ambiguity (β = .104, p = .037) had direct effects on neglect(T1)
in the Phase III cross‐sectional analysis (but not longitudinal). Thus, this
hypothesis was partially supported.
4. the more they with engage in voice responses. Role conflict at baseline was
positively related to voice(T1) (β = .273, p < .007) in the Phase III cross‐
sectional analysis, but not at voice(T2) in the Phase IV longitudinal analysis.
No statistically significant pathways were found from role ambiguity to
voice in all phases of the analysis. Therefore, this hypothesis was partially
supported.
Psychological wellbeing
HC. The more employees experience a heightened sense of psychological wellbeing at
time1,
5. the less they will engage in exit responses. The statistically significant direct
effect of psychological wellbeing to exit(T1) (β = ‐.180, p < .001) in the Phase

143
III cross‐sectional analysis offer partial support for this hypothesis.
However, this path was not statistically significant with exit(T2) as the
outcome variable in the Phase IV longitudinal analysis.
6. the less they will engage in neglect responses. This hypothesis was partially
supported. Psychological wellbeing at baseline was a statistically
significant predictor of neglect(T1) (β = ‐.172, p < .001) at baseline, but not at
the six‐month follow‐up.
7. the more they will engage in voice responses. Psychological wellbeing had no
direct effects on voice in the Phase III cross‐sectional or longitudinal
analyses. Thus, this hypothesis was not supported.
Hypotheses Set D: The Relationship Among Exit, Neglect, and Voice
Exit   neglect & voice
HD. The more employees’ engage in exit at baseline,
1. the more they will engage in neglect responses at the six‐month follow‐up.
2. the less they will engage in voice responses at the six‐month follow‐up.
Exit(T1) had no direct effects on neglect(T1) or voice(T2) in the Phase IV
longitudinal analysis; thereby providing a lack of support for these two
hypotheses.

144
Neglect   exit & voice
HD. The more employees’ engage in neglect at baseline,
3. the more employees’ engage in exit at baseline and at the six‐month follow‐up.
Partial support for this hypothesis was provided by the statistically
significant direct effect from neglect(T1) to exit(T1) (β = .138, p = .005) in the
Phase III cross‐sectional analysis. However, this relationship was not
present at the six‐month follow‐up based on findings from the
longitudinal analysis.
4. the less they will engage in voice at 2. This hypothesis was not supported
due to a lack of statistically significant path between neglect at baseline
and voice responses at the six‐month follow‐up.
Voice   exit & neglect
HD. The more employees’ engage in voice at baseline,
5. The less they will engage in exit at baseline and at the six‐month follow‐up.
This hypothesis was not supported. Voice(T1) had no direct effects on
exit(T1) in the Phase III cross‐sectional or exit(T2) longitudinal analysis.
6. The less they will engage in neglect at baseline and at the six‐month follow‐up.
Based on findings from the cross‐sectional analysis, the statistically
significant pathway from voice(T1) and neglect(T1) does not support the

145
hypothesized relationship due to directionality of the relationship. Yet,
in the Phase IV longitudinal analysis, prior to trimming of the full
longitudinal model, the relationship between voice(T1) and neglect(T2) was
negative as hypothesized; however the result was non‐statistically
significant (p = .070) and therefore not included in the refined model.
However, the shift in directionality is substantively statistically
significant, and therefore post‐hoc analysis was conducted to provide
insight into the (1) positive associated between voice and neglect at
baseline and (2) at the six‐month follow‐up the shift in the directionality
of the relationship over time.
Post‐hoc analysis examined a simple cross‐lagged model between
voice and neglect over two time‐points controlling for statistically
significant relationships from gender, ethnicity, and job tenure to neglect
and voice at baseline and at the six‐month follow‐up. Controlling for
these specific diversity characteristics based on the findings from
hypotheses A.6 & A.7, results revealed that the simple neglect/voice cross‐
lagged model fit the data well (χ
2
= 4.409; df = 8; p = .551; NFI = 1; CFI =
1.0; RMSEA = .000; CN = 651). Most relevant, the statistically significant
positive path from voice(T1) to neglect(T1) (β = −.175; p = .017). indicated that
the more employees engaged in voice, the more they engaged in neglect.

146
This was consistent with findings related to hypothesis D.5, but not in the
anticipated direction. On the other hand, the cross‐lagged relationship
from voice(T1) neglect(T2) was in the predicted direction (β = −.170; p = .043).
Other statistically significant pathways can be found in Appendix E,
Table 20 as they are not directly germane to the purpose of the post‐hoc
analysis.
5.5. INDIRECT EFFECTS
Several statistically significant indirect effects were found among the
diversity characteristics, perceptions of work climate, job stress and psychological
wellbeing and exit/neglect/voice as response to undesirable experiences at work. As
depicted in Figure 4 Chapter 4, extensive combinations of hypothesized indirect
effects were tested. Therefore in this section, only the statistically significant indirect
relationships are detailed.
HE.1. Inclusion in decision‐making will mediate the relationships among employees’
diversity characteristics and their perceptions of supervisory or organizational
support. This hypothesis was partially supported. Employees’ job position
had statistically significant indirect effects on perceptions of supervisory
support (Sobel’s test statistic = ‐7.51, p < .001) and organizational support
(Sobel’s test statistic = ‐7.51, p < .001) through inclusion in decision‐making.
Direct service workers perceived that they experienced higher levels of

147
supervisor and organizational support than managers or supervisors; while
also having lower levels of perceived inclusion in decision‐making.
However, perceived inclusion was positively related to perceived supervisor
and organizational support. No other statistically significant indirect
relationships were found.
HE. Work climate variables (perceived inclusion in decision‐making, supervisory and
organizational support) will mediate the relationships among:
2. diversity characteristics and job stress (role conflict, role ambiguity) and
psychological wellbeing. Three statistically significant results demonstrate
that the perceptions of work climate; specifically, support variables,
mediated the relationship between diversity characteristics and job
stress. First, possession of a graduate degree indirectly affected role
conflict through perceptions of organizational support (Sobel’s test
statistic = 3.93, p < .001) with full mediation. On the one hand, having a
graduate degree is positively associated with role conflict; however, this
relationship is dropped when perceived organizational support was
added as a mediator. Second, being a seasoned employee indirectly
affected role conflict through perceived supervisory support (Sobel’s test
statistic = 2.86, p = .004) as well as perceived organizational support
(Sobel’s test statistic = 5.80, p < .001). Last, as seasoned employees

148
reported less job stress associated with role ambiguity at baseline than
shorter‐tenured workers and this relationship was also partially
mediated by perceived supervisory support (Sobel’s test statistic = 2.86, p
= .004) as well as perceived organizational support (Sobel’s test statistic =
3.78, p < .001).
3. diversity characteristics and exit/neglect/voice responses. Being in a direct
service position indirectly affected the extent to which employees
engaged in voice(T1) through their perception of inclusion in decision‐
making (Sobel’s test statistic = ‐3.76, p < .001).
H E. Job stress—role conflict and role ambiguity—will mediate the relationships among:
4. diversity characteristics and exit/neglect/voice responses. Job tenure indirectly
affected exit(T1) through role ambiguity (Sobel’s test statistic = 2.60, p
=.009) in the Phase III cross‐sectional analysis. Also, African Americans’
perception of role conflict indirectly affected their voice responses
(Sobel’s test statistic = ‐3.34, p = .001) at baseline. No other statistically
significant indirect relationships were found; thus, this hypothesis was
not supported.
5. work climate variables and exit/neglect/voice responses. A myriad of
statistically significant indirect relationship were found from perceptions
of work climate to exit/neglect/voice. Specifically, role conflict mediated

149
the relationships between the following perceptions of work climate and
the outcomes variables at baseline:
i. inclusion in decision‐making and neglect(T1) (Sobel’s test statistic =
2.19, p = 029),
ii. inclusion and voice(T1) (Sobel’s test statistic = 2.85, p = .004),
iii. supervisory support and exit(T1) (Sobel’s test = ‐3.03, p = .002),
iv. supervisory support and neglect(T1) (Sobel’s test = ‐2.93, p = .003),
v. organizational support and exit(T1) (Sobel’s test = ‐2.85, p = .004),
vi. organizational support and neglect(T1) (Sobel’s test = ‐2.76, p = .005),
and
vii. organizational support and voice(T1) (Sobel’s test = ‐4.76, p < .001).
No other indirect effects were found in the Phase III cross‐sectional or
longitudinal analyses with role ambiguity as a mediator.

H E. Psychological wellbeing will mediate the relationships among:
6. diversity characteristics and exit/neglect/voice responses. No statistically
significant indirect relationships were found; thus, this hypothesis was
not supported.
7. work climate variables and exit/neglect/voice responses. Perception of
inclusion in decision‐making had statistically significant indirect effects

150
through psychological wellbeing on neglect(T1)) in the Phase III cross‐
sectional analysis (Sobel’s test statistic = ‐2.11, p = .035. The work climate
variables had no other statistically significant indirect effects on
exit/neglect/voice through psychological wellbeing.
8.   job stress and exit/neglect/voice responses.
Psychological wellbeing mediated the relationships from:
i. role conflict to exit(T1) (Sobel’s test statistic = 2.91, p = .004) and to
neglect(T1) (Sobel’s test statistic = 2.74, p = .006)
ii. role ambiguity to exit(T1) (Sobel’s test statistic = 2.26, p = .024) and to
role ambiguity to neglect(T1) (Sobel’s test statistic = 2.17, p = .030).
There were no other statistically significant indirect effects from role
conflict or ambiguity to voice with psychological wellbeing as a mediator
in any phase of the analyses.

H E. Exit at baseline will mediate the relationship among the model variables at baseline
(i.e., diversity characteristics, work climate variables, job stress and psychological
wellbeing, neglect, and voice) and the following:
9. neglect responses at the six‐month follow‐up.
10. voice responses at the six‐month follow‐up.

151
This hypothesis was not supported due to a lack of statistically significant
indirect paths.
H E. Neglect at baseline will mediate the relationship among the model variables at
baseline (i.e., diversity characteristics, work climate variables, job stress and
psychological wellbeing, and voice) and the following:
11. exit responses at baseline and at the six‐month follow‐up. Neglect at baseline
mediated the relationship from role conflict to exit at baseline, as
hypothesized (Sobel’s test statistic = 2.04, p = .041).
12. voice responses at the six‐month follow‐up. This hypothesis was not
supported due to a lack of statistically significant indirect paths among
the hypothesized relationships.
H E. Voice at baseline will mediate the relationship among the model variables at baseline
(i.e., diversity characteristics, work climate variables, job stress and psychological
wellbeing, and the following:
13. neglect responses. Inclusion in decision‐making (Sobel’s test statistic =
2.260, p = .024) and role conflict (Sobel’s test statistic = 2.54, p = .011)
indirectly affected neglect through voice, although not in the
hypothesized direction. Hence, this hypothesis was not supported.

152
14. exit responses. This hypothesis was not supported due to a lack of
statistically significant indirect paths among the hypothesized
relationships.
5.6. SUMMARY OF RESULTS
This study examined the organizational and individual antecedents of exit,
neglect, voice responses over two time points—baseline and six‐month follow‐up.
Both the cross‐sectional and longitudinal path analytic models did fit the data well
and provides evidence that diversity characteristics, perceptions of work climate,
job stress, and psychological wellbeing are important determinants of employees
exit/neglect/voice responses to their work environment. Tables 10 to 13 offer a simple
summary of significant findings.

153
TABLE 10. RECAP OF SIGNIFICANT FINDINGS FROM HYPOTHESIS SET A
Hypothesis Predictor Variable Direction of
Relationship
Outcome Variable
HA.1 Direct service worker (‐) Inclusion
HA.2 Latino (‐) PSS
Graduate degree (‐)* POS
Job tenure (‐)* PSS
Job tenure (‐)* POS
H A.3 Women (‐)* RA
African American (‐)* RC
H A.4 Job tenure (‐) RA
HA.5 Direct service worker (‐) Exit(T2)
Job tenure (‐) Exit(T1)
HA.6 Latino/a (+) Neglect(T1)
HA.7 African American (‐) Voice(T1)
Latino/a (‐) Voice(T1)
Job tenure (+) Voice(T1)
Gender (+)* Voice(T2)
*Although significant, relationship was not in hypothesized direction.
Legend:
Inclusion = Perception of inclusion in decision‐making
PSS = Perceived supervisory support
POS = Perceived organizational support
RC = Role conflict
RA = Role ambiguity

154
TABLE 11. RECAP OF SIGNIFICANT FINDINGS FROM HYPOTHESIS SET B
Hypothesis Predictor Variable Direction of relationship Outcome Variable
HB.1 Inclusion (+) PSS
Inclusion (+) POS
HB.2 Inclusion (+)* RC
HB.3 Inclusion (+) PWB
HB.4 Not supported
HB.5 Inclusion (‐) Neglect(T1)
HB.6 Inclusion (+) Voice(T1)
HB.7 Supervisory support (‐) RC
Supervisory support (‐) RA
HB.8 Not supported
HB.9 Supervisory support (‐) Exit(T1)
H B.10 Supervisory support (‐) Neglect(T2)
HB.11 Not supported Not supported
HB.12 Org. support (‐) RC
Org. support (‐) RA
HB.13 Not supported
HB.14 Org. support (‐) Exit(T1)
HB.15 Not supported
HB.16 Not supported
*Although significant, relationship was not in hypothesized direction.
Legend:
Inclusion = Perception of inclusion in decision‐making
PSS = Perceived supervisory support
POS = Perceived organizational support
RC = Role conflict
PWB = Psychological wellbeing

155
TABLE 12. RECAP OF SIGNIFICANT FINDINGS FROM HYPOTHESIS SET C
Hypothesis Predictor
Variable
Direction of
relationship
Outcome
Variable
HC.1 Role conflict (‐) PWB
Role ambiguity (‐) PWB
HC.2 Role conflict (+) Exit(T1)
Role conflict (‐)* Exit(T2)
Role ambiguity (+) Exit(T1)
HC.3 Role conflict (+) Neglect(T1)
Role ambiguity (+) Neglect(T1)
HC.4 Role conflict (+) Voice(T1)
HC.5 Psychological
wellbeing
(‐) Exit(T1)
HC.6 Psychological
wellbeing
(‐) Neglect (T1)
HC.7 Not supported
*Although significant, relationship was not in hypothesized direction.
Legend:
PWB = Psychological wellbeing

TABLE 13. RECAP OF SIGNIFICANT FINDINGS FROM HYPOTHESIS SET D
Hypothesis Predictor Variable Direction of
relationship
Outcome
Variable
HD.1 Not supported
HD.2 Not supported
HD.3 Neglect(T1) (+) Exit(T1)
HD.4 Not supported
HD.5 Voice T1) (+)* Neglect(T1)
HD.6 Not supported
*Although significant, relationship was not in hypothesized direction.

