University of Southern California Dissertations and Theses

Racial/ethnic diversity and disparities in the child welfare system: a multilevel perspective

(USC Thesis Other)

Racial/ethnic diversity and disparities in the child welfare system: a multilevel perspective

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content RACIAL/ETHNIC DIVERSITY AND DISPARITIES
IN THE CHILD WELFARE SYSTEM: A MULTILEVEL PERSPECTIVE

by

Ahraemi Kim

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

August 2013

Copyright 2013 Ahraemi Kim

ii
DEDICATION
This dissertation is dedicated to my family members who have provided me with
unconditional love and personal sacrifices: my parents, Seedong Kim and Myungsuk Seo;
my husband, Hyeongwoo Kim; and my daughter, Hailey Daon Kim.

iii
ACKNOWLEDGEMENTS
I would like to express my sincere appreciation to Dr. Michàlle E. Mor Barak,
my mentor, for her extraordinary guidance, support, understanding and patience
throughout the doctoral program at the University of Southern California School of
Social Work. By providing me with a range of exceptional opportunities, she encouraged
me to grow not only as a doctoral student but also as an instructor and an independent
researcher. I am also very grateful to my dissertation committee, Erick G. Guerrero and
Peter J. Robertson, for their guidance and support.
I would like to thank the USC School of Social Work for their profound support
throughout the doctoral program. Especially, I would like to thank Ms. Malinda Sampson,
a doctoral program manager, for everything she did for me. I would like to thank the
great efforts of the research team for the National Survey of Child and Adolescent
Wellbeing. Their efforts and their data have made this dissertation possible.
In addition, I wish to thank my friends and colleagues in the doctoral program for
their incredible support and encouragement: Minah Kim, Erica Lizano, Min-Kyoung
Rhee, Hsin-Yi Hsiao, Juye Ji, Brooklyn Levin, Jina Sang, Jaehee Yi, Kyounghae Lee and
Jennifer Paek Diaz. Their emotional support was vital for me as I was separated from my
family across the Pacific Ocean.
Most importantly, I am very grateful to my parents, my sister, and my parents-in-
law. Their faith, inspiration, support, patience, sacrifices and unwavering love are
undeniably the bedrock upon which my life has been built. I am deeply grateful to my
husband, Hyeongwoo Kim, for his tremendous support, encouragement, patience and

iv
love. Lastly, I would like to thank my baby, Hailey Daon Kim, for helping me complete
the doctoral program by being an easy, brilliant and healthy little girl.

v
TABLE OF CONTENTS
DEDICATION.................................................................................................................... ii
ACKNOWLEDGEMENTS............................................................................................... iii
LIST OF TABLES............................................................................................................ vii
LIST OF FIGURES......................................................................................................... viii
ABSTRACT....................................................................................................................... ix
CHAPTER 1: INTRODUCTION....................................................................................... 1
Background............................................................................................................ 1
Study Aims and Conceptual Model....................................................................... 3
CHAPTER 2: DESCRIPTION OF THE DATA SET......................................................... 6
CHAPTER 3: RACAIL/ETHNIC DISPARITIES IN USE OF NEEDED CHILD
WELFARE SERVICES...................................................................................................... 9
Introduction........................................................................................................... 9
Service Needs of Children from Diverse Racial/Ethnic Backgrounds in Child
Welfare................................................................................................................. 10
Differences in Characteristics and Use of Needed Child Welfare Services........ 12
Method................................................................................................................. 16
Findings................................................................................................................ 20
Discussion............................................................................................................ 30
CHAPTER 4: THE ROLE OF RACIAL/ETHNIC MATCHING OF CASEWORKER-
CLIENT DYADS AND PROFESSIONALISM IN CLIENT SATISFACTION:
MODERATING EFFECTS OF ENGAGEMENT AND STRESS IN CHILD WELFARE
AGENCIES....................................................................................................................... 35
Introduction......................................................................................................... 35
Similarity between Worker and Client................................................................ 37
Considerable Debate: Racial/Ethnic Matching and Professionalism................... 39
Method................................................................................................................. 43
Findings................................................................................................................ 48
Discussion............................................................................................................ 63

vi
CHAPTER 5: THE IMPACT OF ORGANIZATIONAL FACTORS ASSOCIATED
WITH CULTURALLY RESPONSIVE PRACTICES, AND PROFICIENT AND
RESISTANT CULTURES ON OUTCOMES IN THE CHILD WELFARE SYSTEM.. 69
Introduction.......................................................................................................... 69
Racial/Ethnic Diversity and Institutional Demands............................................. 72
Organizational Responses in the Child Welfare System...................................... 73
Proficient and Resistant Cultures in Child Welfare Agencies............................. 75
Child Welfare Services and Outcomes................................................................. 76
Method................................................................................................................. 77
Findings................................................................................................................ 82
Discussion............................................................................................................ 90
CHAPTER 6: CONCLUSIONS AND IMPLICATIONS................................................ 96
Summary of Major Research Findings................................................................ 97
Strengths and Limitations.................................................................................... 99
Implications for Research.................................................................................. 100
Implications for Management and Social Work Practice.................................. 101
REFERENCES............................................................................................................... 104

vii
LIST OF TABLES
Table 3.1. Child, Caregiver, Neighborhood and Service Characteristics by Race/Ethnicity
(N = 1,555)………….....………………………….…………………....…...... 22
Table 3.2. Correlation Matrix for the Explanatory and Outcome Variables (N=1,555)... 24
Table 3.3. Results of Regression Models Predicting Needed Service Use of Children.... 28
Table 3.4. Results of Regression Models Predicting Needed Service Use of Children by
Race/Ethnicity.…..……………………………………………….......……… 29
Table 4.1. Sample Characteristics of Children, Caregivers and Caseworkers………...... 50
Table 4.2. Correlations between Study Variables (N = 702)………................................ 52
Table 4.3. Results of Moderated Regression Modeling for Caregiver Satisfaction…...... 55
Table 4.4. Results of Moderated Regression Modeling for Caregiver Satisfaction–
Empathy.…....................................................................................................... 58
Table 4.5. Results of Moderated Regression Modeling for Caregiver Satisfaction–Service
Process.............................................................................................................. 60
Table 4.6. Results of Moderated Regression Modeling for Caregiver Satisfaction–Service
Provision........................................................................................................... 63
Table 5.1. Sample Characteristics of Child Welfare Agencies and Children……..…….. 84
Table 5.2. Results of the Null Model Analyses................................................................. 86
Table 5.3. Results of Multilevel Regression Models – Use of Needed Services (N =
1,192)………………….…………………………………………..…….....… 88
Table 5.4. Results of Multilevel Regression Models – Behavioral/Emotional Problems (N
= 1,192)………..……………………………...………….…………..........… 89

viii
LIST OF FIGURES
Figure 1.1. Conceptual Model for the Dynamics of Racial/Ethnic Diversity and
Disparities in the Child Welfare System...........……….……..……..…..……. 5
Figure 2.1. NSCAW Study Design.…………...…………………………………….....… 8
Figure 3.1. Hypothesized Model of Study1.………...………………………...………… 16
Figure 4.1. Hypothesized Model of Study 2.………………….…………...….……...…. 43
Figure 5.1. Hypothesized Model of Study 3.………………..…………...…….…...…… 77

ix
ABSTRACT
This dissertation examines racial/ethnic diversity and disparities in the child
welfare system using a multilevel perspective. Given the conspicuous increase in
racial/ethnic diversity among children during the past two decades, child welfare agencies
face tremendous challenges in addressing the service needs of children and their families
from diverse racial/ethnic backgrounds. To date, service disparities in the child welfare
system have become evident and it is a significant challenge for child welfare agencies to
reduce these disparities by implementing organizational practices. However, little is
known about the extent to which child welfare agencies are executing practices to
respond to the service needs of clients from diverse racial/ethnic backgrounds and how
the practices impact service and child outcomes.
Using nationally representative, multilevel longitudinal data from the National
Survey of Child and Adolescent Wellbeing II, in three independent studies, this
dissertation seeks to establish disparities in use of needed services, and to identify
worker– and organizational factors associated with the disparities in a multilevel
perspective: (1) client–level: racial/ethnic disparities in use of needed child welfare
services; (2) worker–level: the role of racial/ethnic matching of caseworker–client dyads
and professionalism in client satisfaction: moderating effects of engagement and stress in
child welfare agencies; and (3) agency–level: the impact of organizational factors
associated with culturally responsive practices, and proficient and resistant cultures on
outcomes in the child welfare system.
Findings in the first study, focusing on client-level factors, indicated significant

x
racial/ethnic disparities in use of needed child welfare services. Hispanic children
received relatively more needed child welfare services than White children. Also, Black
and Hispanic children lived in more risky environments than White children did. The
findings indicate that high-risk children involved with the child welfare system receive
more attention from caseworkers, which in turn resulted in more use of needed services.
In the second study, pertaining to worker–level factors and consistent with the previous
study findings, worker–level interventions, such as racial/ethnic matching of caseworker–
client and caseworker professionalism were not significantly related to client satisfaction.
However, partial interaction effects of racial/ethnic matching, professionalism,
engagement, and stress on client satisfaction were found. In the third study, the question
of whether child welfare agencies implement organizational initiatives to address
racial/ethnic diversity issues on use of needed child welfare services and child
behavioral/emotional problems was not found to have any significant effect. Importantly,
use of needed services decreased child behavioral/emotional problems.
Taken together, this dissertation is the first study of its kind to rely on nationally
representative data in the child welfare system and provide important empirical evidence
to organizational approaches to racial/ethnic diversity and disparities in service and child
outcomes using a multilevel perspective. The findings highlighted the racial/ethnic
disparities in use of needed services and the role of worker–and organizational level
strategies in dealing with such disparities. Further understanding of the role of
organizational practices in providing culturally responsive services can help child welfare
agencies respond to the various needs of clients from diverse racial/ethnic backgrounds.

1
CHAPTER 1: INTRODUCTION
Background
Because racial/ethnic diversity among children involved in the child welfare
system has increased, child welfare agencies face the challenge of meeting increasingly
diverse needs. The literature, however, indicates that child welfare agencies have
responded insufficiently to this significant change and to the various needs of children
from diverse racial/ethnic backgrounds (Chapin Hall Center for Children, 2008). This has
resulted in service disparities among racial/ethnic groups of children (Hurlburt et al.,
2004; Wells, Hillemeier, Bai, & Belue, 2009). Non-White children are more likely to
remain longer in foster care than White children (Cheng, 2010; Foster, Hillemeier, & Bai,
2011) and Black children are less likely to use mental health services (Hurlburt et al.,
2004). Minority children such as Black and Hispanic also represent more than half of the
maltreated children in out-of-home care (U.S. Department of Health and Human Services
[DHHS], Administration on Children, Youth and Families [ACUF], 2006; Smith &
Devore, 2004).
To effectively diminish these service disparities, it is critical to understand the
characteristics of each racial/ethnic group of children involved with the child welfare
system. Although research on racial/ethnic disparities in child welfare has progressively
increased during the past two decades, studies have largely focused on specific services
or a particular race/ethnicity (Alzate & Rosenthal, 2009; Dettlaff & Cardoso, 2010;
Dettlaff, Earner, & Phillips, 2009; Font, Burger,& Slack, 2012; Pinderhughes, 1991;

2
Smith &Devore, 2004); this indicates a need to better understand racial/ethnic disparities
in terms of circumstances, challenges, families, and services (Dunbar& Barth, 2007).
Given the crucial issue of disparities derived from the growing racial/ethnic
diversity, child welfare agencies must respond to barriers impacting the delivery of social
services for children and families from diverse racial/ethnic backgrounds. Literature on
racial/ethnic diversity in human services has indicated the importance of providing
culturally responsive services to improve the delivery of services for racial/ethnic
minority clients (McPhatter, 1997; Guerrero, 2010; Guerrero & Andrews, 2011). Agency
factors as well as caseworkers play a significant role in the implementation of culturally
responsive services in the child welfare system. At the worker level, a caseworker is the
key agent in delivering services, and his/her race/ethnicity or language competence can
be matched with those of the client. In addition to matching strategies, hiring more
qualified caseworkers can play a crucial role in offering culturally responsive services to
children of various cultural backgrounds. Furthermore, leaders of child welfare agencies
can establish and implement organizational policies and procedures (e.g., provision of
cross-cultural training and adoption of performance measures to assess cultural
competence) at the agency level.
Emerging literature in behavioral health services has examined the role of
culturally responsive practices on service utilization (Campbell & Alexander, 2002;
Guerrero, 2010; 2012; Sue, Fujino, Hu, Takeuchi, & Zane, 1991; Snowden, Masland, Ma,
& Ciemens, 2006). However, little is known from the literature about how child welfare
agencies can play a role in addressing the racial/ethnic diversity issues at both the worker

3
and agency levels. The lack of multilevel data makes it difficult for researchers to
demonstrate the association between agency– and worker– level factors related to
providing culturally responsive services and impact on outcomes. This examination is
necessary to inform evidence-based policies in the child welfare system. With the release
of multilevel and longitudinal data from the National Survey of Child and Adolescent
Well-Being, investigating the roles of agency– and worker–related factors in child
welfare services and outcomes has become feasible.
Study Aims and Conceptual Model
In response to the issues related to the racial/ethnic diversity of children and the
research gaps discussed earlier, this dissertation aimed to expand the existing knowledge
by examining racial/ethnic diversity and disparities in the child welfare system using a
multilevel perspective. The aim of this study was achieved by conducting three
interrelated studies. The specific aims and research questions of the three studies are
discussed in the chapter related to each study. The overarching aims of this dissertation
are as follows:

1. Aim 1: To establish racial/ethnic disparities in use of needed child welfare
services;
2. Aim 2: To investigate the role of racial/ethnic matching of caseworker–client
dyads and professionalism in client satisfaction: moderating effects of
engagement and stress in child welfare agencies;

4
3. Aim 3: To examine the impact of organizational factors associated with
culturally responsive practices, and proficient and resistant cultures on
outcomes in the child welfare system.
This study takes a multilevel perspective to provide a comprehensive
understanding of racial/ethnic diversity and disparities in the child welfare system.
Considering the hierarchical structure of child welfare agencies and the dynamics in the
agencies, the present study focuses three levels as agency, caseworker and client. Child
welfare agencies play a role in implementing organizational practices and managing their
workers in order to provide a high quality of child welfare services to clients.
Caseworkers play a primary and direct role in delivering the services to clients. Children
and families in the child welfare agencies are clients and their characteristics need to be
understood to help child welfare agencies and caseworkers meet their needs. It is
necessary to understand the role of each level in racial/ethnic diversity and disparities.
Figure 1.1 features a conceptual model of this dissertation study constructed in terms of
the dynamics of child welfare agencies, caseworkers, and children from diverse
racial/ethnic backgrounds in the child welfare system.

5

Agency
Level

Worker
Level

Client
Level

Figure 1.1. Conceptual Model for the Dynamics of Racial/Ethnic Diversity and
Disparities in the Child Welfare System

* Race/Ethnicity
Child Factors
Caregiver Factors
Neighborhood Factors
Service Outcomes
- Use of Needed Services
- Client Satisfaction
Child Outcomes
- Child Behavioral/
Emotional Problems
Racial/Ethnic Matching
Professionalism
- Graduate Degree
- Social work Degree
Engagement
Stress
Responses to Racial/
Ethnic Diversity Issues
Proficient Culture
Resistant Culture

6
CHAPTER 2: DESCRIPTION OF THE DATA SET
A nationally representative, longitudinal multilevel data set (National Survey of
Child and Adolescent Well-Being (NSCAW) II, Waves 1 and 2) was used for the three
studies of this dissertation. The NSCAW study is the first national study of child welfare
designed to address questions about the child welfare system, including child functioning,
service use and outcomes. It is sponsored by the Administration for Children & Families
(ACF) and the U.S. Department of Health and Human Services (DHHS). The NSCAW
survey contains information on child well-being, their families, and comprehensive child
welfare services on topics including health, mental health, substance abuse, and child
abuse obtained from multiple informants. The informants include children, caregivers,
caseworkers, and directors. The primary research questions of the NSCAW study are: (1)
Who are the children and families that come into contact with the child welfare system?
(2) What pathways and services do children and families experience while in the system?
(3) What are the shorter- and longer-term outcomes for those children? (NSCAW
Research Group, 2011).
The NSCAW II research design mirrors that of the NSCAW I. The NSCAW I
study used a sample from 92 primary sampling units (PSUs) in 97 counties using a two-
stage, stratified cluster sampling method to yield national population estimates. During
the first stage, the United States was divided into nine sampling strata. Eight of these
strata were the eight states featuring the largest child welfare caseloads. The ninth
included the remaining 42 states and the District of Columbia. Ninety-two PSUs were
chosen within these nine strata. During the second stage, 5,501 children from infancy to

7
14 years old were selected from lists of closed investigations or assessments from the
sampled agencies. Sampling within PSUs was stratified by age, type of maltreatment, and
receipt of services. Eighty-one of the original 92 PSUs in 83 counties nationwide agreed
to participate in NSCAW II (NSCAW Research Group, 2011).
The NSCAW II cohort includes 5,873 children, aged birth to 17.5 years old, who
had contact with the child welfare system within a 15-month period that began in March
2008. After 18 months, a second wave of data was collected, from October 2009 to
January 2011. The NSCAW II offers several advantages compared to the NSCAW I; it
features the most recent information on the child welfare system and included updated or
new measures. The current study used samples from different waves of the NSCAW II
for different research questions, as described in later sections. Figure 2.1 displays the
longitudinal study design of the NSCAW.
All tests for the three interrelated studies were weighted to produce estimates for
the national population of children reported for potential maltreatment to child protective
service (CPS) agencies. The NSCAW data set is not based on a simple random sample,
and not all members of the population have an equal probability of selection. Weights are
developed to reduce potential bias due to unequal selection probabilities (Biemer, Christ,
Wheeles, & Wiesen, 2005). For this study, analysis weights were constructed to adjust
for the selection probability of PSUs and individual children within PSUs. More details
regarding the NSCAW sample design and weight derivation are available elsewhere
(Dowd et al., 2003).

8
1999-2000 2008-2009

12months

18months

36months

72months

[NSCAW I] [NSCAW II]
Figure 2.1. NSCAW Study Design
Note. a. Waves I and II of NSCAW II were used for this dissertation.

Wave 2
Wave 3
Wave 4
Wave 5
Wave 2
a

18months
Wave 3 (In Progress) 36months
Wave 1 Wave 1
a

9
CHAPTER 3: RACIAL/ETHNIC DISPARITIES IN USE OF
NEEDED CHILD WELFARE SERVICES
Introduction
In recent decades, racial/ethnic disparities in child welfare services have been
well documented. According to the literature, Black and Hispanic children are more
likely than White children to experience longer waiting periods for adoption and to spend
more time in foster care. (Hill, 2007; Cheng, 2010). Minority children are less likely to be
served in a timely fashion and receive proper services during foster care, kinship care,
and adoption (Font et al., 2012). Research on service disparities has indicated the
importance of providing needed services to children; services have a positive effect on
children and their families, such as an increased likelihood of family reunification (Cheng,
2010; Cheng & Lo, 2012; Choi & Ryan, 2007). Although some studies examined the use
of specific needed services, including mental health and substance abuse among children
involved in the child welfare system (Hurlburt et al., 2004; Marsh, Cao, Guerrero, & Shin,
2009), previous studies have not explored the extent to which children receive needed
child welfare services provided by child welfare agencies or the differences in the use of
needed services by race/ethnicity. Further, empirical evidence that demonstrates the
factors affecting the use of needed services is limited.
Given the need to understand racial/ethnic disparities in use of needed child
welfare services, the aim of this study was to add to the existing knowledge of
racial/ethnic service disparities in child welfare by seeking answers to the following
research questions.