156
TABLE 14. RECAP OF SIGNIFICANT FINDINGS FROM HYPOTHESIS SET E
Hypothesis Predictor
Variable
Direction of relationship
(mediator listed)
Outcome
Variable
HE.1 Direct service worker (‐) Inclusion (+) PSS
Direct service worker (‐) Inclusion (+) POS
HE.2 Graduate degree (‐) POS (‐) RC
Job tenure (‐)* PSS (‐) RC
Job tenure (‐)* POS (‐) RC
Job tenure (‐)* PSS (‐) RA
Job tenure (‐)* POS (‐) RA
HE.3 Direct service worker (‐) Inclusion** (+) Voice(T1)
HE.4 Job tenure (‐)* RA (+) Exit(T1)
African American (‐) RC (+)* Voice(T1)
HE.5 Inclusion (+)* RC (+) Neglect(T1)
Inclusion (+)* RC (+)* Voice(T1)
Supervisory support (‐) RC (+) Exit(T1)
Supervisory support (‐) RC** (+) Neglect(T1)
Org. support (‐) RC (+) Exit(T1)
Org. support (‐) RC (+) Neglect (T1)
Org. support (‐) RC** (+)* Voice (T1)
HE.6 Not supported
HE.7 Inclusion (+) PWB (‐) Neglect(T1)
HE.8 Role conflict (‐) PWB (‐) Exit(T1)
Role conflict (‐) PWB (‐) Neglect(T1)
Role ambiguity (‐) PWB (‐) Exit(T1)
Role ambiguity (‐) PWB (‐) Exit(T1)
HE.9 Not supported
HE.10 Not supported
HE.11 Role conflict (‐) Neglect(T1) (+) Exit(T1)
HE.12 Not supported
HE.13 Inclusion (+) Voice (+)* Neglect(T1)
*Although significant, relationship was not in hypothesized direction.
**Full mediation model. Significant at level .025 with Bonferroni’s
correction. (All other indirect effects not denoted represent partial
mediation).
Legend:
Inclusion = Perception of inclusion in decision‐making
PSS = Perceived supervisory support
POS = Perceived organizational support
RC = Role conflict
RA = Role ambiguity

157
Chapter 6: Discussion
6.1. OVERVIEW
Human service organizations are challenged to cultivate optimal outcomes
for the wellbeing of the individuals, families, and communities that they serve. At
the heart of these organizations are direct service workers, supervisors, and
managers/administrators who conduct the day‐to‐day operations that enable
organizations to fulfill their missions. This study sought to understand the factors
contributing to these employees’ exit/neglect/voice responses. To reiterate, exit,
neglect, and voice are conceptualized in the current study as follows (Farrell &
Rusbult, 1992; Hagedoorn et al., 1999; Rusbult et al., 1988):
♦ Exit: Efforts toward quitting one’s job,
♦ Neglect: Efforts that reflect one’s disengagement from work‐related
tasks and organizational activities, and
♦ Voice: Efforts to improve dissatisfying work conditions or promote
organizational change.
As human service organizations are faced with rapid turnover rates and concerns
regarding employee satisfaction and organizational commitment, gauging
employees’ exit/neglect/voice responses offers a valuable contribution to the
profession of social work and the field of human services.

158
This chapter first presents the key study findings. Next, the strengths and
limitations of the study are discussed. Then, theoretical and practical implications
are discussed. Last, recommendations for further research are provided.
6.2. KEY FINDINGS
The present study examined organizational and individual determinants of
human service workers’ exit/neglect/voice responses by testing a theory‐driven
conceptual model depicting the relationships among the following:
♦ employee personal and professional diversity characteristics;
♦ perceptions of organizational climate;
♦ job stress and psychological wellbeing; and
♦ exit/neglect/voice responses (outcome variables)
The study also tested the model longitudinally with respect to the outcome
variables and examined the relationships among the outcome variables over two
time points. The model was developed from social work, organizational behavior,
and industrial psychology theories to offer a comprehensive multidisciplinary
perspective. Results revealed that the hypothesized model fit the data well in both
the cross‐sectional and longitudinal analyses, respectively explaining 27.6%, 18.8%,
and 29.7% of the variance of exit/neglect/voice at baseline (Phase III cross‐sectional
analysis as well as 45.3%, 37.0% 51.8% of the variance t in exit/neglect/voice at the six‐
month follow‐up (in longitudinal analyses).

159
Key cross‐sectional findings revealed that exit, neglect, and voice were
associated mostly as hypothesized, while other unforeseen relationships emerged.
Specifically, as employees perceived themselves as included in decision‐making,
they were more likely to engage in voice and less likely to exercise neglect; while as
employees experienced supervisory and organizational support, they were less
likely to experience job stress and less likely to engage in exit. Note here that
supervisory and organizational support had indirect effects on neglect through job
stress. These and other results are discussed in more detail in the following sections.
The Role of Diversity Characteristics
For this study, diversity characteristics were categorized based on
employees’ membership in societally‐defined non‐dominant and dominant groups.
Women, racial/ethnic minorities, bachelor‐degreed workers, direct service workers,
and inexperienced employees were classified as non‐dominant group members.
Men, Caucasians, graduate‐degreed workers, managers/supervisors, and seasoned
employees were classified as holding a dominant group status. Key cross‐sectional
findings revealed that non‐dominant and dominant group members differed in
their perceptions of work climate and job stress, but not in their sense of
psychological wellbeing. These workers also differed in their exit/neglect/voice
responses.

160
Diversity characteristics   perceptions of work climate
As hypothesized, some differences were found between dominant and non‐
dominant group members with respect to their perceptions of work climate. These
results are detailed as follows:
♦ As anticipated, direct service workers (a non‐dominant group) reported lower
levels of perceived inclusion in decision‐making compared to
supervisors/managers (a dominant group). This may be attributed to the
supervisors/managers’ hierarchical power within the organizational structure as
compared to those workers who deal directly with clients (Cousins, 2004).
Despite this relationship, no other diversity characteristics served as predictors
of inclusion in decision‐making. This is contrary to research that has found
gender and ethnicity to be significant predictors of employees’ perception of
inclusion (Mor Barak et al., 2003; Mor Barak & Levin, 2002).
♦ As expected, Latinos (a non‐dominant group) perceived that they received
lower supervisory support than Caucasians (a dominant group). This result is in
contrast to Amason, Allen, and Holmes’s (1999) finding that demonstrated a
lack of significant difference between Latinos and Caucasians in felt social
support from supervisors and organizational higher‐ups. This relationship was
not statistically significant for the other non‐dominant groups—African
Americans, Asians, or women.

161
♦ An unexpected finding was that graduate‐degreed employees (a dominant
group) had lower levels of perceived organizational support than those with
bachelor’s degrees (a non‐dominant group). Relatedly, seasoned employees
similarly had lower levels of perceived supervisory and organizational support
than shorter‐tenured employees. Although this is contrary to the hypothesized
relationship, other researchers have had similar findings. Nissly et al. (2005)
found that graduate‐degreed workers (dominant) reported greater levels of
stress than those with bachelor’s degrees (non‐dominant). The difference may be
accounted for by the distinction between personal and professional
characteristics. In the social work profession, as individuals hold more
organizationally‐based hierarchical power, they may be faced with challenges of
supporting others, as well as the demands and expectations of maintaining
accountability to their organization and communities they serve (Cousins, 2004).
Perhaps, in some cases, these demands and expectations foster lower perception
of support.
Diversity characteristics   job stress & psychological wellbeing
Key findings regarding the association among non‐dominant/dominant
group membership and stress are highlighted as follows:

162
♦ Contrary to the hypotheses, women had lower levels of role conflict as
compared to men and African Americans had lower levels of role ambiguity
as compared to Caucasians.
♦ As hypothesized, seasoned employees (a dominant group) experienced
lower levels of role ambiguity than shorter‐tenured employees (a non‐
dominant group). Considering their level of organizational experience,
seasoned employees may be more informed about their roles in the
organization, resulting in lower levels of role ambiguity than shorter‐
tenured employees.
♦ Non‐dominant group and dominant group members did not differ with
respect to psychological wellbeing. This finding was not as predicted based
on the reviewed literature.
Diversity characteristics   exit, neglect, voice
Study findings revealed that employees’ differed in their exit/neglect/voice
responses by gender, ethnicity, job level, and job tenure. No significant differences
were present on the outcomes variables based on educational level. The specific
details related to these findings are as follows:
♦ Seasoned employees (a dominant group) were less likely to engage in cognitions
and behaviors toward quitting their jobs as compared to shorter‐tenured
employees (a non‐dominant group). This finding was not surprising considering

163
that employees with longer tenure may be more invested in their work
organization and, therefore, less likely to leave (Mor Barak, Nissly et al., 2001).
The lack of statistically significant relationships among other diversity
characteristics and exit is aligned with mixed findings in the reviewed literature
(Griffeth et al., 2000; Mor Barak et al., 2006; Mor Barak, Nissly et al., 2001; Nissly
et al., 2005).
♦ Mostly as hypothesized, dominant and non‐dominant group members did not
differ in their neglect responses. However, Latino/s reported having more efforts
that reflect disengagement from work‐related tasks or organizational activities
in comparison to Caucasians, African Americans, and Asians. This finding is
challenging to interpret due to the limits in the literature on the relationship
between non‐dominant group status and neglect in human service organizations.
However, as a point of context, neglect as a construct reflects how workers are
fairing in their work‐related tasks or organizational activities (e.g., showing up
late, missing meetings, working slowly), not how they are interacting with or
treating the clients. Therefore, other factors such as cultural background or
challenges with work/family balance may exacerbate neglect responses. For
example Kowalski (2002) found that work/family conflict was positively
correlated with lateness, although the researcher did not control for ethnicity as
an antecedent. Stone‐Romero et al. (2003) specifically discussed the role of

164
culture among Latinos in relation to their work scripts. The authors stated that
in some cases the act of being late may have differing meanings depending on
the cultural context. In the United States lateness is considered disruptive to
organizational effectiveness; whereas some Latin American countries do not
have such negative connotations. Consequently, the measure of neglect used in
this study may not account for this cultural difference.
♦ As predicted, some non‐dominant groups members engaged in less voice than
their dominant group counterparts; these included African Americans, Latinos,
and shorter‐tenured employees. This finding is consistent with researchers who
posited that those in historical‐ or hierarchal‐based power positions have greater
access to information, latitude in decision‐making, and connections in
interpersonal processes that make certain voice responses more acceptable (Islam
& Zyphur, 2005; Van Dyne & LePine, 1998). Interestingly, however, those in the
other non‐dominant group categories (women, bachelor‐degreed workers, and
direct service workers) did not differ from their dominant group counterparts in
voice responses when this relationship was considered among other predictors.
The Influence of Work Climate Variables
The three work climate variables (inclusion in decision‐making, supervisory
support, organizational support) had differing effects on employees’ job stress and

165
psychological wellbeing as well as their exit/neglect/voice responses. These findings
are detailed as follows:
Inclusion   supervisory and organizational support
Inclusion in decision‐making is reflective of employee perceptions of their
involvement in critical processes affecting their work group and organization.
Findings demonstrated that as employees perceived that they were included, they
were more likely to feel supported, either by their supervisors or organization. This
finding is consistent with empowerment theory and Allen’s et al. (2003) empirical
research. Allen found that perceptions of human resource practices (specifically the
extent employees participated in decision‐making) were a correlate of perceived
organizational support.
Perceptions of work climate   job stress & psychological wellbeing
Paradoxical findings emerged among perceptions of work climate and
wellbeing in the Phase III cross‐sectional analysis. Specifically,
♦ Role conflict
o was positively associated with perceived inclusion in decision‐making
and
o was negatively associated with perceived supervisory and
organizational support.
♦ Role ambiguity

166
o was not affected by perceived inclusion in decision‐making and
o was negatively associated with perceived supervisory and
organizational support.
♦ Psychological wellbeing
o was positively associated with perceived inclusion in decision‐making
and
o was not affected by perceived supervisory and organizational support.
The positive relationship between role conflict and inclusion in decision‐making
was contrary to the hypothesized direction. Conceivably, employee involvement in
decision‐making may offer added responsibilities, which, in turn, may result in
greater job stress as employees deal with competing demands and challenges in
meeting others’ expectations. Further, employees’ perception of inclusion in
decision‐making was significantly and positively related to their psychological
wellbeing. This finding has been supported by other researchers (Mor Barak &
Levin, 2002). In contrast, the more employees’ perceived their supervisors and
organization as supportive, the lower their levels of role conflict and role ambiguity.
Yet, surprisingly perceptions of support were not associated with psychological
wellbeing.
No longitudinal associations were found between work climate, job stress,
and psychological wellbeing variables.

167
Perceptions of work climate   exit/neglect/voice
Results from the cross‐sectional analysis also demonstrated that work
climate variables affected employee exit/neglect/voice responses in differing ways as
detailed below:
♦ Exit
o was not influenced by perception of inclusion in decision‐making and
o was inversely related to perceived supervisory and organizational
support.
♦ Neglect
o was inversely associated with inclusion in decision‐making and
o was not affected by perceived supervisory or organizational support.
♦ Voice
o was positively associated with inclusion in decision‐making and
o was not affected by perceived supervisory or organizational support.
Longitudinal findings revealed that supervisory support at baseline was
inversely related to neglect at the six‐month follow‐up. No other association among
the work climate variables and exit, neglect, and voice were found in the cross‐
sectional or longitudinal analyses.
Empowerment (Perkins & Zimmerman, 1995; Spreitzer, 1995; Spreitzer, De
Janasz, & Quinn, 1999; Zimmerman, 1995) and social exchange theories (Blau, 1964;

168
Cropanzano & Mitchell, 2005) help explain these differing effects of inclusion versus
support on exit, neglect, and voice. For starters, voice and neglect reflect employees’
responses that may directly contribute to employees’ on‐the‐job functioning.
Specifically, these respectively reflect the extent to which employees are engaged
(i.e., voice) or not engaged (i.e., neglect) in task or organizational activities (Dundon,
Wilkinson, Marchington, & Ackers, 2004; Kidwell & Robie, 2003; Withey & Cooper,
1989). As a result, based on empowerment theory, perceived involvement in critical
organizational processes may result in positive worker outcomes (Laschinger &
Havens, 1996; Laschinger, Wong, McMahon, & Kaufmann, 1999; Spreitzer et al.,
1999). This may be particularity relevant for voice‐related behaviors because,
researchers have argued that participation in decision‐making processes are
elements of voice (Dundon et al., 2004). Exit involves employee intentions and
behaviors toward leaving their jobs that may not directly affect their daily activities.
Therefore, support may cultivate an exchange‐based relationship between the
employee and their supervisor or organization. In this relationship, as employees
experience support, they will reciprocate by remaining employed with their current
work organization. However, the experience of support does not necessarily keep
employees engaged in work group or organizational activities; rather those
outcomes are a product of being included in decision‐making.

169
Hopkins (2002) also reported a lack of a statistically significant relationship
between perceived support and voice in her of study children and family service
workers’ organizational citizenship behaviors. The lack of relationship, perhaps,
could be attributed to the meaning of perceived support. Perceived support reflects
the extent to which an employee feels valued and cared about as distinct from, for
example, the extent to which they feel the supervisor or organization is trustworthy
or open to their ideas. For example, Chia, Landau, and Ong (2000) found a positive
relationship between employees perceptions’ of their supervisors as approachable
and responsive and their willingness to engage in voice.
The Role of Job Stress & Psychological Wellbeing
Role conflict, role ambiguity, and psychological wellbeing had differing
effects on exit, neglect, and voice. First, Phase III cross‐sectional results revealed the
following:
♦ Exit
o was positively related to role conflict,
o was positively related to role ambiguity, and
o was inversely related to psychological wellbeing.
♦ Neglect
o was positively related to role conflict,
o was positively related to role ambiguity, and

170
o was inversely related to psychological wellbeing.
♦ Voice
o was positively related to role conflict,
o was not related to role ambiguity, and
o was not related to psychological wellbeing.
However, only one statistically significant relationship among wellbeing and exit,
neglect, and voice emerged over time: Exit at the six‐month follow‐up was inversely
related to role conflict at baseline.
It is curious that role conflict at baseline was positively related to exit at
baseline; yet inversely related at the six‐month follow‐up. Perhaps, increased
responsibilities created extra demands. Over time, however, employees may have
been more invested in the organization and acclimated to this type of stress.
The positive relationship between role conflict and voice suggests that this form
of stress appeared to stimulate the ‘fight’ response. As such, employees may have
engaged in change‐oriented efforts to reduce sources of their role stress. The
relationship between role conflict and neglect was positive. Based on role stress
theory (Biddle, 1986; Kahn et al., 1964), researchers and organizational leaders have
long recognized that job stress may adversely affect worker outcomes, such as job
productivity and counterproductive work behaviors (Jones, 1981).

171
Role ambiguity had significant and positive associations with exit and
neglect, but not voice. Moreover, none of the significant relationships were present
over time. Accordingly, as employees experienced job stress associated with lack of
clarity about one’s job roles, the more they held cognitions and behaviors about
quitting their job and limiting their involvement in work‐related tasks or
organizational activities. Interestingly, role ambiguity had no significant
relationship with voice at either time point (It was, however, inversely related in the
bivariate correlations). Because role ambiguity reflects uncertainty about one’s roles,
‘spiral of silence’ theory (F. Bowen & Blackmon, 2003) offers perspective on this
relationship. Plausibly, uncertainty about one’s role that is a function of role
ambiguity breeds fear or insecurity which may be accompanied by a lack of
willingness to speak up. Under conditions of uncertainty the ‘flight’ response—a
propensity to withdraw from duties or the job itself—may be actualized rather than
a ‘fight’ or willingness to voice.
Finally, as employees had a heightened sense of psychological wellbeing, the
lower their levels of engagement in exit and neglect related efforts. This finding was
in contrast to Mor Barak et al. (2001) findings that psychological wellbeing was not
related to intention to leave or actual turnover in their meta‐analysis of the
antecedents of human services workers turnover. On the other hand, not
surprisingly, in relation to neglect, the inverse relationship between psychological

172
wellbeing and counterproductive work behaviors was supported by others’
research (Boyd, 1997; Martinko et al., 2002; P. E. Spector & Fox, 2005).
Psychological wellbeing had no statistically significant relationship with voice,
which is contrary to researchers’ studies of individuals’ state of being such as self‐
esteem in relation to voice (LePine & Van Dyne, 1998). None of the job stress or
psychological wellbeing variables had relationships to exit, neglect, or voice over
time.
The Relationship Among Exit, Neglect, & Voice over Two Time Points
Intention to leave and employee turnover have been extensively examined
as worker outcome variables in social work and human services research (Mor
Barak, Nissly et al., 2001). This present study examined a related variables (exit or
efforts toward quitting one’s job) alongside constructs that assessed employees’
level of involvement in work‐related tasks or organizational activities (neglect) and
efforts toward improving work conditions or promoting organizational change
(voice). In addition, the study looked at the relationships between all three
constructs (exit/neglect/voice) over two time‐points. Findings revealed that employee
exit responses are determined by neglect, but not voice responses. These are detailed
as follows:
♦ Not unexpectedly, a significant and positive relationship between neglect and
exit was found. Researchers generally agree that employees’ intention to leave

173
and actual turnover may be precipitated by one’s limited engagement in work‐
related tasks and organizational activities (Griffeth et al., 2000; Withey &
Cooper, 1989). Therefore, as employees become disengaged from work‐related
tasks or organizational activities (for example, by showing up late, missing
meetings, not putting forth full effort in their job), they are more inclined to
engage in cognitions and behaviors with regard to quitting their job. However,
in the present study this relationship did not hold up over time.
♦ Voice and neglect were also significantly and positively related in the cross‐
sectional analysis. That is, as employees made efforts to improve dissatisfying
workplace conditions or promote change, they were also more likely to
disengage from work‐related tasks and organizational activities. Upon
reexamination of the literature, the nature of this cross‐sectional relationship
may be, in part, due to employees’ reacting to undesirable experiences at work,
as originally conceptualized in the landmark studies of exit/neglect/voice (Farrell,
1983; Hirschman, 1970; Rusbult et al., 1988) (Farrell, 1983). Within the same time
frame, employees may engage in a variety of options to cope with the
presenting circumstances at work including both voice and neglect. Dalal’s (2005)
hedonism assumption supports this idea. Dalal posited that individuals engage
in specific behaviors to achieve a higher level of satisfaction or good mood.
Neglect has been certainly framed as “a set of behaviors that dissatisfied

174
individuals enact to avoid the work situation” (Hanisch & Hulin, 1990, p.63).
Voice as conceptualized and operationalized in the present study shares a similar
sentiment in which the construct has been partially defined as an employee
efforts to improve dissatisfying conditions (Dundon et al., 2004; Zhou & George,
2001).
♦ Over time, the relationship between voice and neglect changes. In the simple,
cross‐lagged model, employee voice at baseline was inversely related to neglect at
the six‐month follow‐up. Although the data do not demonstrate that voice efforts
at baseline predicts actual changes in neglect responses at the six‐month follow‐
up, the finding suggests that being able to voice may be sufficient to discourage
unfavorable work responses over time.
Indirect Effects
The present study aimed to explore both the direct and indirect relationships
within the comprehensive theory‐driven conceptual model. Cross‐sectional results
revealed an array of significant indirect relationships that contributed to employees’
perceived support, job stress, psychological wellbeing, and choice of response.
However, two statistically significant relationships among the model variables in
which full mediation occurred are discussed as key findings.
♦ Role conflict mediated the relationship between perceived supervisory support
and neglect. The nature of this relationship was as hypothesized. Employees who

175
experienced supervisory support had lower levels of role conflict, and lower
levels of role conflict were associated with lower levels of neglect.
♦ In the same manner, role conflict mediated the relationship between
organizational support and voice responses in that as employees experienced
support from the organization, their role conflict lowered. However, the more
role conflict was experienced, the greater the voice response. Therefore, the
existence of role conflict buffered the relationship between organizational
support and voice, but not in an expected manner.
6.3. STRENGTHS AND LIMITATIONS
Strengths
This study forges new ground in its focus on the relationships among
organizational factors and employeesʹ cognitive and behavioral responses to
undesirable work experiences in a human service setting. First, although
exit/neglect/voice models have been extensively examined in the management and
industrial/organizational psychology literature, this study is the first to consider
these outcomes variables as they pertain to human service workers. Likewise, the
study’s conceptual model draws from social work, organizational behavior, and
industrial psychology theories to offer a comprehensive multidisciplinary
perspective.

176
Second, unique combinations of organizational and individual variables
were examined in this study for their effects on the exit/neglect/voice responses over
time. Accordingly, this study was able to gain insight into the work climate, job
stress and psychological wellbeing factors in combination with diversity
characteristics that enhanced the likelihood of human service workers engaging in
behaviors that benefit the organization (i.e., voice) rather than possibly being a
detriment (i.e. exit or neglect). Third, the conceptual model was analyzed using both
cross‐sectional and longitudinal data enabling predictive factors for choice of
response to be identified. Finally, the findings from this study help provide insight
into the influence of neglect on exit as well as the connection between voice and
neglect over time. As such, this study is one of the few studies that have looked at
the relationships among exit, neglect, and voice over time.
Limitations
This study does have limitations that warrant consideration for interpreting
the findings. First, the data were collected using self‐reported measures at baseline
and at the six‐month follow‐up. Common method variance might occur in self‐
reported data because the same person serves as the source of their responses for all
the variables of interest (Doty & Glick, 1998). Consequently, any correlations many
be an artifact due to common method variance. To determine whether common
method variance is a validity threat, Harman’s one factor test was conducted to

177
assess whether one underlying factor accounted for the majority of the covariance in
the data (Podsakoff & Organ, 1986). Accordingly, all of the items in the key study
variables were entered into an unrotated principal components factor analysis.
Principal components analysis yielded 16 factors with eigenvalues greater than 1.0
with no single factor accounting for a majority of the covariance. Consequently,
common method variance did not emerge as a significant problem in the dataset.
Second, a limitation of the study relates to external validity due to the
availability or convenience sampling of 359 child welfare workers. The study
sample was over‐represented or under‐represented in certain categories as
compared to the study population from which the data were drawn. The sample
was significantly younger, had significantly more women and fewer African
Americans and more Asian/Pacific Islanders than the study population. The sample
was also more educated, included a higher proportion of line workers, and had
significantly shorter agency tenure than the population. While cross‐cutting
workplace issues exist among human service workers (The Annie E. Casey
Foundation, 2003), an expanded study that captures the perspectives of other types
of human service workers would be advantageous to understanding more about
exit/neglect/voice in human services. Hence, to increase the external validity of the
current study, a more representative sample is desirable for future studies.

178
Next, although the study uses cross‐sectional and longitudinal analyses to
assess the relationship among the variables of interest over time, much of the
supported hypotheses came from cross‐sectional data. Reliance on cross‐sectional
limits opportunities to infer statistical causality among antecedent and outcome
variables is problematic (Rubin & Babbie, 2001). A supplemental longitudinal study
that incorporates several different time points would be useful to test for additional
causality in the hypothesized model.
Fourth, the study findings are limited by the use of composite indicators
rather than latent variable constructs to measure inclusion in decision‐making,
perceived supervisory and organizational support, role conflict, role ambiguity,
psychological wellbeing, exit, neglect, and voice as in structural equation modeling.
Netemeyer, Johnston, and Burton (1990) caution against composite indicators
because the models do not account for random measurement error. Given that the
sample size was not sufficient enough to accommodate the use of multiple
indicators per construct, the use of composite indicators was acceptable.
Researchers suggest that 15 cases per observed variable be used in multiple
regression analysis as well as structural equation modeling. The present study did
not meet that criterion due to having a total of 54 observed predictors (7 observed
variables relating to diversity characteristics and 47 observed scale items, equaling 9
scaled‐ constructs) with 359 cases. Furthermore, the scale reliability scores were, by

179
and large, within acceptable ranges. Despite this limitation, path analysis using
composite indicators are commonly used in the research on organizational
literature.
The final set of study limitations are conceptually and methodologically
based. The hypothesized model captured only five diversity characteristics (gender,
ethnicity, educational level, job position, job tenure). Therefore, other personal or
individual characteristics such as personality, personal skills, and life experience, as
well as other non‐dominant or dominant group characteristics (e.g., parental status)
were not assessed. Researchers have cautioned against not considering these types
of characteristics in modeling individual attitudes, cognitions, and behaviors
(Jackson & Joshi, 2004). However, the current study did capture one measure of the
participants’ personal characteristics (psychological wellbeing) that was not work‐
related. The addition of this variable in the model helps to provide insight into
employees’ person‐in‐environment fit as well.
Last, the research study assessed a combination of cognitions and behaviors
as outcome variables. The exit/neglect/voice constructs conceptually and
operationally reflected employees’ cognitions about and behaviors toward quitting
their jobs, disengaging from work‐related and organizational activities, and seeking
to improve dissatisfying work conditions or promote organizational change. The
cognitions represented the employees’ thought process regarding their propensity

180
to engage in such behaviors (e.g., intending to quit, thinking about disengaging
from a work activities; thinking of a new change‐oriented idea). On the other hand,
the behavioral elements reflected the extent to which they actually engaged in these
exit, neglect, or voice responses (e.g., searching for a new job, showing up late, or
talking with a supervisor about a workplace concern). Many researchers on exit/
neglect/voice responses have classified these responses as behaviors; however, the
measurements of the responses are reflective of both cognitions and behaviors (see
for example, Hagedoorn et al., 1999; Turnley & Feldman, 1999; Withey & Cooper,
1989). Further development of the conceptual and operational definitions of the
exit/neglect/voice responses is needed to tease out these differences.
6.4. THEORETICAL IMPLICATIONS
Theoretically, this study’s findings suggest that inclusion should be
conceptually distinguished from supervisory and organizational support and that
role conflict is distinguishable from role ambiguity. Additionally, the constructs of
voice and neglect should be refined and reconceptualized in accordance with the
challenges and demands of a human service setting.
Inclusion vs. Support
Based on this study’s findings, researchers interested in understanding the
effects of work climate factors on employee outcomes should consider further
conceptually distinguishing inclusion from supervisory and organizational support.

181
Inclusion in decision‐making, reflects the extent to which employees feel actively
engaged in critical organizational processes (Mor Barak & Cherin, 1998). Perceived
support, on the other hand, reflects the extent to which employees feel valued and
cared about by their supervisor or organization (Eisenberger et al., 1986;
Eisenberger et al., 2002; Kottke & Sharafinski, 1988; Rhoades & Eisenberger, 2002).
Based on findings from the present study, perceptions of inclusion affected
employees’ efforts that reflect disengagement from work‐related tasks and
organizational activities (neglect) and engagement in improving dissatisfying work
conditions or promoting organizational change (voice), whereas perceived support
more distinctly influenced employees’ efforts toward quitting their jobs (exit).
Further, as employees’ felt included in decision‐making, the more they felt
supported by their supervisor or their work organization. Therefore, findings from
this study broaden the conceptual understanding of inclusion versus support and
the ways in which these constructs may foster or hinder employees’ favorable and
unfavorable exit, neglect, or voice responses.
Role Ambiguity vs. Role Conflict
As distinct forms of job stress, role conflict and role ambiguity had differing
effects on exit, neglect, and voice as well as dissimilar antecedents. Both role conflict
and role ambiguity contributed to the likelihood that employees will engage in
unfavorable behaviors like quitting their job and disengaging from work‐related

182
tasks and organizational activities. However, only role conflict inspired employees
to take steps toward improving dissatisfying conditions and promoting workplace
change.
Additionally, the antecedents to role conflict and role ambiguity differed.
The more employees’ felt included in decision‐making, the greater their s role
conflict; on the contrary, perceptions of support were correlated with lower levels of
both role conflict and role ambiguity. These findings suggest that researchers keep
these conceptually distinguished and not combine them to create a general role
stress construct. McGee, Ferguson, Jr., and Seers (1989) also recommended that
these constructs not be combined into a single measure of role stress.
Voice and Neglect in Human Services
Voice and neglect are complex constructs that comprise a variety of behaviors
(Farrell, 1983). These constructs have been particularly studied in business or
service (e.g, retail, sales) industries (Hagedoorn et al., 1999; Janssen, de Vries, &
Cozijnsen, 1998; Kidwell & Robie, 2003; Luchak, 2003; Rusbult et al., 1988; Rusbult
et al., 1982; Turnley & Feldman, 1999; Withey & Cooper, 1989), but not human or
social services. In many disciplines, scholars generally concurred that voice needs
conceptual clarification (Kowtha et al., n.d.). As measured in this study, voice
addressed employee efforts toward overcoming dissatisfactory work conditions or
promoting organizational change either through, for example, talking to a

183
supervisor or making a determined effort to implement change in the absence of a
particular dissatisfaction. Consequently, further researchers might do well to
examine the multidimensionality of voice as a construct; research on this specifically
singling out change‐related behaviors has started (Avery, 2003; Avery & Quiñones,
2004; Islam & Zyphur, 2005; Kowtha et al., n.d.; Morrison & Phelps, 1999; Sverke &
Goslinga, 2003). However, this research would benefit from further development
specifically geared toward the human and social services.
Similarly, the measure of job neglect as used in this study reflected a single
form of counterproductive work behaviors reflective of organizational concerns
within in a corporate setting (Hagedoorn et al., 1999; Kidwell & Robie, 2003).
However, job neglect might imply two different sets of activities in the human
services. On the one hand, neglect might have an impact on organizational process
through employees missing meetings or not putting full effort into their work. On
the other hand and more seriously, neglect might also reflect failing to attend to
client needs. The latter was not measured in this study, and future researchers
should consider distinguishing or identifying measures focusing on employee
neglect and client need as distinct from organizational processes.
Considering the demands and challenges of working in a human service
environment (e.g., high caseloads, reporting responsibilities, one‐on‐one client
interactions, accountability to governing bodies) conceptually teasing out how both

184
and voice and neglect responses work within the specific realm of human services
provides opportunities for researchers to probe deeper into how employers might
keep workers engaged in constructive work behaviors.
6.5. PRACTICAL IMPLICATIONS
At the level of practice, findings from the present study suggest that
perceptions of inclusion in decision‐making and support—supervisory or
organizational—generally have a differential impact on worker outcomes.
Moreover, job stress (i.e. role conflict and role ambiguity) differs from psychological
wellbeing in its effects on employees’ propensity to engage in exit, neglect, and voice.
Hence, these findings provide guidance to organizational leaders on how to target
interventions to retain employees and keep them fully engaged. Outlined below are
several recommendations based on the findings for those in social work or human
services.
Recommendation 1: Provide opportunities for inclusion in decision‐making
A major component of employee empowerment involves their inclusion and
participation in decision‐making (Laschinger et al., 1997; Parker & Price, 1994;
Zimmerman, 1995). While organizations might consider supporting employees to
prevent turnover and foster decreased neglect over time, they should also consider
involving employees in decision‐making. This may help empower employees to
engage in improving work conditions or promoting organizational change (voice) as

185
well as reducing their disengagement from work‐related tasks and organizational
activities (neglect). With that being said, empowering management practices involve
more than simply delegating tasks to employees (Weiss, 2001), letting employees do
their jobs as they see fit (Randolph, 2000), or asking for employees’ input (Carroll,
1994). Organizational leaders must restructure their policies and practices to
generate and maintain worker empowerment (D. E. Bowen & Lawler, 1995) and to
provide opportunities for meaningful input into decision‐making (Bednar, 2003).
For example, Bowen and Lawler (1995) found that implementing empowerment‐
based practices that increase worker autonomy in performing tasks, addressing
problems, and handling mistakes can, in turn, contribute to employees’
participation in decision‐making.
Recommendation 2: Distinguish between supervisory and organizational support
Perceived supervisory and organizational support reflect the extent to which
employees feel valued and cared about by organizational leaders (Eisenberger et al.,
2001; Eisenberger et al., 1986; Eisenberger et al., 2002; Hutchison, 1997; Kottke &
Sharafinski, 1988; Rhoades & Eisenberger, 2002). As such, employees’ experience of
support is vital when considering people’s desire to make a contribution in their
work and to be appreciated for their efforts (Carroll, 1994). Yet, the two types of
support may affect workers in different ways.