10
Research Question 1: To what extent do racial/ethnic groups of children differ
in use of needed child welfare services?
Research Question 2: What factors are associated with racial/ethnic disparities
in use of needed child welfare services?
Service Needs of Children from Diverse Racial/Ethnic Backgrounds
in Child Welfare
Along with an increase in racial/ethnic diversity in the child welfare system, two
concepts illustrating different service needs by race/ethnicity have emerged:
disproportionality and disparities (Derezotes, Poertner, & Testa, 2005). In the child
welfare context, disproportionality refers to the overrepresentation or underrepresentation
of minority children in the child welfare system, as compared to the number of minority
children in the total population. The overrepresentation of some racial/ethnic groups of
clients in the child welfare system has been evident (Knott & Donovan, 2010; McPhatter,
1997). The U.S. Census Bureau reports that Black children comprise 15% of the total
child population (U.S. Census Bureau, Housing and Household Economic Statistics
Division, Fertility & Family Statistics Branch, 2010), while some estimates indicate 27%
of children in foster care are Black (U.S. DHHS, ACF, ACUF, Children’s Bureau, 2012).
According to the U.S. Department of Health and Human Services (2003), 25% of
individuals who experience child abuse and 45% of children in foster care are Black
(Chibnall et al., 2003).
The literature indicates that the overrepresentation of minority children, including
Black and Hispanic children, is common in many subfields of child welfare such as

11
investigation, placement, service delivery, and aftercare services. In accordance with the
societal trend of increasing racially/ethnically diverse populations due to globalization,
the overrepresentation of minority groups in the child welfare field is expected to
increase (Hasenfeld, 2009); the service needs of minority groups is expected to
correspondingly grow and diversify. Considering the complex needs of children from
diverse racial/ethnic backgrounds, research on the use patterns of various needed child
welfare services is more essential than examining the use of specific services.
Another concept that may explain different service use patterns among
racial/ethnic groups is disparities. Disparities refer to “disparate or inequitable treatment
or patterns of service utilization” (Ortega, Grogan-Kaylor, Ruffolo, Clarke, & Karb, 2010,
p. 236). Disparity is typically illustrated by comparing non-White and White children in
service utilization (Ortega et al., 2010). However, among non-White children, Black and
Hispanic children have different cultural backgrounds. Thus, their characteristics and
needs are distinguishable, and the two racial/ethnic groups should be separately examined
in racial/ethnic diversity and disparities. As discussed earlier, there is extensive evidence
of racial/ethnic disparities in the child welfare system. This implies that child welfare
agencies have not effectively met the needs of minority children such as Black and
Hispanic through their services.
Regardless of the well-documented service disparities among racial/ethnic groups
of children, the extent to which children have their needs met by child welfare services
has remained largely unexplored. A previous study in social services, such as substance
abuse treatment, which is used by more than 70% of child welfare families, has examined

12
matching service need, with service received (Marsh et al., 2009). The study found that
racial/ethnic minorities were insufficiently provided needed treatment services than
White clients. Similarly, a recent study examining racial/ethnic disparities in unmet
service needs among parents who have children involved with the child welfare system
showed that Hispanic parents were more likely to receive needed services from child
welfare agencies than any other race/ethnic groups (Cheng & Lo, 2012). Black caregivers
faced more service access barriers than White caregivers (Cheng & Lo, 2012). However,
this study only focused on parents of children who received needed child welfare services.
Wells and his colleagues (2009) demonstrated racial/ethnic differences in health service
access, but did not include the access to other crucial services such as counseling,
education, or daycare services, and therefore could not explain the overall use of needed
child welfare services across racial/ethnic groups of children. In these contexts, there is a
need to understand racial/ethnic difference among children in use of needed child welfare
services. Therefore, the following hypothesis was constructed:
Hypothesis 3.1. There is a racial/ethnic difference in use of needed child welfare
services.
Differences in Characteristics and Use of Needed Child Welfare Services
Child, caregiver, and neighborhood factors were targeted to demonstrate
racial/ethnic differences and predictors of use of needed services. The literature on
service disparities in child welfare underscores the importance of child characteristics
such as substantiation, setting, type of maltreatment and behavioral/emotional problems
in service use. During the investigation process, when a child is substantiated or indicated

13
for maltreatment, the caseworker is responsible for establishing a case plan and providing
child welfare services for the child (Font et al., 2012). Therefore, children who are
substantiated or indicated may be more likely to use needed child welfare services.
Regarding settings in which children live, the literature suggests that Black and Hispanic
children in foster care are less likely to receive mental health services than White children
(Garland, Landsverk, & Lau, 2003; Zima, Bussing, Yang &Belin, 2000). The literature
also indicates that Black children who have been sexually abused exhibit low levels of
service utilization (McCrae & Frusco, 2010). Type of maltreatment has been regarded as
a crucial factor that may vary based on race/ethnicity and could affect receipt of services.
The Child Behavior Checklist (Achenbach, 1991) has been widely used to measure
emotional and behavioral problems among children when comparing racial/ethnic
differences in need for mental health services and overall functioning (e.g., Dunbar &
Barth, 2007; Hurlburt et al., 2004). Black children were less likely to use services than
White children for all items on the Child Behavior Checklist, indicating that Black
children use fewer services when levels of need among racial/ethnic groups of children
are equal (Dunbar & Barth, 2007). However, the literature lacks an understanding of the
relationship between behavioral/emotional problems and the use of needed services
across racial/ethnic groups of children. Given the differences in the factors by
race/ethnicity, it can be assumed that levels of service needs of children are different
among racial/ethnic groups. In light of these contexts, the following hypotheses were
examined:
Hypothesis 3.2.There is a racial/ethnic difference in child characteristics.

14
Hypothesis 3.3. Child characteristics are associated with use of needed services.

Caregiver factors have received a significant amount of attention in the literature
on racial/ethnic disparities in child welfare due to substantial differences in caregiver
characteristics by race/ethnicity. Black and Hispanic caregivers of children in contact
with child welfare may struggle with poverty (Grella, Needell, Shi, & Hser, 2009; Wells
et al., 2009). The literature indicates that Hispanic caregivers report lower level of
education than other racial/ethnic groups of caregivers (Gandara, 2010; Hines, Lemon,
Wyatt, & Merdinger, 2004). The low socio-economic status of caregiver may negatively
affect service use among children. Primary caregiver gender in the child welfare system
is female. If primary caregiver is male, it has the high potential to living with the male
caregiver or having a mother with severe psychological or health issues. Thus, whether or
not a primary caregiver is female make a difference in family environments. Caregiver
age is also included in the model to see racial/ethnic difference in age. This study
included caregiver factors to examine racial/ethnic difference and the impact on use of
needed services.
Hypothesis 3.4. There is a racial/ethnic difference in caregiver characteristics.
Hypothesis 3.5. Caregiver characteristics are associated with use of needed
services.
Neighborhood factors have been recognized important factors of racial/ethnic
differences. It is safe to say that minority children such as Black and Hispanic children
are often located in the riskier neighborhoods than White children. Font and his

15
colleagues (2012) confirmed the belief by demonstrating that Black children lived in
more problematic neighborhoods with regards to assaults, substance abuse and gang
involvement. Brooks-Gunn, Duncan and Maritato (1997) indicate that Hispanic children
are more likely to reside in low quality neighborhoods that have a number of
neighborhood concerns. The poor quality and problematic neighborhoods are linked to
insufficient service use due to a lack of available services in their communities (Hines et
al., 2004; Roberts, 2002). Children living in an impoverished and high-risk neighborhood
may report low receipt of services. However, examining the neighborhood poor quality
and problem effects on use of needed services has not been attempted. Therefore, this
study seeks to confirm racial/ethnic differences in neighborhood factors and the
associated relationship with use of needed services.
Hypothesis 3.6. There is a racial/ethnic difference in neighborhood characteristics.
Hypothesis 3.7. Neighborhood characteristics are associated with use of needed
services.
In sum, previous research demonstrating predictors of service use and differences
among racial/ethnic groups of children did not consider use of needed services and
associated factors. Further, racial/ethnic differences in children’s use of needed services
and the factors associated with use of needed services have not been explored. In light of
this gap, this study examined the hypothesized model featured in Figure 3.1.

16

Figure 3.1. Hypothesized Model of Study 1
Method
Sample
This study used the first wave of NSCAW II data from interviews with children,
caregivers, and caseworkers. The current study used a subset of children who (1) were
Black, Hispanic, or White; and (2) for whom Child Behavioral Checklist and use of
needed child welfare services data were available. The final study sample included 1,555
children. Male children comprised slightly more than a half of the sample (56.63%). The

Use of Needed Child
Welfare Services
(e.g., Special Education,
Mental health and Health)

Caregiver Factors
- Gender - Age
- Education Level
Child Factors
- Race/Ethnicity
- Gender - Age
- Substantiation
- Abuse Experience
- Placement
- Behavioral/Emotional Problems
- # of Service Needs
Neighborhood Factors
-Neighborhood Poor Quality
-Neighborhood Problem

Racial/Ethnic Differences
Use of Needed Services
Child Factors
Caregiver Factors
Neighborhood Factors

17
average age of children in the sample was 7.81 years (SE = .21). The largest age group of
children was those between 2 and 5 years of age (40.35%), 33.49% were between 6 and
10 years of age, and 26.16% were between 11 and 17 years of age. The racial/ethnic
composition of the sample was 46.96% White, 28.16% Hispanic and 24.86% Black.
Many of the children (82.95%) lived at home with their birth or foster parents. 4.31%
were in kinship care, 1.09% were in foster care, and 3.0% were in other out-of-home
setting such as group homes.
Measures
Dependent variable
Use of needed child welfare services. Caseworkers in PSUs provided
information on the continuous variable of use of needed child welfare services, indicating
the percentage of needed child welfare services each child received. To construct this
variable, need for 13 services were first identified: learning problems/developmental
disability assessment, special education services, substance abuse, counseling for mental
health, counseling for delinquency, Medicaid, health problems, check-ups or
immunizations, dental care, hearing, vision, independent living skills and daycare. The
number of service needs of each child were totaled. Then, the number of received
services of each child were totaled. Lastly, the use of needed child welfare services
variable was made by dividing the number of received services by the number of service
needs. Higher percentages represented greater levels of the needed child welfare services
that were received. Although the variable did not account for patterns of service use
(Cheng & Lo, 2012), the measure of use of needed child welfare services is beneficial in

18
that it allows for a greater understanding of the extent to which child welfare agencies
meet the needs of children involved in the child welfare system.
Independent variables
Child factors. Race/ethnicity include three categories of Black, Hispanic, and
White. The children identified as other racial/ethnic groups were excluded due to small
subsample size. Substantiation was measured using a dichotomous variable (1 =
substantiated) indicating the child welfare system’s official case decision that allegations
of child maltreatment were valid. Placement consisted of a categorical variable (1 = in-
home with parents, 2 = in-home in kin care, 3 = foster care, 4 = group home or other).
Due to a high comorbidity of multiple subtypes among children, six dichotomous
variables representing maltreatment experience for each type were measured: (1) physical
maltreatment, (2) sexual abuse, (3) emotional maltreatment, (4) physical neglect, (5)
neglect (lack of supervision), and (6) abandonment. Behavioral/emotional problems were
measured using the Child Behavior Checklist (CBCL) developed by Achenbach et al.
(1991). CBCL scores have been widely used as indicators of child emotional and
behavioral problems (Casanueva et al., 2012). One hundred items on a 3-point Likert
scale for children aged 2 to 5 years and 103 items on a 3-point Likert scale for children
aged 6 to 17 years were summed to create a total score of child behavioral/emotional
problems. A standardized score of 64 or higher on the total scale was defined as clinical
range indicating clinically significant need for mental health services. The number of
needed services was measured by summing needed services, as were children’s gender
(1= male) and age in years as demographic characteristics.

19
Caregiver and neighborhood factors. Caregiver’s education level (1 = less than
high school, 2 = high school, 3 = college or higher) was included. Caregiver’s gender (1=
male) and age in years were also measured. Caregivers reported neighborhood poor
quality and neighborhood problems were measured as neighborhood factors.
Neighborhood poor quality was measured using four items on a 3-point Likert scale (1 =
better to 3 = worse). Caregivers reported on their neighborhood safety, neighborly
assistance, and caregiver involvement compared to other areas and whether it was a better
place to live than other areas. The Cronbach’s alpha was .89. Higher scores represent
poorer neighborhood quality. A neighborhood problems variable was estimated using
five items on a 3-point Likert scale (1 = not at all to 3 = highly problematic). Caregivers
were asked to rate the extent to which their neighborhood has problems in terms of
assaults/muggings, delinquents/drug gangs, open drug use/dealing, unsupervised children,
or groups of teens hanging out. The reliability of the measure was .78. Higher scores
indicate more problematic neighborhood.
Analysis
Sampling weights were used for all analyses to address the sampling design of
the NSCAW. To explore differences in the characteristics of racial/ethnic groups of
children, chi-square tests and one-way analyses of variance(ANOVAs) for mean
differences were employed. Bivariate analyses were also conducted to investigate
associations between the variables. Using STATA, a weighted three-step hierarchical
ordinary least squares regression analysis was conducted to examine the relationships
between explanatory variables and use of needed child welfare services for each

20
subsample and the total sample. In the first step, child variables were entered; in the
second step, caregiver variables were entered; in the third step, neighborhood variables
were entered. As with other large data sets, NSCAW includes a large amount of missing
data. This resulted in a reduction of sample size for this study. A NSCAW report on non-
response analyses suggests that missing data in NSCAW is “unlikely to be consequential
for most types of analyses” (U.S. DHHS, ACUF, 2005, p. 2-12). Therefore, this study
used listwise deletion and all available data were used for analyses.
Findings
ANOVA and chi-square results are presented in Table 3.1. There was a
significant difference in use of needed child welfare services (F = 5.11, df = 2, p < .01),
providing support for Hypothesis 3.1. Hispanic children received more needed services
than White children. Among child-level variables, experiencing neglect (lack of
supervision) differed significantly by race/ethnicity (χ
2
= 4.61, df = 2, p < .05). Hispanic
children were more likely to experience neglect (lack of supervision) than Black and
White children. In addition, Black and Hispanic children reported more
behavioral/emotional problems than White children (χ
2
= 6.34, df = 4, p < .01). In regard
to service needs, White children had a higher level of service needs than Black and
Hispanic children (F = 19.32, df = 2, p < .001). Thus, the findings partially supported
Hypothesis 3.2. A significant difference in the education level of caregivers was revealed.
Hispanic caregivers had a relatively lower level of education than the other two groups
(χ
2
= 6.23, df = 4, p < .01). The findings partially supported Hypothesis 3.4. In terms of
neighborhood factors, both neighborhood poor quality and problems differed

21
significantly by race/ethnic groups. Black children perceived that they lived in poorer
quality neighborhoods (F = 10.24, df = 2, p < .001) and more problematic neighborhoods
(F = 19.31, df = 2, p < .001) than the other two groups. In addition, Hispanic children
lived in more problematic neighborhood than White children. Hypothesis 3.6 was
supported by the findings.

22
Table 3.1
Child, Caregiver, Neighborhood and Service Characteristics by Race/Ethnicity (N=1,555)
Black Hispanic White Total p
Child
Race/ethnicity
Male (%)
24.88
49.95
28.16
55.96
46.96
60.57

56.63

n.s.
Age (M, SE) 7.51 (.34) 7.81 (.42) 7.97 (.31) 7.81 (.21) n.s.
Substantiated (%) 27.30 24.03 28.33 26.80 n.s.
Placement (%)
In home with parent 81.47 80.55 85.17 82.95 n.s.
In home in kin care 11.90 15.00 9.51 4.31
Foster care 4.27 3.66 4.71 1.09
Other OOHA 2.36 3.30 0.60 3.00
Type of abuse
Physical maltreatment

30.01

29.75

30.52

30.17

n.s.
Sexual maltreatment
Emotional maltreatment
8.64
4.56
8.52
7.52
14.33
10.16
11.28
8.02
n.s.
n.s.
Physical neglect
Neglect (LOS)
Abandonment
20.43
29.08
1.12
19.81
44.61
3.98
15.16
31.91
1.29
17.78
34.81
2.00
n.s.
<.05
n.s.
B/E problems (%)
Normal ≤ 63 29.40 23.43 36.21 30.77 <.05
Clinical range >63
# of services needed
(M, SE)
70.60
2.43 (.12)
76.57
2.85 (.20)
63.79
3.12 (.13)
69.23
2.87 (.11)

< .001
Caregiver
Race/ethnicity
Education level (%)
20.85 24.75 52.29
Less than HS 22.88 47.42 23.41 30.07 < .01
HS 40.03 37.01 45.90 41.95
College or higher 37.09 15.57 30.69 27.98
Male (%)
Age (M, SE)
Neighborhood
Neighborhood poor quality
(M, SE)
Neighborhood problems
(M, SE)
10.29
34.81 (.76)

7.43 (.09)

7.20 (.26)
15.43
35.32 (1.07)

7.12 (.24)

6.80 (.34)
10.45
35.41 (.61)

6.80 (.13)

6.25 (.14)
11.82
35.24 (.44)

7.04 (.10)

6.64 (.15)
n.s.
n.s.

< .001

< .001
Service
# of service received
(M, SE)
0.84 (.09) 1.08 (.29) 0.92 (.07) 0.94 (.11) n.s.
% of use of needed
services (M, SE)
31.16 (.03) 34.39 (.06) 27.23 (.02) 30.23 (.03) < .01
Note. All figures represent weighted percentages, means and standard errors. HS = high school; OOHA =
out-of-home arrangements; LOS = lack of supervision; B/E problems = behavioral/emotional problems.

23
As shown in the correlation matrix in Table 3.2, several factors were
significantly associated with use of needed child welfare services. Of child factors, being
Hispanic increased the use of needed services, corresponding to the ANOVA results (r
= .07, p < .01). Younger children used more needed services (r = -.13, p < .001). Children
who were substantiated for maltreatment were more likely to use needed services (r = .14,
p < .001). Compared to living with parents, being in kin care (r = .08, p < .01), foster care
(r = .18, p < .01),or other out-of-home care settings (r = .10, p < .01) was positively
associated with the use of needed services. Experiencing physical maltreatment (r = -.12,
p< .01) was negatively related to use of needed services, whereas experiencing neglect
(lack of supervision, r =.09, p < .01) and abandonment (r = .17, p < .01) were positively
related to use of needed services. Higher service needs led to higher use of needed
services (r = .11, p < .01). None of the caregiver factors was significantly linked to use of
needed services among children from diverse racial/ethnic backgrounds. Children who
lived in more problematic neighborhoods (r = -.12, p< .01) were less likely to receive
needed child welfare services.

Table 3.2
Correlation Matrix for the Explanatory and Outcome Variables (N=1,555)
1 2 3 4 5 6 7 8 9 10 11
1. Use of needed services 1
2. Black .01 1
3. Hispanic .07** -.36** 1
4. Male -.03 -.01 -.08** 1
5. Age -.13** -.04 -.00 -.11** 1
6. Substantiated .14** .01 -.04 -.10** .08** 1
7. Kin care .08** .00 .07* .00 .08** .10** 1
8. Foster care .18** -.00 -.02 -.04 .01 .14** -.08** 1
9. Other OOHA .10** .07** -.02 .01 .10** .07** -.04 -.02 1
10. Physical maltreatment -.12** -.00 -.01 .07** .14** -.05 -.11** .02 .02 1
11. Sexual maltreatment -.00 -.05 -.05* -.13** .15** -.01 .01 .01 -.02 -.15** 1
12. Emotional maltreatment .00 -.07** -.01 .01 .06* .12** -.01 .05 .00 .13** -.05*
13. Physical neglect .05 .04 .03 .01 -.09** -.02 .07** .04 .04 -.15** -.16**
14. Neglect (LOS) .09** -.07** .13** .04 -.16** .06* .05 .05* .05* -.23** -.08**
15. Abandonment .17** -.04 .09** .01 .03 -.01 .26** .14** .03 -.08** -.05*
16. B/E problems .04 .02 .10** .01 -.18** -.03 -.01 -.01 -.02 -.10** -.10**
17. # of service needs .11** -.14** -.01 .12** .10** .08** .13** .21** .16** .11** .03
18. Caregiver – male .01 -.03 .07** .03 -.01 -.03 -.02 -.04 .04 .05 .05*
19. Caregiver – age .05 -.02 .00 -.01 .42** .14** .52** .23** .08 .01 .12**
20. Caregiver – < high school -.02 -.09** .24** -.02 -.00 -.04 -.03 -.08** .01 -.06* .08**
21. Caregiver – high school .02 -.02 -.06* .09** -.05 -.07** .05* .04 -.00 .02 .02
22. Neighborhood poor quality -.03 .13** .04 -.04 .04 .06* -.11** -.10** -.02 .05* -.04
23. Neighborhood problems -.12** .10** .02 -.10** -.05 .09** -.08** -.08** -.02 -.09** -.02
Note. All figures represent weighted correlation coefficients. OOHA = out-of-home arrangements; LOS = lack of supervision; B/E problems =
behavioral/emotional problems.
* p < .05, ** p < .01
24

Table 3.2
Correlation Matrix for the Explanatory and Outcome Variables (Cont’d)
12 13 14 15 16 17 18 19 20 21 22
12. Emotional maltreatment 1
13. Physical neglect -.05 1
14. Neglect (LOS) -.02 .03 1
15. Abandonment -.02 .13** .10** 1
16. B/E problems -.07** -.01 -.00 -.08** 1
17. # of needed services .14** .0** .06* .28** .22** 1
18. Caregiver – male -.00 -.04 -.01 -.04 .08** .04 1
19. Caregiver – age .02 .02 -.02 .10** -.01 .16** .03 1
20. Caregiver – less high school .05 -.01 .13** -.07** .12** -.03 .06* -.08** 1
21. Caregiver – high school -.12** .07** -.03 .10** -.05 .00 .02 -.03 -.56** 1
22. Neighborhood poor quality -.00 .09** .03 -.02 -.12** -.06* -.00 -.08** .04 .02 1
23. Neighborhood problems .01 .06* .04 -.05 -.08** -.10** -.06* -.21** .05 -.03 .47**
Note. All figures represent weighted correlation coefficients. LOS = lack of supervision; B/E problems = behavioral/emotional problems.
* p< .05, ** p < .01

25

26
The results of hierarchical ordinary least squares regression analysis for the total
sample are shown in Table 3.3. Being Hispanic was found to be a significant predictor of
use of needed services (b =.13 , p < .05), providing support for Hypothesis 3.3. Of other
child factors, younger age (b = -.14, p < .01), foster care (b =.16, p < .001), maltreatment
substantiation (b = .13, p < .01), and abandonment (b = .13, p < .001) were related to
increased use of needed child welfare services. The model for these variables was
significant (F[16, 1054] = 12.09 , p < .001), and they accounted for 12.8% of the variance.
In the second step, caregiver variables were added, and the results showed a
nonsignificant increase in R
2
, although the overall model was significant (F[20, 1050] =
10.17, p < .001). Thus, Hypothesis 3.5 was not supported. In the third step, neighborhood
variables were entered, resulting in a significant increase in R
2
. The regression model was
statistically significant (F[22, 1048] = 9.95, p < .001) and explained 14.7% of the
variance. Level of neighborhood problems was significantly related to use of needed
child welfare services (b = -.15, p < .001). The findings of the regression analysis
partially supported Hypothesis 3.7.
The findings of the regression analyses for each race/ethnic group are displayed
in Table 3.4. For Black children, a number of factors including gender (b = .16, p < .05),
age (b = -.22, p < .01), substantiation (b = .17, p < .05), foster care, other out-of-home
arrangement (b = .08, p < .001), physical neglect (b = -.20, p < .01), and abandonment (b
= .27, p < .01), were significantly associated with use of needed services. Foster care and
other out-of-home arrangements were the strongest predictors among Black children. For
Hispanic children, the strongest factors were substantiation (b = .18, p < .01), and living

27
in less problematic neighborhoods (b = -.25, p < .01), which increased the extent to which
child welfare agencies met their service needs. For White children, being in foster care (b
= .19, p < .001) and experiencing abandonment (b = .10, p < .001) were found to be
predictors of use of needed services.

28
Table 3.3
Results of Regression Models Predicting Needed Service Use of Children
Model 1 Model 2 Model 3
b SE b SE b SE
Child Factors
Black
Hispanic
Male
.03
.13*
-.01
.04
.04
.04
.03
.13*
-.01
.04
.05
.04
.06
.15**
-.03
.04
.05
.04
Age -.14** .00 -.13* .00 -.12* .00
Substantiated .13** .03 .13** .03 .13** .03
Placement
In home in kin care .04 .05 .05 .06 .03 .06
Foster care .16*** .04 .17*** .05 .16*** .05
Other OOHA .07 .14 .02 .13 .01 .13
Physical maltreatment -.06 .04 -.06 .04 -.04 .04
Sexual maltreatment
Emotional maltreatment
Physical neglect
.06
.03
-.00
.05
.06
.05
.05
.03
-.00
.05
06
.05
.03
.02
-.01
.05
.06
.05
Neglect (LOS) .01 .04 .01 .04 .04 .04
Abandonment .13*** .04 .13*** .08 .14*** .08
B/E problems
# of service needs
.04
.05***
.03
.01
.05
.04***
.04
.01
.03
.04***
.03
.01
Caregiver Factors
Male
Age
.00
-.02
.05
.00
-.01
-.01
.05
.00
Education level
Less than HS -.02 .05 -.01 .05
HS .01 .05 .03 .04
Neighborhood Factors
Neighborhood poor quality .04 .01
Neighborhood problems -.15*** .01
R
2
.128 .129 .147
Note. Living in home with parents was the reference category for placement. HS = high
school; OOHA = out-of-home arrangements; LOS = lack of supervision; B/E problems =
behavioral/emotional problems.
* p < .05, ** p < .01, *** p < .001

29
Table 3.4
Results of Regression Models Predicting Needed Service Use of Children by
Race/Ethnicity
Black Hispanic White
b SE b SE b SE
Child Factors
Male .16* .06 -.07 .06 -.07 .05
Age -.22** .01 -.20† .01 -.02 .01
Substantiated .17* .06 .18** .05 .08 .06
Placement
In home in kin care .07 .07 -.02 .14 .03 .06
Foster care .21*** .09 .06 .12 .19*** .06
Other OOHA .08*** .09 - - -.01 .16
Physical maltreatment .03 .08 -.06 .08 -.06 .06
Sexual maltreatment
Emotional maltreatment
Physical neglect
.11
-.15†
-.20**
.09
.11
.07
.05
.09
.19†
.15
.13
.10
-.01
.01
-.01
.05
.07
.05
Neglect (LOS) .04 .08 -.03 .06 .06 .06
Abandonment .27** .20 .06 .15 .10*** .08
B/E problems
# of service needs
.01
-.02
.05
.02
.12
.13
.07
.02
.01
.03
.04
.01
Caregiver Factors
Male
Age
.07
.02
.10
.00
-.03
.01
.10
.01
.01
.01
.06
.00
Education level
Less than HS .04 .07 -.00 .09 -.02 .06
HS .11 .06 .03 .07 .00 .05
Neighborhood Factors
Neighborhood poor quality .05 .02 .08 .01 .01 .02
Neighborhood problems -.09 .02 -.25** .01 -.03 .02
R
2
.230 .273 .114
Note. Living in home with parents was the reference category for placement. HS = high
school; OOHA = out-of-home arrangements; LOS = lack of supervision; B/E problems =
behavioral/emotional problems.
† p < .10, * p < .05, ** p < .01, *** p < .001

30
Discussion
During the past two decades, racial/ethnic disparities in the child welfare system
have been a crucial concern of researchers, managers and practitioners. The literature on
racial/ethnic disparities in child welfare is somewhat limited because most existing
studies have focused on unveiling racial/ethnic disparities in specific services. The
present study extended the efforts of previous researchers by examining racial/ethnic
disparities in the use of needed child welfare services and its predictors in a sample of
children involved with the child welfare system.
The findings indicated that Hispanic children receive more of the needed services
than White children. Moreover, White children use fewer the needed services than Black
children, although White children reported the largest number of service needs compared
to minority children. These findings contradict the existing racial/ethnic disparities
studies that have demonstrated minority children struggle to access child welfare services
(Garland et al., 2000; Garland et al., 2003). However, recent studies using NSCAW data
resulted in findings similar to the current study. Cheng and Lo (2012) focused on the
percentage of needed services not obtained by permanent caregivers and found that White
caregivers are less likely to receive needed services than Hispanic caregivers. Wells and
his colleagues (2009) found that Hispanic children are more likely to access to counseling
and vision services than Black and White children. One tenable explanation of the current
study’s findings is that in the case of children who is currently involved in the child
welfare system, high level of reported neglect (lack of supervision) and
behavioral/emotional problems among Hispanic children led to more use of needed child

31
welfare services. Caseworker recognition of those problems among Hispanic children
may have prompted them to pay more attention to higher-risk children and make more of
an effort to address their issues by providing needed services. Another feasible
explanation is that Hispanics can more easily access services than Whites and Blacks
since they are more likely to live in metropolitan areas with more service providers. In
order to ensure this explanation, future research should examine the impact of urbanicity
on use of needed services.
The present study identified racial/ethnic differences in child, caregiver, and
neighborhood characteristics, partially supporting the hypotheses. In terms of child-
related factors, Hispanic children had a higher percentage of neglect (lack of supervision)
cases than White children. This is consistent with the previous findings (Mennen, Kim,
Sang, & Trickett, 2010). However, in terms of other types of abuse, no significant
differences emerged among racial/ethnic groups. The percentage of behavioral/emotional
problems in clinical range among Hispanic children was significantly higher than those of
Black and White children, indicating that Hispanic children have more service needs to
reduce behavioral/emotional problems.
Hispanic children were also significantly different than other racial/ethnic groups
in terms of their caregivers’ education level. Previous research points to Hispanic
caregivers tend to have low levels of education attainment (Gandara, 2010),
corresponding to this study’s findings. In general, a low level of education among
caregivers is a crucial predictor of being poor (Brooks-Gunn et al., 1997). Therefore,
child welfare workers may need to pay attention to whether Hispanic children need

32
financial assistance.
Examination of neighborhood factors showed significant racial/ethnic differences.
Black caregivers perceived that their children and families lived in poorer quality and the
more problematic neighborhoods than Hispanic and White caregivers. Not surprisingly,
White children lived in the relatively higher quality and less problematic neighborhoods.
Research consistently has shown that children in poorer neighborhoods tend to have high
levels of risk and harm (Hines et al., 2004). Hence, when caseworkers investigate
children for alleged maltreatment, neighborhood factors should be considered.
Regression results predicting use of needed services identified significant child
and neighborhood factors. Substantiated maltreatment was linked to receiving more
needed services. In addition, living in foster care settings led to use of more needed
services. Children in foster care were in more frequently contact with caseworkers due to
a lack of supervision from caregivers, enabling caseworkers to assess and respond to their
needs. Abandonment as a significant predictor of service use can be explained by the
results related to involvement in foster care. Children who experienced abandonment are
more likely to enter in foster care, in which they are prone to receive more needed
services. A higher level of service needs was associated with a higher level of use of
needed services. This finding indicates the important role of needs assessment to meet
children’s needs through child welfare services.
None of the caregiver factors was significantly associated with use of needed
services. This study examined the effects of the demographic characteristics of caregivers
on use of needed services. Future research can explore other characteristics of caregivers,

33
such as the child–caregiver relationship and employment status to explain the effect of
caregiver factors on needed service use. Lastly, living in more problematic neighborhood
environments impeded use of needed services. The literature has demonstrated that when
children and families reside in poor neighborhoods, they often experience a lack of
available services in their geographical area (Hines et al., 2004). This study confirmed
this finding using secondary child welfare data.
Black, Hispanic, and White children exhibited distinct predictors of use of
needed services. Being in foster care and abandonment were determinants of needed
service utilization among White children. Living in foster care and other out-of-home
arrangements were the strongest determinants of use of needed services by Black children.
For Hispanic children, substantiation and neighborhood problems were most likely to
increase the use of needed services. These findings imply that child welfare agencies
should understand racial/ethnic composition of their clients and differentiate their
strategies to meet the service needs of clients by racial/ethnic groups of clients.
This study has both strengths and limitations. First, this study used nationally
representative data, enabling findings to be generalized to children involved with the
child welfare system in the United States. However, this strength also represents a
limitation of this study. Due to the use of secondary data, I could not include some
variables of interest such as family income and trauma experienced by children, which
may be predictors of use of needed services. However, this study did include caregiver
education level, neighborhood problems, and behavioral/emotional problems among
children, which are proxies of some variables that were not included. Further, children

34
aged 1 year or younger were oversampled in NSCAW data, comprising approximately 20%
of the total sample. To avoid any bias that may be caused by oversampling, this study
excluded the youngest children and findings represent children who were 2 years old or
older at the time of the survey.
This study contributed to knowledge on racial/ethnic disparities in needed service
use in child welfare by providing a comprehensive understanding of the extent to which
child welfare agencies meet the needs of children from diverse racial/ethnic backgrounds.
In contrast, by measuring the percentage of use of needed services, this study did not
determine whether child welfare agencies met children’s needs for individual services.
This issue should be addressed in future research. This study only included Black,
Hispanic, and White children in the study sample due to small subsamples of other
racial/ethnic groups. As racial/ethnic diversity of children in the child welfare system
increases, researchers should explore how child welfare agencies provide services to
other racial/ethnic groups of children, such as Asian and Native American children.
Two important implications for managers and practices emerged. First, child
welfare agencies may not consider the importance of service needs when considering
whether to provide child welfare services. White children reported the highest level of
need but exhibited the lowest level of needed service use. Managers in child welfare
agencies should examine their decision-making process to account for service need.
Second, concerning the findings related to racial/ethnic subsamples, child welfare
agencies should strive to equip themselves to be culturally responsive based on their
client population to achieve service equity.

35
CHAPTER 4: THE ROLE OF RACIAL/ETHNIC MATCHING OF
CASEWORKER-CLIENT DYADS AND PROFESSIONALISM IN CLIENT
SATISFACTION: MODERATING EFFECTS OF ENGAGEMENT AND
STRESS IN CHILD WELFARE AGENCIES
Introduction
As racial/ethnic diversity has become salient in the child welfare system,
addressing the issue of culturally responsive services has been paid a great deal of
attention by researchers, practitioners, and policy-makers. The literature suggests that
culturally responsive services are highly dependent upon the capacity of caseworkers to
provide services in clients’ language, or consistent with clients’ cultural norms and
expectations as human services are mainly delivered by caseworkers (Hall, Guterman,
Lee, & Little, 2002; Perry & Limb, 2004; Pinderhughes, 1991; Smith & Donovan, 2003).
In light of this, matching race/ethnicity of caseworker and client has been regarded to be
one of the main efforts in responding to the racial/ethnic diversity of clients (Perry &
Limb, 2004; Font et al., 2012).
Although a large body of literature supports the significance of helping
relationships in culturally responsive services (Chapman, Gibbons, Barth, McCrae, & the
NSCAW research group, 2003; Guerrero, 2010; Mor Barak & Travis, 2009), mixed
findings exist on whether racial/ethnic matching of caseworkers and clients benefits
outcomes of clients from racially/ethnically diverse backgrounds (Cabral & Smith, 2011;
Gamst, Dana, Der-Karabetian, & Gramer, 2004; Manoleas, 1994; Perry & Limb, 2004;
Shin, Chow, Camacho-Gonsalves, Levy, &Leff, 2005). Previous studies show that

36
racial/ethnic matching of caseworker-client plays a role in improving client outcomes
(Hall et al., 2002; Guerrero, 2010). However, some researchers argue that client
satisfaction is determined by how competent caseworkers are in their professions rather
than racial/ethnic matches (Cabral & Smith, 2011). They assert that professionalism is
deemed a key in managing clients from diverse racial/ethnic backgrounds effectively.
Despite considerable debate, little empirical knowledge on the influence of racial/ethnic
matching of caseworker-client dyads and professionalism on client satisfaction as one of
crucial outcomes in child welfare has been accumulated to date.
Research on worker attitudes indicates that workers who perceive a low level of
work engagement and a high level of role stress produce detrimental work outcomes
(Mor Barak, Travis, Pyun, & Xie, 2009; Glisson & Green, 2011; Glisson, Green, &
Williams, 2012). When caseworkers perceive themselves well engaged in their work,
they are inclined to yield positive work outcomes (Mor Barak et al., 2009). It has been
well recognized that the working environment of child welfare workers is quite stressful
(Mor Barak et al., 2009; Weaver, Chang, Clark, & Rhee, 2007). The stressful working
environment has the potential to lead to the unsatisfactory performance of child welfare
workers. As engagement and stress among caseworkers are crucial issues in child welfare
agencies, the relationship between racial/ethnic matching, professionalism and client
satisfaction can be affected by caseworker engagement and stress. This study argues that
engagement and stress may moderate the relationship between racial/ethnic matching of
caseworker-client, professionalism and client satisfaction. When a caseworker is less
engaged in his/her work and highly stressed in his/her working environment, racial/ethnic

37
matching and professionalism may not play a role in promoting client satisfaction.
However, the moderating role of engagement and stress has not been examined in child
welfare research. In these contexts, the present study aims to demonstrate the role of
worker–level factors associated with the racial/ethnic diversity of clients – with particular
attention paid to racial/ethnic matching of caseworker–client dyads and professionalism –
in client satisfaction and the moderating role of engagement and stress in the relationship
between the worker–level factors and client satisfaction. The research questions of this
study are as follows:
Research Question 1. What is the role of racial/ethnic matching of caseworker-
client dyads and caseworker professionalism in client satisfaction?
Research Question 2. How do engagement and stress among caseworkers
moderate the relationship between racial/ethnic matching of caseworker-client
dyads, caseworker professionalism, and client satisfaction?
Theoretical Background: Similarity between Worker and Client
The impact of similarity of dyads has typically been grasped through three
primary theories: Similarity-attraction, social identity theory and relational demography
theories. According to the similarity-attraction theory, which was developed to account
for dyadic relationships, dyads that are similar in terms of gender, race/ethnicity, values
and attitudes tend to have a more successful relationship than those with different
attributes (Byrne, 1971, DuBois & Karcher, 2005; Williams & O’Reilly, 1998).
Empirical research indicates supportive findings that similarity between dyads facilitates
interpersonal liking (Flaskerud & Liu, 1991; Manoleas, 1994; Perry & Limb, 2004).

38
As a major theory that takes into account the role of similarity in groups, social
identity theory assumes that people are likely to classify themselves and others into
certain categories (Hogg, 2006; Tajfel, 1982, Stahl, Maznevski, Voigt, & Jonsen, 2009).
This social categorization creates stereotyping by favoring the members of their own
groups and judging the members of other groups (Stahl et al., 2009). In the context of
caseworker-client relationships, when a client treats his/her caseworker as a member of
his/her racial/ethnic group, the client may favor his/her caseworker, suggesting that it has
the potential of developing into a quality working alliance. Those two theories support
the idea that the racial/ethnic matching of caseworker-client can be an effective way of
developing a quality working alliance.
On the other hand, relational demography theory explains the importance of
relative similarities. People compare their own demographic characteristics with those of
others in their dyadic relationships so as to detect any similarities with their pairs in terms
of demographic characteristics (Tsui, Egan, & O’Reilly, 1992; Tsui& O’Reilly, 1989).
That is, the extent of similarity between themselves and their pairs is determined by their
perceptions regardless of the absolute demographic similarities (Riordan & Shore, 1997).
The theory proposes that the relative similarity has an impact on the quality of their
relationships. Caseworker–client matching in race/ethnicity using the absolute similarity
may not be beneficial in the relationship with a caseworker. Rather, the extent to which
caseworkers are knowledgeable and competent may be a key to improve client
satisfaction (Lee & Ayón, 2004) and service outcomes (Trotter, 2002).

39
Considerable Debate: Racial/Ethnic Matching and Professionalism
In an effort to respond to the racial/ethnic diversity of clients, a large body of
research has been conducted to examine the impact of racial/ethnic matching between
caseworker and client on client satisfaction with caseworker services in human services.
Interestingly, little consensus on the effects of racial/ethnic matching exists. There are
some studies on worker–client matching in race/ethnicity showed improved service
outcomes (Flaskerud & Liu, 1991; Hall et al., 2002; Manoleas, 1994). Flaskerud and Liu
(1991) demonstrated the positive association between therapist-client matching in
ethnicity and treatment utilization, and a decrease in the dropout rates while gender and
language matching did not significantly affect the outcomes. Hall and his colleagues
(2002) also illustrate the positive role of ethnicity matching in lowering dropout rates of
children who attended the treatments. The findings imply that workers and clients of the
same race/ethnicity can easily form a rapport in the initial stage of intervention and share
similar cultures, values, and attitudes that can help in understanding clients. Those
commonalities facilitate interacting with clients and subsequently improving client
outcomes. The current literature has focused on the racial/ethnic matching effects on
service retention (Guerrero, in press; Guerrero & Andrews, 2011) and client functioning
(Guerrero, Campos, Urada, & Yang, 2012), and has not explored the impact of
racial/ethnic matching on client satisfaction. Evaluating client satisfaction is critical in
that client satisfaction has information on how much caseworkers successfully meet
client needs. Yet, little effort has been devoted to understanding the relationship between
racial/ethnic matching of caseworker–client and client satisfaction in the child welfare

40
contexts.
A notable number of studies showed conflicting evidence of the matching effects
(Cabral & Smith, 2011; Gamst et al., 2004; Shin et al., 2005). A meta-analysis of Shin
and his colleagues (2004) involved 10 studies regarding the impact of racial/ethnic
matching of worker–client on outcomes. These authors found no significant effects of
racial/ethnic matching on treatment retention and posttreatment functioning. A
subsequent meta-analysis by Cabral and Smith (2011) summarized 52 studies. They
revealed that clients significantly preferred a therapist of their own race/ethnicity. Yet,
racial/ethnic matching of worker–client did not show benefit to client outcomes. In child
welfare, Gamst and his colleagues (2004) indicate that when other factors were controlled
for, ethnicity matching was not significantly associated with client outcomes in child
welfare. Those studies suggest that professionalism plays a stronger role in client
satisfaction. If workers are well trained on how to interact with clients from different
cultures in the training system, the highly competent workers can work effectively with
their clients of a different racial/ethnicity. Researchers have defined professionalism of
caseworkers using several indices: whether a worker hold a social work degree, a
graduate degree or a professional license (Jones & Okamura, 2000; Perry, 2006) and the
percentage of workers with a graduate degree, a social work degree, or a professional
license in an agency (Healy, 2004). Given that the interest of this study involves
caseworker–level factors, this study defined professionalism as a graduate degree and a
social work degree at the caseworker–level.
Along with the controversial results from the literature and the research gaps, the

41
roles of race/ethnicity matching, and caseworker professionalism in client satisfaction in
the child welfare system are unclear. Hence, an examination of the effects of the two
factors in an analytical model which illustrates which factor is more important in dealing
with children from diverse racial/ethnic backgrounds needs to be carried out. Based on
the literature and the child welfare contexts, this study assumes that racial/ethnic
matching and professionalism are important factors to interact with clients from diverse
racial/ethnic backgrounds effectively. Therefore, this study constructs and examines the
following hypotheses.
Hypothesis 4.1. Racial/ethnic matching between caseworkers and clients is
positively associated with client satisfaction.
Hypothesis 4.2. Professionalism is positively associated with client satisfaction.
Other factors which may affect having quality relationships between children and
caseworkers are engagement and stress among caseworkers. Engagement in their work
may play a role in promoting client satisfaction and stress in their work may interfere
with improving client satisfaction (Aarons & Sawitzky, 2006; Glisson & Green, 2011;
Kim & Stoner, 2008). Child welfare workers’ working environment has been recognized
for being a very stressful one due to excessive workloads and insufficient rewards (Mor
Barak et al., 2009). Literature on child welfare workers consistently demonstrated that
workers perceive a high level of stress and a low level of engagement (Lizano & Mor
Barak, 2012). By considering child welfare working environments, this study pays
attention to moderating effects of engagement and stress on the relationship between the
worker–level factors and client satisfaction. Interactions between racial/ethnic matching,

42
professionalism, engagement and stress may be critical to explain client satisfaction with
caseworker. Based on the existing literature and the current circumstances child welfare
workers confront, those important factors may interact closely with one another. Thus, it
can be assumed that interactions between the factors more effectively affect client
satisfaction than direct effects of each factor. However, there is no study examining the
interaction effects on client satisfaction. In these contexts, this study examines the
following hypotheses.
Hypothesis 4.3. Engagement moderates the relationship between racial/ethnic
matching of caseworker–client and client satisfaction.
Hypothesis 4.4. Engagement moderates the relationship between
professionalism and client satisfaction.
Hypothesis 4.5. Stress moderates the relationship between racial/ethnic matching
of caseworker–client and client satisfaction.
Hypothesis 4.6. Stress moderates the relationship between professionalism and
client satisfaction.
Client satisfaction has been regarded as an important index to measure quality of
interpersonal relationships. Existing studies on helping relationships in child welfare have
mainly used client satisfaction (Boer & Coady, 2007; Lee & Ayón, 2004; Winefield &
Barlow, 1995). In child welfare agencies, the relationship between caseworkers and
clients is more crucial in that caseworkers primarily interact with clients and are
generally a sole venue to provide them with child welfare services. That is, client
satisfaction with caseworker in child welfare is associated with emotional support and

43
services that caseworkers offered. In light of these contexts, three aspects of client
satisfaction with caseworker services are examined for this study: (1) empathy, (2)
service process, (3) service provision. Figure 4.1 shows a hypothesized model of Study 2.