186
Supervisors play a key role as agents of the organization. More specifically,
supervisors symbolize the nature of the organizational culture as well as serve as an
intermediary between workers and administrators (ʺSupervision in child welfare,ʺ
2003). In this role, supervisors relay organizational policies and procedures,
promote the goals of the organization, and use organizational resources to help
support workers (Gustavsson & MacEachron, 2004). Furthermore, supervisors also
monitor and evaluate of the quality of services that their workers provide. Study
findings suggest that supervisory support helps to reduce stress (Kelloway & Day,
2005; Nissly et al., 2005), and helps keep employees from leaving the organization
(Acker, 2004; Hagedoorn et al., 1999; Nissly et al., 2005) or limiting disengagement
in work‐related tasks or organizational activities (Duffy et al., 2002; Hagedoorn et
al., 1999).
Employees may characterize the work organizations by its leaders
(managers and administrators) or the enacted policies and procedures. Hence, work
organizations can send their own distinct messages to employees regarding their
value and contribution to the organization. The findings from this study
demonstrated the importance of organizational support to employees regardless of
hierarchical position. Employees at all levels have a need to be valued and
appreciated by their work organization, not just the supervisor. This in turn, may
contribute to their employee outcomes, specifically employee retention.

187
Recommendation 3: Foster increased wellbeing
Findings from this study suggest that decreased role conflict and role
ambiguity as well as increased psychological wellbeing are critical to employee
functioning at their jobs. Therefore, organizational wellness efforts should be geared
to reducing role‐related stress and benefiting psychological wellbeing (i.e., mental
health status of their employees). Bednar (2003) stressed the importance of
nurturing role clarity in a human service environment in creating a positive work
climate. Mackie (2001) proposed that organizational interventions focusing on
individual behavior change are limited. In contrast, Mackie found that
organizational interventions are most effective and long lasting when employees
develop coping skills and increased their participation in decision‐making as a
result. Related to mental health and consistent with person‐in‐environment fit
(Feldman, 2003), work organizations should consider continuing to provide support
that aid employees in sorting through their work/life challenges (as in EAP
programs) that can indirectly affect their employees’ functioning on the job
(Stewart, Ricci, Chee, & Morganstein, 2003).
Recommendation 4: Use a range of employee assessment measures
Organizations should consider using a range of measures to understand
how workers are fairing. Based on the present study’s findings, exit, neglect, and
voice have differing antecedents and effects. From one perspective, assessments of

188
employees’ exit responses help organizations prevent turnover and retain
competent, well‐trained employees as desired. Assessment of neglect can help work
organizations gauge employee engagement in work‐related tasks and
organizational activities. An assessment of neglect may provide afford work
organizations opportunities to learn about the determinants of employee
participation and effort in their particular settings. The study of voice, in particular,
provides opportunities for work organizations to understand the antecedents and
outcomes of employee mechanisms to express concerns as well as processes geared
toward workplace innovation (Kowtha et al., n.d.). Each of these constructs—exit,
neglect, and voice—offer unique opportunities for human service organizations to
learn about how their workers are doing on the job. This may provide opportunities
for organizations to develop innovative policies and practices to retain employees
and keep them fully involved in work and organizational activities.
Recommendation 5: Encourage constructive voice & validation of change efforts
Employee engagement in voice primarily can serve two functions. Voice can
ease dissatisfaction in the workplace and also enhance organizational effectiveness
through worker innovation and feedback (M. Keeley & Graham, 1991; Morrison &
Phelps, 1999). In the human service environment, voice as a tool for increased
organizational effectiveness is particularly relevant. Hopkins (2002) stated human
and social service systems now rely on employees engaging in extra role activities

189
(e.g., taking on added responsibility, working longer hours) to meet the demands of
the human service environment. Therefore, organizations should consider
encouraging voice; while, at the same times providing opportunities for employees
to gain skills in receiving and giving feedback through voice efforts.
As a part of that process, Chia, Landau, and Ong (2000) suggested that
organizational leaders should consider training supervisors on soliciting input from
and providing validation to employees, even if not all ideas are incorporated. Part
of this effort may consist of opportunities to develop skills in being open and
receptive to change, even with the demands of their own job responsibilities as well
as the requirements of organizational governing bodies. Additionally, supervisors
may benefit from learning new skills that enable them to relay to employees a
rationale for why their ideas were not incorporated. The goal would be to let
employees know that their ideas were considered and to prevent tem from
withdrawing participation in the future.
Organizational leaders should also consider training employees on offering
constructive voice. As discussed in chapter 2, constructive voice differs from
complaining or speaking up with the goal of winning (Kowalski, 2002). Many
employees may feel that if they speak up, they are engaging in a change‐related
behavior. However, organizational leaders may not experience it that way and
associate the employee efforts as the irritability and accusations associated with

190
complaining. Ultimately, these types of interventions would cultivate an
organizational climate that encourages employee innovation and feedback,
meaningful dialogue, and participation in decision‐making (Berry, n.d.).
6.6. SUGGESTIONS FOR FUTURE RESEARCH
The present findings offer support to the role of diversity characteristics,
perceptions of work climate, and job stress and psychological wellbeing in
influencing workers’ exit, neglect, and voice responses to their work environment.
Several questions arising from the study findings would benefit from further
exploration. First, to enhance external validity, a more representative sample of
human service workers is needed. Researchers from other areas of human services
are encouraged to examine the exit, neglect, and voice as outcome variables. This has
the potential to provide valuable insight into differences among fields as well as
how clients may be indirectly affected by worker behaviors. A few key studies have
looked at the relationships between organizational context and client outcomes
(Glisson, 1994; Glisson & Hemmelgarn, 1998; Glisson & James, 1992; Glisson &
James, 2002; Yoo, 2002; Yoo & Brooks, 2005) Hopefully, researchers will continue to
take on this challenge and add exit, neglect, and voice as important outcome and
antecedent variables.
Second, to comprehensively gain perspective on the relationship among exit,
neglect, and voice, a longitudinal study that takes into consideration more than two

191
time points is needed. This study looked at antecedents of and relationships among
exit, neglect, and voice over a six‐month time period, with limited results. Perhaps, a
longer time frame would help address questions about the factors that contribute to
these efforts as well as the relationships among them.
Last, just as this present study recommended that human service
organizations use a range of measures to assess how employees are faring,
researchers are also encouraged to develop and test more valid measures of neglect
and voice. The goal would be to have specific measures of neglect and voice that are
amenable to organizational context within a human and social service setting.
6.7. CONCLUSIONS
Well‐trained and highly qualified workers are invaluable assets to human
service organizations. Yet, these workers face tremendous challenges in doing their
jobs. Further, human service organizations have had challenges retaining these
workers (Ellett & Ellett, 2004; Mor Barak, Nissly et al., 2001; The Annie E. Casey
Foundation, 2003; United States General Accounting Office [GAO], 2003). This
study sought to provide insight into the factors contributing not only to employees’
efforts toward quitting (exit), but the extent to which they disengage from work‐
related tasks and organizational activities (neglect) and attempt to improve
dissatisfying work conditions or promote organizational change (voice). Based on
these findings, researchers are encouraged to continue making conceptual

192
distinctions among support and inclusion as well as forms of job stress; further
explore the role of psychological wellbeing; and refined and reconceptualized the
constructs of voice and neglect in accordance with the challenges and demands of a
human service setting. Finally, through better understanding of organizational and
personal factors affecting workers’ response to undesirable experiences at work,
human service organizations may be better equipped to retain engaged employees
who are committed to proving optimal client outcomes.

193
Glossary
Cognitive and behavioral responses – Cognitions refer to individuals’ thought
processes, while behaviors involve individual actions and responses based one’s
their environment. Cognitions may affect specific behaviors, but also are
influenced by individual mood, attitudes, and prior experiences.
Direct service worker – human service workers who work directly with clients – those
children, families, communities who receive the organization’s services, as
opposed to those who lack client contact are not considered “direct service
workers” (i.e., managers, supervisors, or administrators).
Diversity characteristics – Characteristics that distinguish members of social groups
from one another such as gender, race/ethnicity, age, educational level, work
status, job tenure.
Dominant/non‐dominant group status – Dominant and non‐dominant group status
are reflective of societal norms based on ones socially defined position of power.
For example, in the United States, men, Whites, Christians, heterosexuals, etc.
are all normatively dominant group members based diversity characteristics.
Exit – Employee efforts toward quitting their jobs.
Inclusion in decision‐making – The extent to which employees perceive that they
experienced opportunities to participate in and influence decision‐making
within the work organization and their work groups.

194
Neglect – Employees’ efforts that reflect their disengagement from work‐related
tasks and organizational activities.
Perceived supervisory and organizational support – Employees’ global views about the
extent to which their supervisor or work organization, respectively, values their
contributions and cares about their wellbeing.
Psychological wellbeing – One’s general mental health status or psychological
standing specifically related to levels of happiness, anxiety, stress, or depression
Role ambiguity – The extent to which an employee is unclear about work‐related
expectations.
Role conflict – The extent to which expectation of one or more roles does not mirror
the reality of others’ perceptions of that role.
Voice – Employee efforts to improve dissatisfying conditions or promote
organizational change.

195
References

Abrams, D., Ando, K., & Hinkle, S. (1998). Psychological Attachment to the Group:
Cross‐Cultural Differences in Organizational Identification and Subjective
Norms as Predictors of Workersʹ Turnover Intentions. Personality and Social
Psychology Bulletin, 24(10), 1027.
Abu‐Bader, S. H. (2000). Work satisfaction, burnout, and turnover among social
workers in Israel: a causal diagram. International Journal of Social Welfare, 9(3),
191‐2000.
Acker, G. M. (2004). The Effect of Organizational Conditions (Role Conflict, Role
Ambiguity, Opportunities for Professional Development, and Social
Support) on Job Satisfaction and Intention to Leave Among Social Workers
in Mental Health Care. Community Mental Health Journal, 40(1), 65.
Alexander, J. A., Lichtenstein, R., Oh, H. J., & Ullman, E. (1998). A causal model of
voluntary turnover among nursing personnel in long‐term psychiatric
settings. Research in Nursing & Health, 21(5), 415‐427.
Allen, D. G., Shore, L. M., & Griffeth, R. W. (2003). The role of perceived
organizational support and supportive human resource practices in the
turnover process. Journal of Management, 29(1), 99.
Amason, P., Allen, M. W., & Holmes, S. A. (1999). Social support and acculturative
stress in the multicultural workplace. Journal of Applied Communication
Research, 27(4), 310‐334.
American Public Human Services Association [APHSA]. (2005). Report from 2004
Child Welfare Workforce Survey. Retrieved May, 2005, from
http://www.aphsa.org/Home/Doc/Workforce%20Report%202005.pdf
Anderson, C., & Berdahl, J. L. (2002). The experience of power: Examining the
effects of power on approach and inhibition tendencies. Journal of Personality
and Social Psychology, 83(6), 1362‐1377.
Arbunkle, J. L. (2005). Amos [Computer software] (Version 6.0): SPSS.

196
Aselage, J., & Eisenberger, R. (2003). Perceived organizational support and
psychological contracts: A theoretical integration. Journal of Organizational
Behavior, 24(5, Special Issue), 491‐509.
Avery, D. R. (2003). Personality as a predictor of the value of voice. Journal of
Psychology: Interdisciplinary and Applied, 137(5), 435‐446.
Avery, D. R., & Quiñones, M. A. (2004). Individual Differences and the Voice Effect:
The Moderating Role of Value of Voice. Group & Organization Management,
29(1), 106‐124.
Balfour, D. L., & Neff, D. M. (1993). Predicting and managing turnover in human
service agencies: A case study of an organization in crisis. Public Personnel
Management, 22(3), 473.
Bather, A., & Kelly, M. (2005, September). Whistleblowing: The Advantages of Self‐
Regulation. The University of Waikato Department of Accounting Working Paper
Series, 82, 1‐19.
Bednar, S. G. (2003). Elements of satisfying organizational climates in child welfare
agencies. Families in Society, 84(1), 7.
Bell, E., Edmondson, L. J., Meyerson, D., Nkomo, S., & Scully, M. A. (2003).
Interpreting silence and voice in the workplace: A conversation about
tempered radicalism among Black and White women researchers. Journal of
Applied Behavioral Science, 39(4), 381‐414.
Benkhoff, B. (1996). Catching up on competitors: how organizations can motivate
employees to work harder. The International Journal of Human Resource
Management, 7(3), 736‐752.
Bennett, R. J., & Robinson, S. L. (2000). Development of a measure of workplace
deviance. Journal of Applied Psychology, 85(3), 349‐360.
Bergin, A. E., & Garfield, S. L. (Eds.). (1994). Handbook of psychotherapy and behavior
change (4th ed.). New York: John Wiley & Sons.
Berry, B. (n.d.). Employee Voice in Corporate Governance: The Dynamics of
Organizational Culture and Employee Communication and Reporting Behaviors.
Retrieved July 31, 2006, from http://www.aepp.net/2003/EmployeeVoice.pdf

197
Biddle, B. J. (1986). Recent developments in role theory. Annual Review of Sociology,
12, 67‐92.
Blau, P. M. (1964). Exchange and power in social life. New York,: J. Wiley.
Block, P. (1991). The empowered manager: Positive political skills at work. San Francisco:
Jossey‐Bass Publishers.
Borman, W. C. (2004). The concept of organizational citizenship. Current Directions
in Psychological Science, 13(6), 238‐241.
Bowen, D. E., & Lawler, E. E. (1995). Empowering service employees. Sloan
Management Review, 36(4), 73.
Bowen, F., & Blackmon, K. (2003). Spirals of silence: The dynamic effects of diversity
on organizational voice. Journal of Management Studies, 40(6), 1393‐1417.
Boyd, A. (1997). Employee traps–corruption in the workplace. Management Review,
86(9).
Bronfenbrenner, U. (1979). The ecology of human development : experiments by nature
and design. Cambridge, Mass.: Harvard University Press.
Brooker, A.‐S., & Eakin, J. M. (2001). Gender, class, work‐related stress and health:
Toward a power‐centered approach. Journal of Community & Applied Social
Psychology. Special Issue: Power, control and health, 11(2), 97‐109.
Bussing, A., Bissels, T., Fuchs, V., & Perrar, K. M. (1999). A dynamic model of work
satisfaction: Qualitative approaches. Human Relations, 52(8), 999‐1028.
Byrne, B. M. (2001). Structural equation modeling with AMOS : basic concepts,
applications, and programming. Mahwah, N.J.: Lawrence Erlbaum Associates.
Cabana, S. (1995). Participative design works, partially participative doesn’t. Journal
for Quality and Participation, 18, 10‐19.
Calvert, L. M., & Ramsey, V. J. (1996). Speaking as female and White: A non‐
dominant/dominant group standpoint. Organization, 3(4), 468‐485.
Carmeli, A. (2005). The relationship between organizational culture and withdrawal
intentions and behavior. International Journal of Manpower, 26(2), 177‐195.