Figure 4.1. Hypothesized Model of Study 2

Method
Sample
Wave 2 of the NSCAW II data was used for the analysis. Wave 2 is appropriate
to assess the relationships between caseworkers’ characteristics and client satisfaction
with caseworker than Wave 1 because it has information on service caseworkers who
have directly provided child welfare services to children and caregivers during the past
Professionalism
- Graduate Degree
Client Satisfaction
with Caseworker
- Empathy
- Service Process
- Service Provision

Racial/Ethnic
Matching
Engagement
Professionalism
- Social Work Degree
Stress

44
18 months. In Wave 1, investigative caseworkers who were associated with investigation
only were the informants and their characteristics are not directly associated with the
satisfaction measure. This study includes White, Black and Hispanic caregivers who have
children involved with the child welfare system. The NSCAW study collected
information regarding satisfaction with caseworker services from two main clients -
adolescents 11 years or order and caregivers in the child welfare system. However, the
adolescent sample has a lot of missing cases (N = 197) and a regression model of
adolescent satisfaction did not show a good fit. Thus, caregiver satisfaction with
caseworker services was used to examine the relationship between caseworker
characteristics and client satisfaction. Caregiver satisfaction was originally measured to
assess satisfaction with caseworker services have contacted during the past 18 months. In
order to ensure if caregiver satisfaction is directly linked to the caseworkers in the data,
caregivers who reported that they have contacted with two or more caseworkers during
the past 18 months were excluded. Lastly, 702 caregivers met those criteria and were
included in analysis.
Measures
Dependent variables
Caregiver satisfaction. Caregiver satisfaction with caseworker services was
measured in three aspects; 1) empathy, 2) service process 3) service provision. Caregiver
satisfaction with empathy was assessed with two items on a 4-point Likert scale (1 =
never to 4 = all of the time). The items are “in all your meetings or talks, do you feel the
caseworker listened to your concerns?”; and “do you feel the caseworker treated you with

45
respect?”. The reliability coefficient of the items was .78. Higher scores represent higher
satisfaction with empathy. Caregiver satisfaction with service process was gauged with
three items on a 4-point Likert scale (1 = very dissatisfied to 4 = very satisfied). The
items include “how satisfied have you been with the extent to which the caseworker
maintained contact with you? “how satisfied have you been with the extent to which the
caseworker invited you to relevant meetings?”; and “how satisfied have you been with
the extent to which the caseworker involved you in decision-making regarding the care of
the child?”. The Cronbach’s alpha for the three items is.87. Higher scores indicate higher
levels of caregiver satisfaction with caseworker services in service process. As the third
dimension of satisfaction with caseworker, caregiver satisfaction with service provision
was estimated using three items on a 5-point Likert scale (1 = strongly disagree to 5 =
strongly agree). The items are “I should have been given more time to make the changes
expected of me.”; “the services I was told to use should have been more helpful.”; “I
should have been offered more services.” In order to be consistent with the other two
satisfaction scales, the item scores for service provision were reverse-coded, enabling that
higher scores represent higher satisfaction with service provision.
Independent variables
Racial/ethnic matching. Racial/ethnic matching between caseworker and
caregiver was measured by comparing race/ethnicity of each caseworker with that of
client. (1 = racial/ethnic matching). In order to investigate if the effects of matching
differed by race/ethnicity, three dichotomous variables were created (White-White,
Black-Black, Hispanic-Hispanic).

46
Professionalism. Literature indicates that a graduate degree and a social work
degree are indicators of professionalism in child welfare (Perry, 2006). Graduate degree
was measured by providing information on their level of education (1 = masters’ degree
or higher). Social work degree was assessed with a dichotomous variable indicating
whether or not a caseworker had a social work degree (1 = yes).
Moderating variables
Engagement. Engagement of caseworkers was assessed with two subscales:
personalization with 5 items and personal accomplishment with 6 items. The items were
rated on a 5-point Likert scale (0 = not at all to 4 = a very great extent). Sample items for
personalization and personal accomplishment include “I feel I treat some of the clients I
serve as impersonal objects” – reverse coded and “I have accomplished many worthwhile
things in this job” respectively. The 5 items of personalization were reverse-coded to
make directions of all items for engagement the same. Then, all items were totaled to
form a scale. The items revealed a good internal validity of the scale (Cronbach’s alpha
= .78). Higher scores indicate higher engagement in their work.
Stress. Stress of caseworkers was indicated with three subscales: emotional
exhaustion with 6 items, role conflict with 7 items and role overload with 7 items. The
items were rated on a 5-point Likert scale (0 = not at all to 4 = a very great extent).
Sample items for emotional exhaustion, role conflict and role overload include “I feel like
I am at the end of my rope,” “Interests of the clients are often replaced by bureaucratic
concerns – e.g., paperwork” and “The amount of work I have to do keeps me from doing
a good job” respectively. Responses of all items were summed to create a scale. The

47
items showed a good internal consistency reliability in the sample (Cronbach’s alpha
= .93). Higher scores represent higher levels of stress.
Control variables
Gender matching of worker–client dyads was measured using a dichotomous
variable of whether a caseworker’s gender is the same as that of his/her client (1 = yes).
Child age in years and caseworkers’ working tenure were controlled for their effects on
the outcome variables.
Analysis
Considering the characteristics of the sampling design of the NSCAW, this study
accounts for sampling weights for all analyses. Descriptive and bivariate correlation
statistical analyses were performed to describe the sample characteristics and the
relationships with the variables. As a primary analysis, a weighted two-step hierarchical
OLS regressions were conducted STATA 12.0 to examine the relationships between
racial/ethnic matching of worker–client, caseworker professionalism, engagement and
stress, and client satisfaction. In the NSCAW data, children were nested within
caseworkers. However, the average children per caseworker are less than two children
(range = 1 – 5). In such data structure, examining the cluster impact of caseworker is not
an appropriate approach. Thus, caseworker variables were deemed individual level
variables for analysis.
Moderating models was examined to estimate the main effects of the
independent variables and the interaction effects of racial/ethnic matching,
professionalism, engagement and stress. The moderating variables as continuous

48
variables were mean-centered to avoid multicollinearity. The mean-centered variables
were then used to create the interaction terms (Holmbeck, 1997). Then, a weighted two-
step hierarchical OLS regression was conducted. In the first step, the direct impact of the
independent variables on client satisfaction was examined. In the second step, the
interaction terms were added. Like Study 1, this study used complete case analysis to deal
with missing data.
Findings
Table 4.1 reflects the sample characteristics of children, caregivers and
caseworkers, and descriptive statistics for the primary variables. 42.85% of the children
of the caregivers in the sample were male and the average age of the children was 7.38
years (SE = .17). The largest racial/ethnic group was White (44.49%), followed by
Hispanic (32.58%) and Black (22.93%). The large majority of the child sample was
located in-home care settings with their parents (87.86%). 5.91% of the adolescents were
placed in kin care settings. Foster care comprised 3.52% and other out-of home
arrangements were 1.90%. Of caregiver respondents, only 10.35% of them were male.
The average age of caregivers was 35.38 years (SE = .94). Of the caseworker respondents,
7.84% of the respondents were male. 30.7% of caseworkers were under 30 years of age.
27.40% of them were between 30 and 39 years. 22.16% of them were between 40 and 49
years. 19.76% of them were 50 years or older. The average working tenure of
caseworkers in the field of child welfare was 7.98 years (SE = .90).
As the primary variable of interest, racial/ethnic matching between children and
caseworkers comprised 64.62%. The percentage of gender matching of children-

49
caseworker dyads was 46.67%. For caregivers, 67.20% were matched with caseworkers
in terms of race/ethnicity. 87.83% of the caregivers worked with same gender
caseworkers. This reflects a large female portion in the sample of caregivers and
caseworkers. Regarding professionalism, 20.80% of the caseworker respondents held a
master’s degree or a higher level of degree. 35.12% of the caseworkers held a bachelor’s
or master’s degree in social work. Caseworker respondents were relatively well engaged
in their work with a mean of 2.86 (SE = .06) and were relatively highly stressed from
their work with a mean of 2.24 (SE = .10).
Caregivers were highly satisfied with their caseworkers in empathy (M = 3.43,
SE = .14) and were highly satisfied with their caseworkers in regard to service process (M
= 2.89, SE = .20). Caregivers were moderately satisfied with caseworkers in service
provision (M = 3.00, SE = .31). Among the three dimensions of satisfaction with
caseworker services, empathy was the most satisfactory dimension that caregivers
perceived.

50
Table 4.1
Sample Characteristics of Children, Caregivers and Caseworkers
Children
(N = 702)
Caregiver
(N= 702)
Caseworker
(N = 331)
% Mean (SE) % Mean (SE) % Mean (SE)
Male 42.85 10.35 7.84
Age
< 30
30 – 39
40 – 49
> 50
7.38 (.17) 35.38 (.94)
30.7
27.40
22.16
19.76

Race/ethnicity
Black
Hispanic
White

22.93
32.58
44.49

24.42
19.63
55.95

21.21
11.45
a
64.28

Placement
In home - Parent 87.86 84.39
In home - Kin care 5.91 4.33
Foster care 3.52 4.63
Other OOH arrangement 1.90 .72
Racial/ethnic Matching
White
Black
Hispanic
Total
75.42
41.03
53.65
64.62
81.27
47.53
40.33
67.20

Gender matching 46.67 87.83
Satisfaction w/ caseworker
Empathy
Process
Service Provision

3.43 (.14)
2.89 (.20)
3.00 (.31)

Graduate degree 20.80
Social work degree 35.12
Engagement 2.86 (.06)
Stress
Working tenure
2.24 (.10)
7.98 (.90)
Note. a. Other racial/ethnic groups of caseworkers (3.05%) were included in the sample to
maximize available data for analysis.

51
As presented in Table 4.2, caregiver satisfaction with caseworkers’ empathy was
strongly correlated with caregiver satisfaction with service process that caseworkers
provided (r = .76, p < .01) and was correlated with caregiver satisfaction with service
provision (r = .28, p < .05). Satisfaction with service process was also related to
satisfaction with service provision (r = .44, p < .01). Racial/ethnic matching between
caregiver and caseworker was positively associated with caregiver satisfaction with all
dimensions - empathy (r = .16, p < .01), service process(r = .19, p < .01) and service
provision (r = .29, p < .01). Whether or not caseworkers hold a master’s degree or higher
was not related to the satisfaction with all three dimensions. However, a social work
degree was found to correlate with the satisfaction with empathy (r = .14, p < .01) and
service process (r = .16, p < .01). Caseworkers’ engagement in their jobs was not
associated with caregiver satisfaction with caseworker services in regard to empathy,
service process and service provision. Negative correlations between stress and
satisfaction with empathy (r = -.17, p < .01), service process (r = -.15, p < .05),and
service provision (r = -.08, p < .05) were found. Gender matching of caregivers with
caseworkers was negatively associated with satisfaction with empathy (r = -.14, p < .01)
and satisfaction with service process (r = -.17, p < .01). Older children was linked to
improved caregiver satisfaction with empathy (r = .11, p < .05).

52
Table 4.2
Correlations Between Study Variables (N = 702)
1 2 3 4 5 6 7 8 9 10
1. CS – Empathy -
2. CS – Process .76** -
3. CS – SV .28* .44** -
4. CG- REM .16** .19** .29** -
5. Graduate degree -.02 -.08 -.09 -.21** -
6. SWD .14** .16** -.00 -.02 .01 -
7. Engagement -.02 -.07 -.00 .07 .17** -.06 -
8. Stress
9. CG – GM
10. Working tenure
11. Child age
-.17**
-.14**
-.03
.11*
-.15*
-.17**
-.07
.02
-.08*
-.03
.11*
-.05
-.06
.02
-.06
.14*
-.12*
-.06
.17**
.17**
.02
-.01
.04
-.10
-.48**
-.02
.28**
.23**
-
.06
-.28
-.07

-
.06
.02

-
.00
Note. CS = caregiver satisfaction; SV = service provision; CG = caregiver; REM = racial/ethnic
matching; SWD = social work degree; GM = gender matching
* p < .05, ** p < .01

The results of moderated regression analyses for the total sample are presented in
Table 4.3. In the direct impact model of satisfaction in regard to empathy, gender
matching between caregivers and caseworkers was negatively associated with satisfaction
with empathy (b = -.14, p < .05). A higher level of stress was marginally related to a
lower level of satisfaction with empathy (b = -.23, p < .10). Caregivers who have older
children showed a higher level of satisfaction with empathy (b = .22, p < .05). The model
for these variables was significant (F[8, 50] = 3.09 , p < .01), and they accounted for 13.9%

53
of the variance. In the moderating effect model, interaction terms were added. The results
revealed a non-significant increase in R
2
, although the overall model was significant
(F[14,44] = 4.94, p < .001) and accounted for 20.6% of the variance. The weak, but
significant interaction of caseworker engagement with racial/ethnic matching for
caregivers’ satisfaction with caseworkers’ empathy was found (b = .22, p < .10),
indicating that those who are matched with well engaged caseworkers from the same
racial/ethnic background were more likely to be satisfied with the quality of caseworkers’
empathy.
The direct effect model of caregivers’ satisfaction with service process revealed
similar results with the satisfaction with empathy. Gender matching of caregiver-
caseworker diminished caregivers’ satisfaction level in service process (b = -.18, p < .01).
Likewise, when caseworkers feel stress in their roles, it negatively influenced caregivers’
satisfaction with service process (b = .24, p < .10). But the relationship was marginally
significant. None of the other factors were significant in the direct impact model. The
model for the variables was significant (F[8, 48] = 4.20 , p < .01), and they accounted for
14.2% of the variance. In the moderating impact model, the interaction effects of
caseworkers’ engagement with racial/ethnic matching appeared to be marginally
significant (b = .26, p < .10). Regardless of a non-significant increase in R
2
, the
moderation model was significant (F[14,42] = 2.43, p < .05) and explained 22.7% of the
variance.
The analyses of the models pertaining to caregivers’ satisfaction with service
provision yielded different results from those of other two dimensions of satisfaction.

54
Racial/ethnic matching was the only significant predictor of service provision satisfaction
(b = -.32, p < .01). However, the direction of the relationship was against Hypothesis
4.1.The direct impact model for the variables was significant (F[8,50] = 2.57, p < .05)
and explained 12.8% of the variance. After adding the interaction terms, the moderation
model was significant (F[14,44] = 3.13, p < .05) and explained 16.6% of the variance.
But a non-significant increase in R
2
was found. Racial/ethnic matching was negatively
associated with service provision satisfaction (b = -.29, p < .01) as with the direct impact
model.
In summarizing the direct and interaction model results together, different
predictors of each satisfaction dimension were found. Overall, when caregivers are
racially/ethnically paired with caseworkers, caregivers’ satisfaction with caseworkers
services was not affected not supporting Hypothesis 4.1. Further, neither a graduate
degree nor a social work degree were significantly related to the satisfaction in any of the
three aspects, not supporting Hypothesis 4.2. Providing partial support for Hypothesis 4.3,
racial/ethnic matching weakly interacted with caseworkers’ engagement in their roles in
regard to caregiver satisfaction. The findings did not support Hypothesis 4.5. Caregivers’
stress had a negative impact on caregivers’ satisfaction with empathy and service process
but did not show a moderating role in the relationship between racial/ethnic matching,
professionalism and caregiver satisfaction, not supporting Hypotheses 4.4 and 4.6.

Table 4.3
Results of Moderated Regression Modeling for Caregiver Satisfaction
Caregiver SF
-Empathy
Caregiver SF
-Service Process
Caregiver SF
-Service Provision
Model 1 Model 2 Model 1 Model 2 Model1 Model 2
b SE b SE b SE b SE b SE b SE
Child variable
Child age .22* .04 .22* .03 .06 .07 .06 .05 .07 .07 .04 .08
Caregiver variables
Racial/ethnic matching
Gender matching
.09
-.14*
.46
.33
.12
-.16*
.39
.35
.13
-.18**
.92
.53
.13
-.17**
.68
.46
-.32**
.02
.82
.70
-.29**
.02
.78
.71
Caseworker variables
Graduate degree -.07 .50 -.10 .50 -.08 .93 -.15 .93 .02 .81 .05 .94
Social work degree .14 .43 .11 .40 .16 .79 .18 .66 .07 .85 .04 .87
Engagement
Stress
Working tenure
Moderating variables
R/E matching * engagement
R/E matching * stress
Graduate degree * engagement
Graduate degree * stress
Social work degree * engagement
Social work degree * stress
-.18
-.23†
-.03
.46
.29
.02
-.05
-.20
.03

.22†
.08
.19
.22
-.04
.13
.36
.28
.02

.78
.42
1.14
.90
.69
.42
-.19
-.24†
-.05
.82
.45
.05
-.09
-.18
.04

.26†
.13
.24
.16
-.06
.13
.59
.52
.05

1.49
.80
2.32
1.70
1.77
.86
.06
.05
-.18
.92
.60
.05
.11
.08
-.20†

-.09
-.10
.06
.19
-.03
-.07
.87
.59
.05

1.65
1.11
1.79
1.16
2.13
1.55
R
2
.139 .206 .142 .227 .128 .166
Note. SF = satisfaction; R/E matching = racial/ethnic matching.
† p < .10, * p < .05, ** p < .01, *** p < .001
55

56
Next, the moderation model results for each racial/ethnic group were presented to
compare the effects of the variables by race/ethnicity. Table 4.4 shows the results of the
moderation model analysis of empathy satisfaction by race/ethnicity. Gender matching (b
= -.15, p < .05) was negatively related to empathy satisfaction and child age (b = .18, p
< .05) was positively associated with empathy satisfaction among Black caregivers. In
addition, holding a social work degree (b = .49, p < .001) and stress of caseworker (b = -
.46, p < .001) were the strongest factors of empathy satisfaction for Black caregivers.
Engagement was negatively associated with empathy satisfaction (b = -.34, p < .05) and
working tenure was positively associated with empathy satisfaction (b = .12, p < .05).
None of the interaction effects for Whites were found. The model for Black caregivers
was statistically significant (F[14,56] = 95.76 p < .001) with 61.0% of the variance. For
Latino caregivers, the regression model was significant (F[14,51] = 23.23, p < .001) and
explained 68.8% of the variance. When caseworkers are stressed from their work,
caregivers’ satisfaction is reduced (b = -.21, p < .05). Caregivers with younger children
had a higher level of empathy satisfaction. Three interaction effects in the model for
Latino caregivers were revealed. Caseworkers’ stress moderated the impact of
racial/ethnic matching on satisfaction with empathy (b = -.31, p < .05). Caregivers’
engagement moderated the effect of a graduate degree (b = .68, p < .001) and a social
work degree (b = -.50, p < .05) on empathy satisfaction. For White caregivers, the model
was significant (F[14,51] = 4.06, p < .001) and explained 30.0% of the variance. Gender
matching was negatively linked to empathy satisfaction (b = -.21, p < .05) and caregivers
who have older children showed a higher level of empathy satisfaction (b = .34, p < .001).

57
Altogether, racial/ethnic matching and a graduate degree were not related to
empathy satisfaction for the three racial/ethnic groups. A social work degree was found to
be a significant factor for Black caregivers only. Caseworkers’ stress was linked to
minority caregiver satisfaction with empathy. To some extent, the moderating role of
engagement and stress in the relationship between racial/ethnic matching,
professionalism and empathy satisfaction was found among Hispanic caregivers. The
interaction effect of graduate degree and engagement on empathy satisfaction was the
strongest interaction factor.

58
Table 4.4
Results of Moderated Regression Modeling Results for Caregiver Satisfaction- Empathy
Black Hispanic White
b SE b SE b SE
Child–level variable
Age .18* .03 -.47* .07 .34*** .03
Caregiver–level variables
Racial/ethnic matching
Gender matching
-.01
-.15*
.34
.35
.15
-.12
.33
.41
-.03
-.21*
.35
.42
Caseworker–level variables
Graduate degree -.02 .28 -.01 .69 .06 .49
Social work degree .49*** .15 -.12 .37 .07 .57
Engagement
Stress
Working tenure
Moderating variables
R/E matching * engagement
R/E matching * stress
Graduate degree * engagement
Graduate degree * stress
Social work degree * engagement
Social work degree * stress
-.34*
-
.46***
.12*
-.06
-.01*
.10
.02
.10
.08
.46
.26
.01

.68
.55
.65
.49
.61
.50
-.20†
-.21*
.12

.03
-.31*
.68***
.15
-.50*
-.07
.34
.20
.05

.66
.69
1.11
1.09
1.36
.73
.03
-.16
.21†

.20
.16
.01
.44†
.03
.11
.47
.29
.03

1.02
.55
1.38
1.11
.87
.71
R
2
.610 .688 .300
Note. R/E matching = racial/ethnic matching.
† p < .10, * p < .05, ** p < .01, *** p < .001
Caregiver satisfaction with the service process that caseworkers implemented
was affected by several factors among caregivers from diverse racial/ethnic backgrounds.
The model for Black caregivers was significant (F[14,56] = 8.44, p < .001) and
accounted for 41.7% of the variance. Older children’s caregivers were more satisfied with

59
their caseworkers in terms of service process (b = .25, p < .01). As with the empathy
satisfaction results for Hispanic caregivers, the model was significant (F[14, 50] = 25.97,
p < .001) and accounted for 76.5 % of the variance. Racial/ethnic matching led to an
increase in satisfaction with service process of caseworkers (b = .28, p < .05). Three
interaction effects were found. The interaction between racial/ethnic matching and stress
has an impact on satisfaction with service process (b = -.32, p < .05). The interaction
between a graduate degree and engagement of caseworkers was associated with service
process satisfaction (b = .94, p < .001). When caseworkers held a social work degree and
were highly engaged in their work, caregivers were less likely to satisfy caseworkers’
service process (b = -.59, p < .01). For White caregivers, the regression model was
significant (F[14,49] = 29.78, p < .001) and explained 32.7% of the variance. Gender
non-matching led to an increase in the service process satisfaction (b = -.21, p < .01). The
stress level of caseworkers (b = -.27, p < .10), racial/ethnic matching (b = -.14, p < .10)
and social work degree (b = .18, p < .10) were weakly related to the satisfaction.
Caseworker stress moderated the relationship between racial/ethnic matching and the
satisfaction (b = .35, p < .01). A graduate degree is marginally associated with the
satisfaction.
Regarding service process satisfaction, mixed findings were found by
race/ethnicity. Hispanic caregivers showed a higher level of service process satisfaction
when they are racially/ethnically matched with caseworkers while others did not have a
significant relationship of racial/ethnic matching and service process satisfaction. Not
having a graduate degree and higher levels of engagement were significantly linked to

60
higher levels of satisfaction for Hispanic caregivers only. Significant interaction effects
for Hispanic caregivers were found. The directions of the relationships were different
than expected.
Table 4.5
Results of Moderated Regression Modeling for Caregiver Satisfaction– Service Process
Black Hispanic White
b SE b SE b SE
Child–level variables
Age
Caregiver–level variables

.25**

.06

-.61**

.11

.09

.04
Racial/ethnic matching
Gender matching
.11
-.01
.54
1.27
.28*
-.05
.67
1.47
-.14†
-.21**
.45
.65
Caseworker–level variables
Graduate degree .16† .59 -.16* 1.25 .07 .64
Social work degree .18 1.35 -.10 .84 .18† .60
Engagement
Stress
Working tenure
Moderating variables
R/E matching * engagement
R/E matching * stress
Graduate degree * engagement
Graduate degree * stress
Social work degree * engagement
Social work degree * stress
.00
-.25
.03

-.00
.21
-.05
-.00
-.03
.12
1.47
.95
.10

1.89
1.19
1.23
1.20
1.74
1.20
.11*
-.17
.05†

.22
-.32*
.94***
.34
-.59**
-.07
.89
.56
.07

1.64
1.18
3.04
2.06
2.06
1.14
-.07
-.27†
.07

.13
.35**
-.06
.17
.18
-.05
.85
.53
.06

1.69
.96
2.82
1.81
1.89
1.39
R
2
.417 .765 .327
Note. R/E matching = racial/ethnic matching.
† p < .10, * p < .05, ** p < .01, *** p < .001

61
Distinctive predictors of service provision satisfaction for the three racial/ethnic
groups were found. For Black caregivers, the model was significant (F[14, 56] = 43.62, p
< .001) and explained 68.7% of the variance. A lower engagement of caseworker (b = -
.60, p < .001) and caseworkers with higher tenure (b = .37, p < .01) were the strongest
predictors of service provision satisfaction. A social work degree was positively
associated with satisfaction with service provision (b = .25, p < .01). For Latino
caregivers, the regression model was significant (F[14, 51] = 4.41 p < .001) and
accounted for 74.8% of the variance caseworkers’ engagement was the strongest
predictor of satisfaction with providing services (b = -.48, p < .001). Social work degree
(b = -.28, p < .01) and child age (b = -.55, p < .01) were negatively associated with the
satisfaction. A lower level of stress increased the satisfaction (b = -.27, p < .05).
Regarding the interaction effects, when caseworkers with a non-social work degree are
less engaged in their work, it increased caregiver satisfaction with caseworker’s service
provision (b = -1.17, p < .001). A social work degree and stress showed a significant
interaction effect on client satisfaction as well (b = -.64, p < .001). Racial/ethnic
matching and higher level of stress interactively affected client satisfaction (b = .28, p
< .05). An interaction of a graduate degree of a caseworker and higher level of
engagement in work led to an increase in caregiver satisfaction with service provision (b
= .28, p < .05). For White caregivers, the model was significant (F[14, 51] = 4.41, p
< .001) and accounted for 14.6% of the variance. An interaction effect of a non-graduate
degree and being less stressed from their work on satisfaction with service provision was
found, although the relationship is weak (b = -.33, p < .10).

62
Taken together, racial/ethnic matching was linked to service provision
satisfaction for Black caregivers only. A non-social work degree improved the
satisfaction for Hispanic caregivers while a social work degree enhanced the satisfaction
for Black caregivers. Lower levels of engagement were associated with higher levels of
service provision satisfaction for Black and Hispanic caregivers. Higher levels of stress
reduced service provision satisfaction for Hispanic caregivers. Some interaction effects
on satisfaction with service provision were found for Hispanic caregivers only.

63
Table 4.6
Results of Moderated Regression Modeling for Caregiver Satisfaction– Service Provision
Black Hispanic White
b SE b SE b SE
Child–level variables
Age .20 .08 -.55** .10 -.03 .09
Caregiver–level variables
Racial/ethnic matching
Gender matching
.36*
-.13†
.86
.60
.12
-.04
.53
1.06
.33†
-.07
1.18
.98
Caseworker–level variables
Graduate degree .24 1.24 -.07 .98 -.08 .96
Social work degree .25* .71 -.28** .57 -.05 1.09
Engagement
Stress
Working tenure
Moderating variables
R/E matching * engagement
R/E matching * stress
Graduate degree * engagement
Graduate degree * stress
Social work degree * engagement
Social work degree * stress
-.60**
.00
.37**

.01
.11
-.13†
.16
.02
-.01
.96
.66
.04

1.26
1.04
1.03
1.26
.82
1.15
-.48***
-.27*
.26

-.31†
.28*
.65*
.49
-1.17***
-.64***
.71
.40
.09

1.72
1.01
3.99
3.30
3.50
1.25
.02
.06
.23

.05
.09
-.20
-.33†
.20
.22
.92
.59
.08

2.97
1.89
2.88
1.78
2.70
1.85
R
2
.687 .748 .146
Note. R/E matching = racial/ethnic matching.
† p < .10, * p < .05, ** p < .01, *** p < .001
Discussion
This study investigated the impact of caseworker–level factors including
racial/ethnic matching, professionalism, engagement, and stress on client satisfaction
with caseworker services among caregivers from diverse racial/ethnic backgrounds. The

64
three dimensions of caregiver satisfaction had distinctive predictors. First, a higher level
of stress was slightly related to a decrease in empathy satisfaction. This reflects previous
study findings indicating that caseworkers’ stress negatively affects their empathy
competence (Kinman & Grant, 2011). Gender matching was linked to a decrease in
empathy satisfaction. Given the fact that most of the caregivers and caseworkers are
female, female caseworker – male caregiver and male caseworker – female caregiver
were positively associated with caregiver satisfaction with empathy. This can be
explained by the queen bee syndrome that women have difficulty working with one
another (Staines, Tavris, & Jayaratne, 1974; Ragins & Scandura, 1994). In child welfare
contexts, female caregivers may feel uncomfortable with disclosing their concerns and
issues to female caseworkers. Thus, female caregivers may feel more comfortable with
interacting with male caseworkers. Although racial/ethnic matching and engagement
themselves are not significantly related to empathy satisfaction, the weak but significant
interaction effect of racial/ethnic matching and engagement on empathy satisfaction were
found.
The service process satisfaction model yielded similar results with the empathy
satisfaction model. Non-matching of gender and a lower level of stress were linked to an
increase in service process satisfaction. Likewise, engagement was a weak moderator of
the relationship between racial/ethnic matching and service process satisfaction. A
plausible explanation of the similarities is that the two dimensions of satisfaction are
more likely to focus on assessing caseworkers’ relationship building capacity than service
provision satisfaction.

65
Regarding service provision of caseworkers, racial/ethnic non-matching was
associated with satisfaction. This implies that race/ethnicity itself does not matter when
caseworkers deliver their services to clients. There were no moderating effects of
engagement and stress on the relationship between racial/ethnic matching,
professionalism and caregiver satisfaction with service provision.
The three racial/ethnic groups of caregivers showed their distinctive predictors of
satisfaction with caseworker services. In the Black caregiver model for empathy, holding
a social work degree and stress were the strongest predictors. In the Hispanic caregiver
model, the interaction of a graduate degree and engagement was the strongest factor and
stress was also strongly related to empathy satisfaction. In the White caregiver model,
child age was the strongest predictor among White caregivers. Stress was the strongest
factor affecting service process satisfaction and racial/ethnic matching was marginally
associated with satisfaction with service provision for White caregivers. Regarding
service process satisfaction, child age was the strongest factor for Black caregivers and
the interaction effect of graduate degree and engagement was the strongest factor for
Hispanic caregivers. The interaction effect of racial/ethnic matching and stress was the
strongest one for White caregivers. For White and Hispanic caregivers, stress
significantly moderated the relationship between racial/ethnic matching and satisfaction
with service process but the directions of the relationships were different. In service
provision satisfaction, Black caregivers were likely to be satisfied with their caseworkers
who were matched with race/ethnicity.
In sum, overall, the findings demonstrate that racial/ethnic matching did not

66
matter in caregiver satisfaction with caseworker services, although racial/ethnic matching
mattered service provision for Black caregivers. The results are in line with the relative
demography theory indicating demographic similarities such as race/ethnicity did not
affect interrelation building (Riordan & Shore, 1997). A graduate degree and a social
work degree for caseworkers did not play a crucial role in increasing caregiver
satisfaction in the three dimensions for the total sample. This implies that child welfare
agencies may have a good selection system and train their workers well. However, it
should also be noted that the professionalism factors were significant for some
racial/ethnic groups in some dimensions.
The extent to which caseworkers are engaged in their work and are stressed from
their work was slightly related to caregiver satisfaction with caseworker services. The
effects of caseworker engagement and stress were stronger for some racial/ethnic groups
in some dimensions. The findings are consistent with the literature on engagement and
stress (Glisson & Green, 2011; Lizano & Mor Barak, 2012). The findings showed that
stress had a stronger impact on caregiver satisfaction than engagement. These results can
be understood in that it has been well documented that child welfare workers work in a
highly stressful working environment.
The results of the moderating effects of engagement and stress were interesting.
When a caregiver is paired with a caseworker who is more engaged and has the same
race/ethnicity, caregiver satisfaction with empathy and service process were increased.
For Hispanic caregivers, caseworkers who have a masters’ degree or higher and are more
engaged in their work increased satisfaction with empathy, service process and service

67
provision. The interaction effects across the racial/ethnic groups should be noted when
child welfare agencies consider strategies in responding to the diverse needs of clients
from different cultural backgrounds.
There are several limitations of the present study. First, although the NSCAW
data provided this researcher with essential information on the primary variables so as to
examine the hypothesized model, the data have an innate issue. The NSCAW study
excluded children aged 10 years or younger for measuring child satisfaction with
caseworkers. Thus, only adolescent satisfaction was available but the regression model
for adolescents was not significant. However, this study examined caregiver satisfaction
with caseworker which can be regarded a proxy of child satisfaction. In general, data
collection from children should be used with caution because children’s responses are
sometimes not reliable. As an alternative to collect reliable data, child information is
frequently obtained from caregivers, teachers or caseworkers (Kapp & Propp, 2002). In
fact, the NSCAW study gathered the child data from multiple informants for this reason.
Also, the bivariate correlation analysis in this study showed that adolescent satisfaction
with empathy was strongly associated with caregiver satisfaction with empathy,
indicating that the caregiver satisfaction is a good proxy of child satisfaction in the data.
Moreover, caregivers are also primary clients.
Despite the limitation, the current study significantly expands prior research by
examining the impact of caseworker factors on client satisfaction with caseworker
services using the nationally representative data. To some degree, the analyses revealed
similar predictors of satisfaction in the three dimensions. However, different predictors

68
emerged as well. Overall, racial/ethnic matching and professionalism were not related to
caregiver satisfaction. Rather, stress marginally predicted caregiver satisfaction. More
importantly, the results of this study shed light on the moderating role of stress and
engagement in shaping racial/ethnic matching, professionalism, and client satisfaction
relations. The results have implications for managers seeking strategies to address
racial/ethnic diversity issues. Managers in child welfare agencies should consider
racial/ethnic matching, engagement, and stress simultaneously when they create policies
and procedures for providing culturally responsive services. That is, it is important to
understand that just one of the factors alone may not affect services to children and
families from diverse racial/ethnic backgrounds. Hence, when caseworkers are
racially/ethnically paired with clients, efforts to reinforce engagement and diminish stress
among caseworkers should be made.

69
CHAPTER 5: THE IMPACT OF ORGANIZATIONAL FACTORS
ASSOCIATED WITH CULTURALLY RESPONSIVE PRACTICES, AND
PROFICIENT AND RESISTANT CULTURES ON OUTCOMES IN THE
CHILD WELFARE SYSTEM
Introduction
Child welfare agencies have struggled with responding to the needs of children
from diverse racial/ethnic backgrounds. Extensive research on child welfare identifies the
conspicuous disproportionality of minorities in the child welfare system compared to
their representation in the general population (Chibnall, et al., 2003; Font et al., 2012;
McMillen et al., 2004) and racial/ethnic disparities in service provision to children
(Cheng & Lo, 2012; Dunbar & Barth, 2007; Garland et al., 2003; Hurlburt et al., 2004).
In addition to internal demands to address racial/ethnic disproportionality and disparities,
child welfare agencies are receiving external demands such as health care reform and an
increase in the racial, ethnic, national and language diversity of children and families that
pressure the agencies to respond to issues pertaining to the racial/ethnic diversity of
children. Those demands require child welfare agencies to implement culturally
responsive policies and practices.
To date, researchers have primarily focused on understanding current problems
related to the racial/ethnic diversity of children (Cheng & Lo, 2012; Chibnall et al., 2003;
Font et al., 2012; Wells et al., 2009). Yet, it is imperative to explore how child welfare
agencies react to the racial/ethnic diversity of children and understand the role of
organizational factors in child welfare outcomes. The lack of available multilevel data

70
that include information on clients nested in child welfare agencies has hindered this
process. However, the National Survey of Child and Adolescent Well-being (NSCAW)
data which include multilevel information have recently enabled researchers to examine
the relationship between organizational factors associated with providing culturally
responsive services, and service and child outcomes.
A few studies have examined potential organizational efforts to address the
racial/ethnic diversity issue (Chapin Hall Center for Children, 2008; Earner, 2007;
Guerrero, 2010; Mor Barak & Travis, 2009). These studies have recommended
organizational–level efforts such as educating staff, hiring minority staff, and establishing
organizational policies and procedures to serve culturally responsive services.
Organizational adaptation of those efforts can lead to improved quality of services for
clients of diverse cultural backgrounds (Chapin Hall Center for Children, 2008; Guerrero,
2010; Mor Barak & Travis, 2009). Although the need for culturally responsive services
has been recognized by researchers, managers, and policy-makers, there is no study to
date that has explored the role of organizational efforts in providing culturally responsive
services in the child welfare system. This study, therefore, fills this research gap by
investigating the impact of organizational efforts related to racial/ethnic diversity and
disparities on service and child outcomes in the child welfare system.
Responding to the institutional demands discussed earlier may involve executing
organizational changes to provide culturally responsive services. Research on
organizational change has underscored the notion that organizational culture that inspires
workers to implement new policies and practices is important to making changes in

71
organizations (Glisson, Green, & Williams, 2012; Guerrero & Kim, 2013). Child welfare
agencies are more likely to resist organizational change due to the bureaucratic structure
of child welfare agencies (Aarons, 2004; Jones & Okamura, 2000; Perry, 2006). In
organizations with bureaucratic structures, workers are prone to follow their supervisors’
or leaders’ directions and are not encouraged to pursue creative ideas for organizational
development. Workers in organizations that are more open to change and employ
creativity may efficiently respond to the racial/ethnic diversity of children by offering
culturally responsive services (Guerrero & Kim, 2013). Despite the importance of
organizational culture that encourages organizational change and the characteristics of
child welfare agencies, no study to date has explored the role of proficient and resistant
cultures in service and child outcomes among children from diverse racial/ethnic
backgrounds.
Child welfare agencies aim to improve the well-being of children involved in the
child welfare system through their services. It is assumed that use of needed services can
positively affect child outcomes (e.g., reducing child behavioral/emotional problems).
There is limited knowledge of the role of use of needed services in child outcomes. In
these contexts, this study aims to fill this gap by investigating the associations between
organizational factors associated with culturally responsive practices, and proficient and
resistant cultures, needed service use, and behavioral/emotional problems in the child
welfare system. The research questions that the current study sought to address are as
follows.