198
Carr, E. S. (2003). Rethinking Empowerment Theory Using a Feminist Lens: The
Importance of Process. Affilia, 18(1), 8‐20.
Carroll, A. (1994). Whatʹs Behind the ʺEʺ Word: Myths About Empowerment and Why
You Need It. Retrieved July 9, 2006, from
http://www.interactiondesign.com/downloads/empwrart.pdf
Chia, H.‐B., Landau, J., & Ong, J. S.‐W. (2000). Predictors of voice: The moderating
effect of the general economic climate. Research and Practice in Human
Resource Management, 8(2), 3‐28.
Clark, K., Peters, S. A., & Tomlinson, M. (2005). The determinants of lateness:
evidence from British workers. Scottish Journal of Political Economy, 52(2), 282‐
304.
Coffey, M., Dugdill, L., & Tattersall, A. (2004). Stress in Social Services: Mental Well‐
being, Constraints and Job Satisfaction. British Journal of Social Work, 34(5),
735‐746.
Conlee, M. C., & Tesser, A. (1973). The Effects of Recipient Desire to Hear on News
Transmission. Sociometry, 36(4), 588‐599.
Cook, K. S. (1977). Exchange and Power in Networks of Interorganizational
Relations. The Sociological Quarterly, 18, 62‐82.
Cortina, L. M., & Magley, V. J. (2003). Raising voice, risking retaliation: Events
following interpersonal mistreatment in the workplace. Journal of
Occupational Health Psychology, 8(4), 247‐265.
Cousins, C. (2004). Becoming a social work supervisor: A significant role transition.
Australian Social Work, 57(2), 175‐185.
Coyle‐Shapiro, J. A. M. (2002). A psychological contract perspective on
organizational citizenship behavior. Journal of Organizational Behavior, 23(8),
927‐946.
Coyle, D., Edwards, D., Hannigan, B., Fothergill, A., & Burnard, P. (2005). A
systematic review of stress among mental health social workers. International
Social Work, 48(2), 201‐211.
Creed, W. E. D. (2003). Voice lessons: Tempered radicalism and the use of voice and
silence. The Journal of Management Studies, 40(6), 1503.

199
Cropanzano, R., & Mitchell, M. S. (2005). Social Exchange Theory: An
Interdisciplinary Review. Journal of Management %R
10.1177/0149206305279602, 31(6), 874‐900.
Cummings, T. G. (1978). Self‐Regulating Work Groups: A Socio‐Technical Synthesis.
The Academy of Management Review, 3(3), 625‐634.
Daradkeh, T. K., Ghubash, R., & El‐Rufaie, O. E. F. (2001). Reliability, validity, and
factor structure of the Arabic version of the 12‐item General Health
Questionnaire. Psychological Reports, 89(1), 85‐94.
de Croon, E. M., Sluiter, J. K., Blonk, R. W., Broersen, J. P., & Frings‐Dresen, M. H.
(2004). Stressful work, psychological job strain, and turnover: a 2‐year
prospective cohort study of truck drivers. J Appl Psychol, 98(3), 442‐454.
DeVellis, R. F. (1991). Scale development theory and applications: Applied social research
methods (Vol. 26). Newbury Park, London: Sage Publications.
Dewettnick, K., Singh, J., & Buyens, D. (2003). Psychological Empowerment in the
Workplace: Reviewing the Empowerment Effects on Critical Work Outcomes.
Retrieved August, 2006, from http://ideas.repec.org/p/vlg/vlgwps/2003‐
29.html
Diamantopoulos, A., & Siguaw, J. (2000). Introducing Lisrel : a guide for the uninitiated.
London: SAGE.
Dickinson, N. S., & Perry, R. (1998). Why do MSWʹs Stay in Public Child Welfare?:
Organizational and Training Implication of a Retention Study. Paper presented at
the 11th National Conference of the National Staff Development and
Training Association, New Orleans, Louisiana.
Diener, E. (2000). Subjective well‐being: The science of happiness and a proposal for
a national index. American Psychologist, 55(1), 34‐43.
Dobreva‐Martinova, T., Villeneuve, M., Strickland, L., & Matheson, K. (2002).
Occupational role stress in the Canadian forces: Its association with
individual and organizational well‐being. Canadian Journal of Behavioural
Science, 34(2), 111‐121.
Doty, D. H., & Glick, W. H. (1998). Common methods bias: Does common methods
variance really bias results? Organizational Research Methods, 1(4), 374‐406.

200
Dowding, K., John, P., Mergoupis, T., & Vugt, M. V. (2000). Exit, voice, and loyalty:
analytic and empirical developments. European Journal of Political Research,
37(4), 469‐495(427).
Duffy, M. K., Ganster, D., & Pagon, M. (2002). Social undermining in the workplace.
Academy of Management Journal, 45(2), 331‐351.
Dundon, T., Wilkinson, A., Marchington, M., & Ackers, P. (2004). The meanings and
purpose of employee voice. International Journal of Human Resource
Management, 15(6), 1149‐1170.
Dutton, J. E., Ashford, S. J., OʹNeill, R. M., & Lawernce, K. A. (In Press). Moves that
matter: Issue selling and organizational change. Academy of Management
Journal.
Dutton, J. E., Ashford, S. J., OʹNeill, R. M., & Lawrence, K. A. (2001). Moves that
matter: Issue selling and organizational change. Academy of Management
Journal, 44(4), 716‐736.
Edmondson, A. C. (2003). Speaking Up in the Operating Room: How Team Leaders
Promote Learning in Interdisciplinary Action Teams., Journal of Management
Studies (Vol. 40, pp. 1419‐1452): Blackwell Publishing Limited.
Edmondson, A. C., & Detert, J. R. (2005). The role of speaking up in work‐life
balancing. In E. E. Kossek & S. J. Lambert (Eds.), Work and life integration :
organizational, cultural, and individual perspectives (pp. 401‐423). Mahwah, N.J.:
Lawrence Erlbaum Associates, Publishers.
Eisenberger, R., Armeli, S., Rexwinkel, B., Lynch, P. D., & Rhoades, L. (2001).
Reciprocation of perceived organizational support. Journal of Applied
Psychology, 86(1), 42‐51.
Eisenberger, R., Fasolo, P., & Davis‐LaMastro, V. (1990). Perceived organizational
support and employee diligence, commitment, and innovation. Journal of
Applied Psychology, 75(1), 51‐59.
Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived
organizational support. Journal of Applied Psychology, 71(3), 500‐507.
Eisenberger, R., Stinglhamber, F., Vandenberghe, C., Sucharski, I. L., & Rhoades, L.
(2002). Perceived supervisor support: Contributions to perceived

201
organizational support and employee retention. Journal of Applied Psychology,
87(3), 565‐573.
Eisler, R., & Levine, D. S. (2002). Nurture, nature, and caring: We are not prisoners
of our genes. Brain & Mind. Special Issue: Brain development and caring behavior,
3(1), 9‐52.
Ellett, A., & Ellett, C. D. (2004, February). Employee Retention in Child Welfare: A
Statewide Study with Implications for Higher Education. Paper presented at the
Council on Social Work Education Annual Program Meeting, Anaheim,
California.
Ellickson, M. C., & Logsdon, K. (2001). Determinants of Job Satisfaction of
Municipal
Government Employees. State and Local Government Review, 33(3), 173–184.
Ellinger, A., Watkins, K. E., & Bostrom, R. P. (1999). Managers as facilitators of
learning in learning organizations. Human Resource Development Quarterly,
10(2), 105‐eoa.
Emery, M. (1993). Participative Design for Participative Democracy: Canberra ACT:
Centre for Continuing Education, Australian National University.
Farrell, D. (1983). Exit, voice, loyalty, and neglect as responses to job dissatisfaction:
A multidimensional scaling study. Academy of Management Journal, 26(4), 596.
Farrell, D., & Rusbult, C. E. (1992). Exploring the exit, voice, loyalty, and neglect
typology: The influence of job satisfaction, quality of alternatives, and
investment size. Employee Responsibilities & Rights Journal. Special Research on
Hirschmanʹs Exit, Voice, and Loyalty model, 5(3), 201‐218.
Feldman, R. S. (2003). Development across the lifespan (3rd ed.). Upper Saddle River:
New Jersey: Prentice Hall.
Firth, L., Mellor, D. J., Moore, K. A., & Loquet, C. (2004). How can managers reduce
employee intention to quit? Journal of Managerial Psychology, 19(1/2), 170.
Foster‐Fishman, P. G., & Keys, C. B. (1997). The person/environment dynamics of
employee empowerment: An organizational culture analysis. American
Journal of Community Psychology, 25(3), 345‐369.

202
Fowler, F. J. (2002). Survey Research Methods (3rd ed.). Thousand Oaks, Calif.: Sage
Publications.
Frese, M., Kring, W., Soose, A., & Zempel, J. (1996). Personal Initiative at Work:
Differences between East and West Germany. The Academy of Management
Journal, 39(1), 37‐63.
Germain, C. B. (1978). Social work practice: People and environment. New York:
Columbia University.
Germain, C. B. (1979). Ecology and social work. In C. B. Germain (Ed.), Social work
practice: People and environments (pp. 1‐22). New York: Columbia University Press.
Glisson, C. (1994). The effect of services coordination teams on outcomes for
children in state custody. Administration in Social Work, 18(4), 1.
Glisson, C., & Durick, M. (1988). Predictors of job satisfaction and organizational
commitment in human service organizations. Administrative Science
Quarterly, 33, 61‐81.
Glisson, C., & Hemmelgarn, A. (1998). The effects of organizational climate and
interorganizational coordination on the quality and outcomes of childrenʹs
service systems. Child Abuse & Neglect, 22(5), 401.
Glisson, C., & James, L. (1992). The interorganizational coordination of services to
children in state custody. Administration In Social Work, 16(3,4), 65.
Glisson, C., & James, L. R. (2002). The cross‐level effects of culture and climate in
human service teams. Journal of Organizational Behavior, 23(6), 767.
Goldberg, D., McDowell, I., & Newell, C. (1996). General Health Questionnaire
[GHQ], 12 Item version, 20 item version, 30 item version, 60 item version
[GHQ 12, GHQ 20, GHQ 30, GHQ 60] (1972). In Measuring health: A guide to
rating scales and questionnaires (2nd ed., pp. 181, 225‐236). New York: Oxford
University Press.
Graham, J. W. (1986). Principled organizational dissent: A theoretical essay. In L. L.
Cummings (Ed.), Research in organizational behavior (Vol. 8, pp. 1‐52).
Greenwich, CT: JAI.

203
Greene, A. D., & Latting, J. K. (2004). Whistle‐Blowing as a Form of Advocacy:
Guidelines for the Practitioner and Organization. Social Work, 49(2), 219‐231.
Griffeth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta‐analysis of antecedents
and correlates of employee turnover: Update, moderator tests, and research
implications for the next millennium. Journal of Management, 26(3), 463‐488.
Griffin, M. A., Patterson, M. G., & West, M. A. (2001). Job satisfaction and
teamwork: The role of supervisor support. Journal of Organizational Behavior,
22(5), 537‐550.
Gruys, M. L., & Sackett, P. R. (2003). Investigating the dimensionality of
counterproductive work behavior. International Journal of Selection and
Assessment, 11(1), 30‐42.
Gustavsson, N., & MacEachron, A. E. (2004). When a Child Welfare Client Dies: An
Agency‐Centered Perspective. Child Welfare, 83(4), 317‐340.
Gutierrez, L. M. (1994). Beyond coping: An empowerment perspective on stressful
life events. Journal of sociology and social welfare, 21(3), 201‐219.
Hagedoorn, M., Van Yperen, N. W., Van de Vliert, E., & Buunk, B. P. (1999).
Employeesʹ reactions to problematic events: A circumplex structure of five
categories of responses, and the role of job satisfaction. Journal of
Organizational Behavior, 20(3), 309‐321.
Hanisch, K. A., & Hulin, C. L. (1990). Job Attitudes and Organizational Withdrawal:
An Examination of Retirement and Other Voluntary Withdrawal Behaviors.
Journal of Vocational Behavior, 37(1), 60‐78.
Hanisch, K. A., & Hulin, C. L. (1991). General attitudes and organizational
withdrawal: An evaluation of a causal model. Journal of Vocational Behavior,
39(1), 110‐128.
Hanisch, K. A., Hulin, C. L., & Roznowski, M. (1998). The importance of individualsʹ
repertoires of behaviors: The scientific appropriateness of studying multiple
behaviors and general attitudes. Journal of Organizational Behavior, 19(5), 463‐
480.
Herrenkohl, R. C., Judson, G. T., & Heffner, J. A. (1999). Defining and Measuring
Employee Empowerment. The Journal of Applied Behavioral Science, 35(3), 373‐
389.

204
Hess, P., Folaron, G. & Jefferson, A. (1992). Effectiveness of family reunification
services: An innovative evaluative model. Social Work, 97, 304‐311.
Hirschman, A. O. (1970). Exit, voice, and loyalty; responses to decline in firms,
organizations, and states. Cambridge, Mass.: Harvard University Press.
Honold, L. (1997). A review of the literature on employee empowerment.
Empowerment in Organizations, 5(4), 202‐212.
Hopkins, K. M. (2002). Organizational citizenship in social service agencies.
Administration in Social Work, 26(2), 1‐15.
Hopkins, K. M., & Hyde, C. (2002). The human service managerial dilemma: New
expectations, chronic challenges and old solutions. Administration in Social
Work, 26(3), 1‐15.
Hopkins, W. E., Hopkins, S. A., & Mallette, P. (2001). Diversity and managerial
value commitment: A test of some proposed relationships. Journal of
Managerial Issues, 13(3), 288.
House, R. H., & Rizzo, J. R. (1972). Role conflict and ambiguity as critical variables
in a model of organizational behavior.
Huang, L., Macbeth, G., Dodge, J., & Jacobstein, D. (2004). Transforming The
Workforce In Childrenʹs Mental Health. Administration and Policy in Mental
Health, 32(2), 167‐187.
Hutchison, S. (1997). A path model of perceived organizational support. Journal of
Social Behavior & Personality, 12(1), 159‐174.
Islam, G., & Zyphur, M. J. (2005). Power, Voice, and Hierarchy: Exploring the
Antecedents of Speaking Up in Groups. Group Dynamics: Theory, Research,
and Practice, 9(2), 93‐103.
Jacob, K. S., Bhugra, D., & Mann, A. H. (1997). The validation of the 12‐item General
Health Questionnaire among ethnic Indian women living in the United
Kingdom. Psychological Medicine, 27(5), 1215‐1217.
Janssen, O., de Vries, T., & Cozijnsen, A. J. (1998). Voicing by adapting and
innovating employees: An empirical study on how personality and
environment interact to affect voice behavior. Human Relations, 51(7), 945‐
967.

205
Johnson, P. R., & Indvik, J. (1997). The Boomer Blues: Depression in the Workplace.
Public Personnel Management, 26(3).
Jones, J. W. (1981). Dishonesty, Burnout, and Unauthorized Work Break Extensions.
Personality and Social Psychology Bulletin, 7(3), 406‐409.
Jones, J. W., & Boye, M. W. (1994). Job stress, predisposition to steal, and employee
theft. Am J Health Promot, 8(5), 331‐333.
Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek, J. D., & Rosenthal, R. A. (1964).
Occupational Stress: Studies in Role Conflict and Ambiguity: New York:
Wiley.
Keeley, M., & Graham, J. W. (1991). Exit, voice, and ethics. Journal of Business Ethics,
10(5), 349.
Keeley, M. G., Jill W. (1991). Exit, Voice, and Ethics. Journal of Business Ethics, 10(5),
349.
Kelloway, E. K., & Day, A. L. (2005). Building Healthy Workplaces: What We Know
So Far. Canadian Journal of Behavioural Science, 37(4), 223‐235.
Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and
inhibition. Psychological Review, 110(2), 265‐284.
Kidwell Jr., R. E., & Robie, C. (2003). Withholding Effort in Organizations: Toward
Development and Validation of a Measure. Journal of Business and Psychology,
17(4), 537.
Kidwell, R. E., & Bennett, N. (1993). Employee Propensity to Withhold Effort: A
Conceptual Model to Intersect Three Avenues of Research. The Academy of
Management Review, 18(3), 429‐456.
Kidwell, R. E., & Robie, C. (2003). Withholding effort in organizations: Toward
development and validation of a measure. Journal of Business and Psychology,
17(4), 537‐561.
Kline, R. B. (1998). Principles and Practice of Structural Equation Modeling. New York:
The Guilford Press.

206
Kottke, J. L., & Sharafinski, C. E. (1988). Measuring perceived supervisory and
organizational support. Educational & Psychological Measurement, 48(4), 1075‐
1079.
Kowalski, R. M. (2002). Whining, griping, and complaining: positivity in the
negativity. J Clin Psychol, 58(9), 1023‐1035.
Kowtha, N. R., Landau, J., & Beng, C. H. (n.d.). The Culture of Voice: Exploring the
relationship between employee voice and organizational culture. Retrieved
January, 2006, from http://www.fba.nus.edu.sg/ba/mscphd/0105.pdf
Krausz, M., Koslowsky, M., Shalom, N., & Elyakim, N. (1995). Predictors of
intentions to leave the ward, the hospital, and the nursing profession: A
longitudinal study. Journal of Organizational Behavior, 16(3), 277.
Ladebo, O. J. (2005). Relationship between Citizenship Behaviors and Tendencies to
Withdraw among Nigerian Agribusiness Employees. Swiss Journal of
Psychology, 64(1), 41‐50.
Lait, J., & Wallace, J. E. (2002). Stress at work: A study of organizational‐professional
conflict and unmet expectations. Relations Industrielles, 57(3), 463.
Landsman, M. (2001). Commitment in public child welfare. Social Service Review,
386‐419.
Laschinger, H. K. (1996). A theoretical approach to studying work empowerment in
nursing: a review of studies testing Kanterʹs theory of structural power in
organizations. Nurs Adm Q, 20(2), 25‐41.
Laschinger, H. K., & Havens, D. S. (1996). Staff nurse work empowerment and
perceived control over nursing practice: Conditions for work effectiveness.
Journal of Nursing Administration, 26(9), 27‐35.
Laschinger, H. K., Sabiston, J. A., & Kutszcher, L. (1997). Empowerment and staff
nurse decision involvement in nursing work environments: testing Kanterʹs
theory of structural power in organizations. Research in Nursing and Health,
20, 341‐352.
Laschinger, H. K., Wong, C., McMahon, L., & Kaufmann, C. (1999). Leader behavior
impact on staff nurse empowerment, job tension, and work effectiveness.
JONA: The Journal of Nursing Administration, 29(5), 28‐39.

207
Latting, J. K., & Blanchard, A. (1997). Empowering Staff in aʺ Poverty Agencyʺ: An
Organization Development Intervention. Journal of Community Practice, 4, 59‐
76.
Lau, V. C. S., Au, W. T., & Ho, J. M. C. (2003). A qualitative and quantitative review
of antecedents of counterproductive behavior in organizations. Journal of
Business and Psychology, 18(1), 73‐99.
Lee, M., & Koh, J. (2001). Is empowerment really a new concept? International Journal
of Human Resource Management, 12(4), 684‐695.
LePine, J. A., & Van Dyne, L. (1998). Predicting voice behavior in work groups.
Journal of Applied Psychology, 83(6), 853‐868.
LePine, J. A., & Van Dyne, L. (2001). Voice and cooperative behavior as contrasting
forms of contextual performance: Evidence of differential relationships with
Big Five personality characteristics and cognitive ability. Journal of Applied
Psychology, 86(2), 326‐336.
Lerner, R. M. (2002). Concepts and theories of human development (3rd ed.). Mahwah,
N.J.: L. Erlbaum Associates.
Light, P. C. (2003). The Health of the Human Services Workforce: Center for Public
Service, The Brookings Institution & Wagner School of Public Service, New
York University.
Loehlin, J. C. (2004). Latent variable models : an introduction to factor, path, and
structural equation analysis (4th ed.). Mahwah, N.J.: L. Erlbaum Associates.
Luchak, A. A. (2003). What kind of voice do loyal employees use? British Journal of
Industrial Relations, 41(1), 115‐134.
Lundberg, U. (2000). Workplace stress. In G. Fink (Ed.), Encyclopedia of Stress (Vol. 3,
pp. 684‐692). San Diego: Academic Press.
Lyubomirsky, S., King, L., & Diener, E. (2005). The Benefits of Frequent Positive
Affect: Does Happiness Lead to Success? Psychological Bulletin, 131(6), 803‐
855.
Mackie, K. S., Holahan, C. K., & Gottlieb, N. H. (2001). Employee involvement
management practices, work stress, and depression in employees of a
human services residential care facility. Human Relations, 54(8), 1065‐1092.

208
Marcus, B., & Schuler, H. (2004). Antecedents of Counterproductive Behavior at
Work: A General Perspective. Journal of Applied Psychology, 89(4), 647‐660.
Martinko, M. J., Gundlach, M. J., & Douglas, S. C. (2002). Toward an Integrative
Theory of Counterproductive Workplace Behavior: A Causal Reasoning
Perspective. International Journal of Selection and Assessment, 10(1&2), 36‐50.
Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job Burnout. Annual Review of
Psychology, 52(1), 397‐422.
Mayes, B. T., & Ganster, D. C. (1988). Exit and Voice: A Test of Hypotheses Based on
Fight/Flight Responses to Job Stress. Journal of Organizational Behavior, 9(3),
199.
McGee, G. W., Ferguson, C., & Seers, A. (1989). Role conflict and role ambiguity: do
the scales measure these two constructs? Journal of applied psychology, 74(5),
815‐818.
Meyerson, D. E. (2001). Tempered radicals : how people use difference to inspire change at
work. Boston: Harvard Business School Press.
Meyerson, D. E., & Scully, M. A. (1995). Tempered radicalism and the politics of
ambivalence and change. Organization Science, 6(5), 585‐600.
Michaels, C. E., & Spector, P. E. (1982). Causes of Employee Turnover: A Test of the
Mobley, Griffeth, Hand, and Meglino Model. Journal of Applied Psychology,
67(1), 53.
Michie, S., & Williams, S. (2003). Reducing work related psychological ill health and
sickness absence: a systematic literature review. Occup Environ Med, 60(1), 3‐
9.
Milliken, F. J., & Martins, L. L. (1996). Searching for common threads:
Understanding the multiple effects of diversity in organizational groups.
Academy of Management. The Academy of Management Review, 21(2), 402.
Milliken, F. J., & Morrison, E. W. (2003). Shades of silence: Emerging themes and
future directions for research on silence in organizations. Journal of
Management Studies. Special SPEAKING UP, REMAINING SILENT: THE
DYNAMICS OF VOICE AND SILENCE IN ORGANIZATIONS, 40(6), 1563‐
1568.

209
Milliken, F. J., Morrison, E. W., & Hewlin, P. F. (2003). An exploratory study of
employee silence: Issues that employees donʹt communicate upward and
why. Journal of Management Studies, 40(6), 1453‐1476.
Moorman, R. H., & Blakely, G. L. (1995). Individualism‐collectivism as an individual
difference predictor of organizational citizenship behavior. Journal of
Organizational Behavior, 16(2), 127‐142.
Mor Barak, M. E. (2005). Managing Diversity: Toward a Globally Inclusive Workplace.
Thousand Oaks, Calif.: Sage Publications.
Mor Barak, M. E., & Cherin, D. A. (1998). A tool to expand organizational
understanding of workforce diversity. Administration in Social Work, 22(1),
47‐65.
Mor Barak, M. E., Cherin, D. A., & Berkman, S. (1998). Organizational and personal
dimensions in diversity climate: Ethnic and gender differences in employee
perceptions. The Journal of Applied Behavioral Science, 34(1), 82.
Mor Barak, M. E., Findler, L., & Wind, L. H. (2001). Diversity, inclusion, and
commitment in organizations: International explorations. Journal of
Behavioral and Applied Management, 2(2), 72‐91.
Mor Barak, M. E., Findler, L., & Wind, L. H. (2003). Cross‐Cultural Aspects of
Diversity and Well‐Being in the Workplace: An International Perspective.
Journal of Social Work Research & Evaluation, 4(2), 145‐169.
Mor Barak, M. E., & Levin, A. (2002). Outside of the corporate mainstream and
excluded from the work community: A study of diversity, job satisfaction
and well‐being. Community, Work & Family, 5(2), 133‐157.
Mor Barak, M. E., Levin, A., Nissly, J. A., & Lane, C. J. (2006). Why do they leave?
Modeling child welfare workersʹ turnover intentions. Children and Youth
Services Review, 28(5), 548‐577.
Mor Barak, M. E., Nissly, J. A., & Levin, A. (2001). Antecedents to retention and
turnover among child welfare, social work, and other human service
employees: What can we learn from past research? A review and
metanalysis. Social Service Review, 625‐661.

210
Morris, A., Bloom, J., & Kang, S. (2002). Organizational and Individual Factors Affecting
Consumer Outcomes of Care in Mental Health Services. Paper presented at the
Health Care Organizations Conference, Berkeley, California.
Morrison, E. W., & Milliken, F. J. (2000). Organizational silence: A barrier to change
and development in a pluralistic world. Academy of Management. The
Academy of Management Review, 25(4), 706.
Morrison, E. W., & Milliken, F. J. (2003). Speaking up, remaining silent: The
dynamics of voice and silence in organizations. Journal of Management
Studies, 40(6), 1353‐1358.
Morrison, E. W., & Phelps, C. C. (1999). Taking charge at work: Extrarole efforts to
initiate workplace change. Academy of Management Journal, 42(4), 403.
Morrison, E. W., & Robinson, S. L. (1997). When employees feel betrayed: A model
of how psychological contract violation develops. Academy of Management
Review, 22(1), 226‐256.
National Association of Social Workers. (2002). The Power of Social Work. Retrieved
August 20, 2006, from http://www.naswdc.org/nasw/nasw.pdf
National Association of Social Workers. (n.d.). The Power of Social Work. Retrieved
August 20, 2006, from http://www.naswdc.org/nasw/nasw.pdf
National Institute for Occupational Safety and Health (NIOSH). (n.d.). Stress at
Work. Retrieved June 12, 2006, from Retrieved June 12, 2006, from
http://www.cdc.gov/niosh/stresswk.html
National Organization for Human Services & Council for Standards in Human
Services Education. (2006). The Human Service Worker: A Generic Job
Description. Retrieved July 24, 2006, from
http://www.nationalhumanservices.org/hsworker.html
Near, J. P., & Miceli, M. P. (1996). Whistle‐Blowing: Myth and Reality. Journal of
Management, 22(3), 507‐526.
Netemeyer, R. G., Johnston, M. W., & Burton, S. (1990). Analysis of role conflict and
role ambiguity in a structural equations framework. Journal of Applied
Psychology, 75(2), 148‐157.

211
Nissly, J. A., Mor Barak, M. E., & Levin, A. (2005). Stress, social support, and
workersʹ intention to leave. Administration in Social Work, 29(1), 79‐100.
Noelle‐Neumann, E. (1984). The spiral of silence : public opinion, our social skin.
Chicago: University of Chicago Press.
Noelle‐Neumann, E. (1991). The theory of public opinion: The concept of the Spiral
of Silence. In J. A. Anderson (Ed.), Communication Yearbook (pp. 14, 256‐287).
Newbury Park, CA: Sage.
Olson‐Buchanan, J. B. (1996). Voicing discontent: What happens to the grievance
filer after the grievance? Journal of Applied Psychology, 81(1), 52‐63.
Organ, D. W. (1997). Organizational Citizenship Behavior: Itʹs Construct Clean‐Up
Time. Human Performance, 10(2), 85‐98.
Organ, D. W., & Greene, C. N. (1974). Role ambiguity, locus of control and work
satisfaction. Journal of Applied Psychology, 59(1), 101.
Paré, G., & Tremblay, M. (2004). The Influence of High‐involvement Human Resources
Practices, Procedural Justice, Organizational Commitment and Citizenship
Behaviors on Information Technology Professionalsʹ Turnover Intentions: HEC
Montral, Groupe de recherche en systmes dʹinformation.
Parker, L. E., & Price, R. H. (1994). Empowered managers and empowered workers:
The effects of managerial support and managerial perceived control on
workersʹ sense of control over decision making. Human Relations, 47(8), 911‐
928.
Pearlmutter, S., & Bartle, E. E. (2000). Supporting the move form welfare to work:
What women say. Affilia, 15(2), 153‐172.
Perkins, D. D., & Zimmerman, M. A. (1995). Empowerment theory, research, and
application. American Journal of Community Psychology, 23(5).
Perlow, L., & Williams, S. (2003). Is Silence Killing Your Company? Harvard Business
Review, 81(5), 52‐58.
Peterson, D. K. (2002). Deviant Workplace Behavior and the Organizationʹs Ethical
Climate. Journal of Business and Psychology, 17(1), 47‐61.

212
Peterson, M. F., Smith, P. B., Akande, A., Ayestaran, S., & et al. (1995). Role conflict,
ambiguity, and overload: A 21‐nation study. Academy of Management Journal,
38(2), 429‐452.
Peterson, N. A., & Zimmerman, M. A. (2004). Beyond the individual: Toward a
nomological network of organizational empowerment. American Journal of
Community Psychology, 34(1‐2), 129‐145.
Piderit, S. K. (2000). Rethinking resistance and recognizing ambivalence: A
multidimensional view of attitudes toward an organizational change.
Academy of Management Review, 25(4), 783‐794.
Podsakoff, P. M., & Organ, D. W. (1986). Self‐Reports in Organizational Research:
Problems and Prospects. Journal of Management, 12(4), 531‐544.
Powell, M. J., & York, R. O. (1992). Turnover in county public welfare agencies.
Journal of Applied Social Sciences, 16(2), 111‐127.
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating
indirect effects in simple mediation models. Behavior Research Methods,
Instruments & Computers. Special Issue: Web‐based archive of norms, stimuli, and
data: Part 2, 36(4), 717‐731.
Premeaux, S. F. (2001). Breaking the silence: Toward and Understanding of Speaking up n
the Workplace.Unpublished manuscript.
Premeaux, S. F., & Bedeian, A. G. (2003). Breaking the Silence: The Moderating
Effects of Self‐Monitoring in Predicting Speaking Up in the Workplace*.
Journal of Management Studies (Vol. 40, pp. 1537‐1562): Blackwell Publishing
Limited.
Rhoades, L., & Eisenberger, R. (2001). Affective Commitment to the Organization:
The Contribution of Perceived Organizational Support. Journal of Applied
Psychology, 86(5), 825‐836.
Rhoades, L., & Eisenberger, R. (2002). Perceived organizational support: A review of
the literature. Journal of Applied Psychology, 87(4), 698‐714.
Rhoades, L., Eisenberger, R., & Armeli, S. (2001). Affective commitment to the
organization: The contribution of perceived organizational support. Journal
of Applied Psychology, 86(5), 825‐836.

213
Rissel, C. (1994). Empowerment: the holy grail of health promotion? Health
Promotion International, 9(1), 39.
Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970). Role conflict and ambiguity in
complex organizations. Administrative Science Quarterly, 15(2), 150‐163.
Robbins, T. L., & Fredendall, L. D. (1995). The Empowering Role Of Self‐Directed
Work Teams In The Quality Focused Organization. Organizational
Development Journal, 13, 33‐33.
Robinson, S. L., & Bennett, R. J. (1995). A Typology of Deviant Workplace Behaviors:
A Multidimensional Scaling Study. The Academy of Management Journal, 38(2),
555‐572.
Robinson, S. L., & OʹLeary‐Kelly, A. M. (1998). Monkey see, monkey do: The
influence of work groups on the antisocial behavior of employees. Academy
of Management Journal, 41(6), 658‐672.
Rosse, J. G., & Hulin, C. L. (1985). Adaptation to work: an analysis of employee
health, withdrawal and change. Organ Behav Hum Decis Process, 36(3), 324‐
347.
Rubin, A., & Babbie, E. (2001). Research Methods for Social Work. Belmont:
Wadsworth.
Rusbult, C. E., Farrell, D., Rogers, G., & Mainous, A. G. (1988). Impact of exchange
variables on exit, voice, loyalty, and neglect: An integrative model of
responses to declining job satisfaction. Academy of Management Journal, 31(3),
599‐627.
Rusbult, C. E., Zembrodt, I. M., & Gunn, L. K. (1982). Exit, voice, loyalty, and
neglect: Responses to dissatisfaction in romantic involvements. Journal of
Personality & Social Psychology, 43(6), 1230‐1242.
Saunders, D. M., Sheppard, B. H., Knight, V., & Roth, J. (1992). Employee voice to
supervisors. Employee Responsibilities and Rights Journal, 5(3), 241‐259.
Schmitz, N., Kruse, J., & Tress, W. (1999). Psychometric properties of the General
Health Questionnaire (GHQ‐12) in a German primary care sample. Acta
Psychiatrica Scandinavica, 100(6), 462‐468.

214
Schuler, R. S. (1980). A Role and Expectancy Perception Model of Participation in
Decision Making. The Academy of Management Journal, 23(2), 331‐340.
Somers, M. J., & Birnbaum, D. (2001). Racial differences in work attitudes: What you
see depends on what you study. Journal of Business and Psychology, 15(4), 579‐
591.
Sparks, K., Faragher, B., & Cooper, C. L. (2001). Well‐being and occupational health
in the 21st century workplace. Journal of Occupational and Organizational
Psychology, 74(4), 489‐509.
Spector, P. E. (1986). Perceived control by employees: A meta‐analysis of studies
concerning autonomy and participation at work. Human Relations, 39(11),
1005‐1016.
Spector, P. E., & Fox, S. (2005). The stressor‐emotional model of counterproductive
work behavior. In S. Fox & P. E. Spector (Eds.), Counterproductive Work
Behaviors. Washington, DC: American Psychological Association.
Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions,
measurement, and validation. Academy of Management Journal, 38(5), 1442‐
1465.
Spreitzer, G. M. (1996). Social Structural Characteristics of Psychological
Empowerment. The Academy of Management Journal, 39(2), 483‐504.
Spreitzer, G. M. (1997). Toward a common ground in defining empowerment.
Research in Organizational Change and Development, 10, 31–62.
Spreitzer, G. M., De Janasz, S. C., & Quinn, R. E. (1999). Empowered to lead: The
role of psychological empowerment in leadership. Journal of Organizational
Behavior, 20(4), 511‐526.
Spreitzer, G. M., & Doneson, D. (forthcoming). Musing on the past and future of
employee empowerment. In T. Cummings (Ed.), Handbook of Organizational
Development. Thousand Oaks: Sage.
Stamper, C. L., & Masterson, S. S. (2002). Insider or outsider? How employee
perceptions of insider status affect their work behavior. Journal of
Organizational Behavior, 23(8), 875‐894.

215
Statement of David S. Liederman, Executive Director, Child Welfare League of America,
Inc., House of Representatives One Hundred Third Congress, First Sess.
(1993).
Stewart, W. F., Ricci, J. A., Chee, E., & Morganstein, D. (2003). Lost productive work
time costs from health conditions in the United States: results from the
American Productivity Audit. J Occup Environ Med, 45(12), 1234‐1246.
Stone‐Romero, E. F., Stone, D. L., & Salas, E. (2003). The influence of culture on role
conceptions and role behavior in organisations. Applied Psychology: An
International Review, 52(3), 328‐362.
Supervision in child welfare. (2003). Childrenʹs Service Practice Notes for North
Carolinaʹs Child Welfare Social Workers, 9(1).
Sverke, M., & Goslinga, S. (2003). The consequences of job insecurity for employers
and unions: Exit, voice and loyalty. Economic and Industrial Democracy, 24(2),
241.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Boston,
MA: Allyn and Bacon.
Tait, R. J., French, D. J., & Hulse, G. K. (2003). Validity and psychometric properties
of the General Health Questionnaire‐12 in young Australian adolescents.
Australian & New Zealand Journal of Psychiatry, 37(3), 374‐381.
Tetrick, E. (1992). Mediating effect of perceived role stress: A confirmatory analysis.
Stress & well‐being at work: Assessments and interventions for occupational mental
health, 134–152.
The Annie E. Casey Foundation. (2003). The unsolved challenge of system reform: The
condition of the frontline human services workforce. Retrieved July 24, 2006, from
http://www.aecf.org/initiatives/hswi/report_rev.pdf
Todd, C. M., & Deery‐Schmitt, D. (1996). Factors affecting turnover among family
child care providers: A longitudinal study. Early Childhood Research
Quarterly, 11(3), 351‐376.
Turnley, W. H., & Feldman, D. C. (1999). The impact of psychological contract
violations on exit, voice, loyalty, and neglect. Human Relations, 52(7), 895‐922.

216
United States General Accounting Office [GAO]. (2003). HHS Could Play a Greater
Role in Helping Child Welfare Agencies Recruit and Retain Staff [GAO‐03‐357].
Retrieved May, 2005, from http://www.gao.gov/new.items/d03357.pdf
Valentine, S. R. (2001). A path analysis of gender, race, and job complexity as
determinants of intention to look for work. Employee Relations, 23(2), 130‐146.
Van Dyne, L., Ang, S., & Botero, I. C. (2003). Conceptualizing employee silence and
employee voice as multidimensional constructs. Journal of Management
Studies (Special Issue: Speaking Up, Remaining Silent: The Dynamics Of Voice
And Silence In Organizations), 40(6), 1359‐1392.
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 L. L. Cummings & B. M. Staw (Eds.), Research in Organizational
Behavior: An Annual Series of Analytical Essays and Critical Reviews (Vol. 17, pp.
215‐285). Greenwich, CT: JAI Press.
Van Dyne, L., & LePine, J. A. (1998). Helping and voice extra‐role behaviors:
Evidence of construct and predictive validity. Academy of Management
Journal, 41(1), 108‐119.
Vinokur‐Kaplan, D. (1991). Job satisfaction among social workers in public and
voluntary child welfare agencies. Child Welfare, 70, 80‐91.
Weiss, W. H. (2001). Building morale, motivating and empowering employees.
Supervision, 62(1), 3.
Wilkinson, A. (1998). Empowerment: theory and practice. Personnel Review, 27(1), 40.
Williams, E. S., Konrad, T. R., Scheckler, W. E., Pathman, D. R., & al, e. (2001).
Understanding physiciansʹ intentions to withdraw from practice: The role of
job satisfaction, job stress, mental and physical health. Health Care
Management Review, 26(1), 7.
Withey, M. J., & Cooper, W. H. (1989). Predicting exit, voice, loyalty, and neglect.
Administrative Science Quarterly, 34(4), 521‐539.
Witt, L. A., Andrews, M. C., & Kacmar, K. M. (2000). The role of participation in
decision‐making in the organizational politics‐job satisfaction relationship.
Human Relations, 53(3), 341‐358.

217
Woodside, A. G., Frey, L., & Daly, R. T. (1989). Linking service quality, customer
satisfaction, and behavioral intention. Journal of Health Care Marketing, 9(4), 5‐
17.
Wright, T. A., Bonett, D. G., & Sweeney, D. A. (1993). Mental health and work
performance: Results of a longitudinal field study. Journal of Occupational and
Organizational Psychology, 66(4), 277‐284.
Wuensch, K. L. (2005). Skewness, Kurtosis, and the Normal Curve. Retrieved January,
2006, from core.ecu.edu/psyc/wuenschk/docs30/Skew‐Kurt.doc
Yoo, J. (2002). The relationship between organizational variables and client
outcomes: A case study in child welfare. Administration in Social Work, 26(2),
39‐61.
Yoo, J., & Brooks, D. (2005). The Role of Organizational Variables in Predicting
Service Effectiveness: An Analysis of a Multilevel Model. Research on Social
Work Practice, 15(4), 267‐277.
Zhou, J., & George, J. M. (2001). When job dissatisfaction leads to creativity:
Encouraging the expression of voice. Academy of Management Journal, 44(4),
682.
Zimmerman, M. A. (1990). Taking aim on empowerment research: On the
distinction between individual and psychological conceptions. American
Journal of Community Psychology, 18(1), 169‐177.
Zimmerman, M. A. (1995). Psychological empowerment: Issues and illustrations.
American Journal of Community Psychology. Special Issue: Empowerment theory,
research, and application, 23(5), 581‐599.
Zlotnik, J. L., DePanfillis, D., Daining, C., & Lane, M. M. (2005). Factors Influencing
Retention of Child Welfare Staff: A Systematic Review of Research: Institute for
the Advancement of Social Work Research (in collaboration with University
of Maryland School of Social Work, Center for Families & Institute for
Human Services Policy).

218
Appendix A: Scales from Study Questionnaire
(Note: ‘R’ denotes reverse coded items)
INCLUSION IN DECISION‐MAKING
1 2 3 4 5 6
Strongly
Disagree
Disagree
Slightly
Disagree
Agree
Moderately
Agree
Strongly
Agree

I1. I have a say in the way my work group performed its tasks
I4. I am able to influence decisions that affected my organization
I7. My supervisor often asks for my opinion before making important decisions
I10. I am often invited to voice my opinion in meetings with management higher
than my immediate supervisor
I13. I am often asked to participate in activities not directly related to my job
function

(Mor Barak, 2005; Mor Barak & Cherin, 1998)

PERCEIVED SUPERVISORY SUPPORT
1 2 3 4 5 6
Strongly
Disagree
Disagree
Slightly
Disagree
Agree
Moderately
Agree
Strongly
Agree

PSS1. My supervisor values my contribution to its well‐being
PSS2. My supervisor fails to appreciate any extra effort from me (R)
PSS3. My supervisor would ignore any complaint from me (R)
PSS4. My supervisor really cares about my well‐being
PSS5. Even if I did the best job possible, my supervisor would fail to notice (R)
PSS6. My supervisor cares about my general satisfaction at work.
PSS7. My supervisor shows very little concern for me (R)
PSS8. My supervisor takes pride in my accomplishments at work.

(Eisenberger et al., 2002; Rhoades & Eisenberger, 2001)

219
Perceived Organizational Support

1 2 3 4 5 6
Strongly
Disagree
Disagree
Slightly
Disagree
Agree
Moderately
Agree
Strongly
Agree

POS1. This organization values my contribution to its well‐being
POS2. [This organization] fails to appreciate any extra effort from me (R)
POS3. [This organization] would ignore any complaint from me (R)
POS4. [This organization] really cares about my well‐being
POS5. Even if I did the best job possible, [this organization] would fail to notice (R)
POS6. [This organization] cares about my general satisfaction at work
POS7. [This organization] shows very little concern for me   (R)
POS8. [This organization] takes pride in my accomplishments at work

(Eisenberger et al., 1986; Rhoades, Eisenberger, & Armeli, 2001)

ROLE CONFLICT
1 2 3 4 5 6
Strongly
Disagree
Disagree
Slightly
Disagree
Agree
Moderately
Agree
Strongly
Agree

RC1. I have to do things that should be done differently under different
conditions
RC2. I receive an assignment without the manpower to complete it
RC3. I have to buck a rule or policy in order to carry out an assignment
RC4. I work with two or more groups who operate quite differently
RC5. I receive incompatible requests from two or more people
RC6. I do things that are apt to be accepted by one person and not by others
RC7. I receive an assignment without adequate resources and materials to execute
it.
RC8. I work on unnecessary things

(Rizzo et al., 1970)

220
ROLE AMBIGUITY
1 2 3 4 5 6
Strongly
Disagree
Disagree
Slightly
Disagree
Agree
Moderately
Agree
Strongly
Agree

RA1. I have clear planned goals and objectives for my job (R)
RA2. I know that I have divided my time properly (R)
RA3. I know what my responsibilities are (R)
RA4. I know exactly what is expected of me (R)
RA5. I feel certain about how much authority I have on the job (R)
RA6. Explanation is clear of what has to be done (R)

(Rizzo et al., 1970)

GENERAL HEALTH QUESTIONNAIRE
1 2 3 4
Much
Worse Than
Usual
Worse Than
Usual
Same As
Usual
Better Than
Usual

G1. Have you recently been able to concentrate on whatever you are doing?
G2. Have you recently lost much sleep over worry?
G3. Have you recently felt you were playing a useful part of things?
G4. Have you recently felt capable of making decisions about things?
G5. Have you recently fell constantly under strain?
G6. Have you recently felt you couldn’t overcome your difficulties?
G7. Have you recently been able to enjoy your normal day‐to‐day activities?
G8. Have you recently been able to face up to problems?
G9. Have you recently been feeling unhappy or depressed?
G10. Have you recently been losing confidence in yourself?
G11. Have you recently been thinking of yourself as a worthless person?
G12. Have you recently been feeling reasonably happy, all things considered?

(Goldberg et al., 1996)

221
EXIT/NEGLECT/VOICE
1 2 3 4 5 6
Strongly
Disagree
Disagree
Slightly
Disagree
Agree
Moderately
Agree
Strongly
Agree

Exit

EX1. I have recently spent time looking for another job
EX2. During the next year, I will probably look for a new job outside of this
organization
EX3. When working conditions here decline, I think a lot about quitting
EX4. I often think about quitting

Neglect

NG1. Sometimes when I don’t feel like working I will work slowly or make errors
NG8. Now and then I arrive at work later just because I really am not in the mood
for work that day
NG12. I try to keep out of sight of my supervisor so I can talk to co‐workers, take
breaks, or do other personal business (not work)
NG13. Now and then there are workdays where I just don’t put much effort into my
work
NG15. Sometimes I just don’t feel like working and I call in sick

Voice

VC3. When I think of an idea that will benefit my organization I make a
determined effort to implement it
VC4. I have at least once contacted an outside source (i.e., union, agency,
consultant) to get help in a changing working conditions here
VC5. I sometimes discuss problems at work with my employer
VC7. When things are seriously wrong and the organization won’t act. I am
willing to “blow the whistle”
VC9. I have made several attempts to change working conditions here

222
Appendix B: Supplemental Information—Variable
Transformation
As discussed in Chapter 4, Section 4.6, prior to the testing of the research
hypotheses, all interval/ratio variables were examined for normality. To alleviate
skewness of the distribution, square root transformations (SQRT) of moderately
variables were conducted as recommended by Tabachnick and Fidell (2001). For
positive skewed variables, the formula used was as follows:
NEWVARIABLE = SQRT(X)
For negative skewness, variable reflection was conducted then the square
root transformation was applied. To reflect the variables, a constant (K) was created
by using the largest score in the distribution and adding 1; then subtracting each
score from the constant produced a new variable. In this way, a variable with
negative skewness was converted to one with positive skewness prior to the
transformation. To ease in the interpretation, the transformed negatively skewed
were reflected after the transformation as well. The formula used is as follows:
NEWVARIABLE = SQRT(K‐X)
All variables except voice, showed improvement in skewness and kurtosis
after transformation based on statistical analyses and graphical examination of the
distribution with imposed normal curve. voice, on the other hand, was moderately
skewed prior to the transformation, then the variable became moderately negatively

223
skewed after the transformation, therefore, due to this occurrence, as Tabachnick
and Fidell (2001)voice remained in its original composition. Therefore, so that neglect
and exit remained aligned with voice for the longitudinal scale, items remained in
their original scale form.
TABLE 15. SKEWNESS TRANSFORMATION RESULTS
Prior to transformation Post Transformation
Variable at baseline
N Skewness Kurtosis Skewness Kurtosis

Statistic Std.
Error
Statistic Std.
Error
Statistic Std.
Error
Statistic Std.
Error
Inclusion in
decision‐making
339 0.045 0.132 ‐0.310 0.264 ‐‐ ‐‐ ‐‐ ‐‐
Perceived
supervisory support
338 ‐0.839 0.133 0.276 0.265 0.425 0.133 -0.599 0.265
Perceived
organizational
Support
353 0.196 0.130 ‐0.362 0.259 ‐‐ ‐‐ ‐‐ ‐‐
Role Conflict 347 ‐0.043 0.131 ‐0.582 0.261 ‐‐ ‐‐ ‐‐ ‐‐
Role Ambiguity 356 0.519 0.129 0.019 0.258 0.089 0.129 -0.418 0.258
GHQ 356 ‐0.755 0.129 1.512 0.258 0.300 0.129 1.192 0.258
Exit 355 0.560 0.129 ‐0.663 0.258 0.182 0.129 -1.056 0.258
Neglect 354 0.523 0.130 ‐0.068 0.259 0.109 0.130 -0.716 0.259
Voice 354 0.374 0.130 ‐0.253 0.259 -0.018 0.130 -0.283 0.259

224
Appendix C: Supplemental Path Analysis Results

Legend

Diversity characteristics:
Women Gender dummy variable (Women = 1)
Latino Ethnicity dummy variable (Latino = 1)
Afr Amer Ethnicity dummy variable (African American = 1)
Asian Ethnicity dummy variable (Asian = 1)
Grad deg Educational Level dummy variable (Graduate degree= 1; Bachelor’s = 0)
Direct Srv Wrk Job position dummy variable (Direct service worker = 1; Manager or
supervisor = 0)
Job tenure Job tenure (length of time employed at baseline)

Work climate variables:
Inclusion Inclusion in decision‐making
PSS Perceived supervisory support
POS Perceived organizational support

Job stress and psychological wellbeing:
RC Role conflict
RA Role ambiguity
PWB Psychological wellbeing

Cognitive and Behavioral Responses:
Exit(T1) exit at baseline
Neglect(T1) neglect at baseline
Voice(T1) voice at baseline
Exit(T2) exit at six‐month follow‐up
Neglect(T2) neglect at six‐month follow‐up
Voice(T2) voice at six‐month follow‐up

225
TABLE 16. PATH ANALYSIS RESULTS (PHASE I REFINED MODEL)
Phase I: Diversity Characteristics Work Climate
Outcome
Predictor Estimate S.E. p R
2
Inclusion <‐‐‐ Grad deg ‐0.409 ‐0.174 0.130 0.002
<‐‐‐ Direct Srv Wrk ‐0.637 ‐0.219 0.161 *** 0.056
PSS <‐‐‐ Latino ‐0.075 ‐0.103 0.037 0.04
<‐‐‐ Job tenure ‐0.056 ‐0.199 0.014 ***
<‐‐‐ Inclusion 0.119 0.396 0.015 ***
0.185
POS <‐‐‐ Grad deg ‐0.459 ‐0.197 0.105 ***
<‐‐‐ Job tenure ‐0.357 ‐0.383 0.042 ***
<‐‐‐ Inclusion 0.36 0.364 0.044 ***
0.348
***p < .001
Phase I: Model fit statistics
χ
2
= 18.397, df = 16, p = .301; NFI = .971, CFI = .996; RMSEA = .020; CN = 512; ECVI =.325

226
TABLE 17. PATH ANALYSIS RESULTS (PHASE II REFINED MODEL)
Phase II: Diversity Characteristics Work Climate   Wellbeing
Outcome Predictor Estimate β S.E. P R
2
Inclusion <‐‐‐ Direct Srv Wrk ‐0.487 ‐0.168 0.156 0.002 0.028
PSS <‐‐‐ Latino ‐0.076 ‐0.103 0.037 0.039
<‐‐‐ Job tenure ‐0.056 ‐0.2 0.014 ***
<‐‐‐ Inclusion 0.117 0.389 0.015 *** 0.181
POS <‐‐‐ Grad deg ‐0.466 ‐0.202 0.104 ***
<‐‐‐ Job tenure ‐0.36 ‐0.39 0.042 ***
<‐‐‐ Inclusion 0.362 0.368 0.044 *** 0.333
RC <‐‐‐ Afr Amer ‐0.406 ‐0.157 0.123 ***
<‐‐‐ Inclusion 0.168 0.177 0.051 0.001
<‐‐‐ PSS ‐0.652 ‐0.206 0.165 ***
<‐‐‐ POS ‐0.394 ‐0.408 0.049 *** 0.227
RA <‐‐‐ Women ‐0.079 ‐0.107 0.037 0.03
<‐‐‐ Direct Srv Wrk ‐0.082 ‐0.115 0.045 0.068
<‐‐‐ Job tenure ‐0.098 ‐0.425 0.015 ***
<‐‐‐ PSS ‐0.168 ‐0.207 0.041 ***
<‐‐‐ POS ‐0.061 ‐0.246 0.013 *** 0.165
PWB <‐‐‐ RA ‐0.077 ‐0.146 0.027 0.004
<‐‐‐ RC ‐0.033 ‐0.24 0.007 ***
<‐‐‐ Inclusion 0.019 0.143 0.007 0.006 0.114
***p < .001
Phase II: Model fit statistics
χ2 = 52.043; df = 38; p = .064; NFI = .938; CFI = .981; RMSEA = .032; CN = 368; ECVI = .514

227
TABLE 18. PATH ANALYSIS RESULTS (PHASE III REFINED MODEL)
PHASE III: DIVERSITY CHARACTERISTICS WORK CLIMATE   WELLBEING   E/N/V
Outcome Predictor Estimate β S.E. p R
2
Inclusion <‐‐‐ Direct Srv Wrk ‐0.489 ‐0.169 0.155 0.002 0.029
PSS <‐‐‐ Latino ‐0.075 ‐0.102 0.037 0.041
<‐‐‐ Job tenure ‐0.056 ‐0.199 0.014 ***
<‐‐‐ Inclusion 0.116 0.387 0.015 *** 0.179
POS <‐‐‐ Grad deg ‐0.467 ‐0.203 0.104 ***
<‐‐‐ Job tenure ‐0.36 ‐0.39 0.042 ***
<‐‐‐ Inclusion 0.363 0.369 0.044 *** 0.333
RC <‐‐‐ Afr Amer ‐0.412 ‐0.159 0.123 ***
<‐‐‐ Inclusion 0.167 0.175 0.051 0.001
<‐‐‐ PSS ‐0.648 ‐0.205 0.165 ***
<‐‐‐ POS ‐0.398 ‐0.411 0.049 *** 0.229
RA <‐‐‐ Women ‐0.079 ‐0.107 0.037 0.031
<‐‐‐ Job tenure ‐0.08 ‐0.349 0.013 ***
<‐‐‐ PSS ‐0.168 ‐0.207 0.041 ***
<‐‐‐ POS ‐0.059 ‐0.237 0.014 *** 0.155
PWB <‐‐‐ RA ‐0.077 ‐0.146 0.027 0.004
<‐‐‐ RC ‐0.033 ‐0.242 0.007 ***
<‐‐‐ Inclusion 0.019 0.144 0.007 0.006 0.115
exit(T1) <‐‐‐ Grad deg 0.279 0.094 0.144 0.052
<‐‐‐ Job tenure ‐0.173 ‐0.145 0.064 0.006
<‐‐‐ PSS ‐0.465 ‐0.11 0.209 0.026
<‐‐‐ POS ‐0.247 ‐0.191 0.072 ***
<‐‐‐ RC 0.213 0.159 0.07 0.002
<‐‐‐ RA 0.542 0.104 0.26 0.037
<‐‐‐ PWB ‐1.763 ‐0.18 0.478 ***
<‐‐‐ neglect(T1) 0.209 0.138 0.074 0.005 0.276
neglect(T1) <‐‐‐ Latino 0.3 0.146 0.1 0.003
<‐‐‐ Inclusion ‐0.153 ‐0.183 0.043 ***
<‐‐‐ RC 0.138 0.157 0.047 0.004
<‐‐‐ RA 0.356 0.104 0.171 0.037
<‐‐‐ PWB ‐1.109 ‐0.172 0.331 ***
<‐‐‐ voice(T1) 0.138 0.152 0.049 0.005 0.188
voice(T1) <‐‐‐ Women ‐0.247 ‐0.089 0.128 0.053
<‐‐‐ Afr Amer ‐0.247 ‐0.098 0.122 0.043
<‐‐‐ Latino ‐0.28 ‐0.124 0.11 0.011
<‐‐‐ Job tenure 0.259 0.299 0.041 ***
<‐‐‐ Inclusion 0.159 0.172 0.042 ***
<‐‐‐ RC 0.264 0.273 0.045 *** 0.297
***p < .001
Phase III: Model fit statistics
χ2 = 75.311, df = 61, p = .103; NFI = .936; CFI = .986; RMSEA = .026; CN = 382, ECVI = .719

228
TABLE 19. PATH ANALYSIS RESULTS (PHASE IV REFINED, CROSS‐LAGGED MODEL)
PHASE III: DIVERSITY CHARACTERISTICS WORK CLIMATE   WELLBEING   E/N/V
(T1 & T2)
Outcome Predictor Estimate β S.E. p R
2
Inclusion <‐‐‐ Direct Srv Wrk ‐0.513 ‐0.19 0.2 0.01 0.036
PSS <‐‐‐ Job tenure ‐0.05 ‐0.177 0.021 0.015
<‐‐‐ Inclusion 0.082 0.264 0.023 *** 0.09
POS <‐‐‐ Grad deg ‐0.522 ‐0.205 0.163 0.001
<‐‐‐ Job tenure ‐0.346 ‐0.385 0.058 ***
<‐‐‐ Inclusion 0.306 0.311 0.063 *** 0.286
RC <‐‐‐ Afr Amer ‐0.421 ‐0.155 0.179 0.018
<‐‐‐ Inclusion 0.183 0.187 0.069 0.008
<‐‐‐ PSS ‐0.686 ‐0.218 0.217 0.002
<‐‐‐ POS ‐0.378 ‐0.381 0.068 *** 0.226
RA <‐‐‐ Job tenure ‐0.055 ‐0.246 0.017 0.001
<‐‐‐ PSS ‐0.133 ‐0.168 0.058 0.022
<‐‐‐ POS ‐0.052 ‐0.208 0.019 0.007 0.089
PWB <‐‐‐ RC ‐0.031 ‐0.217 0.01 0.002
<‐‐‐ Inclusion 0.025 0.183 0.01 0.012 0.077
exit(T1) <‐‐‐ PSS ‐0.677 ‐0.162 0.28 0.015
<‐‐‐ POS ‐0.261 ‐0.197 0.088 0.003
<‐‐‐ PWB ‐2.699 ‐0.289 0.612 ***
<‐‐‐ neglect(T1) 0.268 0.173 0.102 0.009 0.235
neglect(T1) <‐‐‐ Latino 0.301 0.145 0.146 0.039
<‐‐‐ Inclusion ‐0.144 ‐0.171 0.061 0.018
<‐‐‐ RC 0.21 0.245 0.062 ***
<‐‐‐ voice(T1) 0.124 0.144 0.065 0.056 0.129
voice(T1) <‐‐‐ Latino ‐0.354 ‐0.147 0.146 0.016
<‐‐‐ Job tenure 0.374 0.423 0.056 ***
<‐‐‐ Inclusion 0.185 0.19 0.06 0.002
<‐‐‐ RC 0.171 0.172 0.062 0.006 0.327
exit(T2) <‐‐‐ Direct Srv Wrk ‐0.602 ‐0.153 0.292 0.039
<‐‐‐ Job tenure ‐0.323 ‐0.242 0.099 0.001
<‐‐‐ RC ‐0.226 ‐0.151 0.091 0.013
<‐‐‐ exit(T1) 0.707 0.63 0.066 *** 0.397
neglect(T2) <‐‐‐ PSS ‐0.421 ‐0.151 0.175 0.016
<‐‐‐ neglect(T1) 0.538 0.519 0.064 ***
<‐‐‐ voice(T2) 0.046 0.048 0.059 0.434 0.315
voice(T2) <‐‐‐ Women 0.293 0.112 0.142 0.039
<‐‐‐ Inclusion 0.066 0.073 0.052 0.199
<‐‐‐ voice(T1) 0.629 0.671 0.053 *** 0.473
***p < .001
Phase IIII: Model fit statistics
χ
2
= 58.402, df = 61, p = .571; NFI = .934; CFI = 1.00 RMSEA = .000; CN = 255; ECVI = 1.916

229
TABLE 20. POST‐HOC PATH ANALYSIS RESULTS: NEGLECT & VOICE CROSS‐LAGGED MODEL
Post–hoc analysis: Simple, Cross‐lagged Neglect/Voice Model
Outcome Predictor Estimate β S.E. P R
2
neglect(T1) <‐‐‐ Latino 0.286 0.138 0.152 0.06
<‐‐‐ voice(T1) 0.15 0.175 0.063 0.017 0.041
voice(T1) <‐‐‐ Women ‐0.168 ‐0.06 0.181 0.353
<‐‐‐ Latino ‐0.35 ‐0.145 0.167 0.036
<‐‐‐ Afr Amer ‐0.165 ‐0.061 0.185 0.373
<‐‐‐ Asian ‐0.036 ‐0.011 0.211 0.865
<‐‐‐ Job tenure 0.425 0.481 0.057 *** 0.284
neglect(T2) <‐‐‐ Latino ‐0.117 ‐0.054 0.136 0.387
<‐‐‐ neglect(T1) 0.568 0.545 0.065 ***
<‐‐‐ voice(T1) ‐0.153 ‐0.17 0.076 0.043
<‐‐‐ voice(T2) 0.175 0.181 0.08 0.03 0.318
voice(T2) <‐‐‐ Women 0.305 0.117 0.144 0.033
<‐‐‐ Latino 0.032 0.014 0.135 0.815
<‐‐‐ Afr Amer ‐0.132 ‐0.053 0.147 0.367
<‐‐‐ Asian ‐0.06 ‐0.021 0.167 0.718
<‐‐‐ Job tenure 0.098 0.119 0.052 0.059
<‐‐‐ neglect(T1) 0.075 0.069 0.059 0.203
<‐‐‐ voice(T1) 0.579 0.622 0.059 *** 0.484
***p < .001
Phase IIII: Model fit statistics
χ
2
= 4.409; df = 8; p = .551; NFI = 1; CFI = 1.0; RMSEA = .000; CN = 651; ECVI (N/A)

Abstract (if available)

Abstract Well-trained and highly qualified workers are invaluable assets to human service organizations

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Examining the impact of job burnout on the well-being of human service workers

PDF

A longitudinal study of job resources, burnout, and turnover among social workers: implications for evidence-based organizational practices

PDF

Broken promises: gaining an understanding of policy-practice and means-ends decoupling within diversity and inclusion

PDF

Changing workplace and inequality in work-family policies

PDF

Social Determinants of Health: working with social workers and social work managers to build capacity to screen and refer in the medical setting

PDF

Post-investigation experiences of families with complex needs at risk of maltreatment: an examination of need, matched services, and changes in need

PDF

Social norms intervention to promote help-seeking with depressed Asian and European Americans: A pilot study

PDF

Transformational leadership theory aligned-practices and social workers' well-being: an exploratory study of leadership practices in the context of stress and job burnout

PDF

Intimate partner violence in the Black community: empowering Black families through education and engagement to increase survivor safety via a strengthened social support network

PDF

Institute on social practice integration research and education (INSPIRE): a workforce development solution to close the health gap

PDF

Working-class social identity and sense of belonging in higher education: a mixed-methods study

PDF

Examining the longitudinal influence of the physical and social environments on social isolation and cognitive health: contextualizing the role of technology

PDF

Exploring the social determinants of health in a population with similar access to healthcare: experiences from United States active-duty army wives

PDF

A paradigm shift approach to change theory: a field study of a selected leadership development and communication training program (volumes I & II)

PDF

Networking in the age of virtual work: women’s experiences and strategies for success

PDF

A multiple comparison case study of Los Angeles Public High Schools: LGBTQA+ policies and facilities, student advocacy, and change in policies and facilities over time