72
Research Question 1.What is the impact of organizational responses to
racial/ethnic diversity in child welfare on needed service use and child
behavioral/emotional problems?
Research Question 2. What is the impact of proficient and resistant
organizational cultures in child welfare agencies on needed service use and child
behavioral/emotional problems?
Racial/Ethnic Diversity and the Institutional Demands
Child welfare agencies have been pressured by institutional demands for
culturally responsive services to react to the racial/ethnic diversity among children in the
system. There is a substantial body of evidence indicating racial/ethnic disproportionality
and disparities in child welfare services (Cheng & Lo, 2012; Foster et al., 2011; Hill,
2007; see also the empirical evidence of racial/ethnic disproportionality and disparities
described in Study 2). The disproportionality and disparities play a role in internal
demands. A number of external demands including regulative and normative forces have
also required child welfare agencies to shape organizational responsiveness. As
regulative forces, health care reform and the Multiethnic Placement Act (MEPA) have
strongly emphasized that child welfare agencies should make efforts to diminish service
disparities among minority clients. Normative forces include globalization and the code
of ethics for social workers. Increasing globalization has increased racial/ethnic diversity
in the United States. This societal transformation with respect to racial/ethnic
composition is expected to further influence the racial/ethnic diversity of clients in child
welfare. Thus, child welfare agencies must embrace and respond to globalization by

73
meeting the service needs of racially and ethnically diverse individuals (Garland et al.,
2003; Graham, Bradshow, & Trew, 2009; Hasenfeld, 2009; McMillen et al., 2004). In
addition to increasing racial/ethnic diversity of children, the code of ethics enacted by the
National Association of Social Workers, which plays a crucial role in providing
guidelines for social work practice, notes that social workers are responsible for
providing services to all needy people regardless of racial and ethnic differences. The
code of ethics also mandates that social workers be responsible for understanding their
clients’ cultures and providing culturally responsive services. In the light of these
normative pressures, child welfare agencies are forced to respond to racial/ethnic
disparities and disproportionality.
In these contexts, there is growing interest in examining how child welfare
agencies address the service disparity issue of minorities in institutional environments.
However, to date, most literature on racial disparities in child welfare has focused on
examining the phenomenon of racial/ethnic disparities in child welfare service use
(Garland et al., 2003; McMillen et al., 2004) and has been limited by an inability to
demonstrate strategies of child welfare agencies to address the disparities. In response to
this research gap, this study examined the impact of organizational responses to
racial/ethnic disparities on service and child outcomes.
Organizational Responses in the Child Welfare System
For years, new institutional theory has intrigued organizational researchers in
their efforts to explain organizational responses to institutional pressures (e.g., DiMaggio
& Powell, 1983; Ingram & Simons, 1995; Kostova & Roth, 2002; Meyer, Scott & Strang,

74
1987). According to new institutional theory, organizations are required to gain
legitimacy through organizational responses to institutional demands to increase their
probability of survival. Organizations need to conform to the standards of propriety that
have become institutionalized in their fields (DiMaggio & Powell, 1983; Oliver, 1991).
According to new institutional theory, the survival potential of organizations relies on the
extent to which their structures reflect and reinforce institutional rules (Baum & Oliver,
1992; Hasenfeld, 2009). In the case of social service agencies, under institutional
demands related to issues of racial/ethnic diversity in service delivery, managers need to
seek strategies for implementing culturally responsive services that are consistent with
institutional expectations (Guerrero, 2010; Guerrero & Kim, 2013).
The literature suggests a variety of responses that child welfare agencies can adapt
in response to racial/ethnic diversity issues (Font et al., 2012; Korbin, 2002; Wells et al.,
2009). Organizational responses to racial/ethnic diversity often involve establishing
organizational policies and procedures. Leaders of child welfare agencies can provide
cultural competence trainings for their workers, add performance measures to agency
appraisal forms to evaluate the extent of culturally responsive services, and implement a
policy to match the race/ethnicity of workers–clients. In addition, child welfare agencies
can promote racial/ethnic diversity in their staffs based on the racial/ethnic composition
of clients and reach out to communities to understand the changing needs of children
from diverse cultural backgrounds. Those responses can have a positive impact on their
ability to meet the diverse needs of clients, which in turn may improve service and child
outcomes.

75
Hypothesis 5.1. Organizational responses to racial/ethnic diversity issues are
positively associated with use of needed services.
Hypothesis 5.2. Organizational responses to racial/ethnic diversity issues are
negatively associated with child behavioral/emotional problems.
Proficient and Resistant Culture in Child Welfare Agencies
Research has underscored the critical role of organizational culture in fostering
and implementing organizational changes (Glisson et al., 2012). Child welfare agencies
vary in organizational culture in work performance based on their characteristics (Webb,
Dowd, Harden, Landverk, & Testa, 2010). Child welfare agencies with an organizational
culture that inspires workers to adopt new changes may be more likely to respond to
concerns regarding racial/ethnic disparities in a timely manner. This study focuses on
proficient and resistant cultures that may affect service and child outcomes among
children from diverse racial/ethnic backgrounds. Proficient cultures have been defined as
“organizational cultures that expect caseworkers will place the well-being of each client
first and that caseworkers will be competent and have up-to-date knowledge” (Glisson et
al., 2012, p.625). Resistant cultures have been defined as “organizational cultures that
expect caseworkers will show little interest in new ways of providing services and
suppress any change effort” (Glisson et al., 2012, p.625). More proficient and less
resistant and cultures can have a positive impact on service and child outcomes. However,
there is no empirical study that has examined the relationship between proficient and
resistant cultures, and service and child outcomes among children from diverse
racial/ethnic backgrounds.

76
Hypothesis 5.3. Proficient culture is positively associated with use of needed
services.
Hypothesis 5.4. Proficient culture is negatively associated with
behavioral/emotional problems.
Hypothesis 5.5. Resistant culture is negatively associated with use of needed
services.
Hypothesis 5.6. Resistant culture is positively associated with
behavioral/emotional problems.
Child Welfare Services and Outcomes
The primary goal of child welfare agencies is to provide effective and client-
centered services to meet the various needs of children, thus improving their well-being.
Research consistently has indicated the need for further research to examine the
association between service outcomes and the well-being of children (Dunbar & Barth,
2007). A study on service use and child outcomes suggested service use of caregiver with
depression improved child outcomes (Burns et al., 2010). With the recognition of the
need, this study focused on the relationship between needed service use and child
behavioral/emotional problems. Improved child outcomes can be affected by the extent to
which children’s needs are met through child welfare services. As such, this study
examined the following hypothesis.
Hypothesis 5.7. Needed service use is negatively associated with child
behavioral/emotional problems.

77

Figure 5.1.Hypothesized Model of Study 3

Method
Sample

For the current research, Wave 1 and Wave 2 data of the NSCAW II were used
to examine the relationship between organizational factors and child outcomes. Weighted
response rates of Wave 2 were 82.8% for children, 86.3% for caregivers, and 93.9% for
caseworkers (Casanueva et al., 2012). Agency information from the local agency director
interview was gathered at Wave 1 only. Agencies with missing cases and children 1 year
old or younger were excluded. Further, children who Black, Hispanics or White, and who
provided information on Child Behavioral Checklist and use of needed child welfare
services were included. Of the second wave data, organizational culture and child- level
information from the 70 primary sampling units (N = 1,192 children) was included for
analysis.
Child Behavioral/
Emotional Problems
Organizational Responses
- Provision of Cultural
Competence Training
- Racial/ethnic Matching
- Performance Measures

Proficient Culture
Resistant Culture
Use of Needed
Child Welfare Services

78
Measures
Dependent variables
Service and child outcomes. As a service outcome, use of needed child welfare
services was measured in the same manner as Study 1. Child outcomes were
behavioral/emotional problems measured using the Child Behavior Checklist (CBCL).
This measure is described in Study 1. Total scale scores at Wave 2 were included in the
analysis.
Independent variables
Organizational responses to racial/ethnic diversity. Using agency-level
information from the NSCAW data, three dummy variables were included as
organizational responses to racial/ethnic diversity: (1) whether or not child welfare
agencies provided trainings to address racial/ethnic or cultural issues (1 = yes); (2)
whether or not child welfare agencies used performance measures to diminish racial
imbalance in placement (1 = yes); and (3) whether or not child welfare agencies
intentionally matched the race/ethnicity of caseworkers and clients (1 = yes).
Organizational culture. Proficient and resistant cultures that may affect child
outcomes were measured using items adapted from Glisson (2012). Responses from
caseworkers were aggregated to gauge agency-level organizational culture. I aggregated
measures of proficient and resistant cultures to the agency level of analysis.
Proficient culture was measured using two sub-scales - responsiveness (7 items)
and competence (8 items). Response options ranged from 0 = not at all to 4 = a very
great extent. For example, items of responsiveness and competence included “Members

79
of my organizational unit are expected to be responsive to the needs of each client” and
“Members of my organizational unit are expected to have up-to-date knowledge,”
respectively. Higher scores indicated higher levels of proficient culture. The reliability of
the proficient culture scale was .79. Interrater reliability (r
wg
) scores for the proficient
culture scale ranged from .66 to 1. Of the 70 agencies, only one child welfare agency
reported scores below the .70 rule of thumb (James, Demaree, & Wolf, 1993). Further,
the average r
wg
score for the scale was .89, indicating sufficient within-system agreement.
The type 1 intraclass correlation value was .29, indicating significant between-group
variance. These allowed for the use of agency-level proficient culture.
Resistant culture was measured using two subscales: apathy (6 items) and
suppression (7 items). Response options ranged from 0 = not at all to 4 = a very great
extent. For example, items of apathy and suppression were “Members of my
organizational unit are expected to not make waves” and “Members of my organizational
unit are expected to be critical,” respectively. Higher scores indicated higher levels of
resistant culture. Cronbach’s alpha of resistant culture was .76. Interrater reliability (r
wg
)
scores for the resistant culture scale ranged from .28 to 1. Of the 70 agencies, three child
welfare agencies reported scores below the .70 rule of thumb (James et al., 1993). Further,
the average r
wg
score for the scale was .83, indicating adequate within-system agreement.
The type 1 intraclass correlation value for that scale was .18, representing significant
between-group variance. These allowed for the use of agency-level resistant culture.
Control variables
Race/ethnicity of children was included as a control variable. (1 = Black, 2 =

80
White, 3 = Hispanic). Child’s age in years and placement (1 = in-home with parents, 2 =
kin care, 3 = foster care, 4 = group home or other) were also controlled for. Required
annual training for workers related to enhancing worker competence in each agency (1 =
none or less than 1day, 2 = 1-3 days, 3 = 4-10 days, 4 = 11 or more days) was also
included as an agency-level control variable.
Analysis
Two-level random intercept model analyses were conducted to test the
hypotheses. Multilevel modeling accounts for the clustering effect of organizational
factors, and allows for examining the impact of organizational factors, including
organizational responses to racial/ethnic diversity, and proficient and resistant cultures on
service and child outcomes. Multilevel regression modeling is a useful analytic tool for
data in which individuals are nested within groups. It allows researchers to take
advantage of the layered structure of the data by simultaneously examining the effects of
higher-level variables on a lowest-level outcome variable while adjusting for possible
correlation between errors that result from layered data (Hox, 1995). In the 2-level model,
level 1 was children and level 2 was agencies. At level 1, the ratio of needed service use
of each child in each agency was modeled as follow.
Y
ij
= π
0j
+ π
pj
X
pij
+ u
oj
+ e
ij

In this model, Y
ij
= ratio of needed service use for child i in agency j; π
0j
= the
mean ratio of needed service use across all children; X
pij
= the p measured factors of
individual children used to predict use of needed services (i.e. race/ethnicity,
substantiation, and placement); π
pj
= the corresponding coefficient that represents the

81
direction and strength of relationship between individual factor X
pij
on Y
ij
; and e
ij
= error
term, the residual from each within-person regression of the needed service use ratio.
Level-2 outcome variables were the π child-specific parameters.
π
00
= β
00
+ β
0q
X
qj
+ γ
0j
π
pj
= β
pj
+ γ
pj
In this model, β
00
= an intercept modeling the agency effect on π
0j
; X
qj
= the q
agency factors (i.e. cultural competence training initiative, racial/ethnic matching
initiative and performance measure initiative); β
0q
= the corresponding coefficient
between agency factor X
qj
on π
0j
; β
pj
= an intercept modeling the agency effect on π
pj
; and
γ
pj
= the effect of unmeasured agency factors that did not vary across children.
The model specification for child behavioral/emotional problems is the same as
that of use of needed services. The model specification above presents the 2-level random
intercept models examined in this study. It was assumed that the effect of child factors on
use of needed services remained across agencies. Analyses were performed using
STATA 12.0.

As recommended by Raudenbush and Bryk (2002), the multilevel analysis was
conducted in three steps. First, a null model to partition the total variability in use of
needed services into its within- and between- agency components was examined. Second,
the effect of children’s characteristics (level-1 variables) on ratio of needed service use
was investigated. Lastly, the impact of organizational factors (level-2 variables) on the
use of needed services was investigated. The multilevel analysis for child
behavioral/emotional problems was conducted using the same steps. As with the two

82
other studies in this dissertation, all analyses were weighted to account for the sampling
design of NSCAW.
Findings
Table 5.1 shows sample characteristics of child welfare agencies and children.
On average, child welfare agencies implemented 1.44 responses (SE = .94) to address
issues of racial/ethnic diversity and disparities. Specifically, 35.71% of the agencies
implemented one initiative, followed by two initiatives (34.29%), no initiatives (20.00%)
and three initiatives (10.00%). In addition, 74.29% of the child welfare agencies provide
their workers with cultural competence training, 40.00% of the agencies used
performance measure to assess the extent to which they addressed racial/ethnic diversity
issues appropriately, but 18.57% of agencies have the initiative of racial/ethnic matching
among children, caregivers and caseworkers. Child welfare agencies had a relatively high
proficient culture, with a mean of 2.69 (SE = .22) and low resistant culture, with a mean
of 1.28 (SE = .25). In addition, 37.14% of the agencies reported that they required 1 - 3
days of training for workers, 35.71% of them reported that 4 - 10 days of training are
required for their workers, 15.7% of them did not require any training for their workers,
14.29% of them required 11 or more days of training for their workers.
Average child age was 8.61 years (SE = .25). White children comprised 46.94%
of the sample. Hispanic children comprised 28.56%, Black children comprised 24.50%.
More than half of the children lived with their parents in home (67.83%), 16.07% of
children lived in foster care settings, 13.73% of children lived with their relatives, and
2.36% had other out-of-home placements such as a group home. In addition, 46.17% of

83
children were substantiated for maltreatment. On average, children received 62% (SE =
3%) of needed services. Their average score of behavioral/emotional problems was 56.86
(SE = .89).

84
Table 5.1
Sample Characteristics of Child Welfare Agencies and Children
Variable n Weighted % M SE
Child variables
Age 8.61 .25
Race/ethnicity
Black 413 24.50
Hispanic 367 28.56
White 358 46.94
Living arrangement
With parents 595 67.83
Kin care 249 16.07
Foster care 292 13.73
Other out-of-home placement 34 2.36
Maltreatment substantiation 889 46.17
Use of needed services .62 .03
Emotional and behavioral functioning 56.86 .89
Agency variables
OR- training 52 74.29
OR - racial/ethnic matching 13 18.57
OR - performance measures 28 40.00
Number of organizational responses 1.44 .94
0 14 20.00
1 25 35.71
2 24 34.29
3 7 10.00
Proficient culture 2.69 .22
Resistant culture 1.28 .25
Training requirement
None
1- 3 days
4 - 10 days
11 or more days
11
26
25
10
15.71
37.14
35.71
14.29

Note. OR = organizational response.

85
Two-level mixed effects models for use of needed service and
behavioral/emotional problems were conducted. Table 5.2 presents the results of the null
model analyses. The first null model indicated that the average rate of use of needed
services was .72 (SE = .01). The residual was .09 at the children level and the variability
of the agency mean was .28. Thus, the intraclass correlation coefficient (ICC) was .24
(.09 / [.09 + .28]), indicating the majority of the variation (.24) in use of needed services
among children was between agencies. In general, ICCs for the null model range
from .05 to .30 (Bliese, 2000) and the log likelihood test was significant (χ
2
= 29.98, p
< .001). These findings supported the use of multilevel modeling for use of needed
services. That is, use of needed services is different depending on which agency they
belong to.
The second null model showed that the average rate of behavioral/emotional
problems was .53.57 (SE = .44). The residual was 2.44 at the children level and the
variability of the agency mean was 22.98. Thus, the ICC was .18 (2.44/ [2.44 + 22.98]),
indicating the majority of the variation (.18) in behavioral/emotional problems among
children was between agencies. The log likelihood test was also significant (χ
2
= 19.85, p
< .001). The results supported the use of multilevel modeling for behavioral/emotional
problems is valid. That is, behavioral/emotional problems is different depending on
which agency they are involved with.

86
Table 5.2
Results of Null Model Analyses
Use of Needed Services Behavioral and Emotional
Functioning
Fixed Effect B SE B SE
Intercept .722*** .013 53.571*** .440
Random Effect B SE B SE
Intercept .086 .011 2.436 .431
Residual .275 .004 11.530 .203
*** p < .001

Results of the two-level random intercept model analyses are displayed in Table
5.3. Regarding use of needed services, analysis of child-level variables in Model 1
showed that living in foster-care and kin-care settings were found to increase use of
needed child welfare services. Living in other out-of-home care was marginally related to
use of needed services. The variance component corresponding to the residual increased
for the use of needed services model, reflecting the fact that including agency-level
variables accounted for some of the variance in the rate of use of needed services. The
estimate was significant, indicating that some variance remained unaccounted for. In
Model 2, agency-level variables were added. The findings showed that none of the three
organizational initiative variables were significantly associated with use of needed
services. The results did not support Hypothesis 5.1. The other two organizational

87
initiative factors, proficient culture, and resistant culture were not significantly related to
use of needed services, not supporting Hypothesis 5.3 and Hypothesis 5.5. When child
welfare agencies require more trainings for workers, use of needed services were
increased. Variance that was unaccounted for remained after the organizational factors
were entered.
In Model 1 for behavioral/emotional problems, living in foster care settings and
other out-of-home care settings were positively related to behavioral/emotional problems
of children (see Table 5.3). Use of needed services decreased child behavioral/emotional
problems, providing support for Hypothesis 5.7. Likewise, the residual was reduced,
indicating that child-level variables explained some of the variance in the
behavioral/emotional problems scores. However, the estimate was significant and some
variance remained unaccounted for. Organizational responses to racial/ethnic diversity,
proficient culture and resistant cultures did not have a significant relationship with
behavioral/emotional problems. The estimate was significant for Model 2, indicating that
some variance remained unaccounted for.

88
Table 5.3
Results of Multilevel Regression Models: Use of Needed Services (N = 1,192)
Model 1 Model 2
Fixed Effect B SE B SE
Intercept .622*** .024 .523*** .170
Child variables
Black -.007 .022 -.009 .022
Hispanic .039† .023 .036 .023
Maltreatment substantiation .027 .021 .030 .021
Foster care .109*** .021 .107*** .021
Kin care .090*** .022 .089*** .022
Other out-of-home placement .092† .051 .089† .051
Agency variables
OR - cultural competence training .019 .030
OR - racial/ethnic matching -.013 .031
OR - performance measures .005 .026
Proficient culture .060 .210
Resistant culture .052 -.310
Training requirement .029* .013
Random Effect B SE B SE
Intercept .066 .012 .058 .013
Residual .277 .006 .278 .006
† p < .01, * p < .05, ** p < .01, *** p < .001

89
Table 5.4
Results of Multilevel Regression Models: Child Emotional and Behavioral Functioning
(N = 1,192)
Model 1 Model 2
Fixed Effect B SE B SE
Intercept 60.749*** 1.253 65.355*** 6.353
Child variables
Black -1.036 .906 -.814 .923
Hispanic -1.465 .934 -1.260 .981
Maltreatment substantiation -1.445† .865 -1.362 .869
Foster care 3.384**** .894 3.338*** .894
Kin care .011 .945 .117 .948
Other out-of-home placement 7.138** 2.165 6.909** 2.173
Use of needed services -6.933*** 1.265 -6.878*** 1.270
Agency variables
OR - cultural competence training -.521 1.109
OR - racial/ethnic matching -.327 1.181
OR - performance measures -.512 .959
Proficient culture -2.802 2.235
Resistant culture 2.251 1.939
Training requirement .215 .498
Random Effect B SE B SE
Intercept 1.642 .566 1.571 .581
Residual 11.751 .258 11.742 .258
† p < .01, * p < .05, ** p < .01, *** p < .001

90
Discussion
By considering the multilevel structure of the child welfare agencies, this study
was the first to examine the impact of organizational responses to racial/ethnic diversity,
proficient culture and resistant culture on child welfare outcomes. One main finding of
this study was the non-significant role of organizational responses to racial/ethnic
diversity in service outcomes. The provision of training to address racial/ethnic diversity
issues was not an effective strategy to meet needs of clients from diverse racial/ethnic
backgrounds. A tenable explanation is that the measure did not include the quality of the
training. Some child welfare agencies may provide training but it may not be sufficient to
affect use of needed services for children and families from diverse racial/ethnic
backgrounds.
The organizational strategy of racial/ethnic matching was also not related to use
of needed services. This corresponds to findings in Study 2. Racial/ethnic matching is
considered an essential strategy to serve clients of diverse racial/ethnic backgrounds
because of an assumption that caseworkers of the same race/ethnicity can better
understand client concerns or problems based on understanding of their culture. However,
previous studies have demonstrated that cultural understanding is not critical to
delivering mental health services (Cabral & Smith, 2011; Shin et al., 2005).
Use of performance measures to estimate the quality of culturally responsive
services was not associated with use of needed child welfare services among children
from diverse racial/ethnic backgrounds. A possible explanation is that the performance
measures child welfare agencies employ may not be effective to measure culturally

91
responsive services. Although the data provided an opportunity to examine the effect of
the use of performance measures, they did not include information on what performance
measures child welfare agencies use and the quality of those measures. Thus, further
research is needed to understand the effect of existing performance measures regarding
culturally responsive services and their effect on service and child outcomes.
Three organizational responses to racial/ethnic diversity issues were not
associated with behavioral/emotional problems of children with diverse racial/ethnic
backgrounds. This finding should be interpreted with caution. Racial/ethnic diversity
issues may not be a top priority in some child welfare agencies. The racial/ethnic
composition of children in child welfare agencies varies by geographical locations.
Controlling for the factor indicating the extent of racial/ethnic diversity in agencies'
communities may allow a better understanding of the relationships between
organizational factors, and service and child outcomes. In addition, the nature of
behavioral/emotional problems should be considered when interpreting the findings.
Unlike service provision, it is difficult to improve such functioning of children. This may
explain why organizational strategies were not related to children’s behavioral/emotional
problems.
The findings indicated that proficient and resistant organizational cultures were
not significantly associated with use of needed services and child behavioral/emotional
problems. A previous study found similar results regarding caseworker attitudes that have
significant potential to affect child welfare outcomes using NSCAW data (Glisson et al.,
2012). These relationships should be further examined by identifying factors that affect

92
the relationships between organizational cultures and child welfare outcomes. This study
found child characteristics that were significantly associated with the two outcomes.
Future research should note child-level factors to identify predictors of the two outcomes.
An important finding of this study was the significant relationship between use of
needed services and child behavioral/emotional problems. To what extent child welfare
agencies meet client needs was positively linked to improvements in functioning. The
findings confirmed the pivotal role of meeting the needs of White and minority children
in achieving child wellbeing. In the contexts of racial/ethnic diversity, child welfare
agencies should strive to identify the specific needs of children from different cultural
backgrounds.
Among control variables, the organizational effort to enhance professional
competence of caseworkers was positively associated with use of needed services. This
implies that child welfare agencies should consider caseworker competence as a key to
improving organizational performance.
This study has some strengths. The findings of this study contributed to child
welfare research by connecting organizational factors with service and child outcomes.
The existing literature has noted the lack of understanding of these relationships and the
need to bridge organizational factors and client outcomes. This study responded to this
need using the nationally representative multilevel data, resulting in implications for
research, management and practice. Second, this study considered a time lag of 18
months to examine the relationship between organizational responses and outcomes. This
somewhat resolved a limitation of cross-sectional designs, which cannot examine causal

93
relationships between variables. It should be noted that proficient and resistant cultures,
and the outcome variables were measured at the same time point.
Some limitations of this study should be considered when interpreting the
findings. First, although the nationally representative multilevel data allowed for
conducting a multilevel modeling analysis, the sample size of agencies was relatively
small (N = 70) to examine the clustering effect of organizational factors. Although the
literature has indicated the sample size of level-2 data (e.g., agencies) of 35 – 60 can be
acceptable (Muthén & Muthén, 1998-2012), a larger sample size would be ideal and
produce more accurate estimations of the clustering effect. The second limitation
concerns the validity of the organizational measure. Organizational responses were
measured using dichotomous variables that did not allow for examining the effects of
their intensity on outcomes.
The findings of the current study have implications for research. First, this study
focused on how organizational responses to racial/ethnic diversity issues affect service
and child outcomes. However, there remains a lack of understanding what make child
welfare agencies respond to issues of racial/ethnic disparities. Therefore, future research
should examine predictors of organizational responses to address racial/ethnic diversity
issues. Second, as discussed earlier, this study sought to overcome a common limitation
of cross-sectional designs by using two-wave data. However, this study did not
demonstrate change patterns in needed service use and child functioning. A construction
of the following data collection of the NSCAW study will enable researchers to examine
change patterns of child functioning affected by organizational factors. This study

94
investigated the direct effects of organizational factors on service and child outcomes.
However, there is the possibility that needed child welfare services played a mediating
role between organizational factors and child outcomes. Organizational factors can affect
how the agencies meet clients’ needs, which in turn can affect child outcomes. Thus,
future research should address this. Lastly, there is a need to establish cultural
competence manual for child welfare agencies. To date, the literature on cultural
competence has attempted to develop modules and processes of cultural competence
(Grant & Haynes, 1995; Hernandez et al., 2009; Lynch & Hanson, 1992). However, the
cultural competence manual considering child welfare contexts should be developed in
future research so as to guide child welfare managers and caseworkers.
The findings also have implications for management and practice. This study
provided empirical evidence of how child welfare agencies respond to racial/ethnic
diversity issues. Given the growing attention paid to service equity in child welfare,
agencies should strive to respond to the diversity issues appropriately. The findings of
this study showed that agencies are less likely to implement organizational responses to
address disparities. Thus, child welfare agencies should initiate or execute new policies
and procedures such as hiring minority staff, offering cultural competence trainings,
developing language services and assessing their communities’ needs for culturally
responsive services in response to diversity issues. These strategies may lead to making
organizational changes. Child welfare agencies can also apply the planned organizational
change model. A meta-analysis study of the model showed that planned change
interventions were effective in terms of improving worker behavior and organizational

95
performance (Robertson, Roberts & Porras, 1993). Lastly, in line with prior research
conducted in child welfare settings, the extent to which the child welfare system meets
service needs played a significant role in child outcomes. In order to meet service needs
of clients effectively, identifying the diverse needs of clients should be prioritized and a
valid tool for needs assessment should be established.

96
CHAPTER 6: CONCLUSIONS AND IMPLICATIONS
During the past two decades, racial/ethnic diversity issues in the child welfare
system have received significant attention from researchers, managers, and practitioners
in terms of accomplishing service equity and meeting the diverse needs of clients from
different backgrounds. Although racial/ethnic diversity issues such as service disparities
have been well documented, organizational efforts to reduce such disparities has not been
examined in research to date. This study was the first to use nationally representative,
multilevel data to investigate the relationships between caseworker and organizational
factors that may help policymakers and practitioners reduce disparities and improve
outcomes in the child welfare system.
This dissertation focused on examining racial/ethnic diversity and disparities in
the child welfare system using a multilevel perspective. The dissertation featured three
independent studies that addressed three different aspects of the child welfare system in
the United States: (1) racial/ethnic disparities in use of needed services and client-level
predictors of needed service use; (2) the effect of caseworker factors – namely
racial/ethnic matching, professionalism, engagement and stress on client satisfaction; and
(3) the impact of organizational responses to racial/ethnic diversity and organizational
culture on service and child outcomes.
The three studies used two-wave National Survey of Child and Adolescent Well-
Being data gathered in the United States. The data set has several notable advantages.
The NSCAW is the first national data set in child welfare that allows generalizability of
findings to the entire U.S. population. Second, the data were gathered from multiple

97
informants, including children, caregivers, caseworkers and directors, resulting in a wide
range of information regarding children, families, and child welfare services. Third, the
hierarchical data structure allowed for multilevel modeling to examine the clustering
impact of agency-level factors. Finally, the longitudinal design of the NSCAW survey
allowed for the determination of causal relationships between organizational factors and
outcomes.
Summary of Major Research Findings
The first study provided an understanding of racial/ethnic disparities in the use of
needed child welfare services and the role of client-level factors in use of needed services.
An important finding of the study was that White children had a lower percentage of use
of needed services than Hispanic children. The finding was surprising given the common
belief that minority children struggle with service use due to factors such as a long
waiting-list for adoption and foster care. However, the finding is consistent with those of
the study recently conducted with similar measures for caregivers (Cheng & Lo, 2012).
The results can be interpreted in the relationship between needs assessment and service
provision. Caseworkers may pay more attention to Hispanic children who have high
levels of need, which in turn results in increased receipt of needed services. The findings
revealed the important role of substantiation, neglect (lack of supervision), abandonment
and foster care, and neighborhood problems in utilization of needed services.
The second study examined how child welfare agencies can respond to
racial/ethnic diversity issues, specifically focusing on caseworker–level factors. A
notable finding of this study was that racial/ethnic matching and professionalism were not

98
critical to increasing caregiver satisfaction with caseworker services. Rather, caseworkers’
stress and engagement had an impact on caregiver satisfaction. More importantly, the
relationships between racial/ethnic matching, professionalism and caregiver satisfaction
were moderated by stress and engagement. This may indicate a need for enhancing
engagement and mitigating stress among caseworkers when children have highly diverse
cultural backgrounds.
The third study shifted the focus of the investigation to the effects of
organizational responses to racial/ethnic diversity and organizational culture on use of
needed services and child behavioral/emotional problems. This study found that
implementation of organizational responses was not significantly related to service and
child outcomes among children from diverse racial/ethnic backgrounds. Further,
proficient and resistant cultures in child welfare agencies did not have an impact on the
two outcomes. An important finding of this study was that the extent to which child
welfare agencies meet the service needs of children from diverse cultural backgrounds
was related to improving their behavioral/emotional problems.
In all, the findings of this dissertation reflect the dynamic and multifaceted nature
of racial/ethnic diversity and disparities in the child welfare system in the United States.
The results of this study indicated that addressing racial/ethnic diversity issues involves
complicated processes because many agents and factors interact with one another.
Regarding racial/ethnic differences in use of needed services found in Study 1,
caseworker– and agency–level efforts to diminish the differences should be considered.
The findings of Study 2 indicated how caseworkers play a role in serving clients with

99
diverse racial/ethnic backgrounds. Specifically, it highlighted the importance of the
interaction effects of racial/ethnic matching, engagement and stress on client satisfaction.
The main findings of Study 3 revealed that the pivotal role of receipt of needed services
in improving child behavioral/emotional problems.
Strengths and Limitations
The dissertation research has some strengths. First, this study contributed to
racial/ethnic diversity research by investigating racial/ethnic diversity and disparities in
child welfare using a multilevel perspective. This enhanced our understanding not only of
racial/ethnic diversity issues but also the roles of caseworker and agency in addressing
these issues in the child welfare system. Second, the current study strived to connect
caseworker– and agency–level factors to service and client outcomes. A large body of
literature on child welfare has consistently suggested that there is a need to shed light on
the effect of worker and organizational factors on client outcomes. Using NSCAW data,
this study responded to this need by connecting caseworkers’ and organizational factors
to client satisfaction, use of needed service use and child functioning.
This study has some limitations. For the purpose of obtaining a sufficient sample
size for between-group comparisons, the present study did not include American Indian
and Asian groups. However, American Indians and Asian children comprise the fourth-
and fifth-largest groups of children in NSCAW data, respectively. Thus, it is necessary to
understand how American Indian and Asian children are involved in the child welfare
system in the future. On a related note, the analyses did not consider the characteristics of
Hispanic subgroups such as Mexicans, Cubans, Puerto Ricans and other Hispanics.

100
However, researchers have argued that significant differences exist in characteristics
among Hispanic groups (Guerrero et al., 2012). Thus, it would be interesting to explore
differences in service and child outcomes by Hispanic subgroups. Second, the NSCAW
study oversampled children aged 1 year or younger, who comprise 20% of the total
sample. This population was therefore excluded from this study to avoid any biased
interpretations of the findings. Thus, the findings of the dissertation research represent
children aged 2 years or older.
Implications for Research
The present study provides insight for developing culturally responsive services
for children of diverse cultural backgrounds. It is true that the current literature mainly
focused on demonstrating the need for culturally responsive services in the child welfare
system in response to the racial/ethnic diversity of clients. Although this dissertation is
one of only a handful of studies that have investigated efforts to solve racial/ethnic
diversity issues such as disparities using a multilevel perspective, further research on
culturally responsive services should be conducted. Using the NSCAW data, this study
relied on dichotomous variables indicating whether child welfare agencies have
organizational initiatives related to training provision, racial/ethnic matching, and
performance measures. Future research should use a standardized scale to appraise the
extent to which agencies provide culturally responsive services.
A great deal of research has tended to use a dichotomous variable to represent
whether or not a client uses a certain service (e.g., Burns et al., 2010; Garland et al., 2003;
Hurlburt et al., 2004). This measure overlooks the service use of clients in need. However,

101
the measure of use of needed child welfare services used in this study allowed this
research to more accurately assess the effectiveness of child welfare service provision by
considering the percentage of service use compared to service needs. This measure can be
further utilized in other contexts (e.g., children in foster care and Asian children). The
lack of studies of the measure highlights the need to demonstrate how child welfare
agencies respond to the needs of clients in specific child welfare contexts.
Another future consideration for research is associated with multilevel modeling.
Taking into consideration the multilevel structure of the child welfare system, this
dissertation research conceptually constructed a hypothesized model for each
independent study. Children and caregivers are generally nested in child welfare agencies.
However, Study 2 could not examine the clustering impact of caseworkers because
caseworkers did not have enough cases to conduct multilevel modeling analysis.
Therefore, there is a need to consider the nested structure of the child welfare system
when designing research and collecting data, thus allowing researchers to examine the
clustering effects of caseworker factors on client outcomes.
Implications for Management and Social Work Practice
Significant racial/ethnic differences in use of services when in need can inform
practitioners on how to promote service effectiveness. Child welfare agencies should
ideally increase the ratio of services receipt to service needs to 100%. To achieve this
goal, child welfare managers should analyze why children did not receive services when
in need. It is possible that services may not be available, that children may be placed on a
waiting-list, or that children refuse services. By understanding these factors through the

102
lens of race/ethnicity, child welfare agencies can better meet the needs of children from
different backgrounds and achieve service equity. Significant risk factors for children
including maltreatment substantiation, neglect (lack of supervision), abandonment, and
neighborhood problem led to increased use of needed services. The underlying
assumption is that caseworkers pay more attention to children at greater risk of negative
outcomes, which in turn may increase provision of needed services. Hence, child welfare
workers need to assess what specific services are required by children who have the risk
factors and to develop client-centered services based on their needs.
One important lesson learned from this study is that caseworkers play an
important role in addressing racial/ethnic diversity issues through the interaction effects
of caseworker factors. For managers to establish strategies to make their agencies
culturally responsive, they need to take into consideration creating more engaged and less
stressful working environments with implementing racial/ethnic matching and promoting
caseworkers’ professional knowledge. Another substantial lesson learned from this study
is that child welfare agencies are less likely to implement agency-level initiatives with
regard to mitigating racial/ethnic disparities. Given the prevalence of racial/ethnic
diversity and disparities in the child welfare system, these initiatives appear to be a
necessary strategy to providing culturally responsive services. Therefore, managers
should assess how much they need these various initiatives and implement the necessary
initiatives. In this regard, it is important to note that the quality of the initiatives may
make a difference in their effectiveness. Thus, managers should establish and execute
well-designed initiatives.

103
Lastly, managers in child welfare agencies should create a multidimensional plan
to properly respond to racial/ethnic diversity issues based on the multilevel dynamics in
child welfare agencies. This dissertation informed the role of caseworker– and agency–
level factors in reducing such disparities. In considering the findings, it is important to
implement all areas of the multidimensional plan simultaneously in order to maximize the
effectiveness of the policy and procedures. By so doing, child welfare agencies can move
toward becoming highly efficient agencies that provide culturally responsive services,
accomplish service equity and mitigate outcome disparities.

104
REFERENCES
Aarons, G. A. (2004). Mental health provider attitudes toward adoption of evidence-
based practice: The Evidence-Based Practice Attitude Scale (EBPAS). Mental
Health Services Research, 6(2), 61–74.
doi:10.1023/B:MHSR.0000024351.12294.65
Aarons, G. A., & Sawitzky, A. C. (2006). Organizational climate mediates the effect of
culture on work attitudes and staff turnover in mental health services. Admin
Policy Ment Health, 33(3), 289–301. doi:10.1007/s10488-006-0039-1
Achenbach, T. M. (1991). Manual for the child behavior checklist/4-18 and 1991 profile.
Burlington, VT: University of Vermont, Department of Psychiatry.
Alzate, M. M., & Rosenthal, J. A. (2009). Gender and ethnic differences for Hispanic
children referred to child protective services. Children and Youth Services Review,
31(1), 1–7. doi:10.1016/j.childyouth.2008.05.002
Baum, J. A. C., & Oliver, C. (1992). Institutional embeddedness and the dynamics of
organizational populations. American Sociological Review, 57, 540–559.
Biemer, P., Christ, S., Wheeless, S., & Wiesen, C. (2005). National survey of child and
adolescent well-being statistical user’s manual. Retrieved from
http://www.ndacan.cornell.edu/NDACAN/Datasets/UserGuidePDFs/092_Intro_to
_NSCAW_Wave_1.pdf
Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability
implication for data aggregation and analysis. In Klein, K. J. (Eds.), Multilevel
theory, research and methods in organization (pp. 349–381). San Francisco, CA:
Jossey-Bass.
Boer, C., & Coady, N. (2007). Good helping relationships in child welfare: Learning
from stories of success. Child and Family Social Work, 12, 32–42.
doi:10.1111/J.365-2206.2006.00438.x
Brooks-Gunn, J., Duncan, G. J., & Maritato, N. (1997). Poor families, poor outcomes: the
well-being of children and youth. In G. J. Dunan, J. Brooks-Gunn, Consequences
of growing up poor (pp. 1–17). New York: Russell Sage.
Burns, B. J., Mustillo, S. A., Farmer, E. M. Z., Kolko, D. J., McCrae, J., Libby, A. M. &
Webb, M. B., Dowd, K., Harden, B. J., Landsverk, J., & Testa, M. (2010).
Caregiver depression, mental health service use, and child outcomes. In Webb, M.
B., Dowd, K., Harden, B. J., Landsverk, J., & Testa, M. F. (Eds.), Child welfare
and child well-being (pp. 351–379). New York: Oxford University Press.
Byrne, D. (1971). The attraction paradigm. New York: Academic Press.

105
Cabral, R. R., & Smith, T. B. (2011). Racial/ethnic matching of clients and therapists in
mental health services: A meta-analytic review of preferences, perceptions and
outcomes. Journal of Counseling Psychology, 58(4), 537–554.
doi:10.1037/a0025266
Campbell, C. I., & Alexander J. A. (2002). Culturally competent treatment practices and
ancillary service use in outpatient substance abuse treatment. Journal of
Substance Abuse Treatment, 22(3), 109–119.
doi:10.1016/S0740-5472(02)00221-0
Casanueva, C., Wilson, E., Smith, K., Dolan, M., Ringeisen, H., & Horne, B. (2012).
NSCAW II wave 2 report: Child well-being. OPRE Report #2012-38, Washington,
DC: Office of Planning, Research and Evaluation, Administration for Children
and Families, U.S. Department of Health and Human Services.
Chapman, M. V., Gibbons, C. B., Barth, R. P., McCrae, J. S., & the NSCAW Research
(2003). Parental views of in-home services: What predicts satisfaction with child
welfare workers? Child Welfare, 82(5), 571–596.
Chapin Hall Center for Children. (2008). Understanding racial and ethnic disparity in
child welfare and juvenile justice. Chicago: Chapin Hall Center for Children at
the University of Chicago.
Cheng, T. C. (2010). Factors associated with reunification: A longitudinal analysis of
long-term foster care. Children and Youth Services Review, 32(10), 1311–1316.
doi:10.1016/j.childyouth.2010.04.023
Cheng, T. C., & Lo, C. C. (2012). Racial disparities in access to needed child welfare
services and worker–client engagement. Children and Youth Services Review, 34,
1624–1632. doi:10.1016/j.childyouth.2012.04.021
Chibnall, S., Dutch, N.M., Jones-Harden, B., Brown, A., Gourdine, R., Smith, J., et al.
(2003).Children of color in the child welfare system: Perspectives from the child
welfare community. Washington, DC: U.S. Department of Health and Human
Services, Administration for Children and Families and the Children’s Bureau.
Choi, S., & Ryan, J. P. (2007). Co-occurring problems for substance abusing mothers in
child welfare: Matching services to improve family reunification. Children and
Youth Services Review, 29(11), 1395–1410. doi:10.1016/j.childyouth.2007.05.013
Derezotes, D., Poertner, J., & Testa, M. (2005). Race matters in child welfare: The
overrepresentation of African American children in the system. Washington, DC:
Child Welfare League of America
Dettlaff, A. J., & Cardoso, J. B. (2010). Mental health need and service use among latino
children of immigrants in the child welfare system. Children and Youth Services

106
Review, 32(10), 1373–1379. doi:10.1016/j.childyouth.2010.06.005
Dettlaff, A. J., Earner, I., & Phillips, S. D. (2009). Latino children of immigrants in the
child welfare system: Prevalence, characteristics and risk. Children and Youth
Services Review, 31, 775–783. doi:10.1016/j.childyouth.2009.02.004
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional
isomorphism and collective rationality in organizational fields. American
Sociological Review, 48, 147–160.
Dowd, K., Kinsey, S., Wheeless, S., Thissen, R., Richardson, J., Suresh, R., Lytle, T.
(2003). National Survey of Child and Adolescent Well-being: Combined waves 1-
3 data file user’s manual. Ithaca, NY: Cornell University, National Data Archive
on Child Abuse and Neglect.
Dunbar, K., & Barth, R. P. (2007). Racial disproportionality, race disparity, and other
race-related findings in published works derived from the National Survey of
Child and Adolescent Well-Being. Annie E. Casey Foundation.
DuBois, D. L., & Karcher, M. J. (2005). Handbook of youth mentoring. Thousand Oaks,
CA: SAGE.
Earner, I. (2007). Immigrant families and public child welfare: Barriers to services and
approaches to change. Child Welfare, 86(4), 63–91.
Flaskerud, J. H., & Liu, P. Y. (1991). Effects of an Asian client-therapist language,
ethnicity and gender match on utilization and outcome of therapy. Community
Mental Health Journal, 27(1), 31–42. doi:10.1007/BF00752713
Font, S. A., Berger, L. M., & Slack, K. S. (2012). Examining racial disproportionality in
child protective services case decisions. Children and Youth Services Review, 34,
2188–2200. doi:10.1016/j.childyouth.2012.07.012
Foster, E. M., Hillemeier, M. M., & Bai, Y. (2011). Explaining the disparity in placement
instability among African-American and White children in child welfare: A
Blinder–Oaxaca decomposition. Children and Youth Services Review, 33(1), 118–
125. doi:10.1016/j.childyouth.2010.08.021
Gandara, P. (2010). The Latino education crisis. Educational Leadership, 67(5), 24-30.
Garland, A. F., Hough, R. L., Landsverk, J. A., McCabe, K. M., Yeh, M., Gander, W. C.,
& Reynolds, B. J. (2000). Racial and ethnic variations in mental health care
utilization among children in foster care. Children’s Services: Social Policy,
Research and Practice, 3(3), 133–146. doi:10.1207/S15326918CS0303_1
Garland, A. F., Landsverk, J. A., & Lau, A. S. (2003). Racial/ethnic disparities in mental
health service use among children in foster care. Children and Youth Services

107
Review, 25(5-6), 491–507. doi:10.1016/S0190-7409(03)00032-X
Gamst, G., Dana, R., Der-Karabetian, A., & Kramer, T. (2004). Ethnic match and
treatment outcomes for child and adolescent mental health center clients. Journal
of Counseling Development, 82, 457–465.
doi:10.1002/j.1556-6678.2004.tb00334.x
Glisson, C., & Green, P. (2011). Organizational climate, services and outcomes in child
welfare systems. Child Abuse & Neglect, 35, 582–591.
doi:10.1016/j.chiabu.2011.04.009
Glisson, C., Green, P., & Williams, N. J. (2012). Assessing the organizational social
context (OSC) of child welfare systems: Implications for research and practice.
Child Abuse & Neglect, 36(9), 621–632. doi:10.1016/j.chiabu.2012.06.002
Graham, J. R., Bradshow, C., & Trew, J. L. (2009). Addressing cultural barriers with
Muslim clients: An agency perspective. Administration in Social Work, 33(4),
387–406. doi:10.1080/03643100903172950
Grant, D., & Haynes, D. (1995). A developmental framework for cultural competence
training with children. Children Schools, 17(3), 171–182. doi:10.1093/cs/17.3.171.
Grella, C. E., Needell, B., Shi, Y. F., & Hser, Y. I. (2009). Do drug treatment services
predict reunification outcomes of mothers and their children in child welfare?
Journal of Substance Abuse Treatment, 36(3), 278–293.
doi:10.1016/j.jsat.2008.06.010
Guerrero, G. E. (2010). Managerial capacity and adoption of culturally competent
practices in outpatient substance abuse treatment. Journal of Substance Abuse
Treatment, 39(4), 329–339. doi:10.1016/j.sat.2010.07.004
Guerrero G., E., & Andrews, C. (2011). Cultural competence in outpatient substance
abuse treatment: Measurement and relationship with wait time and retention.
Drug Alcohol Depend, 119, 13–22. doi:10.1016/j.drugalcdep.2011.05.020
Guerrero, E. G. (2012). Organizational characteristics that foster early adoption of
cultural and linguistic competence in outpatient substance abuse treatment in the
United States. Evaluation and Program Planning, 35, 9–15.
doi:10.1016/j.evalprogplan.2011.06.001
Guerrero, E. G., Campos, M., Urada, D., & Yang, J. C. (2012).Do cultural and linguistic
competence matter in Latinos’ completion of mandated substance abuse treatment?
Substance Abuse Treatment, Prevention, and Policy, 7(34), 1–7.
doi:10.11866/1747-597X-7-34
Guerrero, E. G. (In press). Enhancing access and retention in substance abuse treatment:
The role of Medicaid payment acceptance and cultural competence. Drug and

108
Alcohol Dependence. doi:10.1016/j.drugalcdep.2013.04.005
Guerrero, E. G., & Kim, A. (2013).Organizational structure, leadership and readiness for
change and the implementation of organizational cultural competence in addiction
health services. Evaluation and Program Planning, 40, 74-81.
doi:10.1016/j.evalprogplan.2013.05.002
Hasenfeld, Y. Z. (2009). Human services as complex organizations. Newbury Park, CA:
Sage.
Hall, J., Guterman, D. K., Lee, H. B., & Little, S. G. (2002). Counselor-client matching
on ethnicity, gender and language: Implications for counseling school-aged
children. North American Journal of Psychology, 4(3), 367–380.
Healy, K. (2004). Social workers in the new human services marketplace: Trends,
challenges and responses. Australian Social Work, 57(2), 103–114.
Hernandez, M., Nesman, T., Mowery, D., Acevedo-Polakovich, I. D., & Callejas, L. M.
(2009). Cultural competence: A literature review and conceptual model for mental
health services. Psychiatric Services, 60(8), 1046–1050.
Hill, R. B. (2007). An analysis of racial/ethnic disproportionality and disparity at the
national, state, and county level. Washington, DC: Casey Family Programs.
Hines, A. M., Lemon, K., Wyatt, P., & Merdinger, J. (2004). Factors related to the
disproportionate involvement of children of color in the child welfare system: a
review and emerging themes. Children and Youth Services Review, 26(6), 507–
527. doi:10.1016/j.childyouth.2004.01.007
Hogg, M. A. (2006). Social identity theory. In J. Burke (Ed.), Contemporary social
psychological theories. Stanford. CA: Stanford University Press.
Holmbeck, G. N. (1997). Towards terminological, conceptual, and statistical clarity in the
study of mediators and moderators: Examples from the child-clinical and pediatric
psychology literatures. Journal of Counseling and Clinical Psychology, 65, 599–
610.
Hox, J. J. (1995). Applied multilevel analysis. Amsterdam: TT-Publicaties.
Hurlburt, M. S., Leslie, L. L. K., Landsverk, J., Barth, R. P., Burns, B. J., Gibbons, R. D.,
Slymen, D. J., & Zhang, J. (2004). Contextual predictors of mental health service
use among children open to child welfare. Arch Gen Psychiatry, 61, 1217–1224.
doi:10.1001/archpsyc.61.12.1217
Ingram, P., & Simons, T. (1995). Institutional and resource dependence determinants of
responsiveness to work-family issues. Academy of Management Journal, 38(5),
1466–1482. doi:10.2307/256866

109
James, L. R., Demaree, R. G., & Wolf, G. (1993). R
wg
: An assessment of within-group
interrater agreement. Journal of Applied Psychology, 75(2), 306-309.
doi:10.1037/0021-9010.78.2.306
Jones, L. P., & Okamura, A. (2000). Reprofessionalizing child welfare services: An
evaluation of Title IVE training program. Research on Social Work Practice,
10(5), 607–621.
Kapp, S. A., & Propp, J. (2002). Client satisfaction methods: Input from parents with
children in foster care. Child and Adolescent Social Work Journal, 19(3), 227–245.
doi:10.1023/A:1015580015223
Kim, H., & Stoner, M. (2008). Burnout and turnover intention among social workers:
Effects of role stress, job autonomy and social support. Administration in Social
Work, 32(3), 5–25. doi:10.1080/03653100801922357
Kinman, G., & Grant, L. (2011). Exploring stress resilience in trainee social workers: The
role of emotional and social competencies. British Journal of Social Work, 41,
261–275. doi:10.1093/bjsw/bcq088
Knott, T., & Donova, K. (2010). Disproportionate representation of African-American
children in foster care: Secondary analysis of the National Child Abuse and
Neglect Data System, 2005. Children and Youth Services Review, 32, 679–684.
doi:10.1016/j.childyouth.2010.01.003
Korbin, J. E. (2002). Culture and child maltreatment: cultural competence and beyond.
Child Abuse & Neglect, 26(6-7), 637–644. doi:10.1016/S0145-2134(02)00338-1
Kostova, T., & Roth, K. (2002). Adoption of an organizational practice by subsidiaries of
multinational corporations: Institutional and relational effects. Academy of
Management Journal, 45(1), 215–233. doi:10.2307/3069293
Lee, C.D. & Ayón, C. (2004) Is the client–worker relationship associated with better
outcomes in mandated child abuse cases? Research on Social Work Practice,
14,351–357. doi:10.1177/1049731504265833
Lizano, E. L., & Mor Barak, M. E. (2012). Workplace demands and resources as
antecedents of job burnout among public child welfare workers: A longitudinal
study. Children and Youth Services Review, 34(9), 1769–1776.
doi:10.1016/j.childyouth.2012.02.006
Lynch, E. W., Hanson, M. J. (1992). Developing cross-cultural competence: A guide for
working with young children and families. Baltimore, MD: Bookes Publishing.
Manoleas, P. (1994). An outcome approach to assessing the cultural competence of MSW
students. Journal of Multicultural Social Work, 3(1), 43–57.
doi:10.1300/J285v03n01_05

110
Marsh, J. C., Cao, D., Guerrero, E., & Shin, H. (2009). Need-service matching in
substance abuse treatment: Racial/ethnic differences. Evaluation & Program
Planning, 32(1), 43–51. doi:10/1016/j.evalprogplan.2008.09.003
McCrae, J. S., & Fusco, R. A. (2010). A racial comparison of family group decision
making in the USA. Child & Family Social Work, 15(1), 41–55.
doi:10.1111/j.1365-2206.2009.00636.x
McMillen, J. C., Scott, L. D., Zima, B. T., Ollie, M. T., Munson, M. R., Spitznagel, E.
(2004). Use of mental health services among older youths in foster care.
Psychiatry Service, 55, 811–817. doi:10.1176/appi.ps.55.7.811
McPhatter, A.R. (1997). Cultural competence in child welfare: What is it? How do we
achieve it? What happens without it? Child Welfare, 76(1), 255–278.
Mennen, F. E., Kim, K., Sang, J. & Trickett, P. K. (2010). Child neglect: definition and
identification of youth’s experiences in official reports of maltreatment. Child
Abuse and Neglect, 34(9), 647-658. doi:10.1016/jchiabu.2010.02.007
Meyer, J. W., Scott, R., & Strang, D. (1987). Centralization fragmentation and school
district complexity. Administrative Science Quartely, 32, 186–201.
Mor Barak, M. E., & Travis, D. J. (2009). Diversity and organizational performance. pp.
341 – 379. In Y. Hasenfeld (Eds)., Human Services as Complex Organizations.
Thousand Oaks, CA: SAGE.
Mor Barak, M. E., Travis, D. J., Pyun, H., & Xie, B. (2009). The impact of supervision
on worker outcomes: A meta-analysis. Social Service Review, 83(1), 3–32.
doi:10.1086/599028
Muthén, L. K., &Muthén, B. O. (1998-2012). Mplus User’s Guide. Seventh Edition. Los
Angeles, CA: Muthén & Muthén.
NSCAW Research Group. (2011). National survey of child and adolescent well-being:
Overview of NSCAW and NSCAW II, and main findings of NSCAW. Cornell
University, Research Triangle Institute.
Oliver, C. (1991). Strategic responses to institutional processes. Academy of Management
Review, 16, 145–179. doi:10.5465/AMR.1991/4279002
Ortega, R., Grogan-Kaylor, A., Ruffolo, M., Clarke, J., & Karb, R. (2010). Racial and
ethnic diversity in the initial child welfare experience: Exploring areas of
convergence and divergence. In Webb, M. B., Dowd, K., Harden, B. J.,
Landsverk, J., & Testa, M. F. (Eds.), Child welfare and child well-being (pp. 236–
271). New York, New York: Oxford University Press.
Perry, R. E. (2006). Do social workers make better child welfare workers than non-social

111
workers? Research on Social Work Practice, 16(4), 392–405.
doi:10.1016/j.childyouth.2004/05.001
Perry, R., & Limb, G. E. (2004). Ethnic/racial matching of clients and social workers in
public child welfare. Children and Youth Services Review, 26(10), 965–979.
Pinderhughes, E. (1991). The delivery of child welfare services to African American
clients. American Journal of Orthopsychiatry, 61, 599–605.
doi:10.1037/h0079293
Ragins, B. R., & Scandura, T. A. (1994). Gender differences in expected outcomes of
mentoring relationships. Academy of Management Journal, 37(4), 957–971.
doi:10.2307/256606
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and
data analysis methods (2
nd
ed.), Sage.
Riordan, C. M., & Shore, L. M. (1997). Demographic diversity and employee attitudes:
An empirical examination of relational demography within work units. Journal of
Applied Psychology, 82(3), 342–358. doi:10.1037/0021-9010.82.3.342
Roberts, D. E. (2002). Racial disproportionality in the US child welfare system:
Documentation, research on causes, and promising practices. Baltimore, MD:
The Annie E. Casey Foundation.
Robertson, P. J., Roberts, D. R., & Porras, J. I. (1993). Dynamics of planned
organizational change: Assessing empirical support for a theoretical model.
Academy of Management Journal, 36(3), 619–634. doi:10.2307/256595
Shin, S., Chow, C., Camacho-Gonsalves, T., Levy, R. J., & Leff, H. S. (2005). A meta-
analytic review of racial-ethnic matching for African American and Caucasian
American clients and clinicians. Journal of Counseling Psychology, 52(1), 45–56.
doi:10.1037/0022-0167.52.1.45
Smith, B. D., & Donovan, S. E. F. (2003). Child welfare practice in organizational and
institutional context. Social Service Review, 77(4), 541–563. doi:10.1086/378328
Smith, C., J., & Devore, W. (2004). African American children in the child welfare and
kinship system: From exclusion to over inclusion. Children and Youth Services
Review, 26, 427–466. doi:10.1016/j.childyouth.2004.02.005
Snowden, L., Masland, M., Ma, Y., & Ciemens, E. (2006). Strategies to improve
minority access to public mental health services in California: Description and
preliminary evaluation. Journal of Community Psychology, 34(2), 225–235.
doi:10.1002/jcop.20092
Staines, G., Tavris, C. & Jayaratne, T. (1974). The queen bee syndrome, In Travis, C.

112
(Ed.), The female experience. Del Mar, CA: CRM Books.
Stahl, G. K., Maznevski, M. L., Voigt, A., & Jonsen, K. (2009). Unraveling the effects of
cultural diversity in teams: A meta-analysis of research on multicultural work
groups. Journal of International Business Studies, 1–20. doi:10.1056/jibs/2009.85
Sue, S., Fujino, D., Hu, L., Takeuchi, D., & Zane, N. (1991). Community mental health
services for ethnic minority populations: A test of the cultural responsiveness
hypothesis. Journal of Consulting and Clinical Psychology, 59, 533–540.
Tajfel, H. (1982). Social identity and intergroup relations. New York: Cambridge
University Press.
Trotter, C. (2002). Worker skill and client outcome in child protection. Child Abuse
Review, 11(1), 38-50. doi:10.1002/car.719
Tsui, A. S., Egan, T. D., & O’Reilly, C. A. (1992). Being different: Relational
demography and organizational attachment. Administrative Science Quarterly, 37,
549–579.
Tsui, A. S., & O’Reilly, C. A., III. (1989). Beyond simple demographic effects: The
importance of relational demography in superior-subordinate dyads. Academy of
Management Journal, 32, 402–423. doi:10.2307/256368
U.S. Department of Health and Human Services, Administration for Children, Youth, and
Families. (2005). National survey of child and adolescent well-being (NSCAW);
CPS sample component wave 1 data analysis report. Washington D.C.: U.S.
Government Printing Office.
U.S. Department of Health and Human Services, Administration on Children, Youth and
Families. (2006). Child maltreatment 2004. Washington DC: U.S. Government
Printing Office.
U.S. Census Bureau, Housing and Household Economic Statistics Division, Fertility &
Family Statistics Branch. (2010). America’s families and living arrangements:
2011. Retrieved from http://www.census.gov/population/www/socdemo/hh-
fam/cps2010.html
U.S. Department of Health and Human Services, Administration for Children and
Families, Administration on Children, Youth and Families & Children’s Bureau.
(2012). Preliminary Estimates for FY 2011, 19, 6. Retrieved from
http://www.acf.hhs.gov/sites/default/files/cb/afcarsreport19.pdf
Wells, R., Hillemeier, M. H., Bai, Y., & Belue, R. (2009). Health service access across
racial/ethnic groups of children in the child welfare system. Child Abuse &
Neglect, 35(5), 282–292. doi:10.1016/j.chiabu.2008.11.003

113
Weaver, D., Chang, J., Clark, S., & Rhee, S. (2007). Keeping public child welfare
workers. Administration in Social Work, 31(2), 5–25.
doi:10.1300/J147v31n02_02
William, K. Y., & O’Reilly, C. A. (1998). Demography and diversity in organizations: A
review of 40 years of research. In B. M. Staw & L. L. Cummings (Eds), Research
in organizational behavior, Vol. 20 (pp. 77-140). Greenwich, CT: JAI Press.
Winefield, H. R., & Barlow, J. A. (1995). Client and worker satisfaction in a child
protection agency. Child Abuse & Neglect, 19(8), 897–905.
doi:10.1016/145-2134(95)00052-A
Zima, B. T., Bussing, R., Yang, X., & Belin, T. R. (2000). Help-seeking steps and service
use for children in foster care. Journal of Behavioral Health Services and
Research, 27(3), 271–285. doi:10.1007/BF02291739

Abstract (if available)

Abstract This dissertation examines racial/ethnic diversity and disparities in the child welfare system using a multilevel perspective. Given the conspicuous increase in racial/ethnic diversity among children during the past two decades, child welfare agencies face tremendous challenges in addressing the service needs of children and their families from diverse racial/ethnic backgrounds. To date, service disparities in the child welfare system have become evident and it is a significant challenge for child welfare agencies to reduce these disparities by implementing organizational practices. However, little is known about the extent to which child welfare agencies are executing practices to respond to the service needs of clients from diverse racial/ethnic backgrounds and how the practices impact service and child outcomes. ❧ Using nationally representative, multilevel longitudinal data from the National Survey of Child and Adolescent Wellbeing II, in three independent studies, this dissertation seeks to establish disparities in use of needed services, and to identify worker- and organizational factors associated with the disparities in a multilevel perspective: (1) client-level: racial/ethnic disparities in use of needed child welfare services

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

Strength-Based Reporting: a trauma-informed practice for mandated reporters, to address behavioral health concerns in children at risk of child welfare involvement

PDF

Applying best practices to optimize racial and ethnic diversity on nonprofit boards: an improvement study

PDF

Assessing implementation of a child welfare system practice change

PDF

Father engagement in the child welfare system: a promising practice study

PDF

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

PDF

Immigrants at a loss: the need for services that promote child well-being among Latino families with child welfare contact

PDF

An examination of ethnic minority physician representation in clinical medicine to reduce health disparities among minoritized groups

PDF

A multilevel model of work-family conflict in a global context: a comparative study across 24 countries

PDF

Fairness in machine learning applied to child welfare

PDF

Racial and ethnic disparities in delays of surgical treatment for breast cancer

PDF

Mental health service use by children and youth in the child welfare system: a focus on need and predisposing factors and caregiver type

PDF

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

PDF

Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions

PDF

Interagency collaboration: cultivating resources and strengthening substance use service delivery in child welfare

PDF

Racial and gender gaps in executive management: a retrospective examination of the problem cause and strategies to address disparities

PDF

Racial disparities in U.S. K-12 suburban schools: an evaluation study

PDF

Beyond commitments: a qualitative examination of the persistent disparities faced by Black women in executive leadership roles post the 2020 crisis and beyond

PDF

County of San Diego Child Welfare Services Hotline Redesign

PDF

Racial/ethnic variation in care preferences and care outcomes among United States hospice enrollees

Asset Metadata

Creator Kim, Ahraemi (author)

Core Title Racial/ethnic diversity and disparities in the child welfare system: a multilevel perspective

School School of Social Work

Degree Doctor of Philosophy

Degree Program Social Work

Publication Date 08/08/2013

Defense Date 05/24/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag child welfare system,disparities,multilevel perspective,OAI-PMH Harvest,organizational response,racial/ethnic diversity,racial/ethnic matching

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Mor Barak, Michàlle E. (committee chair), Guerrero, Erick G. (committee member), Robertson, Peter John (committee member)

Creator Email ahraemi.kim@gmail.com,ahraemik@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-320268

Unique identifier UC11293317

Identifier etd-KimAhraemi-1992.pdf (filename),usctheses-c3-320268 (legacy record id)

Legacy Identifier etd-KimAhraemi-1992.pdf

Dmrecord 320268

Document Type Dissertation

Format application/pdf (imt)

Rights Kim, Ahraemi

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA