TESTING A MODEL OF HEALTHY MARRIAGE/HEALTHY RELATIONSHIPS:
THE PREDICTION OF PARENTING AND CHILD WELL-BEING
By
Krista M. Shambleau

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Human Development and Family Studies
2010

ABSTRACT
TESTING A MODEL OF HEALTHY MARRIAGE/HEALTHY RELATIONSHIPS:
THE PREDICTION OF PARENTING AND CHILD WELL-BEING
By
Krista M. Shambleau
Federally funded Healthy Marriage Initiative (HMI) programs provide marriage
education as well as other services to low-income diverse individuals and couples at many points
along the marital continuum with improving children‘s well-being as the overarching purpose.
These programs need appropriate measures of healthy marriage for couples with children that
relate to well-being. Purposes of this research were to examine the factor structure of a healthy
marriage measurement developed by Manpower Demonstration Research Corporation (MDRC)
and how it varied across (a) gender, (b) race (African American and Caucasian), and (c) marital
status (married and unmarried). An additional purpose was to test a measurement model,
developed by Child Trends, Inc., in which healthy marriage is hypothesized to predict parenting
and child well-being.
This cross-sectional quantitative survey research included 343 Head Start parents (151
males and 189 females) involved in a marriage/relationship. Overall, the unidimensional healthy
marriage measurement fit for both males and females and the addition of two marital virtues,
namely forgiveness and teamwork, also loaded well on this measurement. The factor loadings
were significant for both males and females and results of chi-square difference tests indicated
that the factor loadings for the great majority of variables were invariant across gender.
However, the factor loading for relationship stability (one‘s assessment that their relationship is
not in trouble) was larger for males whereas child commitment (one‘s assessment of their
spouse/partner‘s commitment to their child) was larger for females.

Path analysis demonstrated that the Child Trends, Inc. healthy marriage measurement
model fit differed by gender and this was supported by further SEM analysis. The effect size for
the direct path between risk and healthy marriage was significantly larger for females than for
males. The effect size of this path tended to be greater for African American women and
unmarried women. Results using fit indices showed that the fit of the measurement model was
appropriate for the diverse sample. Mediation results revealed that for both males and females,
healthy marriage partially mediates the effect that depression and stress have on parenting. For
females, parenting partially mediates the effect between risk and children‘s social competence as
well as between healthy marriage/healthy relationships and children‘s social competence.
Findings demonstrate support for the Child Trends, Inc. measurement model that healthy
marriage/healthy relationships may enhance children‘s social competence through positive
parenting. Study results may inform federal and state healthy marriage and family initiatives as
well as marriage and relationship education in terms of program delivery and evaluation.

Copyright by
Krista M. Shambleau
2010

DEDICATION
To the loves of my life, Monty, Brandon, and the little baby on the way.

v

ACKNOWLEDGEMENTS
I acknowledge God‘s great presence in my life and am thankful for many blessings. I
would like to thank my parents, Carmela and Stan Surowiec, for their unconditional love and
support throughout my life. They continue to cheer me on and their encouragement is so dear to
me. They have helped me in countless ways so that I may live a full life and pursue my dreams.
I thank my wonderful husband, Monty Shambleau, for being by my side every step of the way.
His loyalty, belief in me, and love is a special blessing. Sharing marriage and parenting with him
is a joy. I am truly grateful for my son Brandon whose hugs, kisses, playfulness, and insight
have made every day brighter. I recognize my in-laws as well as my other family members and
friends for their constant thoughtfulness and kind gestures.
I am sincerely thankful to my committee chairperson, Dr. Holly Brophy- Herb who has
been a mentor to me for the past twelve years. She is a true inspiration and a great scholar from
whom I have learned so much. I would also like to thank my graduate committee members, Dr.
Deborah Johnson, Dr. Robert Griffore, and Dr. Pennie Foster-Fishman for their wisdom,
contributions in helping to improve the project‘s quality as well as for their integral role in my
graduate studies. A special thank you goes to Jason Huang, a consultant from MSU‘s Center for
Statistical Training and Consulting, who was completely responsive, patient, and thorough in
answering all of my statistical questions for the dissertation project.
I am ever so grateful to Wayne County Health and Family Services Head Start for the
opportunity to work with them on many meaningful projects. I want to thank Ms. Carolyn Gray,
Ms. Cecelia Mobley, and Ms. Maria Cross for their excitement about this dissertation research
and for allowing me to work with the Head Start families. Their full support and interest in this
project is greatly appreciated. I would also like to thank the Head Starts specialists, directors,
vi

site leaders and teachers for helping to make this project possible. A special recognition goes to
all of the parents who took the time to participate in this research by so generously sharing with
me aspects of their family life.

vii

TABLE OF CONTENTS

LIST OF TABLES.........................................................................................................

x

LIST OF FIGURES.......................................................................................................

xii

CHAPTER 1
INTRODUCTION.........................................................................................................
Statement of the Problem...................................................................................
Significance of the Study...................................................................................
Conceptual Framework......................................................................................
Ecological Systems Theory................................................................................
Research Questions............................................................................................

1
3
5
7
10
14

CHAPTER 2
LITERATURE REVIEW..............................................................................................
Critique of ―Healthy Marriage‖ and ―Marital Quality‖ Measures.....................
Healthy Marriage Initiatives (HMIs).................................................................
Limited Research on Low-Income Couples.......................................................
The Focus on Low-Income Couples..................................................................
Current Trends for Low-Income Individuals along the Marriage Continuum...
Risk Factors among Low-Income Couples........................................................
Parenting and Child Well-Being........................................................................
Defining Healthy Marriage................................................................................
Healthy Marriage Constructs and Associations to Children‘s Well-Being........

15
15
22
23
23
25
32
33
36
40

CHAPTER 3
METHODS....................................................................................................................
Participants.........................................................................................................
Procedures..........................................................................................................
Measures............................................................................................................

51
51
56
60

CHAPTER 4
RESULTS......................................................................................................................
Preliminary Analyses.........................................................................................
Scale Analysis....................................................................................................
Exploratory Factor Analyses..............................................................................
Results for Study One.........................................................................................
Results for Study Two........................................................................................
Results for Study Three......................................................................................
Summary of Results...........................................................................................

68
68
70
81
82
88
96
109

viii

CHAPTER 5
DISCUSSION................................................................................................................
Overview............................................................................................................
Measurement of Healthy Marriage.....................................................................
Discussion for Study One...................................................................................
Discussion for Study Two..................................................................................
Discussion for Study Three................................................................................
Summary............................................................................................................
Implications........................................................................................................
Limitations..........................................................................................................
Future Research..................................................................................................
Conclusion..........................................................................................................

111
111
113
116
117
119
132
133
135
138
141

REFERENCES...............................................................................................................

143

ix

LIST OF TABLES
Table 1.

Cherlin et al., 2008 Study Results..............................................................

27

Table 2.

Healthy marriage definitions.....................................................................

38

Table 3.

Factors encouraging union stability..........................................................

49

Table 4.

Racial and Marital Status Demographics by Parent Gender.......................

53

Table 5.

Other Demographics by Parent Gender......................................................

54

Table 6.

Correlations Between Partner Scores on HM/HR Measure........................

70

Table 7.

Ranges, Means, and Standard Deviations for the Total Sample.................

73

Table 8.

Frequencies and Percentages of Dichotomous Risk Variables for the
Total Sample...............................................................................................

73

Table 9.

Means and Standard Deviations by Parent Gender.....................................

74

Table 10.

Frequencies and Percentages of Dichotomous Risk Variables by Gender.

75

Table 11.

Means and Standard Deviations by Race and Marital Status Within
Gender: Female Subgroup...........................................................................

76

Frequencies and Percentages of Dichotomous Risk Variables by Race
and Marital Status Within Gender: Female Subgroup................................

77

Means and Standard Deviations by Race and Marital Status Within
Gender: Male Subgroup...............................................................................

78

Frequencies and Percentages of Dichotomous Risk Variables by Race
and Marital Status Within Gender: Male Subgroup....................................

79

Correlations between Social Competence, Parenting, Healthy Marriage,
and Risk for Total Sample...........................................................................

83

Table 16.

Factor Matrix of the Healthy Marriage Measure........................................

85

Table 17.

Factor Matrix of the Healthy Marriage Measure for Males by Race and
Marital Status..............................................................................................

86

Factor Matrix of the Healthy Marriage Measure for Females by Race and
Marital Status.......................................................................................

86

Table 12.

Table 13.

Table 14.

Table 15.

Table 18.

x

Table 19.

Table 20.

Table 21.

Table 22.

Standardized and Unstandardized Coefficients for Males and Females on
HM/HR Model............................................................................................

93

Fit Indices of Three Nested Models of Multi-group Confirmatory Factor
Analysis.......................................................................................................

94

Equivalent and Nonequivalent Factor Loadings of Variables across
Gender..........................................................................................................

95

Unstandardized, Standardized, and Significance Levels for Child Tends
Model for Males and Females by Race and Marital Status.........................

101

Table 23.

Bootstrapped Point Estimates and Confidence Intervals (CIs) for Indirect
Effects of Mediators by Gender................................................................... 104

Table 24.

t and z Values for Normal Theory Mediational Tests by Gender................ 104

Table 25.

Equivalent and Nonequivalent Factor Loadings of Structural Paths across
Parental Gender........................................................................................... 108

Table 26.

Standardized and Unstandardized Coefficients for Males and Females on
SEM Model.................................................................................................. 108

xi

LIST OF FIGURES
Figure 1.

Measurement framework for conceptualizing and defining healthy
marriage.......................................................................................................

9

Figure 2.

Unconstrained CFA healthy marriage model for males..............................

91

Figure 3.

Unconstrained CFA healthy marriage model for females..........................

92

Figure 4.

Unconstrained standardized estimates for males and females....................

98

Figure 5.

Hypothesized structural SEM model..........................................................

106

xii

CHAPTER 1: INTRODUCTION
The transformations in American marriage and family structure since the 1960s have
decreased the average level of adult and child well-being, increased child poverty, and have left a
sizeable financial burden on our society (Amato, 2008). Set out to address these changes, efforts
to strengthen marriage and increase the percentage of children raised in healthy two-parent
families has become an important public policy goal. Federally funded ―healthy marriage‖
initiatives have the ultimate purpose of improving child well-being. The Healthy Marriage
Initiative was launched by the Administration for Children and Families (ACF) in 2002 ―to help
couples who have chosen marriage gain greater access to marriage education services, on a
voluntary basis, where they can acquire the skills and knowledge necessary to form and sustain a
healthy marriage‖ (ACF- HMI mission statement). The plan for the initiative was not solely to
produce more marriages or longer-lasting marriages, but rather to promote healthier marriages.
Local Healthy Marriage Initiatives extend to singles, unmarried couples and married couples by
encouraging the health of the individual, the health of relationships for those in marriages and or
other types of partnerships, the strength of parenting, and the well-being of children. The
understanding is that children‘s well-being is enhanced when their parents are healthy and if they
are married or in a relationship, that these relationships are also healthy (National Healthy
Marriage Resource Center, 2010).
Low-income couples with diverse backgrounds are likely the focus of these initiatives
(Moore et al., 2007). Accordingly, it is important to develop constructs and measures
appropriate for these populations (Moore et al., 2007). Federal marriage initiative programs
most often utilize marriage education interventions initially created for and tested with white,
middleclass, well-educated engaged or already married couples (Dion, 2005). There is some
1

question whether these standard programs can address the needs, interests, and circumstances of
low-income couples (McClintock-Comeaux, Anderson, & Kuvalanka, 2004). Low-income
couples indeed have special challenges along the marital continuum including high marital
instability rates (Fein, 2004; Garfield & McLanahanm, 2000; McGinnis, 2003), increases in
nonmarital childbearing (Ellwood & Jencks 2002), children by multiple partners (Mincy, 2002),
lower levels of education and employment, and marital quality (Fein, 2004) as well as the
associated risks for children. Research on the relationship dynamics of low-income couples
suggests that particular issues make healthy marriage more difficult (Cherlin et al., 2004; Fein,
2004; Gibson-Davis, Edin, & McLanahan, 2005; Karney, Garvan, & Thomas, 2003; Mincy,
2002; Mincy, Pouncy, Reichert, & Richardson, 2004). These differences have implications for
the content and delivery of marriage and relationship skills education for lower-income couples
(Dion, 2005) as well as implications for measuring the impacts of these projects. The evaluation
of marriage initiatives serving low-income couples must include appropriate measures (BronteTinkew et al., 2003; Moore et al., 2007).
In their commentary on measurement issues in family research, Hofferth and Casper
(2007) note that to date, data collection systems have not been adequately modified to reflect
changes in U.S. families. This is problematic as, ―effective economic and social policy depends
in part on the ability to understand and address the changing shapes and needs of American
families. This in turn requires valid and reliable data on marriage, divorce, remarriage,
cohabitation, and the family situations of children over time at the national, state and local
levels‖ (Hofferth & Casper, 2007, pp. ix – x).
Additional measures of healthy marriage designed and validated specifically for a lowincome diverse population are needed for therapists, educators, researchers and policy-makers.
2

In addition, the realities and successes of local marriage programs must be revealed as well as
how they impact children‘s development in order for continuous quality improvement and
sustainability. Although The Office of Planning, Research and Evaluation at ACF is funding
respected evaluation projects for healthy marriage initiatives, evaluation at the local level is also
necessary. Local programs are committed to meeting their grant objectives related to healthier
relationships and stronger family functioning and must rely on easy to use measurements.
Statement of the Problem
Low-income individuals place a high value on marriage but face many barriers and
disincentives to marriage (Edin 2000; Gibson-Davis, Edin, & McLanahan, 2005; Smock,
Manning & Porter, 2005) and their marriages are less likely to survive (Roberts, 2007). Lowincome couples are likely to experience external stressors such as poverty, isolation and lack of
social support, unemployment, and poor health. Coping with setbacks is made even more
difficult with fewer economic and personal resources. They also experience higher rates of
personal problems and have greater difficulty receiving help for these problems. All these factors
place great stress on couple relationships and on child well-being (Ooms, 2007).
Complimentary to this contextual perspective, marital interaction researchers share the
belief that high levels of marital instability mostly stem from quality of the relationship.
Understanding healthy marriage for low-income couples and developing effective intervention
strategies requires consideration of the interactional processes, individual strengths and
vulnerabilities, along with the stressful events and contexts of the poor (Bradbury & Karney,
2004).
Adaptations to healthy marriage curricula are underway to adjust the content for the
needs of diverse low-income couples (e.g., the Survival Skills for Healthy Families/Family
3

Wellness, Relationship Enhancement (RE), and Prevention and Relationship Enhancement
Program (PREP)) and researchers frequently point to this need (Ooms, 2007); however, the
measurement tools used must also be proven reliable and valid for low-income couples.
Comprehensive measures to assess healthy marriage for low-income couples enrolled in these
initiatives are limited. Research on the marital quality of couples with children typically has two
purposes: (a) to study the influence of marital quality on children‘s well-being and (b) to
examine associations between family processes and marital quality. The relationship between
parental conflict and child well being is well-documented; however, relationships between other
aspects of marital quality and child well being are limited (Bronte-Tinkew et al., 2003). The
topic of marriage and couple relationships remains a strong interest of researchers and policy
makers. Efforts to measure aspects of the quality of marital relationships began more than 50
years ago, and instruments continue to be developed and refined. However, there are many
conceptual issues regarding the measurement of relationship quality (ACF: OPRE, retrieved
2009).
Low-income couples have unique needs and circumstances that effect quality of
marriage. Healthy Marriage Initiatives (HMIs) can help provide support with marriage
education and other services to unmarried and married couples as well as to singles. Local HMIs
targeting low-income diverse populations lack available brief, easy to use, appropriate measures
of healthy marriage for couples with children that relate to children‘s well-being. This type of
measure is a necessary piece to the evaluation of such initiatives even at the local level.
Specifically, there is a need to use a comprehensive measurement tool for the following purposes
in local evaluation: to understand the individuals being served, to design and plan appropriate
interventions, to assess quality and effectiveness of services, to inform decisions for continuous
4

quality improvement, to measure grant objectives, to understand healthy marriage for lowincome couples, to capture program successes and importantly, to relate results to the ultimate
goal of children‘s well-being.
Significance of the Study
The Child Trends, Inc.‘ (2003) project to extensively review the ―state of the art‖ in
measuring couple relationships has yielded important results that point to the gaps in the
literature on marriage as well as the needed areas for measuring healthy marriage. They found
few studies that singly examine psychometric issues among lower-income or low-education
married couples. The report suggests that research on measures of relationship quality among
couples with children would benefit from:


Development and validation of constructs such as ‗commitment,‘ ‗intimacy‘ and
‗communication‘ patterns among couples with children, and in particular, testing
whether these constructs are linked to child well-being;



Development of measures that can be used for unmarried as well as married couples;



Testing the applicability of measures for race and ethnic subgroups; and



Testing the applicability of measures for low-income and low-education subpopulations.
In addition, the Child Trends, Inc. literature review (Bronte-Tinkew et al., 2003)

recommends that concepts predicting marital outcomes should be included in marriage studies.
For instance, positive attitudes about marriage and marital expectations (or plans) are associated
with higher relationship stability among cohabiting couples (Axinn & Thornton, 1993;
Clarkberg, Stolzenberg, & Waite, 1995; McGinnis, 2003).

5

Creating an index might be more appropriate given the challenge of developing measures
for non-traditional family forms. In addition, little attention in the literature is given toward
examining brief versions of existing scales and comparing different approaches for assessing a
particular construct. This project explores how the aggregation of constructs into a summary
measure of healthy marriage applies to low-income couples with children.
The Abt Associates Guide to Data Sources on the Determinants of Marriage and
Cohabitation (2003) also highlight several research issues to the study of marriage and
cohabitation. One issue is the need to understand the effects of personal characteristics
(antecedents) on couple interaction within social and economic environments. Relatively stable
background characteristics (i.e. education, ethnicity, and early childhood experiences) affect
marital outcomes indirectly through expectations, values, and appraisals and, ultimately, on
couple interaction. Personality characteristics such as depression and stress can also hinder
positive interaction. Understanding the role of such factors is essential in preventing and treating
relationship distress. Researchers have begun to study the direct and moderating effects of
personal characteristics among middle -class married couples but there is little research on lowincome couples or informal unions. Low-income couples deserve special attention because their
lower levels of education and higher levels of depression, mental illness, and substance abuse
suggest major challenges to sustaining healthy relationships.
Another issue that is addressed in the Abt Associates guide is in regards to the contextual
influences on interaction. Most environmental factors including demographic, economic, and
socio-cultural can influence couple interaction and moderate the effects of interaction on union
outcomes. Understanding these linkages is critical to identifying strategies for helping couples
function better in challenging environments. For example, approaches to conflict resolution may
6

vary by the norms and values that are shaped by culture. In addition, the contextual influence of
financial stress has been found to have strong negative effects on marital quality and stability.
Both chronic and acute stresses (i.e. job loss, car breakdown, or eviction) can be destructive to
couple interaction, especially for low income people.
Purposes of this research were to examine the factor structure of a healthy marriage
measurement developed by Manpower Demonstration Research Corporation (MDRC) and how
it varied across (a) gender, (b) race (African American and Caucasian), and (c) marital status
(married and unmarried). An additional purpose was to test a measurement model, developed by
Child Trends, Inc., in which healthy marriage is hypothesized to predict parenting and child
well-being for a diverse group of low-income couples with children along the marital continuum.
Conceptual Framework
The conceptual framework chosen for this study is a product of the Child Trends, Inc.
work group of scholars who reviewed decades of marriage research and analyzed approximately
50 journal articles pertaining in some way to marital strengths (Moore et al., 2004). This model
was chosen because it is used by Healthy Marriage Initiatives. The model considers a person‗s
background down to two levels of ecological strata, which are society and family. The fivecomponent model includes a background into which both individuals are born (i.e., family and
social context) and possible personal antecedents, that can affect or influence marriage. The
presence (i.e. unemployment) or absence (i.e. religiosity) of antecedents does not necessarily
mean that a couple by definition has an unhealthy marriage. Both the background and
antecedents contribute to the ten characteristics of healthy marriages that they identified from
the literature review. These include: (a) commitment of the couple, (b) satisfaction, (c)
communication, (d) conflict resolution, (e) lack of domestic violence (f) fidelity (understood as
7

sexual), (g) interaction and time together, (h) intimacy and emotional support, (i) commitment to
the children, and (j) duration and legality of marital status. These will be explored in greater
detail later with a look at how each is conceptualized. The characteristics, in turn, are linked to
certain consequences or outcomes with regard to adult and child well-being (see Figure 1). The
framework posits that improvements in marital interaction will have a positive net effect on adult
and child well-being.
Moore et al. (2004) recognize the causes and effects of healthy marriage as complex with
some factors appearing to be both causes and consequences of marriage. They also consistently
report that an evaluation of healthy marriage must be sensitive to the context and issues unique to
couples. They suggest a next step to the conceptual model, ―to develop concrete measures to
bring the model to reality especially developed for and tested among low-income couples‖
(Moore et al., 2004, p. 5).

8

Figure 1. Measurement framework for conceptualizing and defining healthy marriage.
Background

Antecedents

Definition

Social context,
macrolevel factors

Employment,
income

Healthy
Marriage/
Relationships
____________

Family
background
characteristics

Education
Physical
health
Mental health
Stress
Social support
Social skills
Substance use

Consequences
Adult Wellbeing
______________
Employment,
income, wealth

Child Wellbeing
_______________
Socio-emotional
outcomes

Communication

Physical health,
mortality

Interaction/time
together

Mental health

Cognitive
attainment and
educational
achievement

Satisfaction

Commitment to
the couple

Incarceration

Intimacy/
emotional
support

Family
background
characteristics

Conflict/conflict
resolution

Children from
a prior
relationship

Lack of
domestic
violence

Community
context

Social support
Satisfaction and
happiness

Fidelity

Attitudes/
values
Religiosity

Risk-taking,
substance use,
illegal activities
Parenting
Attitudes/values
Religiosity

Commitment to
the children

Health/safety
Attitudes/values
toward marriage
and childbearing
Dating behavior
Sexual activity
Relationship
skills
Marital stability
of offspring in
adulthood

Duration/
legal marital
status

Assumptions Regarding Marriage
The assumptions regarding marriage for this study are those shared by Child Trends, Inc.
(Moore et al., 2004) and were chosen because they set the groundwork for evaluation.
1.

Healthy marriage is a continuum. It is not a yes/no construct that couples have or
do not have, but a scale or index with varying degrees.

2. The elements of a healthy marriage can be learned. Thus, healthy marriage
9

constructs can be affected by interventions of marriage education.
3. Healthy relationships should be assessed differently for different populations. This
study focuses on low-income couples with children- a different population than
those without children. But even low-income couples with children vary in
terms of their family structure and living arrangements. Measures are needed
that will work for various family subgroups: two-resident, biological parents;
married or cohabiting couples; (re)married or cohabitating couples with
stepchildren; nonresident, estranged couples with children in common; and
visiting couples who are parents.
4. Healthy marriage has multiple elements. This study takes a comprehensive view of
healthy marriage, encompassing ten related yet distinct domains that will be
discussed in detail below.
Ecological Systems Theory
The Child Trends, Inc. measurement framework of healthy marriage provides an
ecological basis because it considers a child‘s well-being within environmental contexts.
However, because it is a rather linear model, it may be critiqued for its limited consideration to
bi-directional processes especially how children may influence adult well-being and healthy
marriage. The focus of the Child Trends, Inc. measurement model is on the definition of
―healthy marriage‖: a concept that defines marriages as healthy when it promotes the individual
well-being of both partners and their children (Halford, Kelly, and Markman‘s, 1997). The
measurement model predicts children‘s development by proximal and distal environmental
effects including parenting, marriage, as well as a set of antecedents (i.e. parent mental health).
This dissertation tests the model as a measurement of healthy marriage and one that predicts
10

well-being. This dissertation does not test a model of ecological theory, but does consider some
aspects of ecological theory including aspects of several principals. The ecology of human
development includes several principals. Principal number 4 as stated in Bronfenbrenner‘s
(2005) article, ―Ecological Systems Theory‖ is:
No characteristic of the person exists or exerts influence on development
in isolation. Every human quality is inextricably embedded, and find both
its meaning and fullest expression, in particular environmental settings,
of which the family is a prime example. As a result, there is always an
interplay between the psychological characteristics of the person and of a
specific environment; the one cannot be defined without reference to the other (p.
146).
Principal number 11 extends this fundamental concept:
Each member of a microsystem influences every other member.
In terms of research design, it is therefore important to take into
consideration the influence of each relationship on other relationships:
for example, within the family the effect of the husband-wife relationship
on the parent-child relationship, the effect of the mother-child
relationship on the father-child relationship, and vice versa. The
appropriate design for this purpose is a process-person-context model
in which each relationship is treated as a context for processes taking
place in the others (p. 161).
As applied in this study, children‘s well-being is examined in the context of marriage/partnership
as well as the parent-child relationship. However, the measurement model is limiting in that is
11

does not specifically include effects of child characteristics on parenting and marriage nor the
effect of the parent-child relationship on marriage. The purpose of the measurement model
instead is prediction of well-being from healthy marriage/healthy relationships.
The healthy marriage model does place great consideration on examining the many
influences on child well-being of the characteristics of significant others (parents) in the child‘s
life. The assumption is that a parent‘s background influences their marriage/partnership and
their behavior as parents. The developmentally relevant characteristics of parents who
participate in children‘s home environment are important to ecological systems theory
(Bronfenbrenner, 2005).
This study permits analysis of variations in a child‘s developmental processes and
outcomes as a joint function of the characteristics of the environment and of the person. This is
an assessment not only of children‘s well-being but also of the effectiveness of their parent(s)‘
marriage/intimate partnership and parenting that produce this outcome. The design reveals how
both child well-being and processes vary as a joint function of the characteristics of the child and
of the environment (i.e. antecedents, healthy marriage, parent well-being), thus permitting the
detection of synergetic effects (effects of joint forces that are greater than the sum of the
individual effects). (Bronfenbrenner, 1992; 2005). Thus, the combination of child
characteristics and parent(s)‘ marriage/partnership quality with parent mental health may have a
greater impact than separate effects of each. Allowing for the analysis of processes considers
possible differences in child development.
This study focuses primarily on the microsystem, ―a pattern of activities, roles and
interpersonal relations experienced by the developing person in a given face-to-face setting with
particular physical and material features and containing other persons with distinctive
12

characteristics of temperament, personality and systems of belief‖ (Bronfenbrenner, 2005, p.
148). Of particularly interest in this study is the home environment. Parent(s)‘ marriage/
partnerships and parenting are examined as to how they predict child well-being. This study
does not examine the mesosytem because the linkages and processes taking place between home
and other settings (i.e. preschool) containing the child are not directly analyzed. Although the
exosystem is recognized, it is not a major focus in this study because it encompasses the linkages
and processes between two or more settings, at least one of which does not ordinarily contain the
child, but in which events occur that influence processes within the home.
The macrosytem is not a focus in this study although macrosystemic domains in which
processes are apt to function differentially can be identified in the study by proxy variables in the
form of, for example, ethnicity and family structure (i.e. married parents, cohabitating unmarried
partners, stepparents). Note the special importance of these variables as being characteristics of,
not necessarily the developing child, but of the significant persons (parents) in the child‘s life.
Thus, a child‘s well-being can vary substantially as a function of the personal attributes of their
parents. ―The scientific power of a microsystem model is therefore further increased to the
extent that such characteristics are assessed‖ (Bronfenbrenner, 2005, p. 160).
Finally, the degree of stability, consistency, and predictability are critical for the effective
operation of the system. Results from research conducted in the last two decades demonstrate
growing evidence that children who grow up with two parents who are in a stable relationship
are more likely to receive the parenting and financial supports that promote their well-being
(Amato, 2000; McLanahan & Sandefur, 1996). However, questions remain about the particular
influences that are most significant, the optimal role for public policy, and whether and why

13

marital relationships may be especially beneficial for children (Acs, 2007; Kamp Dush, 2009;
Sweeney, Hongbo, & Videon, 2009).
Research Questions
The following questions are those that guide this research and are divided into three major
study parts. Please note that terms ―healthy marriage‖ and ―healthy relationship‖ (HM/HR) are
used interchangeably in discussing relationship quality of both married and unmarried couples
with children.
Study One
1. What is the factor structure for the HM/HR measure?
2. How do antecedents (i.e. depression and stress) relate with the HM/HR measure?
Study Two
3. Does the HM/HR factor structure vary across groups (a) gender, (b) race (African
American and Caucasian), and (c) marital status (married and unmarried)?
Study Three
4. How are responses on the HM/HR measure related to parenting and child outcomes?

14

CHAPTER 2: LITERATURE REVIEW
This review begins with a critique of existing healthy marriage measures. Next, a
description of Healthy Marriage Initiatives (HMIs) is given in order to provide a context and
purpose for the measurement of healthy marriage/healthy relationships. It then focuses on lowincome couples, for whom the measurement is intended. The review outlines the study‘s
accepted assumptions regarding marriage and recognizes different definitions of healthy
marriage while selecting one particular conceptual model by Child Trends, Inc. Finally, a review
of the literature related to each of the ten healthy marriage constructs is provided and how these
domains relate to parenting and children‘s well-being.
Critique of “Healthy Marriage” and “Marital Quality” Measures
To demonstrate the limitations with existing healthy marriage measures and to build a
case for a more comprehensive measure, the following critique is provided. The measure of
healthy marriage/healthy relationships is a multi-dimensional construct and a simple measure on
a single dimension (i.e. satisfaction) does not fully consider the actual complexity.
Communication, satisfaction, and a lack of conflict are commonly mentioned elements of a
healthy couple relationship; however, a diversity of other constructs is also important. For
example, commitment is often left out of measures on healthy marriage.
In a comprehensive, empirical study of the marital research literature, for example,
Fowers, Bucker, Calbeck, and Harrigan (2003) found that concepts and measures of marital
quality seldom went beyond individual benefits such as satisfaction, intimacy, and sexual
satisfaction and couple communication. Moreover, satisfaction is discussed and measured 12-26
times as frequently as ―commitment‖. The study demonstrated that scholarship on marriage is
15

dominated by a communication-satisfaction model of relationships in which initial attraction plus
high quality communication leads to intimacy and satisfaction, which form the basis for
relationship commitment. This type of model excludes other crucial aspects of healthy
marriage/healthy relationships. Fowers (2003) argues for a more multi-dimensional model of
healthy marriage beyond the individualistic communication-satisfaction model. She explains
how the communication-satisfaction model uses ―personal satisfaction‖ as the definition of
marital quality and represents commitment as individual fulfillment by the relationship. She
writes, ―in the absence of satisfaction, commitment can only be seen as a less than legitimate
constraint on the individual‘s freedom to pursue a more satisfying bond‖ (p. 93).
Stanley (2007) explains that whereas marital satisfaction dominated the early days of
marital research, the focus shifted in the mid-1970s to a more concentrated focus on
communication and conflict. Today the trend is toward fuller constructs of couple formation and
function which ―have the potential to uncover deeper values and beliefs that affect marriages and
how people behave in them‖ (p. 88). This shift in marriage conceptualization and evaluation is
in part influenced by a newer direction of ―healthy‖ marriage in public policy (e.g., Ooms, 2007;
Stanley, 2004). Set out to understand and conceptualize ―healthy marriage‖, Child Trends, Inc.
identified ten constructs: commitment of the couple; satisfaction; communication; conflict
resolution; lack of domestic violence; fidelity; interaction/time together; intimacy/emotional
support; commitment to the children; and duration/legal marital status. The next step after the
conceptual model development is to test the model using a comprehensive measure, which is just
now beginning to be accomplished. There are only two measures that can be found utilizing the
Child Trends, Inc. (CTI) model. The first is one developed for CTI by MDRC for the Supporting
Healthy Marriage Evaluation project, which is the one being used in this study. The second is a
16

measure of marital quality incorporated in the Marital and Relationship Survey (MARS)
constructed by Lichter and colleagues (2005). Even with these very notable endeavors there are
some missing pieces to the measures.
Marital and Relationship Survey (MARS)
In the recent Lichter and Carmalt (2009) study on religion as a source of marital strength
and stability for low-income couples, marital quality is measured using the CTI ―healthy
marriage‖ model via the MARS survey. Specifically, seven of these healthy marriage
dimensions are measured including: (1) commitment of the couple; (2) overall satisfaction; (3)
communication; (4) intimacy/emotional support; (5) commitment to the children; (6) positive
conflict resolution processes; and (7) positive conflict behaviors. The dimension of ―positive
conflict behaviors‖ is the inverse of three items that measure the presence of emotional abuse
(i.e. yelling, treating like an inferior, blaming). However, other aspects of domestic violence, for
example physical abuse, are not included in this ―marital quality‖ MARS measure.
The CTI healthy marriage model is clear that the two elements that pose obvious threats
to healthy marriage are violence and infidelity. Therefore, more dangerous forms of violence
must be clearly identified because any presence of emotional or physical abuse would innately
reveal an unhealthy relationship. The other ―deal-breaker construct left out of this measure, but
specified in the CTI ―healthy marriage model‖ is a measure of couple fidelity. A simple item
like the one used on the SHM measure is thought to address this measurement need: “I worry
about my spouse/partner cheating on me”. Finally, although the dimension of ―time together‖ is
used to separate religious participation from other secular behaviors, ―quality interaction/time
together‖ is not included as a dimension of marital quality. Having positive interactions and
enjoying time together is a key component to the model of healthy marriage. In reverse, if
17

marriage/couple relationships are characterized by negative interactions there is risk of divorce
or break-up (Gottman & Notarius, 2000).
Although the MARS measure is missing some constructs, there are several commendable
advantages. First, it views marriage as a multi-dimensional concept, using only 25 items to
measure seven dimensions. Short-form measures are preferred over lengthy measures because of
time constraints and participant motivation to return completed forms. Second, all items except
for the satisfaction item correspond to a single 4-point Likert scale (1 = strongly disagree, 4 =
strongly agree). Items can be easily combined to form subscales as well as an overall measure of
marital quality. Third, the measure is intended for low-income couples. Testing the applicability
of measures for low-income and low-education sub-populations continues to be a goal. In
addition, the measure is used independently by both partners, rather than depending on proxy
reports form one partner: ―This allows us to gain background information for both partners, to
evaluate the extent of differences in spousal perceptions of relationship quality, and to identify
key factors that can account for divergent views of relationship quality‖ (Lichter & Carmalt,
2009, p. 171-172). Fourth, the measure was developed to be used as a web-based survey
wherein participants are recruited at random. Finally and importantly, the use of this measure is
used to test the CTI model where ―healthy marriage‖ is conceptualized as mediating the
relationship between a variety of ‗‗antecedents,‖ including religiosity, and individual and family
outcomes (e.g., mental health, achievement of children, etc.). For example, some additional
analyses were completed to examine whether religion buffered the deleterious effects of poor
mental and physical health (an adult well-being outcome).

18

Supporting Healthy Marriage (SHM)
The Supporting Healthy Marriage (SHM) measure does at least consider all of the
healthy marriage constructs outlined within the CTI model. However, it does not address the
important role that ―marital virtues‖ play. Hawkins, Fowers, Carroll, and Yang (2007) argue that
measures of marital quality stress individualism (i.e. happiness and fulfillment) whereby the
benefits individuals derive from marriage are emphasized over what partners contribute to the
quality of the relationship in terms of virtues (i.e. generosity). Addressed above as the
communication-satisfaction model, the individualistic view is not sufficient enough in describing
marriage and should only be considered as ―one framework among many‖ (p. 69). In addition,
they argue for a conceptualization of marriage that emphasizes the corporate aspects of marriage
defined by some as ―we-ness‖ (Buehlman, Gottman, & Katz, 1992; Carrere, Buehlman,
Gottman, Coan, & Ruckstuhl, 2000; Hargrave, 2000) or ―partnership‖ (Fowers, 2000). They
propose that the personal strengths and virtues that individuals offer to the relationship may be
strongly related to both parenting and child well-being and may be an even stronger predictor of
parenting behavior than other measures of marital/relationship quality. For these reasons, both
―forgiveness‖ and ―teamwork‖ are included in this study.
The concept of ―forgiveness‖ in marriage research is gaining some attention, although the
integration of forgiveness into more comprehensive theories of marriage is limited. Forgiveness
has been associated with higher levels of dyadic adjustment (e.g., Gordon & Baucom, 2003;
Paleari, Regalia, & Fincham, 2005), more constructive communication and lower levels of
psychological aggression (Fincham & Beach, 2002), more positive cognitions regarding self and
partner, as well as greater intimacy and closeness in the relationship (Finkel, Rusbult,
Kumashiro, & Hannon, 2002; Gordon & Baucom, 2003). Similarly, lower levels of forgiveness
19

are associated with poorer relational adjustment, intimacy, and closeness, and higher levels of
psychological aggression (Fincham & Beach, 2002; Finkel et al., 2002; Gordon & Baucom,
2003). More recent findings from Gordon, Hughes, Tomcik, Dixon, and Litzinger (2009)
suggest that forgiveness of a marital betrayal (i.e. infidelities, major lies, drastic unilateral
financial decisions, and other similar humiliations and betrayals) is significantly associated with
marital satisfaction, the parenting alliance, and children‘s perceptions of parental marital
functioning.
Whereas the concept of ―forgiveness‖ is gaining momentum in marital research, the
concept of ―teamwork‖ is nearly missing (Fowers et al., 2003). Teamwork, however, is a critical
component because it captures the relationship itself: the elements of the relationship that are
distinct from the partners and that transcend interactions. This view of the relationship as a third
entity is important because it is not reduced to the individuals‘ personal experiences and ways of
relating. Rather, this mutual concept of the relationship provides the context for the couple‘s
interactions, as well as definitions of appropriateness, expectations, and rules of engagement.
Further, this concept is the foundation for the other subjective evaluations of the relationship (i.e.
marital satisfaction, happiness, intimacy). Satisfaction, for example, is a function of how well
the relationship fits with the individual and joint definitions (Fowers et al., 2003).
In addition to some missing features, the SHM measure of healthy marriage is suited only
for married couples. Realistically, measures are needed that will work for various family
subgroups: two-resident, biological parents; married or cohabiting couples; (re)married or
cohabitating couples with stepchildren; nonresident, estranged couples with children in common;
as well as visiting couples who are parents. This research makes adjustments to the SHM
measure so that it is appropriate for these other types of couples with children.
20

Finally, to date there are no multi-dimensional measures of healthy marriage that are
being used to fully test the Child Trends, Inc. ―healthy marriage‖ model of predicting parenting
and child well-being. Although there is a great deal of research that relates specific aspects of
marriage with parenting and child development, this research adds to the field by studying the
association between a comprehensive measure of healthy marriage/relationships, parenting and
child well-being.
Much research relates marital conflict with outcomes. For example, Kitzmann (2000)
found a causal link between conflict and disrupted parenting. In particular, fathers showed
significantly less support/engagement toward their children (study included sons only) after a
conflictual discussion with their spouse (mothers) (M = 3.56 on the 5-point scale, SD = 0.99)
than after a nonconflictual discussion (M = 3.95, SD = 0.75), r(39) = 2.34, p < .05. In addition,
the level of expressed negativity between couples during an observed marital discussion
(regardless of topic) was significantly correlated (p <.05, one-tailed) with numerous qualities of
a subsequent family interaction with the child: specifically, lower family cohesion, r(80) = -.31;
lower support/engagement by fathers, r(80) = -.25; lower support/engagement by mothers, r(80)
= -.49; more family negativity, r(80) = .27; lower family warmth, r(80) = -.27; and less
democratic parenting, point-biserial r(80) = -.28. Furthermore, family cohesion after an
enjoyable topic was discussed was significantly higher when mothers reported high marital
satisfaction (M = 4.20, SD = 0.70) than when mothers reported low marital satisfaction (M =
3.25, SD = 0.72). Also, family warmth after the pleasant discussion was significantly higher
when mothers reported high marital satisfaction (M = 4.00, SD = 0.79) than when mothers
reported low marital satisfaction (M = 3.40, SD = 0.75). In addition to testing specific domains
of healthy marriage/healthy relationships (i.e. marital conflict), measures of ―marital quality‖
21

(Morrill, Hines, Mahmood, & Cordova, 2010) are still not comprehensive enough and do not
necessarily consider multiple contexts.
Healthy Marriage Initiatives (HMIs)
Healthy marriage promotion is on the public policy agenda as a result of the reauthorized
welfare reform bill. Specifically, the purposes of the Temporary Assistance for Needy Families
(TANF) include the promotion of marriage, prevention and reduction of out-of-wedlock
pregnancies, and the encouragement of two-parent families. The TANF program was renewed in
the Deficit Reduction Act (DRA) of 2005 (S. 1932), and signed into law in February 2006 by the
President. Developed under the Administration for Children and Families (ACF), the Healthy
Marriage Initiative (HMI) is sponsoring a variety of demonstrations to examine the effectiveness
in strengthening marriage in the United States (Dion, 2005).
Federally funded HMI programs serve individuals and couples at many points along the
marital continuum. They include unmarried parents who may have plans to marry and who may
be cohabitating. They also include single custodial and non-custodial parents, as well as
increasing numbers of high school youth and other individuals and couples who are ―interested
in marriage.‖ This expansion of the target population means that HMI programs are often not
marketed as ―promoting marriage‖ per se but rather as providing education regarding the general
relationship skills important to many types of couples (Ooms, 2007- CLASP Policy).
Improving children‘s well-being is the overarching purpose of welfare reform, and its
marriage initiative is one of many vehicles for achieving this goal. While there are many
questions about the causal effects of marriage on well-being, the changes in family behavior
should be considered. The strategy focuses on improving the skills of individuals to
communicate and resolve conflicts effectively with partners and spouses, to parent well as a
22

couple, to handle financial conflicts and financial management, and to understand the lasting
benefits of marriage. Research demonstrates strong associations between stable marriages and a
range of positive outcomes for adults and children (Amato, 2005; Carlson & McLanahan, 2006;
McLanahan, 2002; Waite & Gallagher, 2000). Besides the well-being of children, adults, and
families, the promotion of marriage also recognizes the desire of many couples to want healthy
marriages, the disincentives to marry built into the nation‘s tax and transfer programs (Carasso &
Steuerle, 2005), alternatives to welfare dependence (Lichter, Graefe, & Brown, 2003; Rector,
Pardue, & Noyes, 2003; Thomas & Sawhill, 2005), and the call to respond to the decline of
marriage, especially in the African American and low-income communities (Lerman, Acs, & Bir,
2007). In order to strengthen marriage for the entire population, policies and services tailored to
the needs and circumstances of low-income families are required (Ooms, 2002).
Limited Research on Low-Income Couples
Federal healthy marriage initiatives have increased the need for improved information on
low-income married couples (Fein, 2004). Few empirical studies have examined the etiology of
marriage among low-income populations (Carlson, McLanahan, England, 2004; Lichter, Graefe,
Brown, 2003). Although there is an emergent body of literature about couple relationships and
marriage among blacks there is little data about poor whites, Native Americans, or Latinos, or
about the differences between urban and rural poor families (Ooms, 2002).
The Focus on Low-Income Couples
The Office of Head Start Healthy Marriage Initiatives serve low-income individuals and
couples who have unique needs. An assessment of a healthy marriage model must begin with an
understanding for whom the model is conceptualized. A large percentage of single low-income
mothers would like to be married at some time and seek healthy marriages. However, many
23

barriers to marriage exist for low-incomes individuals, such as, low-quality and unsafe housing,
lack of quality education, joblessness, and poor health care that present challenges to stable and
healthy home environments for children as well as place stress on couples who choose to marry.
There are however, many couples with children who do not choose to marry (cohabitators and
nonresidential unmarried couples). The decline in marriage among the poor and near-poor is
influenced by a more unique and complex array of factors than in the rest of the population
(Ooms, 2002). Low-income individuals and those of disadvantaged minority groups are more
likely to live in fragile families (unmarried households with children). Since 1970, there have
been large increases in nonmarital childbearing among women in the bottom two-thirds of the
educational distribution, whereas women in the top one-third have experienced virtually no
increase (Ellwood & Jencks, 2002). Similarly, merely 22% of white children were born to
unmarried parents in 1999, compared to 69 and 42% for African American and Hispanic
children, respectively (Ventura & Bachrach, 2000).
Certain issues may stand in the way of a healthy marriage for low-income couples. First,
some unmarried parents set an ―economic bar‖ as a precondition to marriage that could be
unrealistic, and many have trust, fidelity, and commitment issues (Gibson, Edin, & McLanahan,
2005). Second, the prevalence of childhood sexual abuse may be higher among disadvantaged
individuals making it more difficult to form healthy adult relationships (Cherlin, et al., 2003).
Third, couples who conceive a child soon after beginning to date need more time to form a
relationship (Mincy, et al., 2003). Fourth, both unmarried and married low-income couples often
struggle with issues related to having children by multiple partners (Mincy, 2002). Fifth,
compared with the general population, lower income couples tend to be less well educated and to
have had less success in school, to have lower literacy levels, and to be members of minorities
24

and come from diverse cultural backgrounds (Fein, 2004; Karney, Garvan, & Thomas, 2004).
The model of healthy marriage must be inclusive of background characteristics and possible risk
factors, as well as the demographics of those who are served by Healthy Marriage Initiatives.
Current Trends for Low-Income Individuals along the Marriage Continuum
Healthy Marriage Initiatives serving low-income couples must be cognizant of current
marriage trends. For example, most low-income unmarried couples with children would like to
marry but are restricted by barriers (Edin & Reed, 2005). Some couples choose to cohabitate
with intentions to marry, and for others cohabitation is an alternative to marriage. Because of
this trend, any model of healthy marriage for low-income populations must extend this view to
unmarried couples. For those married with children, relationships may be unstable and include
children prior to the marriage. A model predicting child well-being from healthy marriage and
healthy relationships must understand the different trajectories that are at play. The following
trends offer a description of who is realistically being served by Healthy Marriage Initiatives:
both unmarried and married couples.
The Married Poor
David Fein (2004) offers a fundamental profile of economically disadvantaged married
couples with his review of results from published reports, as well as tabulations of data from the
Current Population Survey (CPS) and two surveys of married couples. The CPS is a large, highquality survey conducted by the Bureau of the Census for the Bureau of Labor Statistics that
offers rich detail on demographic and economic characteristics. Seventy thousand married adults
aged 18-59, representing 89 million married adults nationwide were included in the March 2003
CPS sample. Low-income married couples participating in Healthy Marriage Initiatives may
have the following descriptions according to Feins‘ review (2004) of the married poor:
25

•

People with economic disadvantages are just as likely to marry as other people, but their
marriages are substantially more unstable.

•

Whereas the vast bulk of first transitions to parenthood among upper middle class
couples follow first marriages, first births among disadvantaged newlyweds are far more
likely to precede marriage.

•

Disadvantaged married couples are comparatively young and disproportionately
Latino. African Americans constitute a relatively small share of this population.

•

Low-income married couples have substantially lower levels of education and
employment than higher-income couples.

•

Millions of married couples use a wide variety of government assistance programs.

•

The characteristics of major racial and ethnic groups within the low-income married
population vary by age, age of youngest child, education, region of the U.S., and urbanrural residence.

•

There is a marital quality gap between low-income and other couples, but it is not as
large as might be expected based on differences in marital disruption rates.

Marital Attitudes of Low-Income Women
Cherlin, Cross-Barnet, Burton, and Garrett-Peters (2008) tested three propositions
regarding mothers‘ attitudes toward childbearing, marriage, and divorce using data on lowincome mothers in Boston, Chicago, and San Antonio. They found that having children outside
of marriage carries little stigma among low-income women. This finding was supported with
earlier research by (Edin & Kefalas, 2005). In addition, there was little cultural support in lowincome neighborhoods for the idea that the best ages to have children is during the teenage
years. Finally, divorce was not found to be stigmatizing or embarrassing for the large majorities
26

of low-income mothers in the study. There was a lack of support for the idea that most lowincome women have a fear of divorce leading them to avoid marriage. See Table 1 for detailed
results of this study.
Table 1
Cherlin et al., 2008 Study Results
Stigma of children outside of
marriage:
About 80% of the mothers disagreed
that out-of-wedlock childbearing is
embarrassing and hurts a women‘s
potential for future marriage. About
80% agreed that a woman should
have children if she wants, even if
unmarried and about 83% agreed
that marriage does not need to
precede child bearing.

Best ages to have children
and marry:
Most women in the sample
viewed childbearing and
marriage as both starting
during one‘s 20‘s. A
majority of the mothers
thought the best ages to start
childbearing and to marry are
within 5 years of each other.

Fear of divorce:
Almost two thirds of the
sample agreed that
‗‗Most people do not care
if a woman gets a
divorce,‘‘ and only about
one fourth agreed that
‗‗Getting divorced is
embarrassing for a
woman.‘‘

These attitudes may explain the reasons for why Healthy Marriage Initiatives recruit not only
married couples, but also many unmarried individuals and couples with children.
The Non-Married Poor and Unwed Child-Bearing
Larger percentages of poor women have not married by age 35 than women who are near
poor or not poor among all race and ethnic groups aged 35-44, but especially true for poor nonHispanic black women (53% of poor non-Hispanic black women compared with 19% of poor
Hispanic and non-Hispanic white women). In addition, larger percentages of poor non-Hispanic
black men aged 35-44 years have not married by age 35 than not-poor non-Hispanic black men
(200% or more above poverty) (Goodwin, McGill, & Chandra, 2009).
Besides the married poor, there is current research on non-married low-income
individuals including those who cohabitate. For example, qualitative work by Furstenberg
27

(2001) and Edin and Reed (2005) suggest that low-income women face many barriers to
marriage, including gender mistrust, abuse, infidelity, and multiple-partner fertility. Over five
years, Edin and Reed (2005) interviewed 162 low-income mothers living in eight poor
neighborhoods and found that poor women may have personal characteristics (e.g., low
education, mental health problems, or out-of-wedlock births) that reduce their opportunities for
marriage or make it difficult to maintain healthy relationships. (Graefe & Lichter 2002; Qian,
Lichter, & Mellott 2005). This implies that Healthy Marriage Initiatives and healthy marriage
models need to consider the personal characteristics of participants, including for example, areas
of mental health (i.e. depression, stress).
Low-income women, especially African-American women are more willing to hold off
on marriage than on child bearing. (Edin & Kefalas, 2005; Cherlin, Cross-Barnet, Burton, &
Garrett-Peters, 2008). Edin and Kefalas (2005) claimed that although finding a suitable partner
to marry remains difficult, young women do not hold off on childbearing because they place
such a high value on children and may have children when they are still teenagers. In addition,
many low-income women prefer to marry when they are older, for example, in their thirties.
While Edin and Kefalas (2005) argue that women are hesitant to marry because of the stigma of
possible divorce, Cherlin, Cross-Barnet, Burton, Garrett-Peters, (2008) found that divorce was
not found to be stigmatizing or embarrassing for the large majorities of low-income mothers in
their study. In their study on the marriage patterns among unwed mothers before and after
PRWORA, Graefe and Lichter (2008) found that unwed childbearing is associated with lower
marriage rates and marital quality. In addition, nonmarital birth mothers were less likely to
marry "economically attractive" men in the post-welfare reform period compared to other
women. Many low-income children live primarily in environments with a single parent, or with
28

their unmarried biological parents, or with one biological parent and that parent‘s intimate
partner.
Low-Income Cohabitation
In recent years, researchers have had much interest in cohabitation; however, patterns of
cohabitation in the general population and in low- income populations in particular is less
understood (Ooms, 2002). Research shows that some cohabitating unions end and others remain
as an alternative to marriage.
Cohabiting unions among the poor are not sustained. About one-half end within one
year, and over 90% end by the fifth year. Break through research by Lichter, Qian, & Mellott
(2006) found that most cohabiting unions end by dissolution of the relationship rather than by
marriage. The higher dissolution fits with other research showing that cohabiting couples have
lower quality relationships than currently married couples (Brown, 2000).
In addition, poor women are less likely to marry, especially minority women; less than
one-third marry within five years. Poor women are also more likely than nonpoor women to
choose cohabitation as a long-term alternative or substitute for traditional marriage (Lichter,
Qian, & Mellott, 2006). Despite the low rates of marriage for cohabitating couples among
African Americans found in prior research (Manning & Smock, 1995), it appears that Black men
and women have higher odds of beginning their cohabitation with ―marital intentions‖ than
White men and women, with odds of 1.48 and 1.23, respectively (Guzzo, 2009). Guzzo (2009)
suggests that African Americans may hold the intention to marry as a sufficient level of
commitment, and may even display this with an engagement ring. Edin and Kefalas (2005)
observed a new trend in poor urban communities where individuals become engaged as a way of
making a public commitment but without actually marrying and having the risk of divorce. In
29

addition, some African Americans may continue to believe in marriage but a committed
cohabitating relationship may be an acceptable ―budget‖ alternative (Furstenberg, 1996) until
other social requisites for marriage are met (Edin, 2000). If this is true, then cohabitation will be
an alternative to marriage among African Americans and especially Hispanics. In a
disproportionately poor and minority sample from the Fragile Families and Child Wellbeing
Study, a mere 15% of cohabiting couples became married over the following year (Carlson,
McLanahan, & England, 2004).
Most unmarried parents have positive attitudes towards marriage, although not as positive
as those of married parents (Waller & McLanahan, 2005). Half of cohabiting mothers ‗strongly
agree‘ that it is better for children if their parents are married and 90% report that their chances
of marriage are ―fifty-fifty or better.‖ Non-cohabiting parents also hold positive views toward
marriage although they rate their chances of marriage much lower. Research suggests that many
want to marry but face barriers in doing so (Guzzo, 2009; Edin, 2000; Edin & Kefalas, 2005).
Some believe that cohabitation has become a normal ―stepping-stone to marriage‖.
About half of all first unions before age 25 are cohabitations (Bumpass & Lu, 2000). However,
because of findings by Lichter, Qian, and Mellott (2006) that most cohabiting relationships end
through dissolution instead of marriage, serial cohabitation may becoming standard as cohabiting
unions develop and break up several times before marriage. Cohabitation may then be an intense
form of dating where couples are quick to live together but not necessarily to marry (Sassler,
2004). These trends and needs of cohabitating couples must be a consideration of Healthy
Marriage Initiatives serving these couples and may be an area of focus for policy interventions
and implications for children.

30

Rhoades, Stanley, and Markman (2009) found that the degree to which individuals report
cohabiting to ―test their relationships‖ is associated with more negative couple communication
and more physical aggression as well as lower relationship adjustment, confidence, and
dedication. Putting the relationship to a test is also related to higher levels of attachment
insecurity and more signs of depression and anxiety. Men are more likely than women to
support testing their relationships and less likely to support ―convenience‖ as a reason for
cohabiting. Couples who cohabit because they have doubts about a successful marriage may be
most at risk for divorce, if they marry. Besides ―testing the relationship‖, men deciding to
cohabitate must weigh the possible economic costs of caring for their partner's coresidential
children from a prior relationship (Lichter & Graefe, 2001). Cohabiting women whose partners
have children must share their partner's time and finances with children from another woman
residing somewhere else. Garfield & McLanahan (2000) caution that instable cohabitating
relationships may be more detrimental to children than being raised by a stable single parent.
A large percentage of women, both welfare recipients and non-recipients, believe that
parents should marry, but also that single parents can raise children as well as married couples
(Mauldon, London, Fein, & Bliss, 2002). Pickhardt (2003) notes that single parents have many
strengths including but not limiting to being highly committed (dedicated to the welfare of their
children); being able to manage diverse family functions (relinquishing sex role stereotypes);
creating networks of support; and maintaining family values (attention on children).
Marital Dissolution Among The Poor
Low-income individuals have become increasingly likely to have children before
marriage, and divorce for low-income married couples remains very high while declining for
more affluent couples (Ellwood & Jencks, 2004; Martin, 2006; Raley & Bumpass, 2003).
31

Particularly, low educated married couples are more likely split (Fein, 2004; Martin, 2006). A
couple of explanations have been offered by Martin (2006) including age at marriage, premarital
births, and race and ethnicity. First, higher educated women are delaying marriage more than
lower educated women and delaying marriage is a predictor of lower marital dissolution rates. A
second explanation is that first marriages are likely to follow a non-marital birth among lower
educated women (Upchurch, Lillard, & Panis, 2001). A third explanation is that without
statistical controls, racial trends in marital dissolution rates may be conflated with educational
trends (Sweeney & Phillips, 2004). The increasing class differences in divorce rates intensify the
effects of socioeconomic inequality which transmits inequality across generations effecting adult
and child outcomes (Amato, 2000).
Many low-income couples despite their race or ethnic background do not follow
conventional stages of courtship (cohabitation), marriage, and childbearing. Instead family
formation often begins with the (typically unplanned) birth of a child (Ooms, 2002). Many times
if there is marriage, divorce follows. Families formed in this way are more complicated with
joint decision-making, co-parenting, and divisions of labor. In addition, relationships between
the couple and the child‘s non-residential parent are challenging (Ooms, 2002). An
understanding of these stages is important to Healthy Marriage Initiatives as well as models of
healthy marriage/healthy relationships that predict child well-being.
Risk Factors among Low-Income Couples
Couple relationships cannot be understood fully without including other influencing
factors. Rather than addressing healthy marriage/healthy relationships, parenting and child wellbeing in isolation, risks are often interrelated, occur within the context of economic
disadvantage, and may require a comprehensive approach for supporting families. For example,
32

Bratter and Eschbach (2005) found that family income and employment are negatively
associated with levels of psychological distress, whereas status as unmarried is positively related
to distress scores. Distress scores are more likely to be higher among women, the least
economically stable, the unmarried, and those with a chronic illness. In addition, low SES
pertains to less distress for African Americans compared to whites, which was similar to
previous studies (Williams, Takeuchi, & Adair, 1992). Bratter and Eschbach (2005) suggest that
intergroup dynamics, such as racial discrimination and acculturation, as well as structural issues,
such as wealth inequality, may explain differences in health and psychological well-being in
more meaningful ways than individual indicators of SES (Williams & Collins, 1995). Lasting
effects of perceived discrimination and inability to translate high education into higher net worth
may influence the ability of African Americans to handle stressful life events at higher ends of
the socioeconomic continuum. Addressing extrinsic factors is especially important in marriage
and relationship education programs for low-income couples, because these factors likely
produce greater difficulties in preserving healthy relationships.
Parenting and Child Well-Being
It is important to note the dynamic conceptions of individuals in poverty such as
permanent and transitory income, cumulative and persistent poverty, and the number and length
of poverty spells. However, even these measures fail to portray key aspects of a child‘s history
of exposure to economic deprivation; for example, there is no distinguishing between families
with improving economic circumstances and those with deteriorating or stable economic
circumstances. This study does not examine the effect of poverty on children; rather its aim is to
study effects of healthy marriage/healthy relationships and effects of parenting on children in
poverty, as defined by their enrollment in the Head Start program.
33

Healthy Marriage Initiatives have the ultimate goal of positively influencing the wellbeing of children in poverty. McLoyd‘s (1990) family stress model is helpful to consider
because it describes how poverty places stress on the psychological well-being of parents.
McLoyd (1990) argues that ―psychological distress‖ is a significant mediator between economic
hardship and parenting. This distress derives from negative life events, undesirable chronic
conditions, and the absence and disruption of marital bonds. Other assumptions of the model
are: (a) poverty and economic loss reduce the ability for supportive, consistent, and involved
parenting; (b) economic loss and poverty indirectly affect children through their influence on the
parent‘s behavior toward the child; and (c) under the conditions of economic hardship, fatherchild relations depend on the quality of relations between the mother and father.
In a more recent study, Raver, Gershoff, and Aber (2007) also found clear evidence for
the family stress model. Namely, lower income was associated with increased hardship, higher
stress, lower parenting behavior, and lower social skills in kindergarten teachers‘ ratings.
Interestingly, they compared White, Black, and Hispanic 6-year-olds, using the Early Childhood
Longitudinal Study – Kindergarten Cohort (ECLS – K). Negative associations between material
hardship, parents‘ stress, and parenting behavior differed in magnitude across the three ethnic
groups. While pathways from income to material hardship fit similarly across all three groups,
higher material hardship was more strongly associated with higher levels of parents‘ stress for
Black families (β = .33) than for Hispanic families (β = .19) and White families (β = .24). Higher
stress, in turn, was more strongly related to lower levels of positive parenting behavior for Black
and Hispanic families (β = -.77 and β = -.82 respectively) than for White families (β = -.64).
Lastly, coefficients for paths between positive parenting behavior and child social competence

34

were larger for White families (e.g., β = 1.93) than for families of color (β = 1.31 for Black
families and β = .99 for Hispanic families).
Unresponsive and harsher, more punitive parenting occurs more often among low-income
families, beginning as early as infancy (Magnusson & Duncan, 2002; McLoyd, 1998). In a
national study, 85% of American parents above the poverty line were responsive to their young
children (from newborns to three-year olds), compared with 74% of parents in low-income
homes (Bradley, Corwyn, McAdoo, & Coll, 2001). Furthermore, a meta-analysis demonstrated a
strong and consistent negative relation (d + = .48, .55, for cross-sectional and longitudinal
studies, respectively) between socioeconomic disadvantage and harsh, less responsive parenting
(Grant, et al., 2003).
Research on how parenting influences the development of children in poverty is
extensive (McGroder, 2000; Yeung, Linver, & Brooks-Gunn, 2002). The search for explanatory
processes of poverty's impacts on children is commendable. Popular interventions for children
in poverty include targeting the child‘s early childhood environments including home and
school. Owens and Shaw (2003) have identified several child and family factors associated with
the positive adjustment of impoverished children. For example, they found that positive
adjustment at age 8 was associated with secure attachment manifested at 18 months, low
maternal aggressive personality at 1.5 to 3.5 years, and positive mother-child relationships at 5 to
6 years. This finding suggests that a critical intervention to children in poverty would include
concentrated attention on promoting positive mother-child relationships, especially for those
children experiencing chronic poverty (Owens & Shaw, 2003). This dissertation study
appreciates father involvement by including males (both married and unmarried men as well as
biological and non-biological fathers who may play a role in the child‘s life) to appraise their
35

own marriage/partnerships as well as on their parenting. Current research by Johnson, Shears,
and Perry (in press) shows that African American fathers in the Fragile Families Study are
actively engaged in cognitive, physical play, socialization and cognitive activities with their
children. Multivariate analysis reveal that a variety of fathering activities were significantly
correlated ranging from .43 - .66, suggesting various forms of participation with their children.
In addition, there was a significant correlation between cognitive activities and Bayley scores (r
= .33, p <.01). The results support father involvement by African American males and validate
benefits for children.
Healthy Marriage Initiatives attend to building healthy marriage/healthy relationships as a
source of social support system for families in poverty, and use educational approaches to
encourage healthy family functioning and positive parenting. These initiatives are not in any
way considered an immediate solution to lifting all families out of poverty. Definitions and
evaluations of healthy marriage for low-income couples need to consider the complexity of
family relationships while including possible risk factors that may influence these relationships.
Defining Healthy Marriage
The first step toward operationalizing outcomes is to define ―healthy marriage.‖ This
project recognizes various definitions but chooses the definition and conceptual model from
Child Trends, Inc. as a guide for this research. ACF‘s definition of ―healthy marriage‖ is a good
start as it provides a framework for the purpose and design of federally funded HMIs. This
definition highlights two common characteristic that healthy marriages share: ―First, they are
mutually enriching, and second, both spouses have a deep respect for each other. It is a
mutually satisfying relationship that is beneficial to the husband, wife and children (if present).
It is a relationship that is committed to ongoing growth, the use of effective communication
36

skills and the use of successful conflict management skills‖ (ACF-HMI homepage retrieved
November 2009). Furthermore, ACF‘s Community Healthy Marriage Initiative uses a concept
of healthy marriage guided by Lewis and Gossett (1999), who define eight essential
characteristics of a ―healthy marriage‖:


Both partners participate in the definition of the relationship



There is a strong marital bond characterized by levels of both closeness and autonomy



The spouses are interested in each other‘s thoughts and feeling



The expression of feelings is encouraged



The inevitable conflicts that do occur do not escalate or lead to despair



Problem-solving skills are well developed



Most basic values are shared



The ability to deal with change and stress is well developed.
A noteworthy definition by Stanley and Markman (Stanley, 2004; 2007) proposes three

dimensions to ―healthy marriage‖ extending from two types of safety found in successful
relationships and marriages (Stanley, Markman, & Whitton, 2002). The three essential elements
of healthy marriage are defined as:
1). Interaction safety- Being able to talk openly and well (enough) about key issues without
repeated negative interactions (escalation of conflict, criticism, put-downs, withdrawal,
contempt, and so forth).
2). Personal safety- Mutual respect and understanding, and freedom from fear of physical or
emotional harm and intimidation
3). Commitment safety- Security of mutual support both now and in the future.

37

Table 2 below includes additional definitions featured in the memos to Child Trends, Inc.
(Jekielek et al., 2003) gathered from professionals to provide insight on the conceptualization
and measurement of ―Healthy Marriages‖:
Table 2
Healthy Marriage Definitions
Researcher(s)
Cowan & Cowan
(2003)

Fagan (2003)

Fowers (2003)

HM Definition
Definition of a healthy couple relationship:
• Partners have and foster positive perceptions of the
relationship and the self in relationship.
• Partners have and foster interactions that promote the
development of family members and reduce the risks of both
individual and relationship psychopathology.
In a healthy couple relationship, the partners have reasonably
good mental health, they communicate clearly, they are
connected with their families of origin and people and institutions
outside the family, and, if they are parents, they function
reasonably effectively in rearing their children.
A healthy marriage is one where the couple always belongs to
each other and to their children, no matter how difficult this
belonging may be at times. At minimum that means they never
reject each other.
In the positive, the central dynamic of a healthy, stable marriage
is that each spouse serves the other. This dynamic and the
capacity to live in this way needs to be assessed, both
behaviorally and cognitively.
Fowers proposes a six-dimension model to capture the
multidimensional nature of long-term romantic
relationships:
• Individual Benefits, such as satisfaction, and experiences of
emotional and physical intimacy
• Individual Contributions, such as commitment and personal
investment into a relationship
• Couple Interaction, as indicated by specific communicative
behaviors
• Relationship Entity, such as a sense of the relationship as a
team, or a sense of ―we‖
• Family Functioning, the degree to which the relationship
38

Table 2 (cont‘d).

Heyman (2003)

Nock (2003)

Olson &
Knutson (2003)

enables fulfillment of family roles
• Relationship Context, such as social, legal, extended family
and friendship networks which might offer support for the
relationship.
A healthy relationship is one that safeguards partners’ physical
integrity, minimizes their risk of preventable physical and
mental disorders, and promotes their physical and mental
well-being.
Similarly, when partners are parents, a healthy relationship
safeguards children‘s physical integrity, minimizes their risk of
preventable physical and mental disorders, and promotes their
physical and mental development and well-being.
A healthy marriage (or relationship) is one in which partners are
sufficiently committed to their union to sustain the relationship in
those times when love and affection cannot.
A healthy marriage can be conceptualized as a loving,
committed relationship where both partners are satisfied with
their relationship. Two cornerstones of a happy marriage are
strong communication and conflict resolution skills.
This definition is based on the ―National Survey of Marital
Strengths‖ of 21,501 married couples who took the ENRICH
Couple Inventory.

These definitions share commonalities, each with a slightly different focus. Combined, these
definitions of healthy marriage most generally refer to a mutually committed relationship in
which partners participate and interact (i.e. with positive communication and conflict resolution)
to promote safety, development, sustainability, and well-being for the family.
The healthy marriage definition written for Child Trends, Inc. by Moore et al. (2004) that
was outlined in Chapter 1 is the chosen definition which includes ten healthy marriage
constructs.

39

Healthy Marriage Constructs and Associations to Child Well-being
This section more specifically describes the ten constructs of healthy marriage identified
by Moore et al. (2004) that are used in framing this study and describes how the constructs relate
to other aspects of relationship quality and child well-being.
Couple Commitment
This construct refers to viewing one‘s relationship with a future in mind, planning to pull
through difficulties, and staying committed to caring for the other person- a sense of ―us‖ or
―we-ness‖ (Levinger, 1979; Fowers, 2003). Commitment is an important aspect of relationship
quality for adults (Amato & Rogers, 1999) as well important for child well-being (Amato &
Booth, 1997; McLanahan & Sandefur, 1996). Moreover, an essential type of safety
characteristic of healthy relationships and marriages is the security of a mutual supportive
relationship both now and in the future (Stanley, Markman, & Whitton, 2002).
Stanley, Whitton, and Markman (2004) noted that literature on commitment primarily
focused on the institution of marriage rather than on dedication between partners (e.g.,
Ciabattari, 2004). Partners with stronger dedication (interpersonal commitment) tend to have a
stronger sense of couple identity and to report a stronger desire for a future together. Committed
couples place high priority on the needs of their partners and the relationship and are willing to
make sacrifices for one another (Stanley & Markman, 1992; Whitton, Stanley, & Markman,
2002). Rather, constraints refer to those things that keep individuals together despite personal
dedication and help to explain why some individuals stay in unhappy relationships (Stanley &
Markman, 1992). In his response to Child Trends, Inc. (Jekielek, Moore, Carrano, & Matthews,
2004) on the Conceptualization and Measurement of Healthy Marriages and Positive
Relationships, Steve Nock writes that for both parties, the imagined commitment of one‘s partner
40

is by far the most important factor in producing commitment and an indication of obligation to
that person.
Commitment is accepted as vital to healthy relationships and is directly related to marital
satisfaction and survival, although researchers differ as to the precise definition and nature of the
construct (Stanley, Markman, & Whitton, 2002). Stanley noted in the findings of a survey of
2,300 divorced residents in Oklahoma that 85% believed ―lack of commitment‖ was the major
reason for divorce (2002). Those who report higher levels of commitment were less likely to
report having serious thoughts about alternative partners, were less likely to report feeling
trapped, and were more likely to report relationship satisfaction (Stanley, Markman, & Whitton,
2002).
Marital Satisfaction
This construct is defined as the individual happiness regarding various aspects of the
marriage and with the overall marriage. Marital satisfaction significantly influences the broader
family context of children. Analyses of data from 1,101 participants in the National Survey of
Families and Households indicated that mothers‘ and fathers‘ reports of marital satisfaction were
negatively associated with child symptomatology (Fishman & Meyers, 2000). Moreover,
dissatisfied wives were less involved with their children, which sequentially was associated with
greater child psychological distress. This finding is compatible with the spillover hypothesis,
which predicts that mothers who deal with problematic marital relationships are unable to give as
much time to their children.
Participation with various relationship education classes can improve couples satisfaction
for a while (Carroll & Doherty, 2003). Involvement in premarital education was associated with
higher levels of satisfaction and commitment in marriage and lower levels of conflict, in addition
41

to reduced chances of divorce. These projected effects were robust across race, income
(including among the poor), and education levels, suggesting the great benefit of premarital
education (Stanley, Amato, Johnson, & Markman, 2006). Satisfaction is the most frequently
studied component of marriage, followed by couple communication (Fowers, 2003; Fowers,
Bucker, Calbeck, & Harrigan, 2003).
Communication
A popular area of study and a focus for marital interventions (Ragan, Einhorn, Rhoades,
Markman, & Stanley, 2009), communication is another element of healthy marriage that some
view as the substance of relationships (Dindia & Timmerman, 2003). Communication and
relationship quality are related (Kelly, Fincham, & Beach, 2003; Weigle, 2003). Most
importantly is the quality and nature of communication. Marriage education produces positive
effects on relationship quality and marital communication (Carroll & Doherty, 2003; Hawkins et
al., 2008) which is maintained beyond 6 months (Blanchard, Hawkins, Baldwin, & Fawcett,
2009). Marriage educators may promote teaching communication techniques to prevent the
erosion of marital satisfaction (Laurenceau, Stanley, Olmos-Gallo, Baucom, & Markman, 2004).
Conflict Resolution
A prominent topic in public policy (Stanley, 2004- testimony on healthy marriage to the
US Senate), conflict resolution reflects the ability to handle disagreements and problems that can
weaken a relationship. This involves successful problem solving and may involve a respectful
agreement to tolerate the opposing position or the mutual recognition of outside sources to the
conflict. Conflict is a risk factor for marital distress and divorce (Clements, Stanley, &
Markman, 2004) and poorer child adjustment (Grych & Fincham, 2001; Bradbury, Fincham, &
Beach, 2000). Conflict resolution skills can be changed (Fagan, 2003) and is thus a goal for
42

marriage education. Fincham, Stanley & Beach (2007) demonstrate how the study of conflict
has been center stage in psychological research on marriage and propose a shift to include a more
comprehensive view of marriage.
Lack of Domestic Violence
This incorporates both physical assaults and psychological abuse (Strauss, 1992; Strauss
& Gelles, 1990) with distinctions between various types of violence on the basis of the control
contexts (Johnson, 2003) against a partner or child. Johnson (2006) describes four major types
of intimate partner violence: (a) situational couple violence, (b) intimate terrorism, (c) violent
resistance, and (d) mutual violent control. These types are conceptualized by the control motives
of the perpetrator that indicate an attempt to use general control over one‘s partner (Johnson,
2006). He has demonstrated, using a variety of data sets, the different causes, different
developmental trajectories, and different consequences of these types of violence (Johnson,
1999, 2001, 2006; Johnson, Conklin, & Menon, 2002; Johnson & Leone, 2005; Leone, Johnson,
& Cohan, 2003; Leone, Johnson, Cohan, & Lloyd, 2001).
Couple violence and child abuse are obvious indicators of an unhealthy relationship.
Intimate partner violence is recognized as a serious public health problem (Garcia-Moreno,
Jansen, Ellsberg, Heise, & Watts, 2006). Although declining rates of partner violence have been
suggested (Hitt, 2001), the prevalence of domestic violence is extremely high. In the United
States, nearly one in four women reports experiencing violence by a current or former
spouse/partner at some time in her life (Center for Disease Control and Prevention, 2008). Over
15 million children in the United States live in families in which partner violence occurred at
least once within the year, and seven million children live in families with severe partner

43

violence (Whitfield, Anda, Dube, & Felittle, 2003). A true understanding of healthy marriage
must consider the prevalence and nature of violence.
A violent relationship is inherently unhealthy. In addition, relationship violence is
associated with a variety of other relationship health indicators, such as marital satisfaction and
marital stability (Arias, Lyons, & Street, 1997; Fincham, Bradbury, Arias, Byrne, & Karney,
1997; Lawrence & Bradbury, 2001; Marshall, Weston, & Honeycutt, 2000; Rogge & Bradbury,
1999; Testa & Leonard, 2001). A recent study by Afifi et al., (2009) found an association
between male and female victims of interpersonal violence (IPV) and past child abuse. Also,
IPV was strongly associated with some psychiatric disorders, with females experiencing a wider
range of poor mental health outcomes compared to males.
Most nonfatal intimate partner victimizations of women (two-thirds) in the United States
occur in the home environment (Catalano, 2007). Young children under age 12 live in
households with intimate partner violence (IPV) in 38 percent of incidents involving female
victims (Catalano, 2007). The effects of IPV on children‘s adjustment have also been well
documented (Bancroft & Silverman, 2004; Graham-Bermann & Edleson, 2001; Jaffe, Baker, &
Cunningham, 2004; Attala, Bauza, Pratt, & Vieira, 1995; Brandwein, 1998; Campbell &
Lewandowski, 1997; Kolbo, Blakely, & Engleman, 1996; Widom, 1989; Wolak & Finkelhor,
1998). Violence has an independent effect on children‘s adjustment and is significantly more
harmful than high levels of marital conflict (McNeal & Amato, 1998).
Fidelity
Being faithful to one‘s spouse continues to be a key element in conceptualizing healthy
marriage. Infidelity, defined as ―a secret sexual, romantic, or emotional involvement that
violates the commitment to an exclusive relationship‖ (Glass, 2002, p. 489) can have many
44

negative effects on the individuals and relationship as a whole. Infidelity is the number one
cause of divorce (Amato & Previti, 2003), and many times leads to depression (Cano & O‘Leary,
2000) among partners of unfaithful individuals. Qualitative work by Smock and Manning (2003)
indicates that infidelity is a ―deal breaker.‖ In other words, the betrayal of trust caused from
infidelity most often ends relationships.
Sexual mistrust and infidelity combined are the most common relationship problems and
are often associated with breakups among unmarried parents (Reed, 2005). Mistrust and
infidelity are among the top marriage barriers for low-income women (Edin & Reed, 2005).
Sexual mistrust and infidelity fears result from real knowledge of previous infidelity, women‘s
general distrust of men (Edin et al., 2003), and sustained social contact with the parent of a
partner‘s children from a prior relationship. Even the presence of a former intimate partner may
put the current romantic relationship in danger (Johnson, 2003). Ongoing co-parenting with an
ex-partner can challenge trust and fidelity in one‘s current relationship especially for unmarried
persons, whose relationship did not end in divorce (a common pattern among low-income
couples), and yet that couple maintains contact on behalf of the children.
Hall and Fincham (2006) place importance on the role of forgiveness (not necessarily
reconciliation) as the best possible outcome for couples affected by infidelity. Forgiveness can
have significant emotional and physical health benefits even for couples who split (Toussaint
&Webb, 2005; Toussaint, Williams, Musick, & Everson, 2001; Witvliet, Ludwig, & Vander
Laan, 2001).
Interaction and Time Together
This applies especially to quality of the interaction and enjoyment of time together, but
also to the amount of time together and what couples do jointly. Marriage education programs
45

focus on improving the quality of marital interaction — behavioral exchanges and associated
thoughts, feelings, and physiological responses (Knox & Fein, 2008). Marital interaction
includes day-to-day activities and social activities (Booth, Amato, Johnson, & Rogers, 2003).
The amount of time couples enjoy spending together also contributes to marital success (Aron,
Norman, McKenna, & Heyman, 2000; Presser, 2000).
Alternatively, negative interaction between partners (more expressions of negative affect)
is negatively associated with every aspect of relationship quality, positively associated with
considerations and talk of divorce (Stanley, Markman, & Whitton, 2002), related to lower
relationship satisfaction and higher rates of divorce or break-up (Gottman & Notarius, 2000).
Intimacy and Emotional Support
This includes feelings of trust, caring, love, and physical affection. Emotional intimacy,
a feeling of closeness supported by trust and mutual feelings of vulnerability, is another form of
positive interaction (Cordova, Gee & Warren, 2005). High levels of intimacy is associated with
improved psychological, physical, and relational well-being (Hassebrauck & Fehr, 2002), greater
marital satisfaction (Laurenceau, Feldman Barrett, & Rovine, 2005), and reduced divorce risk
(Firestone & Firestone, 2004; Schneller & Arditti, 2004). Some research implies that trust is
dependent on intimacy and necessary for a successful relationship (Larson, Hammond, &
Harper, 1998). Romantic love performs a commitment role through intimacy, connection, and a
mutual planning for the future (Diamond, 2003; Ellis & Malamuth, 2000). Wilcox and Nock
(2006) conclude that the ―functions, character, and stability of contemporary marriages are
intimately tied to their emotional well-being‖ (p. 1340).
Feelings of intimacy may be different for men and women (Laurenceau, Feldman Barrett,
& Rovine, 2005). Compared with husbands, wives‘ increases in intimacy are more powerfully
46

dependent on feeling understood, validated, accepted, and cared for while husbands‘ felt
intimacy is dependent on self-revealing disclosure. One important action identified in the
qualitative study on male‘s perception of intimacy by Patrick & Beckenbach (2009) was that of
sharing. Intimacy involved several levels of sharing, including emotions, thoughts, words, and
physical expressions. In addition, sharing produces feeling of safety, and in turn, honesty.
Participants indicated that intimacy, as it relates to sharing, is a ―place to be yourself, and
accepted by the other in an atmosphere of trust‖ (p. 52). In an atmosphere of intimacy, physical
affection was also seen as an intensely emotional, connecting experience. Demonstrations of
affection and positive emotion predict stability and satisfaction (Gottman, Coan, Swanson, &
Carrerre, 1998).
Commitment to Children
Commitment to children is part of the ―healthy marriage‖ definition, given the focus of
children‘s well-being. A high-quality couple relationship is linked to better co-parenting
(Carlson & Högnäs, 2009). Cowan and Cowan (2003) indicate that an important dimension of
healthy relationships for couples regarding caring for children is: (a) the ability to co-parent
effectively; (b) joint decision making; (c) satisfaction with the division of labor; (d) and
supportive parenting. Healthy marriage initiatives should include co-parenting as a topic,
considering the positive effects on child development and well-being (an important topic for
future research). Such interventions show potential for strengthening family ties among lowincome couples (Cowan et al., 2009). Bearing in mind the influence of children on marriage,
Anderson (2003) emphasizes the fulfillment of paternal responsibilities.
For couples who have and live with only their biological children, commitment to
children encompasses financial, social, and emotional support over time. Among co-resident
47

households, co-parenting is related to both marital behavior (Schoppe-Sullivan, Mangelsdorf,
Frosch, & McHale, 2004) and child wellbeing (Schoppe, Mangelsdorf, & Frosch, 2001) and may
mediate between the former and the latter (Katz & Low, 2004; Margolin, Gordis & John, 2001).
Within two-parent families, cooperative co-parenting has been linked to mother and father‘s
increased responsive parenting with infant and school-age children (Caldera & Lindsey, 2006;
Floyd, Gilliom, & Costigan, 1998; Margolin et al., 2001).
For divorced or split-up couples, or for visiting couples (those who are romantically
involved and have a child together, but who do not share a common residence), the quality of
their commitment to their children, is more complex. There has been limited attention to coparenting among cohabiting unmarried parents with children. Carlson & Högnäs (2009) address
this limitation with their recent research on co-parenting among unmarried parents. They found
that parent depression diminishes positive co-parenting, as does fathers‘ prior incarceration,
whereas having a high-quality couple relationship is linked to better co-parenting. Further they
suggest that individual mental health and more positive relationship health between parents may
increase co-parenting among unmarried parents, a potentially important goal for social policy.
Duration and Legal Marital Status
These constructs originated in research on family structure and child development.
Compared to children residing with married parents, children with unmarried cohabiting parents
have been found to have more behavior problems, poorer school performance, and higher levels
of psychological distress (Artis, 2007; Brown, 2004; Dunifon & Kowaleski-Jones, 2002;
Manning & Lamb, 2003).
As mentioned earlier, Osborne & McLanahan (2007) found that partnership instability is
very common among children born to unmarried parents, especially those born to mothers who
48

are not living with their child's father. By age 3, between 25% and 30% of children born to
noncohabiting mothers go through three or more partnership transitions. Partnership instability
is positively associated with aggressive and anxious/depressive behavior in children at age 3.
One partnership change has a moderate effect on child outcomes, while multiple transitions yield
large effects for children (Osborne & McLanahan, 2007). Furthermore, Carlson & Högnäs
(2009) report that under two-fifths of unmarried couples are together five years following the
birth of their child. Their summary of factors that encourage union stability among unmarried
parents with children is listed in Table 3 below:
Table 3
Factors Encouraging Union Stability
Findings

Sources

Essential elements for marriage after a
nonmarital birth include men‘s
earnings, wages, and employment.
The key economic factor for women is
education, increasing the likelihood of
marriage.
Aspects of culture (widely shared
beliefs and practices) affect decisions
and behavior around family formation.
Positive attitudes toward and
expectations about marriage, and
religiosity play a role in union stability.
Gender distrust and sexual jealousy
(especially by women toward men),
along with multi-partnered fertility, are
key marriage barriers.
Partners‘ perceptions of the emotional
quality of their marriage affects
whether they stay together or part.

Carlson, McLanahan, & England, 2004;
Gibson-Davis, 2009; Harknett, 2008; Harknett
& McLanahan, 2004
Carlson et al., 2004; Harknett, 2008

Axinn & Thornton, 2000; Clarkberg,
Stolzenberg, & Waite, 1995; Nock, 1995
Carlson et al., 2004; Waller & McLanahan,
2005; Wilcox & Wolfinger, 2007
Carlson et al., 2004; Edin & Kefalas, 2005;
Hill, 2007; Monte, 2007

Cowan, Cowan, Schulz, & Heming, 1994;
Gottman, 1994; Karney & Bradbury, 1995
Among unmarried couples with childrenCarlson et al., 2004

49

Federally funded HMI programs serve low-income diverse individuals and couples at
many points along the marital continuum with improving children‘s well-being as the
overarching purpose. These programs have increased need for improved information on lowincome couples, to better serve the population and for a better understanding of healthy marriage
among the poor. HMIs utilize education as a vehicle to enhance marriages by helping to identify
unhealthy aspects of marriages and promote many elements of a healthy marriage that can be
learned and practiced. Unhealthy aspects of marriages have negative influences on children‘s
well-being including but not limited to child psychological distress, behavior problems including
aggressive and anxious/depressive behavior, and poorer school performance. With a family
process perspective in mind, it is important to evaluate how healthy marriage programs influence
the well- being of adults and how this in turn relates to children‘s outcomes.

50

CHAPTER 3: METHODS
In this chapter I describe the participants, research procedure, and instruments used in the
present study. The first purpose of this research was to examine the factor structure of a newly
developed healthy marriage/healthy relationships measure (HM/HR) by Manpower
Demonstration Research Corporation (MDRC) and other measures of personal characteristics,
parenting, and child well-being for a group of low-income couples with children. The second
purpose was to examine how the factor structure of the HM/HR measure varied across (a)
gender, (b) racial group (White and African American) and (c) marital status (married and
unmarried). The third purpose was to test a measurement model of healthy marriage developed
by Child Trends, Inc. that predicts child well-being. This model is used by Federal Healthy
Marriage Initiative Demonstration Projects.
The primary variables in the healthy marriage/healthy relationships measure were couple
commitment, commitment to children, relationship happiness, quality time together, conflict
frequency, lack of domestic violence, communication, stability, emotional intimacy, as well as
the two added marital virtues of forgiveness and teamwork. Primary variables included in the
Structural Equation Model (SEM) were: Risk (stress and depression), Healthy Marriage/Healthy
Relationships (inclusive of the 11 variables stated above), Parenting Behavior, and Preschoolers
Social Competence.
Participants
Adults
The full sample (n = 343) consisted of 151 men (44%), 189 women (55%), and 3 (1%)
not reporting their gender. There were 168 (49%) African American participants, 131 Caucasian
participants (38%), and 35 (10%) participants who identified themselves as either Hispanic (n =
51

9), Arab American (n = 21), American Indian (n=1), or other (n = 4). Nine participants did not
report on their race. Of the participants, 142 (42%) were unmarried and 196 (58%) were
married. Unmarried persons include those who are single (never married) (n = 42), cohabiting (n
= 72), separated (n = 2), common law marriage (n = 17), and divorced (n = 9). Married persons
are those who are experiencing either a first marriage (n = 172) or remarriage (n = 24). Five
participants did not report on marital status. Table 4 displays demographics for gender by the
two dominant racial groups (African American and Caucasian) and marital status (unmarried and
married).
The participants ranged in age from 20 to 62 years, with a mean age of 33.3 (SD = 7.3)
and a median age of 32 years. Eighteen participants did not report their age. Additional
demographic information such as education levels, income, children from a previous relationship
(multiple fertility), residence with the Head Start child, and marital status length is shown in
Table 5. The dataset was split by gender instead of using couple data as a unit of analysis. The
purposes of this study were to establish the factor structure and model fit for males and females.
After this initial work on factor structure, future research might examine the data from a dyadic
perspective.
Children
Participants who were parents to the Head Start child were asked to complete questions
regarding their parenting behaviors and their child‘s social competence. Those who were not
parents to the Head Start child were asked to skip these questions. In some cases, for example, a
biological residential mother had a non-residential boyfriend who did not participate in the
raising of the child and therefore, did not complete the parenting and child portions of the

52

questionnaires. Instead, the mother was asked to complete the parenting and child information.
There were 334 participants who completed the parenting scales (97%).
Participants reported on their children which collectively reflected 156 boys (45%) and
167 girls (49%); 20 (6%) children had no gender reported. Ages ranged from 3 to 6 years, with a
mean age of 4.5 years (SD = .70). Participants identified 36 children (11%) as having a physical,
mental, or emotional condition or behaving in ways that demand extra attention. In addition, 13
children (4%) were identified by their parents as having a condition or behaving in ways that put
severe stress on their marriage/relationship.
Table 4
Racial and Marital Status Demographics by Parent Gender
____________________________________________
N
%
____________________________________________
Males
Race
African American
72
55%
Caucasian
59
45%
Marital Status
Unmarried
Married
Females
Race
African American
Caucasian

60
90

40%
60%

96
70

57.8%
42.2%

Marital Status
Unmarried
81
43.5%
Married
105
56.5%
____________________________________________

53

Table 5
Other Demographics by Parent Gender
_________________________________________________________
N
%
_________________________________________________________
Males
Education
Not yet completed HS/GED
Completed high school/GED
Some College or VoTech Degree
College: 4 year degree
Completed graduate degree
Not reported

16
44
70
12
4
5

10.6%
29.1%
46.4%
7.9%
2.6%
3.3%

Females
Education
Not yet completed HS/GED
Completed high school/GED
Some College or VoTech Degree
College: 4 year degree
Completed graduate degree
Not reported

12
52
100
15
7
3

6.3%
27.5%
52.9%
7.9%
3.7%
1.6%

Males
Income
No Income
Under $4,999
$5,000 - $9,999
$10,000 - $14,999
$15,000 - $19,999
$20,000 - $29,999
$30,000 - $39,999
$40,000 - $49,999
More than $49,999
Not reported

9
16
22
21
16
34
16
5
6
6

6.0%
10.6%
14.6%
13.9%
10.6%
22.5%
10.6%
3.3%
4.0%
4.0%

Females
Income
No Income
Under $4,999
$5,000 - $9,999
$10,000 - $14,999

26
40
33
32

13.8%
21.2%
17.5%
16.9%
54

Table 5 (cont‘d).
$15,000 - $19,999
$20,000 - $29,999
$30,000 - $39,999
$40,000 - $49,999
More than $49,999
Not reported

13
17
11
5
4
8

6.9%
9.0%
5.8%
2.6%
2.1%
4.2%

Males
Children from a previous relationship
Yes
No
Not reported

56
95
0

37.1%
62.9%
0%

Females
Children from a previous relationship
Yes
No
Not reported

84
104
1

44.5%
55.0%
.5%

Males
Residing with Head Start Child
Yes
No
Not reported

127
14
14

82%
9%
9%

Females
Residing with Head Start Child
Yes
185
97.9%
No
0
0%
Not reported
4
2.1%
________________________________________________________________________
N
%
Range
M
SD
________________________________________________________________________
Males
Marital length
85
55% 1 - 33 years 7.1
5.8

Females
Marital length

101

55

53%

1 - 33 years

7.4

5.5

Procedures
Parent and child participants were drawn from Wayne County Health and Family
Services (WCHFS) Head Start. WCHFS Head Start enrolls 3,723 children within 29 sites across
five delegate agencies in out-Wayne County, Michigan. Wayne County Health and Family
Services Head Start received a Healthy Marriage Initiative (HMI) 5-year grant through the
Office of Head Start and serves singles, unmarried couples as well as married couples. Data
collection was planned to be conducted at one or two sites per delegate agency to get a wide
spread across the County. In addition to collection at Head Start site locations, data collection
was also scheduled to take place at Healthy Marriage/Healthy Relationship (HM/HR)
conferences for Head Start parents. During the project‘s proposal stage, the investigator sought
approval of the research by the grantee administrative team, delegate directors, Policy Council,
and education team of grantee and delegate specialists. Specifically, a presentation was made to
the grantee administrative team on April 7, 2010 to education specialists on April 15, 2010,
followed by a presentation at the directors‘ meeting on May 4, 2010 and to Policy Council on
May 5, 2010. The project investigator explained the purpose of the research and methods of data
collection as well as answered questions that were raised. Unanimous support of the research
was received by all parties through formal voting procedures.
Site Data Collection
Following IRB approval of the project with an ―exempt‖ status on May 10, 2010, the
investigator began contacting delegate directors to select participating sites. Seven sites were
chosen (one site each for three delegates and two sites each for the other two delegates). Sites
that had many ELL families were not considered because of potential literacy challenges with
completing the questionnaire. Data collection was grouped into two waves based on the last day
56

of school. The first wave included those sites with a last day of school in May and the second
wave included sites with a last day of school in June.
The investigator met individually with each of the Head Start site leaders. At this
meeting, data collection procedures were discussed as well as plans for communicating with
teachers. A letter from the researcher to teachers was written to explain the project. The letter
was delivered to every teacher by either placing the letter in their site mailbox or by giving the
teachers the letter at a site staff meeting. After letter distribution, the researcher met individually
with every teacher to answer any questions before data collection began.

Data collection at the site began with hanging research announcement flyers around the
school building. The flyer included the research project eligibility criteria: (1) involvement in an
intimate relationship (marriage/romantic partnership) during the past 12 months (to ensure the
study of relatively stable relationships); (2) parent (biological/step) of a Head Start enrolled child
for at least one partner in the intimate relationship; (3) at least one of the Head Start parents must
live with the Head Start child and must not have lived apart from the child for four months or
more, and (4) participation of willing partner- individual questionnaire completion by both
partners.
Data collection procedures for the site continued with the placement of two bundled
questionnaire packets (one per person in the couple relationship) in each of the parent mailboxes
by the researcher. The questionnaire packets were accompanied with the written instruction that
if parents were eligible, interested, and willing to participate they were to individually complete
the questionnaire at home in private and return it to the classroom in the marked ―questionnaire
drop-box‖. Parents who decided not to participate were asked to return their incomplete packets
in the marked drop-box as well so that the questionnaire could be used for a different potential
57

participant. The intention was that incomplete packets returned during wave 1 would be
potentially recycled and used for wave 2. However, the majority of parents did not return
incomplete packets and additional production of packets for wave 2 was required.
After packets were placed in site parent mailboxes, the researcher was available to
parents who may have needed additional clarification regarding the project. This was performed
by being present in the halls during child drop-off/pick-up times, attending a parent meeting, or
attending a parent end-of the year celebration. The entire site data collection method produced
141 completed questionnaire packets collectively across sites.
HM/HR Conference Data Collection
In addition to the site data collection method, data was also collected at two Head Start
sponsored HM/HR events held on May 22, 2010 and August 21, 2010. Time was reserved for
the researcher to collect data at the event. All parents attending the event were gathered together
in one location where the researcher distributed questionnaire packets, described the research as
well as the consent process and invited those who were eligible, interested and willing to
complete the questionnaire packet on-site during the allotted time. Participants were encouraged
to sit separately from their partner to prevent the sharing of responses. Questionnaire completion
ranged from 20-25 minutes. The researcher was present in the room while participants
completed the questionnaires. All questionnaires, whether completed or not-completed, were
returned to the stationed ―questionnaire drop-boxes‖. The first larger training event produced
completed questionnaires by 170 participants and the second smaller training event yielded 32
complete questionnaires.
With site data (141 packets) and event data (202 packets), a grand total of 343
questionnaires were received. Although parents were asked to participate with a willing partner,
58

41 of the 343 people submitted questionnaires without their partner‘s questionnaire fulfillment.
This occurred with both site and event data collection methods. These questionnaires were
retained when it was decided that analyses for the project would not rely on couple-based data
(mean scores of couple reports) as a unit of analysis because of the independence of partner
scores on some of the healthy marriage/healthy relationship measures.
Incentives
Respondents were informed through the announcement flyer, within the invite attached to
the questionnaire packet, on the questionnaire and on the consent form that by returning the
enclosed raffle ticket, they would be entered into a drawing to win one of ten $50 gift cards. Per
IRB guidelines, participation in the study was not necessary for entry in the drawing. The raffle
ticket, containing the person‘s name and contact information was returned in a separate raffle
ticket drop-off envelope to ensure anonymity of questionnaire data. Separate drawings were
made for those who entered the drawing at the sites and for those who entered the drawing at the
parent events. Winners were randomly selected. Raffle ticket winners from the sites were
contacted using information they provided on the ticket and arrangements were made to receive
the gift cards. The drawings for the parent events took place on location and winners received
their gift cards immediately following the drawing since they were present at that time.
Questionnaire Packet
The entire questionnaire packet included:
(a) Consent Form for respondent retention
(b) Set of three Questionnaires:
Self-Administered Questionnaire- SAQ
Background Information Form- BIF
59

Child Information Form- CIF
(c) Raffle Ticket
(c) Resource List (i.e. domestic violence, substance abuse)
Measures
Background Information Form
Participants completed an investigator-created demographic form which asked about the
participant‘s age, gender, race, employment, income, family and household composition and
marital/relationship status.
Self Administered Questionnaire
The primary measure utilized in this research was an instrument developed by MDRC for
the federal ―Supporting Healthy Marriage‖ (SHM) Project. This measure included the risk scales
and healthy marriage scales. In addition to this main measurement, an additional parenting scale
was included.
Risk measures. The two risk variables that ended up in the main analyses were
depression and stress. However, additional risk variables were used for exploratory analysis.
Depression. To measure depression, the Kessler K6 Nonspecific Psychological Distress
Scale was used (Kessler, et al., 2002). This scale has demonstrated excellent internal consistency
and reliability ( = .89). It also has consistent psychometric properties across major sociodemographic sub samples. The K6 scale was developed with support from the U.S.
government's National Center for Health Statistics for use in the redesigned U.S. National Health
Interview Survey (NHIS). Respondents are asked, ―During the past 30 days, how often did you
feel . . . (so sad that nothing could cheer you up?, nervous?, restless or fidgety?, hopeless?, that
everything was an effort?, worthless?). These 6-items were scored on a 5 point Likert scale
60

ranging from ―none of the time‖ (scored as zero) to ―all of the time‖ (scored as four). The mean
of the 6-items was used with higher scores representing greater nonspecific psychological
distress.
Perceived stress. Two items from the Perceived Stress Scale (PSS) were used to measure
stress. The Perceived Stress Scale is the only empirically established index of general stress
appraisal. It measures the degree to which situations in one‘s life are appraised as stressful
(Cohen, et al., 1983; p. 385). The scale has demonstrated internal reliability ( = .78). The two
items from the four-item scale were: In the past 30 days, ―how often have you felt that you were
unable to control the important things in your life?‖ and ―how often have you felt difficulties
were piling up so high that you could not overcome them?‖ Responses on a 4-point Likert scale
range from ―never‖ (scored as zero) to ―always‖ (scored as three). The two items are averaged
with higher scores representing greater perceived stress.
Childhood happiness. Childhood happiness was measured with a single item: ―On a
scale from 1 to 7, where 1 is ―completely unhappy‖ and 7 is ―completely happy‖, while you
were growing up, how happy was the relationship between the parents or parent figures you
reported in Question 14?‖ Responses ranged from 1 = completely unhappy to 7 = completely
happy. After reverse coding, higher scores represent less happiness.
General health. General health was measured with a single item: ―Would you say your
health in general is… (excellent = 5, very good = 4, good = 3, fair = 2, poor = 1). After reverse
coding, higher score indicated worse general health.
Religiosity. Religiosity was measured with two items: (a) ―In the past 12 months, about
how often have you attended a religious service?‖ and (b) ― In the past 12 months, about how
often have you attended a religious service with your spouse/partner?‖ Responses ranged from
61

1 = never to 4 = once a week or more = 4). After reverse coding, the two items were averaged
with higher scores representing less religious attendance.
Health condition. Health condition was measured with one dichotomous item: ―Do you
have a physical or mental health problem now that limits the amount or kind of work or
activities that you can do in or outside the household? (yes = 1, no = 0).
Child stress on relationship. Child stress on the relationship was measured with a single
dichotomous question: ―Do any of your children have a condition or behave in ways that put
severe stress on your marriage or relationship?‖ (yes = 1, no = 0)
Alcohol dependency. Three questions from the CAGE questionnaire (Ewing, 1984)
were used to measure alcohol dependency: (a) ―Have you felt you should cut down on your
drinking?‖ (b) ―Have people annoyed you by complaining about your drinking?‖ and (c) ―Have
you ever felt bad or guilty about your drinking?‖ Responses were 1 = yes, 0 = no, or 0 = don‘t
drink. The three items were averaged with higher scores indicating higher probability of
alcohol dependence.
Drug dependency. The three above questions that were used to measure alcohol
dependency were adapted to measure drug dependency: (a) ―Have you felt you should cut down
on your drug use?‖ (b) ―Have people annoyed you by complaining about your drug use?‖ and
(c) ―Have you ever felt bad or guilty about your drug use?‖ Responses were 1 = yes, 0 = no, or
0 = don‘t use drugs. The three items were averaged with higher scores indicating higher
probability of drug dependence.
Social support. Two items were used to measure social support: (a) ―Have you and your
current spouse/partner ever attended a marriage education class, workshop, or counseling?
Please fill in all the boxes that apply: Yes, before we got married = 1, Yes, since we got
62

married = 1, No = 0‖ and (b) ―Of the following, who could you turn to if you had an emergency
and needed help? Please fill in all the boxes that apply: My spouse = 1, My partner = 1
Someone else = 1, No one, = 0‖. After reverse coding, higher scores indicated less social
support.
Healthy marriage/healthy relationship measures. Healthy marriage/healthy
relationships was measured using the following eleven constructs from the measurement tool
provided by MDRC: couple commitment, commitment to children, communication, conflict
frequency, conflict resolution process, emotional intimacy, lack of domestic violence, quality
time, relationship happiness, relationship quality, and relationship stability.
Child commitment. Two questions were used to construct the measure of commitment to
children: (a) ―My spouse/partner is the type of parent I want for my child(ren)‖ and (b) ―My
spouse/partner is completely committed to being there for our child(ren).‖ Response options
ranged from 1 = strongly disagree to 4 = strongly agree. Responses were averaged with higher
scores indicating greater commitment of one‘s spouse/partner to children.
Communication. Communication was measured using the following two questions: (a)
―It is hard for me to talk with my spouse/partner about the important things in our life‖ (reverse
coded), with a rating from 1 = strongly agree to 4 = strongly agree and (b) ―My spouse/partner
listens to me when I need someone to talk to;‖ with a rating from 1 = none of the time to 4 = all
of the time. Responses were averaged with higher scores indicating better communication.
Conflict frequency. Conflict frequency was measured with nine items: ―How often do
you and your spouse/partner have arguments about. . .‖ (i.e. Household chores? Sex? Spending
time together? Managing money, bills and debt?). Responses to these items were on a 4-point

63

scale from 0 = never to 3 = often. Responses were averaged with higher scores representing
higher conflict frequency.
Conflict resolution process. Conflict resolution processes were measured with the
following two statements: (a) ‗‗Our agreements get very heated (revere coded);‖ and (b) ‗‗I am
satisfied with the way we handle our problems and disagreements‖. Responses to these
statements ranged from 1 = none of time to 4 = all of the time. Responses were averaged with
higher scores indicating more positive conflict resolution processing.
Couple commitment. Two questions were used to construct the measure of commitment:
(a) ‗‗I view our relationship as lifelong;‖ and (b) ‗‗I believe this relationship can stay strong even
through the hard times‖. Response options for each item ranged from 1 = strongly disagree to 4
= strongly agree. Responses were averaged with higher scores indicating higher commitment to
the relationship.
Emotional intimacy. Emotional intimacy in the relationship was measured using the
following three questions: (a) ―I can count on my spouse/partner to be there for me‖, with ratings
from 1 = strongly disagree to 4 = strongly agree; (b) ―My spouse/partner and I have similar
views about what is important in life‖ (c) and ―My spouse/partner expresses love and affection
toward me‖, both using response options that ranged from 1= none of the time to 4 = all of the
time. Responses were averaged with higher scores representing greater emotional intimacy.
Fidelity. Fidelity was measured by a single item: ―I worry about my spouse/partner
cheating on me‖ with responses ranging from 1 = strongly disagree to 4 = strongly agree. After
reverse coding this item, higher scores reflect greater perceived fidelity.
Lack of domestic violence. The lack of domestic was measured with three questions: ―In
the past year, how often has your spouse/partner… (a) Yelled or screamed at you? (b) Blamed
64

you for his/her problems? (c) and How often have your arguments become physical?‖
Responses ranged from 0 = never to 3 = often. After reverse coding, responses were averaged
with higher scores reflecting less negative behaviors.
Quality time. Quality time was measured with two questions: (a) ―My spouse/partner
and I get along well together‖ and (b) ―We enjoy doing even ordinary, day-to-day things
together‖. Responses to these questions ranged from 0 = none of the time to 4 = all of the time.
Responses were averaged with higher scores representing more quality time together.
Relationship satisfaction. Relationship satisfaction was measured in two ways: first, by
a measure of relationship happiness; and second, by a measure of relationship quality.
Relationship happiness was measured with one question: ―How happy are you with your
marriage/relationship?‖ Response options ranged from 1 = completely unhappy to 7 =
completely happy with higher scores representing greater couple relationship happiness.
Relationship quality was measured with two questions: (a) ―How would you rate the good side of
your marriage/relationship?‖, with ratings from 1 = not at all good to 5 = completely good and
(b) ―How would you rate the bad side of your marriage/relationship?‖, with ratings from 1 =
completely bad to 5 = not at all bad. Responses for these two relationship quality items were
averaged with higher scores representing greater perceived quality.
Relationship stability. Relationship stability was measured with a single dichotomous
item instead of using marital status and duration as proxies for stability: ―In the past year, have
you ever thought your marriage/relationship was in trouble?‖ Response options were 1 = yes or
0 = no. After reverse coding, a ―1‖ represented greater relationship stability.
Healthy marriage measures added to the HM/HR measure. In addition to the eleven

65

healthy marriage/healthy relationships constructs from the MDRC measurement tool, two
additional measures of marital virtues were included.
Marital virtues. Two virtues (two subscales) from the ―Marital Virtues Profile‖
(Hawkins, Fowers, Carroll, & Yang, 2007) were added to the healthy marriage measure:
forgiveness (3 ―forgiveness‖ items of the 7-item generosity subscale: My partner is forgiving of
my mistakes) and teamwork (entire 3-item subscale: Our relationship is based on a deep sense of
teamwork). The MVP was developed to capture a profile of individual and relationship virtues
that contributes to marital quality and stability. Items for the forgiveness subscale were scored
on a 6-point Likert scale ranging from ―almost never‖ to ―almost always‖. Items for the
teamwork subscale were scored on a 5-point Likert scale ranging from ―strongly disagree‖ to
―strongly agree‖. Higher scores represent greater presence of the relationship virtue. Alpha
reliability coefficients for the teamwork subscale are .82 for wives and .80 for husbands. Alpha
reliability coefficients for the forgiveness items range from .77 to .88 for wives and .74 to .79 for
husbands.
Parenting behavior measure added to the HM/HR measure. The PBI: Parent
Behavior Inventory (20-item) was used to measure parenting behaviors. The PBI (Lovejoy,
Weis, O‘Hare, & Rubin, 1999) is a 20-item parent report of parent child-interaction and common
disciplinary practices used with preschool to school-aged children. This measure was used by
Head Start parents in this study. The scale consisted of two factors: Hostile/Coercive (10 items:
―I grab or handle my child roughly‖, ―When I‘m disappointed in my child‘s behavior, I remind
him/her about how much I‘ve done for him/her‖) and Supportive/ Engaged (10 items: ―My child
and I hug and /or kiss each other‖, ―I thank or praise my child‖). Items were scored on a 6-point
Likert-type scale ranging from 0 (not at all true) to 6 (very true). The supportive/engaged
66

parenting construct reflects parental warmth whereas the hostile/coercive parenting construct
reflects hostility as well as certain maladaptive control techniques (Lovejoy, Weis, O‘Hare, &
Rubin, 1999). The PBI has demonstrated strong reliability and validity. The initial validation
test of the PBI (Lovejoy, Weis, O‘Hare, & Rubin, 1999) demonstrated a Cronbach's alpha at .81
for the Hostile/Coercive scale and .83 for the Supportive/Engaged scale. The average interitem
correlation was .32 for the Hostile/Coercive scale, with a range of .09 to .52; the average
interitem correlation was .36 for the Supportive/Engaged scale, with a range of .13 to .58. In
addition, the PBI has shown moderate relationship with measures of parental affect, parental
stress, and child behavior problems. Items on the Hostile/Coercive subscale were reverse scored.
Items from both subscales were averaged with higher scores indicating greater positive (i.e.
supportive, engaged, loving) parenting behaviors. This scale was chosen because of its
usefulness in predicting child outcomes.
Child Information Form
Social emotional development. The Social Competence Scale (SCS) -parent version
(Conduct Problem Prevention Research Group [CPPRG], 1995) was used as an assessment of
child well-being. The measure consisted of 12 items that assess children‘s positive social
behaviors, including emotion regulation, prosocial behaviors, and communication skills. Parents
were asked to rate how well 12 statements describe their child on a five-point scale from 0 (not at
all) to 4 (very well). Internal consistency for the total measure is high with  = .92 for the
community sample, and .88 for the high-risk sample. The total measure was used by taking the
average of all scale items: prosocial/communication and emotion regulation.

67

CHAPTER 4: RESULTS
Preliminary Analyses
Missing Data Analysis
The missing value analysis procedure for PASW18.0 was used to identify the percentage
of missing data for the risk, healthy marriage, parenting, and social competence items. All items
had missing data less than 4%. Specifically, missing data was less than 1.25% for the risk items
(i.e. depression and stress), less than 2.5% for the healthy marriage items, less than 1.2% for the
parenting items and less than 4% for the social competence items. Participants who were not
Head Start parents (i.e. partner of a Head Start parent) were asked to skip the parenting and
social competence items if they considered themselves a non-participant in the parenting of the
Head Start child. These parents were removed from analyses that use the parenting measures.
Data that was missing at random were imputed using Expectation Maximization (EM) algorithm,
a maximum likelihood approach, prior to final analyses (Dempster, Laired, & Rubin, 1977).
Normality
One of the key assumptions of CFA is that the data follow a multivariate normal
distribution. Therefore, item distributions were assessed for normality. Statistical tests of
skewness, kurtosis, and mulivariate normality were employed in combination with visual
screening of the variable item histograms. Variable items were considered to be within a normal
range of distribution (per Tabachnick & Fidell, 1996). The data were also screened for outliers
by using graphical methods applied to histograms as recommended by Tabachnick and Fidell
(1996). Because no outliers were identified, all cases were analyzed.

68

Differences in Responses by Data Collection Method
In order to check for possible differences in data that were collected from the sites and
that which was collected at the parenting events, independent t-tests were calculated for the risk
(stress and depression), healthy marriage (11 variables), social competence, and parenting
behavior scores. There were no statistically significant differences between site and event data
on any of the variables. Because ―relationship stability‖ is a healthy marriage dichotomous
variable, it was tested using a Pearson chi-square statistic. Stability did not differ by data
collection method, X2(4, N = 343) = 7.85, p = .10.
Assessing the Degree of Non-Independence
Correlations were calculated to find the degree of nonindependence between partner
scores on the HM/HR measure and to determine the confidence interval (see Table 6).
Correlations of .50 and higher indicate that the variables measure one concept and can be taken
together (Briggs & Cheek, 1986). First, to assess partner correlations for each of the healthy
marriage variables, the data set was structured in a pairwise fashion so that each individual‘s
scores were associated with his or her partner‘s scores. Although all of the variables between
partner scores were statistically correlated at p = .00, only seven of the eleven HM/HR variables
had correlations of .50 and higher. These nonindependent variables were: emotional intimacy,
communication, quality time, teamwork, conflict frequency, stability and lack of domestic
violence (DV). Those variables with correlations less than .50 were: couple commitment, child
commitment, forgiveness, and relationship happiness. Although analyzing the couple is favored,
the HM/HR variable measures were not taken together because not all eleven variables were
found to be nonindependent. Also, the factor structure needed to be analyzed prior to analyzing
the data dyadically.
69

Table 6
Correlations Between Partner Scores on HM/HR Measure

Emotional Intimacy
Couple Commitment
Child Commitment
Communication
Quality Time
Forgiveness
Teamwork
Conflict Frequency
Relationship Happiness
Stability
Lack of DV

R
.494
.273
.313
.557
.586
.496
.546
.483
.437
.620
.569

Scale Analysis
Scales were calculated from items to measure the components outlined in the conceptual
model. Cronbach‘s alphas were computed, along with inter-item and item-total correlations to
evaluate the internal consistency of the measurement scales.
Reliability Analyses for Exogenous Variables
The risk and healthy marriage items were all explored for inclusion in this study.
Risk subscales. To assess the reliability of each of the subscales Cronbach‘s alpha was
computed for the religiosity, perceived stress, depression, alcohol dependence, drug dependence,
and stressful life events subscales. All subscales demonstrated acceptable reliability (the cutoff
value being .70) with coefficient alphas ranging from .70 (alcohol dependency) to .89
(religiosity) except for stressful life events ( = .37). Therefore, the stressful life events subscale
was removed from further analyses.
70

Healthy marriage subscales. To assess the reliability of each of the subscales
Cronbach‘s alpha was computed for the couple commitment, commitment to children,
communication, conflict frequency, conflict resolution process, emotional intimacy, forgiveness,
lack of domestic violence, quality time, relationship happiness, relationship quality, and
teamwork. All subscales had acceptable reliability (over the cutoff value of .70) except for the
three item emotional intimacy measure ( = .67), conflict resolution process ( = .53) and
relationship quality ( = .51). A decision was made to retain the three-item emotional intimacy
measure because no other measure could be used as a proxy for this valuable variable and
because deleting any of the three items would not improve reliability. However, conflict
resolution process and relationships quality were dropped from further analysis. Because the
conflict frequency subscale had satisfactory reliability ( = .79) this subscale was used as a
measure of the conflict construct instead of the conflict resolution process subscale. Because the
satisfaction construct was being tested by MDRC with the relationship quality and relationship
happiness subscale measures, the one –item relationship happiness measure was used instead of
the relationship quality subscale. Finally, although the three-item forgiveness subscale had a
reliability coefficient alpha score of .77, the third reverse scored item was deleted to increase
reliability ( = 84).
Reliability Analyses for Endogenous Variables
Parenting behavior scale. The coefficient for the parenting scale was .79 for the overall
sample indicating internal consistency reliability. However, there was concern regarding the
supportive/engaged item, ―I hold or touch my child in an affectionate way.‖ This concern came
from questions that were asked by respondents during their completion of the questionnaire on
how to interpret the term ―affectionate‖. In addition, there was inconsistency with how
71

respondents answered this item with the item ―My child and I hug and /or kiss each other.‖ A
decision was made to delete the item of concern, thus slightly increasing alpha for this scale to
.80 for the overall sample, .72 for males and .83 for females.
Social competence scale. Cronbach‘s coefficient alpha was computed to examine the
internal consistency of the social competence scale. The scale demonstrated high internal
consistency (.92 for the overall sample, .93 for males and .91 for females).
Ranges, Means, and Standard Deviations
Scale means, standard deviations, ranges, frequencies and percentages were examined for
the risk variables, HM/HR constructs, parenting and child well-being measures and are presented
in Table 7 and Table 8. In addition, Table 9 and Table 10 summarize means and standard
deviations for males and females. Table 11 and Table 12 summarize these same statistics for the
female subgroup by race and marital status and finally Table 13 and Table 14 display results for
the male subgroup by race and marital status. The n‘s noted in the tables do not include the full
sample, rather, they reflect individuals who were parenting the child and therefore were asked to
complete the parenting measures.

72

Table 7
Ranges, Means, and Standard Deviations for the Total Sample (n = 334)

TOTAL Social Competence
SC- prosocial/communication
SC- emotion regulation
TOTAL Parenting Behavior
Parenting- hostile/coercive (reversed)
Parenting- supportive/engaged
Childhood happiness (RISK)
General health (RISK)
Religiosity (RISK)
Depression (RISK)
Perceived stress (RISK)
Alcohol Dependency (RISK)
Drug Dependency (RISK)
Social Support (RISK)
Intimacy (HM/HR)
Couple commitment (HM/HR)
Child commitment (HM/HR)
Communication (HM/HR)
Quality time (HM/HR)
Forgiveness (HM/HR)
Teamwork (HM/HR)
Conflict frequency (HM/HR)
Relationship happiness (HM/HR)
Fidelity (HM/HR)
Lack of DV (HM/HR)

Range
3.50
3.83
3.83
3.28
4.30
5
6
4
3
4
3
1
1
1
3
2.50
3
3
3
5
4
3
6
3
3

Minimum Maximum
0.50
4
0.17
4
0.17
4
1.67
4.95
0.70
5
0
5
0
6
0
4
0
3
0
4
0
3
0
1
0
1
0
1
1
4
1.50
4
1
4
1
4
1
4
0
5
1
5
0
3
1
7
1
4
0
3

M
2.23
2.43
2.04
3.95
3.46
4.44
2.13
1.58
1.76
0.94
0.90
0.07
0.04
0.02
3.08
3.42
3.48
2.94
3
3.50
3.88
1.88
5.66
3.12
1.94

Table 8
Frequencies and Percentages of Dichotomous Risk Variables for the Total Sample

Health condition (RISK)
Child stress on relationship (RISK)
Relationship stability (HM/HR)
n = 334

f
68
59
159
73

%
20.4%
17.7%
47.6%

SD
0.77
0.82
0.83
0.55
0.71
0.72
1.88
0.98
1.08
0.84
0.76
0.20
0.17
0.12
0.64
0.62
0.58
0.82
0.68
1.32
0.92
0.62
1.50
0.94
0.80

Table 9
Means and Standard Deviations by Parent Gender

TOTAL Social Competence
SC- prosocial/communication
SC- emotion regulation
TOTAL Parenting Behavior
Parenting- hostile/coercive (reversed)
Parenting- supportive/engaged
Childhood happiness (RISK)
General health (RISK)
Religiosity (RISK)
Depression (RISK)
Perceived stress (RISK)
Alcohol Dependency (RISK)
Drug Dependency (RISK)
Social Support (RISK)
Intimacy (HM/HR)
Couple commitment (HM/HR)
Child commitment (HM/HR)
Communication (HM/HR)
Quality time (HM/HR)
Forgiveness (HM/HR)
Teamwork (HM/HR)
Conflict frequency (HM/HR)
Relationship happiness (HM/HR)
Fidelity (HM/HR)
Lack of DV (HM/HR)

Males
(n = 143)
Means
SD
2.24
0.75
2.43
0.80
2.04
0.81
3.85
0.49
3.39
0.71
4.32
0.67
2.03
1.70
1.59
1.00
1.81
1.12
0.81
0.79
0.82
0.73
.10
.25
.07
.21
0.01
0.09
3.10
0.56
3.49
0.54
3.58
0.50
2.96
0.71
2.99
0.63
3.42
1.24
3.89
0.86
1.90
0.60
5.98
1.17
3.14
0.92
1.96
0.73

74

Females
(n = 188)
Means
SD
2.25
0.77
2.46
0.82
2.04
0.83
4.05
0.56
3.52
0.70
4.57
0.65
2.20
2.01
1.57
0.98
1.71
1.04
1.03
0.87
0.20
0.40
.04
.02
0.02
3.08
3.38
3.42
2.93
3.01
3.56
3.87
1.87
5.43
3.11
1.90

.14
.13
0.12
0.68
0.66
0.62
0.89
0.71
1.38
0.96
0.63
1.65
0.95
0.84

Table 10
Frequencies and Percentages of Dichotomous Risk Variables by Gender
Males
(n = 143)
Health condition (RISK)
Child stress on relationship (RISK)
Relationship stability (HM/HR)

F
33
22
75

%
23.1%
15.4%
52.4%

75

Females
(n =188)
f
34
37
83

%
18.1%
19.7%
44.1%

Table 11
Means and Standard Deviations by Race and Marital Status Within Gender: Female Subgroup

TOTAL Social Competence
SC- prosocial/communication
SC- emotion regulation
TOTAL Parenting Behavior
Parenting- hostile/coercive*
Parenting- supportive/engaged
Childhood happiness (RISK)
General health (RISK)
Religiosity (RISK)
Depression (RISK)
Perceived stress (RISK)
Alcohol Dependency (RISK)
Drug Dependency (RISK)
Social Support (RISK)
Intimacy (HM/HR)
Couple commitment (HM/HR)
Child commitment (HM/HR)
Communication (HM/HR)
Quality time (HM/HR)
Forgiveness (HM/HR)
Teamwork (HM/HR)
Conflict frequency (HM/HR)
Relationship happiness (HM/HR)
Fidelity (HM/HR)
Lack of DV (HM/HR)
*reversed

African
American
(n = 95)
Means SD
2.27 0.79
2.45 0.84
2.09 0.87
4.07 0.54
3.51 0.69
4.63 0.63
1.94 1.87
1.42 0.95
1.47 1.04
0.93 0.83
0.15 0.36
.07
.02
0.03
3.10
3.32
3.39
2.93
3.00
3.38
3.84
1.91
5.34
2.93
1.91

.18
.12
0.13
0.73
0.72
0.64
0.91
0.75
1.50
0.99
0.70
1.70
1.04
0.89

76

Caucasian
(n = 70)
Means SD
2.17 0.77
2.42 0.86
1.91 0.78
4.04 0.57
3.49 0.78
4.60 0.55
2.69 2.14
1.90 0.97
2.02 0.98
1.19 0.88
0.30 0.46
.00
.03
0.03
3.01
3.38
3.41
2.93
2.96
3.70
3.84
1.81
5.30
3.30
1.82

.04
.17
0.12
0.65
0.61
0.61
0.91
0.65
1.18
0.95
0.53
1.75
0.81
0.81

Married
(n = 105 )
Means SD
2.40 0.76
2.57 0.82
2.23 0.83
4.08 0.56
3.55 0.75
4.60 0.57
2.13 2.02
1.50 1.03
1.54 1.15
0.90 0.78
0.22 0.42
.03
.01
0.01
3.10
3.45
3.53
2.96
3.04
3.73
3.96
1.88
5.62
3.13
2.00

.11
.10
0.08
0.66
0.62
0.58
0.87
0.66
1.26
0.93
0.58
1.62
0.91
0.80

Unmarried
(n = 80 )
Means SD
2.02 0.70
2.28 0.79
1.75 0.73
4.03 0.54
3.51 0.63
4.55 0.73
2.24 2.01
1.66 0.87
1.94 0.82
1.18 0.94
0.18 0.38
.06
.04
0.04
3.05
3.27
3.26
2.89
2.95
3.36
3.74
1.86
5.15
3.10
1.79

.18
.17
0.15
0.72
0.69
0.64
0.92
0.78
1.51
1.01
0.70
1.69
1.00
0.88

Table 12
Frequencies and Percentages of Dichotomous Risk Variables by Race and Marital Status Within
Gender: Female Subgroup

Health condition
Child stress on relationship
Relationship stability

African
American
(n =95)
f
%
18 18.9%
14 14.7%
41 43.2%

Caucasian
(n =70)
f
%
14
20%
21
30%
29 41.4%

77

Married
(n = 105)
f
%
13 12.4%
23 21.9%
51 48.6%

Unmarried
(n =80)
f
%
20 25.0%
14 17.5%
31 38.8%

Table 13
Means and Standard Deviations by Race and Marital Status Within Gender: Male Subgroup

TOTAL Social Competence
SC- prosocial/communication
SC- emotion regulation
TOTAL Parenting Behavior
Parenting- hostile/coercive*
Parenting- supportive/engaged
Childhood happiness (RISK)
General health (RISK)
Religiosity (RISK)
Depression (RISK)
Perceived stress (RISK)
Alcohol Dependency (RISK)
Drug Dependency (RISK)
Social Support (RISK)
Intimacy (HM/HR)
Couple commitment (HM/HR)
Child commitment (HM/HR)
Communication (HM/HR)
Quality time (HM/HR)
Forgiveness (HM/HR)
Teamwork (HM/HR)
Conflict frequency (HM/HR)
Relationship happiness (HM/HR)
Fidelity (HM/HR)
Lack of DV (HM/HR)
*reversed

African
American
(n = 68)
Means SD
2.26 0.73
2.45 0.78
2.07 0.79
3.86 0.53
3.36 0.75
4.35 0.74
1.86 1.60
1.44 0.97
1.57 1.19
0.65 0.69
0.09 0.29
0.14 0.27
0.07 0.20
0.01 0.12
3.07 0.57
3.46 0.61
3.58 0.51
2.99 0.70
3.03 0.62
3.32 1.31
3.83 0.98
1.92 0.63
5.94 1.16
3.12 0.94
2.05 0.73

78

Caucasian
(n = 56)
Means SD
2.19 0.76
2.40 0.83
1.98 0.79
3.90 0.43
3.45 0.68
4.34 0.55
2.19 1.81
1.73 0.98
2.06 1.02
0.85 0.68
0.25 0.44
0.06 0.21
0.07 0.23
0.01 0.07
3.16 0.53
3.57 0.42
3.64 0.45
3.00 0.70
2.99 0.59
3.56 1.09
3.94 0.74
1.87 0.55
6.12 1.02
3.14 0.92
1.85 0.73

Married
(n = 89)
Means SD
2.40 0.73
2.57 0.81
2.23 0.78
3.84 0.50
3.40 0.70
4.29 0.71
1.90 1.61
1.75 1.01
1.66 1.18
0.79 0.81
0.17 0.38
0.11 0.27
0.04 0.20
0.01 0.05
3.07 0.56
3.52 0.52
3.55 0.48
2.94 0.69
2.94 0.61
3.45 1.20
3.86 0.86
1.81 0.59
6.06 1.13
3.25 0.88
1.92 0.77

Unmarried
(n = 53)
Means SD
1.96 0.70
2.20 0.75
1.73 0.77
3.89 0.47
3.38 0.74
4.40 0.57
2.19 1.83
1.32 0.92
2.08 0.99
0.78 0.61
0.11 0.32
0.08 0.21
0.12 0.24
0.02 0.14
3.16 0.55
3.43 0.56
3.62 0.55
3.01 0.76
3.08 0.65
3.42 1.24
3.96 0.87
2.06 0.59
5.84 1.25
2.96 0.96
2.05 0.64

Table 14
Frequencies and Percentages of Dichotomous Risk Variables by Race and Marital Status Within
Gender: Male Subgroup

Health condition
Child stress on relationship
Relationship stability

African
American
(n = 68)
f
%
17
25%
6
8.8%
34
50%

Caucasian
(n = 56)
f
%
12 21.4%
14
25%
32 57.1%

Married
(n =89)
f
%
22 24.7%
15 16.9%
48 53.89%

Unmarried
(n =53)
f
%
10 18.9%
6
11.3%
27 50.9%

Mean Differences
To analyze mean differences between gender as well as within gender between race and
marital status, several independent sample t-test were calculated for the following main
variables: total preschoolers‘ social competence; social competence- prosocial/communication;
social competence- emotion regulation; total parenting behavior; parenting behavior- supportive
engaged; parenting behavior- hostile coercive; depression; stress; and the HM/HR composite.
Females‘ scores were significantly higher than males‘ scores on total parenting behavior t (329)
= -3.29, p = .001, parenting behavior- supportive engaged t (329) = -3.51, p =.001), and on
depression t (329) = -2.37, p < .05. Please see means and standard deviations in Table 9.
Females. Within the female dataset, no significant differences were found between
African American females and Caucasian females. Significant differences were, however, found
between married and unmarried females. Married females had significantly higher scores than
unmarried females on total preschoolers‘ social competence t (183) = 3.52, p = .001, social
competence- prosocial/communication t (183) = 2.44, p < .05, and social competence- emotion
regulation t (183) = 4.10, p < .001. Unmarried females scored significantly higher than married
79

females on depression t (183) = -2.18, p <. 05. Please see means and standard deviations in
Table 11.
Males. Within the male dataset, no significant differences were found between African
American males and Caucasian males. Married males scored their children‘s social competence
significantly higher than unmarried males. Specifically, significantly higher scores were
reported by married males on total preschoolers‘ social competence t (140) = 3.50, p = .001,
social competence- prosocial/communication t (140) = 2.76, p < .01, and social competenceemotion regulation t (140) = 3.70, p < .001. See means and standard deviations in Table 13.
Control Variables
Preschoolers‘ social competence scores as reported by their parents were examined for
differences by gender and age. An independent t-test was used to test for gender differences for
social competence. There were no differences on the total social competence score (t = -.042, p
= .966) between boys (M = 2.25, SD = .81) and girls (M = 2.25, SD = .73). In addition, a oneway ANOVA was used to test for differences among the four age groups (3, 4, 5, and 6 year
olds). Preschoolers‘ social competence differed significantly across the ages, F (3, 303) = 4.02,
p = <. 01. Tukey post-hoc comparisons of the four groups indicate that the three year-old group
(M = 1.57, 95% CI [1.30, 1.83]) had significantly lower total social competence scores than the
five year old group (M = 2.32, 95% CI [2.20, 2.44]), p < .05 as well as the six year old group (M
= 2.40, 95% CI [2.04, 2.76]), p < .05. Comparisons between all the other age groups were not
statistically significant at p < .05. Because child age was missing from 36 respondents, it was
not used as a control variable to preserve sample size.

80

Exploratory Factor Analyses
Exploratory Factor Analysis of the Risk Measures
The SAQ measure provided by MDRC included several risk measures. Decisions about
which risk variables to include in subsequent model testing needed to be explored as well as
whether the variables measured a latent risk variable. These risk variables were religiosity,
childhood happiness, general health, health condition, relationship stress from a child condition,
depression, stress, alcohol dependency, and drug dependency. The first step was to enter all of
the items for each variable into a factor analysis using SPSS 18.0. Principal axis factoring with
direct oblimin rotation was performed to determine how the selected pool of items clustered
together. The factor analysis demonstrated that items loaded onto their associated variables.
This produced six factors with the first factor contributing 25.71% of the variance with an
eigenvalue of 5.14. This first factor included all six of the depression items as well as the two
perceived stress items. The second extracted factor included the three drug dependency items,
contributing 11.33% of the variance with an eigenvalue of 2.27. The third extracted factor
included the three alcohol dependency items, contributing 9.57% of the variance with an
eigenvalue of 1.91. The fourth, fifth, and sixth extracted factors were religiosity (8.70%
variance, eigenvalue = 1.74), general health and health condition (5.72%, eigenvalue = 1.14), as
well as childhood happiness and relationship stress from a child condition (5.53% variance,
eigenvalue = 1.11). An examination of the scree plot suggested one dominant factor.
In addition, each of the nine risk variables were examined using factor analysis. Several
well-recognized criteria for factorability were used. First, only 4 of the nine risk variables
correlated at least .3 with at least one other variable, suggesting possible issues with factorability.
Specifically, general health correlated with health condition (r = .337, p < .001) and depression
81

correlated with perceived stress (r = .646, p < .001). Second, the Kaiser-Meyer-Olkin measure
of sampling adequacy was .64, only slightly above the recommended value of .6, although
Bartlett‘s test of sphericity was significant (2 (36) = 332.20, p = .000). Principal axis factoring
with direct oblimin rotation was performed to determine how the selected pool of variables
clustered together. Four factors were extracted. The initial eigen values showed that the first
―depression ad stress‖ factor explained 24.33% of the variance while the second (childhood
happiness, having a child with a condition that puts stress on the marriage/relationship,
religiosity), third (alcohol and drug dependency) and fourth factors (general health and health
condition) explained 13.66%, 12.21%, and 11.21% of the variance. The second factor did not
make theoretical sense and was dismissed as a factor. Because of the smaller sample size and
after an examination of the scree plot, a decision was made to choose the first risk factor
(depression and stress) for subsequent CFA and SEM analysis.
Results for Study One
The research questions for study one were: (1) What is the factor structure for the
Healthy Marriage/Healthy Relationships (HM/HR) measure? and (2) How do respondent
antecedents (risk factors) relate with the HM/HR measure? This section of the results describes
the exploratory factor analysis and confirmatory factor analysis of the Healthy Marriage
Measure.
Exploratory Factor Analysis of the HM/HR measure
An initial step in exploring the HM/HR measure was to examine inter-item correlations
from responses to all of the Healthy Marriage measures subscales. In addition, correlations were
generated between the HM/HR measure subscales, risk factor scales (depression and stress),
parenting scales and social competence scales (Table 15). Because the HM/HR measure
82

revealed moderate to strong intercorrelations among a majority of items, it was suspected that the
variance would also be shared rather than unique. The one-item fidelity measure was removed
from further analyses given its lower correlation with the other HM/HR subscale items as well as
lower non-significant correlations with the parenting and social competence subscales.
Table 15
Correlations between Social Competence, Parenting, Healthy Marriage, and Risk for Total
Sample (n = 334)
1
1
.754**
.936**
.221**
.095
.207**
.044
.111*
.060
.096
.027
.128*
.146**
.044
.121*
-.005
.087
.033

2

3

4

5

6

7

8

1
.937**
.209**
.182**
.255**
.083
.148**
.134*
.072
.104
.151**
.188**
.119*
.124*
.030
.130*
.067

1
.230**
.148**
.247**
.068
.138*
.104
.090
.070
.149**
.178**
.087
.131*
.013
.116*
.054

1
.176**
.768**
.183**
.185**
.206**
.207**
.177**
.163**
.183**
.157**
.171**
.042
.059
.086

1
.765**
.247**
.175**
.135*
.299**
.266**
.224**
.253**
.239**
.144**
.103
.133*
.235**

1
.280**
.235**
.223**
.330**
.289**
.252**
.285**
.258**
.206**
.094
.126*
.209**

1
.612**
.600**
.763**
.712**
.576**
.691**
.550**
.594**
.135*
.427**
.528**

1
.708**
.522**
.536**
.425**
.566**
.385**
.553**
.136*
.366**
.346**

19 TOTAL HM/HR

.109*

.162** .144** .217** .287** .329** .853** .725**

20 Depression

-.054

-.130* -.098

-.119* -.235** -.231** -.296** -.251**

21 Stress

.001

-.118* -.063

-.067

-.164** -.151** -.280** -.224**

22 TOTAL RISK

-.029

-.136* -.089

-.103

-.220** -.210** -.317** -.262**

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

Social Comp-PC
Social Comp-ER
TOTAL Social Comp
Parenting-SE
Parenting-HC
TOTAL Parenting
Emotional Intimacy
Couple Commitment
Child Commitment
Communication
Quality Time
Forgiveness
Teamwork
Conflict Frequency
Rel. Happiness
Fidelity
Stability
Lack of DV

83

Table 15 (cont‘d).
9
9
10
11
12
13
14
15
16
17
18

Child Commitment
Communication
Quality Time
Forgiveness
Teamwork
Conflict Frequency
Rel. Happiness
Fidelity
Stability
Lack of DV

10

11

12

13

14

15

1
.502**
.491**
.387**
.518**
.379**
.511**
.123*
.320**
.372**

1
.626**
.563**
.630**
.521**
.571**
.173**
.452**
.555**

1
.601**
.683**
.539**
.580**
.167**
.435**
.532**

1
.618**
.386**
.424**
.149**
.403**
.550**

1
.482**
.543**
.117*
.464**
.499**

1
.414**
.145**
.441**
.597**

1
.076
.480**
.433**

19 TOTAL HM/HR

.699** .815** .816** .707** .808** .699** .742**

20 Depression

-.260** -.353** -.335** -.276** -.327** -.373** -.323**

21 Stress

-.184** -.311** -.336** -.232** -.261** -.335** -.348**

22 TOTAL RISK

-.245** -.366** -.370** -.280** -.324** -.390** -.370**

Table 15 (cont‘d).

16 Fidelity
17 Stability
18 Lack of DV

16
17
18
1
.179** 1
.129* .465** 1

19

20

21

22

19 TOTAL HM/HR

.184** .644** .719** 1

20 Depression

-.041

-.252** -.389** -.421** 1

21 Stress

-.083

-.294** -.371** -.389** .649** 1

22 TOTAL RISK

-.068

-.301** -.418** -.446** .908** .908** 1

*Correlation is significant at the 0.05 level (2-tailed).
** Correlation is significant at the 0.01 level (2-tailed).
Total HM/HR and Total Risk are composite scores

With the Kaiser–Meyer– Olkin (KMO) measure of sampling adequacy at 0.926 and
Bartlett's test of Sphericity being significant (χ2 = 2178.738, p <.001), the appropriateness of the
ensuing factor model was ensured. Using the principal axis factoring method of extraction and
84

direct oblimin, the 11 HM/HR subscales were subjected to exploratory factor analyses (EFA)
using PASW 18 for Windows.
The EFA of the HM/HR subscales revealed a dominant first factor. This factor explained
56.5% of the variance in responses to the 11 subscale items in the HM/HR measure. This first
factor had an eigen value of 6.2. The other factors had eigen values less than 1. Only one factor
was extracted and all variables loaded well on this single factor (see Table 16). In addition Table
17 displays the factor matrix for males by race and marital status and Table 18 displays the factor
matrix for females by race and marital status.
Table 16
Factor Matrix of the Healthy Marriage Measure
Factor 1
Emotional Intimacy
Couple Commitment
Child Commitment*
Communication
Quality Time
Forgiveness**
Teamwork
Conflict Frequency
Relationship Happiness
Stability
Lack of DV

Total
.862
.695
.662
.804
.809
.685
.801
.643
.706
.576
.668

*Child commitment is a measure of one‘s appraisal of their partner‟s commitment to the child
**Forgiveness is a measure of one‘s appraisal of their partner‟s acts of forgiveness

85

Table 17
Factor Matrix of the Healthy Marriage Measure for Males by Race and Marital Status
Factor 1
Emotional Intimacy
Couple Commitment
Child Commitment*
Communication
Quality Time
Forgiveness**
Teamwork
Conflict Frequency
Relationship Happiness
Stability
Lack of DV

Males
.814
.616
.507
.783
.767
.724
.766
.535
.681
.630
.648

African
American
.741
.629
.474
.699
.725
.789
.771
.615
.596
.583
.676

Caucasian

Married

Unmarried

.876
.646
.588
.818
.800
.681
.788
.535
.783
.678
.669

.790
.637
.599
.819
.778
.713
.747
.596
.763
.614
.688

.846
.645
.399
.724
.743
.761
.797
.426
.594
.666
.562

*Child commitment is a measure of one‘s appraisal of their partner‟s commitment to the child
**Forgiveness is a measure of one‘s appraisal of their partner‟s acts of forgiveness

Table 18
Factor Matrix of the Healthy Marriage Measure for Females by Race and Marital Status
Factor 1
Emotional Intimacy
Couple Commitment
Child Commitment*
Communication
Quality Time
Forgiveness**
Teamwork
Conflict Frequency
Relationship Happiness
Stability
Lack of DV

Females
.886
.721
.729
.811
.838
.682
.819
.705
.723
.543
.714

African
American
.918
.715
.741
.814
.852
.680
.808
.788
.744
.550
.725

Caucasian

Married

Unmarried

.815
.714
.688
.821
.839
.741
.834
.683
.709
.586
.699

.866
.723
.751
.775
.851
.676
.804
.653
.725
.634
.706

.920
.707
.699
.856
.841
.695
.834
.769
.713
.423
.727

*Child commitment is a measure of one‘s appraisal of their partner‟s commitment to the child
**Forgiveness is a measure of one‘s appraisal of their partner‟s acts of forgiveness

86

Because relationship stability was measured as dichotomous variable, while the other ten
were measured as scale variables, the factor structure was further examined to ensure its‘
appropriate inclusion. After removing this dichotomous variable and performing another EFA
with the ten scale variables, results once again supported a unidimensional measure with a
dominant first factor. This factor explained 58.7% of the variance in responses to the ten
subscale items in the HM/HR measure. This factor had an eigen value of 5.9. Whereas utilizing
EFA with a great number of dichotomous variables could be deemed inappropriate (Shapiro,
Lasarev, & McCauleym, 2002), results for this study showed that inclusion of a single
dichotomous variable amongst many other scale variables could still contribute to the factor. It
was important to include a measure of stability to help define the healthy marriage/healthy
relationships construct because of its value to couple relations and influence on children (as
explained in the literature review).
In order to find how respondent antecedents (risk factors) were related with the HM/HR
measure, a multiple regression was performed whereby scores on depression and stress were
used to predict the HM/HR composite score. A composite score was used because it was found
that the eleven HM/HR variables could be combined into a single measure. Specifically, a
composite score for the healthy marriage measure was formed with unit-weighted z scores of
constituent tests (Ackerman & Cianciolo, 2000) because each variable was scored using different
scales. This composite was determined reliable ( = .92 for the overall sample,  = .90 for
males and  = .93 for females). Depression ( = -.29, p < .001) and stress ( = -.20, p =.002)
were both significantly related to the HM/HR composite, F(2, 331) = 41.50, p < .001, R = .45.
Multiple regressions were also performed separately for males and females. For males,
depression ( = -.10, p = .322) was not significantly related to the HM/HR composite but stress
87

( = -.24, p =.022) was significantly related, F(2, 140) = 7.791, p = .001, R = .32. For females,
depression ( = -.41, p < .001) was significantly related to the HM/HR composite where stress (
= -.16, p =.06) approached significance, F(2, 185) = 34.66, p < .001, R = .52.
Results for Study Two
The research question for study 2 was: does the HM/HR factor structure vary across
groups (a) gender, (b) race (African American and Caucasian), and (c) marital status (married
and unmarried)? This section describes the results from multiple group confirmatory factor
analysis (CFA) as well as nested model comparisons. Multiple group CFA could not be
reasonably used to test for variation by race or marital status given that the dataset was split by
gender, and subdividing this further produced small sample sizes.
Multiple Group Confirmatory Factor Analysis (CFA) of the HM/HR Measure
Next, AMOS 18.0 was used to perform a multiple group CFA on the Healthy
Marriage/Healthy Relationships (HM/HR) measure simultaneously for males (n = 143) and
females (n = 188). Multiple group CFAs were not used to compare fit by race or marital status
due to small sample sizes within these subgroups. Maximum likelihood estimation was chosen
because the data were normally distributed. The rationale for splitting the dataset by gender was
to reduce an increase in Type 1 error in the interpretation of these individual level variables since
both males and females appraised their couple relationship individually. In addition, a multiple
group analysis was appropriate to test the equal factor structure of the healthy marriage construct
(or measurement invariance) for the effect indicator model, further implying the equivalent
construct validity across gender. Each item was constrained to load on to the healthy marriage
latent variable identified by the EFA. No cross-loadings were specified. One parameter estimate
of the latent variable was constrained to one to identify the model (Kline, 1998). First, an
88

unconstrained baseline model was tested. Then, models were fit with constraints across gender
specified for measurement weights (factor loadings) and next by measurement residuals (fixing
the covariances and variances of the errors to be the same). This process of model fitting yielded
a nested hierarchy of models and a chi-square difference test was used to test whether the
equality constraints were upheld.
In order to determine if the hypothesized model should be accepted or rejected, the
following goodness of fit tests were used: the chi-square statistic (Jöreskog & Sörbom, 1984), the
Comparative Fit Index, (CFI; Bentler, 1990), the Tucker-Lewis Index (TLI; Bentler & Bonett,
1980), the Root-Mean-Square Error of Approximation (RMSEA; Browne & Cudeck, 1993) and
the Standardized Root-Mean-Square Residual (SRMR; Bentler, 1995). Generally, a lower chisquare statistic indicates a better fitting model. According to Hu and Bentler (1998), the chisquare statistic is particularly sensitive to sample size and thus should be examined with the
additional fit indices listed above. This is especially important given the smaller sample size in
this dissertation research. The CFI compares the fit for the model of interest with the null model
(i.e. a model that assumes the latent variables are uncorrelated). The TLI measures the
proportion by which the hypothesized model improves the fit of the null model. Scores for the
CFI are ―normed‖ so that their values cannot be below 0 or above 1 and scores for the TLI are
considered ―nonnormed‖ because, on occasion, they may be larger than 1 or slightly below 0.
The RMSEA is an index of lack of fit. Adequate fit was designated as CFI and TLI values of
>.90 (Bentler, 1992) and RMSEA values of <.10, whereas excellent fit was designated as CFI
and TLI values close to .95, and RMSEA values close to .06 (Hu & Bentler, 1999). The SRMR
is a measure of the standardized difference between the observed and predicted covariance.

89

Values of .06 or lower suggest a good fitting model (Hu & Bentler, 1998). Modifications to the
model (i.e. correlating error terms) were only made for substantive reasons (Jo¨reskog, 1993).
Unconstrained model. This initial unconstrained Healthy Marriage model showed
adequate fit of the data χ2(88) = 286.326, p = .000, CFI = .91, TLI = .88, RMSEA = .08, SRMR
= .07. Modification indices suggested eight ways by which fit could be improved with
covariances, and one of these changes was made. Treating the covariance between child
commitment and couple commitment as a free parameter, would reduce the discrepancy by at
least 45.42. Allowing these two error terms to covary improved fit, χ2(86) = 199.328, p = .000,
CFI = .95, TLI = .93, RMSEA = .06, SRMR = .06. The rationale for correlating the two
commitment items is that couple commitment increases when partners share children (Stanley &
Markman, 1992) and healthy marriage is often defined as involving the unconditional
commitment to both one‘s partner and children (Fagan, 2002; Waldo, 2002). Figure 2 shows the
final CFA with standardized parameter estimates for males and Figure 3 shows the final CFA
with standardized parameter estimates for females. Using standardized coefficients allows
different variables to be directly compared when they have different scales (Klein, 2005).
Standardized and unstandardized coefficients for males and females are shown in Table 19. All
factor loadings were significant for males and females. The squared multiple correlation (SMC)
values indicating the reliability of the Healthy Marriage measure are also provided in Figures 2
and 3. For example, in the male model, emotional intimacy (.67) and commitment to the child
(.22) have the highest and lowest SMC respectively. In the female model, emotional intimacy
(.80) and relationship stability (.28) have the highest and lowest, respectively.

90

Figure 2. Unconstrained CFA healthy marriage model for males.

For interpretation of the references to color in this and all other figures, the reader is referred
to the electronic version of this dissertation.

91

Figure 3. Unconstrained CFA healthy marriage model for females

92

Table 19
Standardized and Unstandardized Coefficients for Males and Females on HM/HR Model
Males
Variable
B
SE
β
Emotional Intimacy
.819
.458
.039
Relationship Stability
.625
.309
.039
Lack of DV
.649
.463
.056
Communication
.794
.566
.051
Conflict Frequency
.532
.314
.048
Quality Time
.775
.480
.045
Forgiveness
.734
.900
.092
Teamwork
.767
.652
.062
Relationship Happiness
.677
.793
.089
Couple Commitment
.577
.323
.042
Child Commitment
.468
.252
.041
*All factor loadings were significant for males and females.

β
.893
.527
.701
.814
.703
.844
.685
.821
.717
.723
.728

Females
B
.608
.260
.590
.720
.444
.601
.940
.786
1.183
.473
.448

SE
.039
.034
.055
.054
.041
.042
.090
.058
.106
.042
.039

Nested model comparisons. Table 20 presents summary fit indices of the three nested
models for the multiple group (males and females) confirmatory factor analysis. The
unconstrained model was the baseline model, which relaxed all equality constraints. As
mentioned above, this model tested the factorial structure of the instrument across gender
simultaneously with no cross-group constraints imposed. The fit indices for the baseline model
indicated that the model of the healthy marriage measure exhibited acceptable fit across gender.
The measurement weights model tested the invariance of factor loadings across gender by
placing equality constraints on these parameters. The fit indices for this measurement weight
model were χ2(97) = 224.698, p = .000, CFI = .94, TLI = .93, RMSEA = .06, SRMR = .10. A
cut off of .10 for the SRMR statistic was used. When the measurement weight model was
compared to the baseline model, a significant difference was noted, 2 (11) = 25.533, p = .008.
This suggests that the factor loadings vary across gender groups. A detailed exploration of
which loadings were different across gender is provided below (see partial measurement
93

invariance below). In the measurement residual model, all parameters were specified equally
across gender. The chi-square difference test again yielded a statistically significant value of
49.16 with 22 degrees of freedom at the .001 level. These findings suggested that the equality
constraints of factor loadings, covariances, and error covariances were not upheld across gender.
That is, the assumption of an equivalent factor structure was not supported across gender.
Table 20
Fit Indices of Three Nested Models of Multi-group Confirmatory Factor Analysis (n=151)
Model

χ2 (df)

χ2/df

CFI

TLI

RMSEA SRMR

199.328(86)* 2.318 .947 .932
.083
.055
Unconstrained Model
Measurement Weight
.100
224.698(97)* 2.316 .940 .932
.081
Measurement Residual
.093
247.879(108)* 2.295 .934 .933
.079
Model
* p < .001
Models
Partial measurement invariance. The results of the measurement weights model
analysis above indicated that some equality constraints of factor loadings did not hold across
gender. In an effort to identify factor loadings of which items were equivalent and which were
nonequivalent across gender, chi-square difference tests were conducted on an indicator-byindicator basis in the context of partial measurement invariance where equality constraints were
imposed on some but not all of the factor loadings (Byrne, Shavelson, & Muthen, 1989). In the
context of the partial measurement model, a model was fit first by placing equality constraints on
all the factor loadings (see measurement weights model in Table 27); then, a less restrictive
model was fit by relaxing the equality constraint of the regression weights (factor loadings) of
the item of interest. A chi-square difference test was conducted between the less restrictive
model and the measurement weights model to investigate whether the factor loading of that item
was invariant across gender. A non-significant chi-square difference test indicated that the factor
loading of that item was not statistically different across gender. This process was repeated item
94

by item until all of the nonequivalent items were identified. Results of chi-square difference
tests indicated that the factor loadings of two variables were nonequivalent: relationship stability
and child commitment and the factor loadings of other variables were invariant across gender.
Table 21 displays the results of chi-square difference tests for equivalent factor loadings and
nonequivalent factor loadings of variables across gender.
Table 21
Equivalent and Nonequivalent Factor Loadings of Variables across Gender
Variable
Emotional Intimacy
Relationship Stability
Lack of DV
Communication
Conflict Frequency
Quality Time
Forgiveness
Teamwork
Relationship Happiness
Couple Commitment
Child Commitment
*p < .05

2

 (df =1)

p

1.232
5.416
.123
.196
.924
.018
1.991
.073
2.853
.001
4.420

.267
.020*
.726
.658
.336
.892
.158
.787
.091
.981
.036*

The results suggested that although factor loadings were significant for both males and
females, the healthy marriage measure was a partially invariant measurement across gender.
Specifically, the factor loading for relationship stability was larger for males and the factor
loading for child commitment was larger for females. However, the great majority of variables
were invariant.

95

Results for Study Three
Path Analyses
The final purpose of this research was to test a model of healthy marriage developed by
Child Trends, Inc. that predicts adult and child well-being. The last part of the results section
focuses on the research question: How are responses on the HM/HR measure related to
parenting and child outcomes? In order to test the relationship between responses on the
HM/HR measure to parenting and child outcomes across gender, race and marital status, several
path analyses were modeled in AMOS version 18.0. Path analyses were chosen so that the
model could be tested for variations in race and marital status within gender (both with small
sample sizes). This is a confirmatory technique for testing theoretical causal relationships
between variables and was chosen for the analysis because 1) it is one of the only approaches
that allows for concurrent tests of all relationships in a complex model including multiple
dependent variables, 2) it allows for the incorporation of mediating variables, 3) and it is more
rigorous for parsimonious model testing than multiple regression analysis (Tabachnick & Fidell,
2001). Maximum likelihood estimation was used in all analyses. In all figures showing path
models, standardized path coefficients (βs) are displayed because these coefficients serve as
effect sizes for each path within the context of the model.
First, observed variables for risk and healthy marriage were developed by calculating
composite scores. Although results from study two revealed that the healthy marriage measure
was a partially invariant measurement across gender, a composite score was still used because
the focus was on examining the structural model. Specifically, to provide more stable measures
of the underlying abilities, composites of risk and healthy marriage were formed with unitweighted z scores of constituent tests (Ackerman & Cianciolo, 2000). Both the risk composite
96

( = .79 for the overall sample,  = .78 for males and  = .79 for females) and the healthy
marriage composite ( = .92 for the overall sample,  = .90 for males and  = .93 for females)
scales were determined reliable.
Parent gender differences in the Child Trends, Inc. model. In order to test for
equivalence of the Child Trends, Inc. measurement model across parent gender, a path analysis
was conducted separately for both genders (see Figure 4), and these models were compared to
investigate possible gender differences. This linear model was tested because it is the conceptual
model as outlined by Child Trends, Inc. Specifically, the analysis contrasted the fit of the
models with the weights unconstrained by gender to models where the beta weights were
constrained to be equal for both genders. Chi-square tests were then used to compare the
models. Results showed that the overall unconstrained fit of the model was good, χ2 = 9.912, df
= 6, p = .128, CFI = .97, TLI = .94, RMSEA = .05, and SRMR = .06. This unconstrained model
differed significantly from the model with constrained beta weights χ2 = 21.191, df = 9, p = .012,
CFI = .91, TLI = .88, RMSEA = .06, SRMR = .08, Δχ2 = 11.279, Δdf = 3, p = .010 indicating
that path coefficients significantly varied by gender. Because the unconstrained model provided
a significantly better fit, the relationships between the variables differed between males and
females. All relationships were in the hypothesized directions for males and females but the
effect sizes for the path between risk and healthy marriage was significantly larger for females
than for males (see Figure 4). In addition, the path estimates from risk to healthy marriage were
especially larger for unmarried females and for African American females. Standardized and
unstandardized coefficients for males and females are shown in Table 22.

97

Figure 4. Unconstrained standardized estimates for males and females.
Females

Males

98

Racial differences in the Child Trends, Inc. model. Next, equivalence of the Child
Trends, Inc. measurement model across race was examined. Two multiple group path analyses
were performed. The first path analysis compared the model for African American males (n =
68) and Caucasian males (n = 56) while the second path analysis compared the model for African
American females (n = 95), and Caucasian females (n = 70).
Male by race. Results for the male group showed that the overall unconstrained fit of the
model was adequate, χ2 = 3.319, df = 6, p = .768, CFI = 1.00, TLI = 1.18, RMSEA = .00 and
SRMR = .08. This unconstrained model did not differ significantly from the model with
constrained beta weights χ2 = 4.234, df = 9, p = .895, CFI = 1.00, TLI = 1.18, RMSEA = .00,
SRMR = .09, Δχ2 = .915, Δdf = 3, p = .822 indicating that path coefficients did not significantly
vary for males by race.
Female by race. Results for the female group also showed that the overall unconstrained
fit of the model was adequate, χ2 = 3.691, df = 6, p = .718, CFI = 1.00, TLI = 1.06, RMSEA =
.00 and SRMR = .07. This unconstrained model also did not differ significantly from the model
with constrained beta weights χ2 = 5.450, df = 9, p = .793, CFI = 1.00, TLI = 1.06, RMSEA =
.00, SRMR = .07, Δχ2 = 1.759, Δdf = 3, p = .624 indicating that path coefficients did not
significantly vary for females by race. Standardized and unstandardized coefficients for males
and females by race are shown in Table 22.
Marital status differences in the Child Trends, Inc. model. A final set of multiple
group path analyses was performed to examine the model by marital status. Two multiple group
path analyses were performed. The first path analysis compared the model for Married males (n
= 89) and Unmarried males (n = 53) while the second path analysis compared the model for
Married females (n = 105), and Unmarried females (n = 80).
99

Male by marital status. Results for the male group showed that the overall unconstrained
fit of the model was good, χ2 = 6.361, df = 6, p = .384, CFI = .99, TLI = .98, RMSEA = .02 and
SRMR = .08. This unconstrained model did not differ significantly from the model with
constrained beta weights χ2 = 8.954, df = 9, p = .442, CFI = 1.00, TLI = 1.00, RMSEA = .00,
SRMR = .09, Δχ2 = 2.593, Δdf = 3, p = .459 indicating that path coefficients did not significantly
vary for males by marital status.
Female by marital status. Results for the female group also showed that the overall
unconstrained fit of the model was good, χ2 = 8.645, df = 6, p = .195, CFI = .97, TLI = .95,
RMSEA = .05 and SRMR = .05. This unconstrained model also did not differ significantly from
the model with constrained beta weights χ2 = 9.442, df = 9, p = .397, CFI = .995, TLI = .99,
RMSEA = .02, SRMR = .04, Δχ2 = .798, Δdf = 3, p = .850 indicating that path coefficients did
not significantly vary for females by marital status. Standardized and unstandardized
coefficients for males and females by marital status are shown in Table 22.

100

Table 22
Unstandardized, Standardized, and Significance Levels for Child Tends Model for Males and
Females by Race and Marital Status (Standard Errors in Parentheses)
Structural Model Estimates
Males
Direct Effects
B
p
β
Full Sample
Risk Composite  HM Composite -1.201(.308) -.311 ***
HM Composite  Parenting
.025(.005) .365 ***
Parenting Social Competence
.220(.126) .145 .081
African American Sample
Risk Composite  HM Composite -1.495(.513) -.336 .004
HM Composite  Parenting
.026(.008) .357 .002
Parenting Social Competence
.355(.164) .256 .030
Caucasian Sample
Risk Composite  HM Composite -1.802(.527) -.419 ***
HM Composite  Parenting
.016(.008) .263 .043
Parenting Social Competence
.310(.236) .174 .188
Married Sample
Risk Composite  HM Composite -1.260(.374) -.338 ***
HM Composite  Parenting
.028(.007) .403 ***
Parenting Social Competence
.132(.155) .091 .394
Unmarried Sample
Risk Composite  HM Composite -1.029(.658) -.211 .118
HM Composite  Parenting
.018(.009) .264 .048
Parenting Social Competence
.465(.198) .309 .019
* significantly different from zero at the .001 level (two-tailed)

B

Females
β

p

-2.520(.309) -.512
.022(.004) .347
.440(.095) .320

***
***
***

-2.888(.435) -.565
.020(.005) .355
.492(.142) .337

***
***
***

-2.056(.522) -.428
.022(.007) .337
.394(.156) .290

***
.003
.012

-2.375(.425) -.481
.021(.006) .322
.396(.127) .292

***
***
.002

-2.558(.478) -.515
.022(.006) .379
.551(.132) .425

***
***
***

Mediation Analysis
Mediation was tested by assessing the significance of the cross product of the coefficients
for the IV to mediator relation (the a path), and the mediator to DV relation controlling for IV
(the b path). The a paths tested in the model were: (a) risk(X) – healthy marriage(M); (b) risk(X)
– parenting(M); (c) and healthy marriage(X) – parenting(M). The b paths tested in the model
were: (a) healthy marriage(M) – parenting(Y) controlling for risk(X); (b) healthy marriage(M) –
social competence(Y) controlling for risk(X); (c) parenting(M) – social competence(Y)
controlling for risk(X); and (d) parenting(M) – social competence(Y) controlling for healthy
101

marriage(X). An ab cross product test is a well recognized method to test mediation
(MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002) because it tests the statistical
significance of the difference between the total effect (c path), and the direct effect (c′ path),
which is the impact of the IV on the outcome controlling for the effect of the mediator. The c
paths tested in the model were: (a) risk(X) – parenting(Y); (b) risk(X) – social competence(Y);
(c) healthy marriage(X) to social competence(Y). The c′ paths tested in the model were: (a)
risk(X) – parenting(Y) controlling for healthy marriage(M); (b) risk(X) – social competence(Y)
controlling for healthy marriage(M); (c) risk(X) – social competence(Y) controlling for
parenting(M); and (d) healthy marriage(X) to social competence(Y) controlling for
parenting(M). MacKinnon, Warsi, and Dwyer (1995) demonstrated that in normal least squares
models, ab is algebraically equivalent to c−c′. Therefore, the significance of the cross product of
coefficients directly tests mediation. This is different from inferences made using the causal
steps approach by Baron and Kenny (1986).
The non-parametric method used in the current study (Preacher & Hayes, 2008a;
Preacher & Hayes, 2008b) is through bootstrapping, in which 1,000 samples of the original size
are taken from the obtained data and mediational effects are calculated in each sample. The
point estimate of the indirect cross product (Table 23) is the mean for these 1,000 samples,
ensuring stability of the analyses; the bias corrected and accelerated 95% confidence intervals
are comparable to the 2.5 and 97.5 percentile scores of the acquired distribution over the
samples, but with z-score based corrections for bias due to the original distribution (Preacher &
Hayes, 2008a; Preacher & Hayes, 2008b). If the confidence intervals do not contain zero, the
point estimate is significant at the level indicated.

102

In the present analysis, bias-corrected 95% confidence intervals showed that for males,
the only significant mediation effect was healthy marriage mediated the relationship between risk
and parenting (see Table 23). The a, b, and c paths were significant; and c′ paths were not
significant indicating a partial mediation rather than complete mediation (z = -2.84, p = .005).
See Table 24 for the t values and significance of the a, b, c, and c′ paths as well as the z scores
which represent significance of the indirect effects using Sobel tests (defined as the effect
divided by its standard error). There were no significant effects for males from risk to social
competence or from healthy marriage to social competence that could be mediated.
The mediation analysis for females did demonstrate significant effects from risk to
parenting, from risk to social competence, and from healthy marriage to social competence.
Healthy marriage was found to partially mediate the relationship between risk and parenting (z =
-3.20, p = .001), parenting partially mediated the relationship between risk and social
competence (z = -2.89, p = .004), and parenting partially mediated the relationship between
healthy marriage and social competence (z = 3.14, p = .002 ). However, healthy marriage did not
mediate the relationship between risk and social competence (z = -1.50, p = .133). This was
because healthy marriage did not have a significant effect on social competence when controlling
for risk (t = 1.52, p = .130).

103

Table 23
Bootstrapped Point Estimates and Confidence Intervals (CIs) for Indirect Effects of Mediators by
Gender
Effect

Point
Estimate

SE

Bootstrapping 95%
Lower CI
Upper CI

Male Indirect Effects
Risk HMParenting
-.028
.010
-.052
-.011
Risk HMSocial Comp.
-.009
.012
-.040
.014
Risk  Parenting  Social Comp.
-.014
.010
-.038
-.001
HM ParentingSocial Comp.
.006
.004
-.001
.015
Female Indirect Effects
Risk HMParenting
-.043
.015
-.080
-.016
Risk HMSocial Comp.
.029
.018
-.063
.008
Risk  Parenting  Social Comp.
-.036
.013
-.070
-.014
HM ParentingSocial Comp.
.009
.003
.004
.017
Level of Confidence for Confidence Intervals: 95; Number of Bootstrap Resamples: 1000
Table 24
t and z Values for Normal Theory Mediational Tests by Gender
X-M
(a path)

Male Indirect Effects
Risk HMParenting
Risk HMSocial Comp.
Risk  Parenting  Social Comp.
HM Parenting Social Comp.
Female Indirect Effects
Risk HMParenting
Risk HMSocial Comp.
Risk  Parenting  Social Comp.
HM Parenting Social Comp.
*p < .05; **p < .01; ***p < .001

M(X)-Y
(b path)

-3.88***
-3.88***
-2.32*
4.66***

4.10***
0.88
2.13*
1.74

-8.13***
-8.13***
-4.00***
5.05***

3.46***
1.52
4.15***
3.98***

104

X-Y
(c path)

X(M)Y
(c‘ path)

Indirect
Effect
(normal
test z score)

-2.32*
1.70
1.70
0.31

-1.05
1.89
2.10*
-.3473

-2.84**
-0.87
-1.58
1.64

-4.00***
-2.13*
-2.13*
2.40*

-1.76
-1.06
-0.96
0.95

-3.20**
-1.50
-2.89**
3.14**

Structural Equation Modeling
Metric invariance. First, structural equation modeling was used to examine a metric
invariant model (measurement weights) against an unconstrained model using AMOS 18.0.
Maximum likelihood parameter estimation was chosen over other estimation methods (weighted
least squares, two-stage least squares, asymptotically distribution- free [ADF]) because the data
were distributed normally (Kline, 2005). In the first, unconstrained multiple group analysis, no
measurement weights nor structural paths were fixed for either gender. In the second, more
restrictive multiple group analysis, the measurement weights were fixed to be equal for males
and females. This comparison demonstrated that difference in fit approached significance
2

between the unconstrained and measurement model ( (11) = 18.15, p = .08).
Structural invariance. A structural equation model (SEM) was used to assess the
relationships among the exogenous latent variable (Risk) with the endogenous latent variable
(Healthy Marriage) and the two endogenous observed variables (Parenting and Social
Competence). The hypothesized SEM is described graphically in Figure 5. A multiple group
SEM analysis based on data from males (n = 143) and females (n = 188) with AMOS 18.0
statistical package was performed. Circles represent the two latent variables (risk and healthy
marriage) and the rectangles represent the measured variables. Maximum likelihood parameter
estimation was chosen over other estimation methods (weighted least squares, two-stage least
squares, asymptotically distribution- free [ADF]) because the data were distributed normally
2

(Kline, 2005). The hypothesized model appeared to be an adequate fit with the data (χ =
331.280, df = 176, p = .000, CFI = 0.94, TLI = .93, RMSEA = .05, SRMR = .06). Post-hoc
modifications were not conducted because of the good fit of the data to the model.
105

Figure 5. Hypothesized structural SEM model.

106

Multiple group analysis. A multiple group analysis of model invariance compared the
fit of the model factor structure across the two genders; thereby garnering evidence as to whether
the three structural pathways illustrated in Figure 5 differed significantly between males and
females. In the less restrictive multiple group analysis, the values of the measurement weights
for the healthy marriage latent variable and the two measurement weights for the risk latent
variable were fixed to be equivalent so as to potentially suggest an invariant measurement model
for the two groups. In the second, more restrictive multiple group analysis, all of the
measurement weights for the healthy marriage latent variable, the two measurement weights for
the risk latent variable as well as all structural main paths were constrained to be invariant, thus
fixed to be equal for both groups. During both analyses, factor loadings were held invariant
between the two groups so that constructs would be measured similarly for male and female
groups. This comparison revealed a significant difference in fit between the measurement and
2

structural models ( (3) = 9.423, p = .024), indicating that the main path structural coefficients
significantly differ between males and females (see Table 26 for coefficients of the SEM model).
2

This structural model fit the data adequately well (χ = 358.857, df = 190, p = .000, CFI = 0.93,
TLI = .92, RMSEA = .05, SRMR = .08).
To determine which of the three structural paths were significantly different between
males and females, chi-square difference tests were conducted on an item by- item basis where
equality constraints were imposed on some but not all of the paths. A model was fit first by
placing equality constraints on all measurement and structural weights, and then a less restrictive
model was fit by relaxing the equality constraint of the examined path. A chi-square difference
test (Table 25) was conducted between the less restrictive model and the structural weights
107

model to investigate whether the path was invariant across parental gender. This process was
repeated item by item until all of the nonequivalent items were identified. Results of chi-square
difference tests indicated that risk to healthy marriage was nonequivalent while healthy marriage
to parenting as well as parenting to social competence were equivalent across parental gender.
Table 25
Equivalent and Nonequivalent Factor Loadings of Structural Paths across Parental
Gender
2
 (df =1)
7.222
.242
1.944

Path
RiskHM
HMParenting
ParentingSC
**p < .01

p
.007**
.622
.163

Direct effects. Risk was related negatively to Healthy Marriage/Healthy Relationships
(HM/HR) for males (standardized coefficient = -.33) and for females (standardized coefficient =
-.59). Healthy Marriage was related positively to Parenting Behavior for males (standardized
coefficient = .37) and for females (standardized coefficient = .36). Parenting Behavior was
positively related to Social Competence for males (standardized coefficient = .15) and females
(standardized coefficient = .32). Please see Table 26 for direct effect coefficients.
Table 26
Standardized and Unstandardized Coefficients for Males and Females on SEM Model
Males
Unconstrained Model
Direct Effects
Risk HM
HM Parenting
Parenting Social Comp.
***p < .001

β

B(SE)

Females
p

-.334 -.271(.081)
.368 .399(.092)
.145 .220(.126)

108

***
***
.081

R

2

β

.112 -.588
.136 .363
.021 .320

B(SE)

p

.462(.072) ***
.334(.066) ***
.440(.095) ***

R

2

.346
.132
.102

Summary of Results
Preliminary results showed that females had significantly higher scores than males on
parenting behaviors (especially supportive/engaged behaviors) as well as on depression. Both
married males and married females had significantly higher scores on children‘s social
competence scores than their unmarried counterparts. In addition, unmarried females scored
significantly higher on depression than married females. There were no significant differences
between African Americans and Caucasians within gender. Results from the EFA in study one,
showed that the measure used to assess healthy marriage was unidimensional. In addition,
depression was significantly related to the HM/HR composite for females but not for males
whereas stress was significantly related to the HM/HR composite for males and approached
significance for females. Study two results demonstrated support for the factorial validity of the
healthy marriage measure. Overall, the healthy marriage measurement fit for both males and
females. The factor loadings were significant for both males and females although results of the
chi-square difference tests indicated that the factor loadings of two variables were nonequivalent
across gender: relationship stability (larger for males) and child commitment (larger for females).
This difference was only in the strength of the loading as both loaded significantly on the healthy
marriage latent construct. The factor loadings of the majority of variables were invariant across
gender. Study two results demonstrated convergent validity. The healthy marriage variables for
both males and females were specified to measure a common underlying healthy marriage factor.
Study three results demonstrated that the Child Trends, Inc. healthy marriage
measurement model can be appropriately used for diverse groups. There was a significant effect
of risk (depression and stress) on healthy marriage for both males and females. The strongest
path between risk and healthy marriage was for African American women and unmarried
109

women. In addition to the direct effect of risk (depression and stress) on healthy marriage for
males and females, risk had a significant direct effect on parenting and healthy marriage had a
significant direct effect on parenting. Results showed that for both males and females, healthy
marriage is part of the mechanism through which risk (depression and stress) influences
parenting (Baron & Kenny, 1986). For females, risk had an indirect effect on children‘s social
competence through parenting and healthy marriage had an indirect effect on children‘s social
competence through positive parenting. This study found that for males, healthy marriage had a
significant influence on parenting; however, healthy marriage did not significantly influence the
ratings of children‘s social competence.

110

CHAPTER 5: DISCUSSION
Overview
Although there have been major contributions in defining ―healthy marriage‖ there is still
much to be learned about how healthy marriage as a construct is measured as well as how the
model of healthy marriage predicts adult and child well being. The need for measuring healthy
marriage is a practical one given the many (about 300) federal and state healthy marriage
initiative demonstration projects that exist and the empirical evidence needed to guide future
directions. There is great interest in evaluating the effects of these federal and state marriage and
relationship education programs with the ultimate goal being children‘s well-being.
The first generation of marriage and relationship evaluation (MRE) research, from 1975
through the mid-2000s showed success in improving relationship quality, improving couple
communication as well as increasing stability (Hawkins & Ooms, 2010). However, these studies
were essentially void of examining improvements in parenting or child well-being and certainly
did not focus on lower-income, less educated, more distressed couples (Ooms & Wilson, 2004).
Beginning in 2002, during the second generation of MRE research, programs and evaluation
began to more specifically target low-income, less educated couples and evaluation of these
programs is ongoing. Hawkins and Fackrell (2010) completed a meta-analysis searching for
MRE program outcomes evaluation studies targeted primarily to low-income couples. They
found that these studies demonstrated positive outcomes for participants in regard to relationship
quality, commitment, relationship stability, and communication skills. However, parenting and
child outcomes have been less of a focus. The forthcoming large-scale rigorous Supporting
Healthy Marriage study by Manpower Demonstration Research Corporation (MDRC) expected
in 2012 will include direct assessments of child health and well-being.
111

MDRC provided the baseline tool that they are using for the Supporting Healthy
Marriages study for this dissertation research so that the measure of healthy marriage and its
associated antecedents could be studied together with child well-being at the local level.
Whereas the SHM study includes married couples only, this research extends to unmarried
couples. In addition, local healthy marriage programs have an evaluation need to demonstrate
child outcomes in a feasible way, and quite possibly through parent self-reporting methods. This
research tested two simple short-form parent report questionnaires to assess parenting behaviors
and child well being, specifically children‘s social competence. Because the goal of federal
healthy marriage interventions is to enhance outcomes for children, it is helpful to measure all of
these constructs in the same survey and prospectively examine their individual and cumulative
association with parenting and child outcomes. The benefit of using the SHM measure of
healthy marriage is that it encompassed all ten of the Child Trends, Inc. ―healthy marriage‖
constructs with about 30 items. This was the only measurement of healthy marriage that the
researcher found appropriate for use because its brevity did not compensate for measuring
healthy marriage as an all-encompassing concept.
The present dissertation research is not one of program evaluation; rather it set out to
initially establish a measure and structural model of healthy marriage and to examine structural
fit across gender, race, and marital status. The research examined how each healthy marriage
construct contributed to the measure of healthy marriage and tested additional constructs for
inclusion. Once the measurement was tested, the next step was to test the Child Trends, Inc.
healthy marriage measurement model of predicting adult and child well-being. Testing the
healthy marriage measure in this context provided results on mediation to further explore how

112

risk (depression and stress), healthy marriage, and parenting play a role in predicting children‘s
social competence.
Measurement of Healthy Marriage
Healthy Marriage Measured Variables
Each healthy marriage variable was analyzed for reliability as well as for its relationship
to the other healthy marriage, parenting and preschooler social competence variables.
Preliminary analysis demonstrated that the two-item conflict resolution process variable and the
two-item relationship quality variable had poor reliability ( = .53 and .51 respectively) for this
sample. Because of this, these two variables were dropped from further analysis. The one-item
fidelity variable was also dropped because it did not correlate well with the other healthy
marriage, parenting, and social competence measures. The nine-item conflict frequency variable
and three-item domestic violence variable replaced the conflict resolution process variable for a
measure of conflict. The two conflict resolution items asked participants to rate how often (using
a 4-point Likert scale) the following happened: ―Our arguments get very heated‖ (reverse-coded)
and ―I am satisfied with the way we handle our problems and disagreements‖. Lichter and
Carmalt (2009) utilized these same two items in their research on marital quality with an
additional third item, ―When we are having a problem, my spouse often gives me the silent
treatment.‖ Besides adding the third item to the conflict resolution scale, responses were on a 4item Likert agreement scale instead of the 4 –item frequency scale used in that research. The
scale used by Lichter and Carmalt (2009) had better reliability ( = 0.60 for husbands and 0.66
for wives) than this study ( = .53).
The conflict resolution variable was dropped with the confidence that the domestic
violence variable captured the negative resolution of conflicts. Specifically, the domestic
113

violence variable was able to capture conflict resolution process with its three items. Participants
reported how often in the past year their spouse/partner: yelled or screamed; blamed them for
his/her problems; and how often their arguments became physical. This domestic violence
measure demonstrated satisfactory reliability ( = .76). This study considers both conflict
frequency and conflict resolution as valuable measures of healthy marriage especially in terms of
possible negative spillover from interparental negativity to parent-child negativity. The conflict
frequency scale and complimentary domestic violence scale were both used for this research.
In addition, the one-item relationship happiness variable replaced the two-item
relationship quality variable as a measure of ―satisfaction‖. Relationship satisfaction is the most
common measure of marital quality yet it remains a problematic construct to measure (Fincham
& Beach, 2010). It may be measured by overall happiness with marriage such as the one item
measure used in this study which asks respondents on a scale from 1 (completely unhappy) to 7
(completely happy) how happy they are with their marriage/relationship. At the time that the
questionnaire was offered to the researcher, MDRC was testing both the happiness and
relationship quality measures for satisfaction. The research findings for this study suggested that
the one-item relationship happiness measure is superior to the two-item relationship quality
measure which asks respondents to rate the good side of their marriage (from 1 ―not at all good‖
to 5 ―completely good‖) and then to rate the bad side of their marriage (from 1 ―completely bad‖
to 5 ―not at all bad‖). Quite possibly the lack of reliability for this measure may be understood
with an explanation by Reis and Gable (2003) who argue that ―happiness [is] not simply the
absence of distress and distress [is] not simply the absence of happiness‖ (p. 131). The use of
this scale was not found in the literature and combining these two items did not prove to be a
satisfactory measure of satisfaction. There is confidence in using the one-item relationship
114

happiness measure as Goodwin (1992) has shown that a single-item assessment of marital
happiness correlates highly with total marital adjustment and differentiates between adjusted and
distressed couples. The item used in this current study is much like the first item from the
Marital Adjustment Test (MAT) (Locke & Wallace, 1959) to assess satisfaction and captures a
great deal of information about individuals‘ relationship satisfaction (Funk & Rogge, 2007).
This item was recently used by Doss, Rhoades, Stanley, & Markman (2009). Single item
indicators of marital satisfaction are relatively common in research on marital satisfaction,
happiness, and quality (e.g., Schoen, Rogers, & Amato, 2006).
Unfortunately, the measure of fidelity could not be used in this research due to its low
correlations with the other variables. Fidelity is an important construct of healthy marriage and
infidelity is positively related to divorce (Atkins et al., 2001; Previti & Amato, 2004). The oneitem measure in this study asked participants to rate how strongly they agree (1 ―strongly
disagree‖ to 4 ―strongly agree‖) with the statement ―I worry about my spouse/partner cheating on
me‖. An alternative item offered in the Child Trends, Inc. Summary Report on Cognitive
Interviews for Healthy Marriage Item Development is ―I trust my partner to be faithful to me.‖
The single-item measure used in the 1998 General Social Survey conducted by the National
Opinion Research Center at the University of Chicago asks respondents, ―have you ever had sex
with someone other than your spouse while you were married?‖ However, this item would need
to be revised if used with unmarried couples, reports on one‘s own infidelity behaviors, and is
specific to sexual faithfulness while not considering other forms of infidelity (i.e. online
infidelity and cybersex). In addition, the sensitive nature of fidelity as a self-report item may
also present some methodological challenges (Schroder, Carey & Vanable, 2003).

115

Discussion for Study One
Results from the EFA, showed that the measure used to assess healthy marriage was
unidimensional with emotional intimacy, communication, teamwork, and quality time loading
best on the healthy marriage measure. The healthy marriage measure encompassed eleven
important domains and was still able to be used as a unitary construct. Healthy marriage, as a
general construct may be measured using a single composite score ( = .923). This implies that
researchers and evaluators could compute and interpret an overall healthy marriage score. This
was consistent with results reported to the researcher verbally by a representative of MDRC. In
Holman‘s memo to Child Trends, Inc. (2003) regarding the conceptual, measurement and
method issues of healthy marriage, he offers a description of the RELATionship Evaluation
(RELATE) project. Like the measure used in this current study, The RELATE approach is to
measure a large number of concepts with scales that generally have 3-5 items with an effort to be
comprehensive but not exhaustive. He relays the message that one of the drawbacks with many
surveys is having a long list of items, and thus not being able to measure enough concepts before
respondents become fatigued. He poignantly remarks that the RELATE measure ―is a reflection
of our interest in looking at the tip of many icebergs, assuming the tip will give us adequate
information on the whole, rather than understanding all the dimensionality of a very few
icebergs‖ (p. 171). Although the healthy marriage variables assessed in this current study can be
used individually, as confirmed by satisfactory reliability of each, they may also be used together
in a reliable composite to represent the latent variable of healthy marriage. This is especially
helpful when assessing, overall, how healthy marriage/healthy relationships is modeled with
other background characteristics, antecedents, and consequences. The advantage of utilizing the
unidimensional measure is that all of the healthy marriage constructs outlined by Child Trends,
116

Inc. can be taken together to give a full picture of healthy marriage. A potential disadvantage of
the unidimensional measure is that, for example, the one variable measuring relationship stability
is a single dichotomous variable and may lack variability. This variable was used as a proxy for
marital duration or marital status as it was able to capture instability in even long-lasting marital
relationships.
Results from multiple regressions showed that depression and stress are related to the
HM/HR composite differently for males and females. A discussion of this finding is presented
further under study three.
Discussion for Study Two
Study two results demonstrated support for the factorial validity of the healthy marriage
measure. Overall, the healthy marriage measurement fit for both males and females. The factor
loadings were significant for both males and females yet results of chi-square difference tests
indicated that the factor loadings of two variables were nonequivalent across gender: relationship
stability (larger for males) and child commitment (larger for females). This difference was only
in the strength of the loading as they both were significant to the measure of healthy marriage.
However, the factor loadings of the majority of variables were invariant across gender.
Relationship stability was measured with the reverse of a single dichotomous item: In the past
year, have you ever thought your marriage/relationship was in trouble. Child commitment was
comprised of two items: my spouse/partner is the type of parent I want for my child(ren) and my
spouse/partner is completely committed to being there for our children. Males and females can
both relate to these two concepts, however, the level of importance may vary. Research has
shown that males tend to report that no changes are needed in their relationships, (Christensen &
Miller, 2006; Henry et al., 2005), thus having fewer perceptions of relationship problems
117

compared to their wives. For females, the level of importance placed on ―father involvement‖ is
evident by stronger ratings of their partner‘s commitment to the child. A 2005 report of the
Motherhood Study, surveying a nationally representative sample of more than 2,000 mothers
revealed that low-income mothers with fewer years of education are more likely to report that
they want greater father involvement. They are more likely to be unmarried and have added
responsibility for daily care of the children as well as the financial support of the family. This
may explain a possible different level of importance women place on their partner‘s commitment
to children.
The measurement model provided a powerful tool for explicating some gender
differences. Relationship stability and child commitment may not be equated across gender and
may have different meanings for males and females, however, these non-equivalent items
constituted only a small portion of the model. Even though the two healthy marriage variables
were not equivalent across gender, individuals can still be assessed on a common healthy
marriage measurement scale. One might choose to scale the invariant variables with common
weights, but with different weights for ―relationship stability‖ and ―child commitment‖.
Study two results demonstrated convergent validity. The healthy marriage variables for
both males and females were specified to measure a common underlying healthy marriage factor.
For both males and females, emotional intimacy had the greatest association with the latent
healthy marriage measure. The significance of emotional intimacy to the concept of healthy
marriage is supported theoretically as well. Emotional intimacy in this research includes
counting on your partner, having similar views and expressing love/affection. Conventional
definitions of intimacy encompass factors related to closeness, disclosure, activity sharing,
sexuality, and affection (Clark, 1998; Heller & Wood, 1998; Patrick et al., 2007). Previous
118

research examining the association between intimacy and relationship satisfaction has found that
there is a strong correlation shared between the two constructs (Patrick, Sells, Giordano, &
Tollerud, 2007). In another study of variables that can be used to predict marital satisfaction,
Eckstein and Goldman (2001) found that intimacy significantly predicts marital stability. Failure
to obtain satisfactory levels of intimacy in a marriage was the most frequent reason given by
couples for their divorce (Moss & Barry, 1993; Waring, 1988). However, emotional intimacy is
given less weight in therapy and education. Fowers (2001) argues that marital researchers and
therapists believe that improvements in communication skills are best for improving and
maintaining marriages. While teaching communication skills and even conflict resolution skills
certainly does help couples improve satisfaction and stability, this great deal of emphasis on
technique seems to overshadow other aspects of marital quality, such as emotional intimacy. It
is through intimacy that secure bonds and attachments are strengthened (Johnson, 1996). This
current research demonstrates that the definition, promotion, and measurement of healthy
marriage for both males and females must include emotional intimacy.
Discussion for Study Three
This study found that the healthy marriage constructs most significantly related to
children‘s social competence were: couple commitment, child commitment, forgiveness,
teamwork, relationship happiness, and stability. Although these correlations were low (r < .20)
the findings shed some light on the healthy marriage construct in the context of child outcomes.
Conflict resolution and communication processes have received most attention in marriage
research and is the focus of marriage and relationship education curriculum. However, these
other constructs may facilitate a direction that results in more positive attributions for a partner's
behavior (McCall, 1995). Fincham, Stanley, and Beach (2007) suggest that these types of
119

healthy marriage constructs (i.e. commitment, forgiveness) function as self-repair or selfregulatory processes in healthy relationships. For example, two satisfied spouses may have a
heated disagreement that could potentially harm their relationship. Afterward, these spouses
may forgive one another (or engage in other positive processes) and repair the damage, thus
bringing the marriage back to harmony. Likewise, committed couples often make self-sacrifices
for the sake of the marriage. Greater commitment is associated with more constructive,
accommodative responses to negative partner behavior (Rusbult et al., 1991, Rusbult et al.,
1998), and may provide a more positive and stable environment for children. What this
dissertation research suggests is that when couples believe in the future of their relationship,
view problems as solvable, are committed to their children, let go of negative feelings toward
each other, and work together as a team, their children, for example, are better able to work out
problems, understand other peoples‘ feelings, and control their tempers during a disagreement.
Important to this research was the examination of two marital virtues, namely forgiveness
and teamwork, that are not included the Child Trends, Inc. measurement model of healthy
marriage. Consequently, these two healthy marriage constructs correlated (although correlations
were small) the most with children‘s social competence thus demonstrating that they may be
stronger predictors of child well-being than other measures of marital/relationship quality.
Adding marital virtues to the definition of healthy marriage are important because the factors that
contribute to positive relationship adjustment can help professionals develop more effective
interventions for couples who may be experiencing marital distress, and more effective marriage
education programs. Marital virtues may be further explored in studies on marriage and
relationships to gain a better understanding of their role in relationship adjustment. In their work
on Affective Reconstruction (2002), Snyder and Schneider hypothesize that what might mainly
120

set healthy and unhealthy relationships apart is the degree to which partners are self-aware while
being aware of their partner. They also suggest that another important factor in relationship
health is the ability and willingness to set aside one‘s own gratification for the sake of the other.
The importance of other-centeredness and generosity is worth including in a definition of healthy
marriage/healthy relationships.
The healthy marriage constructs correlated slightly more with the parenting behavior
outcomes (reverse of hostile/coercive parenting behaviors as well as supportive/engaged
behaviors) than the social competence outcomes (prosocial/communication and emotion
regulation). In fact, all of the healthy marriage constructs (especially communication) correlated
with parenting behavior. For example, conflict frequency was correlated more with parenting
behaviors than children‘s social competence. This may be due to the simultaneous role that
marital partners and parents play. Parents with high conflict may, for example, lose their
tempers more with their children and may even use physical punishment.
A description for marital influences on parenting is the ―spillover‖ hypothesis, the direct
transfer of mood, affect, or behavior from one setting to another that explains the interrelatedness
of the marital/partner relationship to the parent-child relationship. Erel and Burman (1995)
found a positive association between marital quality and parenting (the composite mean effect
size was .46). Overall, healthy marriages tend to be associated with more positive parenting
behaviors (i.e. supportive/engaged). Brody, Arias, and Fincham (1996) demonstrated this
association from infancy through adolescence. On the other hand, a marriage characterized by
high marital discord may result in parents' being less involved with their children and
implementing harsh, less consistent and less communicative disciplinary practices in comparison
with couples in healthier marriages (Cummings & Davies, 2002). These studies suggest that the
121

emotional support parents gain from their spouses and the overall quality of their marital
relationship can be expressed in the parent-child relationship. Thus, the behaviors and feelings
generated in a stressed marriage/relationship can "spill over" and negatively affect both parenting
and child development. The stronger association of healthy marriage to parenting may be a
reflection of how partner behaviors first and more directly influence a parent‘s ability to parent,
thereby affecting children‘s development. For example, a meta analytic review found that
marital conflict was associated with harsh parenting (with an overall average weighted effect size
of -.62), demonstrating support for the spillover hypothesis (Krishnakumar & Buehler, 2000).
The implication is for marriage and relationship education to be inclusive of parenting training
components, especially with the ultimate goal being child well-being.
Overall, the Child Trends, Inc. healthy marriage measurement model showed support for
the appropriate use with the diverse sample. There was a significant effect of risk (depression
and stress) on healthy marriage for both males and females. Depression and stress place a
burden on healthy marriage/healthy relationships for both genders, and this is especially true for
females compared to males. To note, depression was significantly higher (t = -2.55, p = .011) for
females (M =1.03, SD = .86) than for males (M = .80, SD = .77), thus contributing greater to the
effect on healthy marriage/healthy relationships. Koball, Moiduddin, Henderson, Goesling &
Besculide (2010) suggest that healthier individuals may be better able, for example, to
communicate and enjoy greater interactions with their partners, leading to better marital quality.
On the other hand, a partner with poor health may strain a marriage to the point of divorce. Path
analysis demonstrated that the Child Trends, Inc. healthy marriage measurement model differed
by gender and this was supported by further SEM analysis. Specifically, the path between risk
(depression and stress) and healthy marriage is nonequivalent for males and females.
122

A recent study by Heidemarie, Hyoun, and Capaldi (2009) found that women‘s
depressive symptoms were more closely linked to couple interactions (i.e. positive engagement,
psychological aggression, withdrawal) through various pathways, whereas the association
between men‘s depressive symptoms and couples‘ behavior was more limiting. Specifically,
women‘s use of positive conflict resolution skills was a protective factor for both themselves and
their partners against depression, and served as a powerful indicator of the couple‘s resilience.
They suggest that especially for women, helping to improve overall relationship quality as well
as to build a positive sense of self within the relationship could prevent depressive problems in
the future.
The effect size between risk and HM/HR tended to be larger among unmarried females (β
= -.52) and unmarried females were also more likely to have significantly higher scores on
depression than married females. This is consistent with hundreds of other studies indicating
that married individuals have better mental and physical health than the unmarried (see Waite &
Gallagher, 2000 for a review). Explanations point to greater social support, lower economic
strain and the lower likelihood to engage in risky health behaviors for married compared to
unmarried individuals (Waite & Gallagher, 2000). However, understanding the mechanisms
through which depression and stress effect marriage is challenging. Teitler and Reichman (2008)
explain with their research that although the hypothesized mediators (human capital, relationship
quality, partner selection, and substance abuse) explain some of the mental illness effect on
marriage, most of the effect remains unexplained.
DeKlyen, Brooks-Gunn, McLanahan, and Knab (2006) used data from the Fragile
Families and Child Wellbeing Study to examine mental health across marital status (married,
cohabiting, not cohabiting but romantically involved, and not romantically involved). After
123

controlling for ages, education, race, immigrant status, multiple fertility, and Medicaid support,
they found that overall, unmarried parents (particularly women) had more mental health and
behavior problems than married parents. However, fathers who were not romantically involved
with their child‘s mother reported the greatest prevalence of depression and anxiety, a subgroup
not analyzed in this dissertation research. Children of unmarried parents may be more at risk for
poor parenting and poor developmental outcomes. The married fathers in their study reported
the lowest prevalence of depression and anxiety. The implication of their research supports the
results of this dissertation research pointing to the need for government initiatives, including the
Healthy Marriage Initiative, to consider the mental health of parents, especially for unmarried
women.
Although, African American female depression/stress scores were not significantly
different than Caucasian depression/stress scores, the strongest path between risk and healthy
marriage was for African American women. These results suggest that although depression may
not be particularly higher for African American females, the process in which depression/stress
affects their marriage may be quite unique. This particular nuance deserves greater exploration.
Koball, Moiduddin, Henderson, Goesling & Besculide (2010) point out that until recently, most
research on the link between marriage and health aggregates data across racial groups instead of
examining this relationship for African Americans in particular (Wood et al., 2007) due to
possible small samples of African Americans in many data sets. While controlling for race, the
unique relationships between marriage and health among African Americans cannot be
investigated. Therefore, little is known about how marriage and health are linked for African
Americans, let alone for African American women (Wood et al., 2007). What this dissertation

124

research suggests is that the mechanism by which mental health may affect the healthy
marriage/healthy relationships may differ for African American women.
Bender (2005) explains the particular vulnerability of depression and stress for African
American women and calls for an understanding of societal, environmental, and biological
factors (such as stress related to racial discrimination). Koball, Moiduddin, Henderson, Goesling
& Besculide‘s (2010) highlight several recent articles produced as a response to a request for
proposals by HHS through Mathematica to study links between marriage and health in African
American communities. One study in particular by Lincoln and Chae (2010), highlights the
importance of high-quality marriages for managing stressors that many African Americans face.
Their analysis of cross-sectional National Survey of American Life data show that financial
strain and unfair treatment, for example racial discrimination, result in lower levels of marital
satisfaction as well as higher levels of psychological distress (measured with the same Kessler K6 as used in this dissertation research). Specifically, greater reports of unfair treatment were
associated with lower levels of marital satisfaction (b = -08, SE = 04, p = .04) as well as financial
strain (b = -.06, SE = .02, p = .003) with a significant covariate in the model being gender. In
both models, women reported lower levels of marital satisfaction compared with men. When
including both unfair treatment and financial strain in the model, unfair treatment was no longer
significant (b = -06, SE = 04, p = .10), whereas financial strain remained significant (b = -.05, SE
= .02, p = .007) in predicting marital satisfaction. They suggested that economic strain may be a
―master stressor‖ in the lives of married African Americans beyond the influence of unfair
treatment. Interestingly, high levels of marital satisfaction protected couples from experiencing
distress in the context of these stressors. Specifically, the interaction between unfair treatment
and marital satisfaction (b = -.05, SE = .03, p = .07) and the interaction between financial strain
125

and marital satisfaction (b = .22, SE = .11, p = .04) were both significant. This protective effect
was lessened among those with high levels of financial strain compared to those with low levels
of financial strain.
Bryant, Wickrama, Bolland, Bryant, Cutrona, and Stanik (2010) propose a model for
understanding marital relationships of African Americans and suggest studies should examine
multiple factors including stressors, individual characteristics, psychosocial resources, and
couple interaction simultaneously as well as examining their interaction. Their literature review
points to research showing that African Americans‘ fewer socioeconomic opportunities lead to
lower levels of socioeconomic status; thus leading to greater exposure of a massive amount of
stressors (Duncan & Magnuson, 2005; Mills & Combs, 2002). The review describes significant
sources of stress for African Americans including economic stress, adverse work conditions,
family obligations (including children), racial discrimination, and minority stress (Bryant &
Wickrama, 2005; Cutrona, Russell et al., 2003; Murry, Brown, Brody, Cutrona, & Simons,
2001). In terms of emotional distress, Bryant et al., (2010) use the frustration–aggression
hypothesis (Berkowitz, 1989), to explain how stressful experiences have psychological
consequences. Specifically, when couples‘ experience financial strain, they may become
depressed and pessimistic (Vinokur, Price, & Caplan, 1996) and this distress may be a factor
leading to marital instability and poor marital quality (Bryant et al., 2010).
In addition to the significant direct effect of risk (depression and stress) on healthy
marriage for males and females, risk had a direct effect on parenting, and healthy marriage had a
significant direct effect on parenting. Poverty‘s influence on the psychological well-being of
parents may explain the effect of depression and stress on parenting (see McLoyd, 1990). There
is a need for greater consideration of risk within models of healthy marriage and parenting. In
126

addition, models, policies, and interventions of healthy marriage can highlight the influences of
healthy marriage on parenting. Whereas a significant direct effect of parenting on social
competence was present for females, this effect only approached significance for males.
Possible reasons are explained below.
Discussion on Mediation
Further, mediation analysis was conducted to allow for a better understanding of the
healthy marriage/healthy relationships model. In Kenny‘s (2008) ―Reflection on Mediation‖ he
emphasizes the importance of mediation by explaining how it provides the investigator with a
―story‖ about a sequence of events. For this research, mediation analysis tells part of the story of
the influences of risk (depression and stress), healthy marriage, and parenting and how they play
a role in predicting children‘s social competence. Results showed that for both males and
females, healthy marriage is part of the mechanism through which risk (depression and stress)
influences parenting (Baron & Kenny, 1986). Healthy marriage, as a mediator can cause
variation in parenting itself and is caused to vary by risk (Last, 1988). This implies that when
marriages/relationships are weighed down by possible stress and depression, positive parenting
can be made more difficult. The effects of risk on parenting can partly be explained through
healthy marriage. Father–child interactions (Lamb, 2004) as well as mother-child interactions
are embedded in a larger ecology including the mother-father relationship as well as family
human and economic resources (Lamb, 2004). Social support, in this case taking the form of a
healthy marriage/healthy relationship, is important in facilitating positive parenting especially
when depression and stress are present. The social support that exists in a healthy
marriage/healthy relationship offers benefits to parenting. Interventions focusing on improving
healthy marriage/healthy relationships consequently may be helping to improve parenting even
127

for parents in distress. Interventions aimed at increasing positive parenting behaviors for couples
in distress, would benefit by focusing attention on strengthening healthy marriage/relationships.
Risk-parenting-social competence. The next question is how does this translate to
effects on children? For females, risk had an indirect effect on children‘s social competence
through parenting. Therefore, parenting is an important mediator between the potentially
harmful effects of depression and stress on children. These results are not different from
previous well established links between stressed parents and poor child functioning (Crnic &
Low, 2002), although the more popular hypothesis is that parenting behaviors likely mediate this
association between stress and child functioning. Belsky (1984) is well known for theoretical
processes of competent parental functioning. The Belsky model focuses on factors affecting
parental behavior and how such factors affect child-rearing, which in turn influences child
development. At the family level, concentration, like Bronfenbrenner's, is primarily on parent
and child interactions. Belsky found that parental personality and psychological well-being were
the most influential of the determinants in supporting parental functioning. This dissertation
research demonstrates that it is through parenting that depression and stress effects children‘s
social competence and not through healthy marriage/healthy relationships alone. The effect of
risk on social competence through parenting is plausible given that mother‘s behaviors are more
proximal to the child‘s experiences. Supportive and engaged rather than hostile and coercive
parents are likely to model positive ways to manage stress including their regulation of emotional
responses to stress (Power, 2004) and inappropriate behavior (Halberstadt, Crisp, & Eaton,
1999). Moreover, they may facilitate children‘s social emotional development by providing a
more stable environment (Brody & Ge, 2001) and by protecting children from exposure to
potentially stressful events (Power, 2004).
128

Healthy marriage-parenting-social competence. For females, healthy marriage may
also enhance children‘s social competence through positive parenting as results from this study
demonstrated this indirect effect. While healthier marriages/relationships may boost children‘s
social competence through positive parenting, less healthy marriage/relationships may harm
children‘s social competence through negative parenting behaviors. A supportive spouse/partner
and a healthy marriage/relationship likely support the parenting process, enabling better and
more competent parenting, thus influencing children‘s social competence. On the other hand, a
less healthy marriage/relationship may harm children‘s social competence through negative
parenting behaviors. In an examination of the data from the 1998 National Survey of Families
and Households for children ages 2–11, revealed that parental harsh discipline helped to explain
the relation between marital conflict and child outcomes (Buehler & Gerard, 2002). Women in
healthy marriages/relationships would likely have more resources to focus on consistent
discipline practices, and would possibly reduce negative spillover from stress to parent–child
negativity. Therefore, if Healthy Marriage Initiatives want to ultimately influence child wellbeing, they must including a parenting component because, especially for women, it is through
parenting that healthy marriage/healthy relationships influence children‘s social competence.
For example, mothers may interact with their children in more supportive and engaged ways
when they have a stronger sense of teamwork and commitment with their spouse/partner. In
summary, the process by which maternal stress and depression effects children is not by healthy
marriage/healthy relationships alone. Rather the process explained by the Child Trends, Inc.
measurement model receives support: healthy marriage/healthy relationships work through
parenting to influence children‘s social competence.

129

It is important to note however, that parenting processes vary across ethnicity and culture.
The parenting behavior measure used in this study included both components of
supportive/engaged behaviors (i.e. hugging/kissing; thanking/praising; comforting) as well as
hostile/coercive behaviors (i.e. grabbing roughly; using physical punishment; threatening).
Research by Le, Ceballo, Chao, Hill, Murry, and Pinderhughes (2008) highlights the importance
of placing parenting in context by disentangling ethnic differences in parenting behaviors from
their cultural influences, for a richer understanding of parenting processes in diverse families.
Although preliminary results from this current study did not show significant differences in
parenting behaviors between African Americans and Caucasians, other research finds differences
across SES and ethnicity. For example, parents from lower socioeconomic backgrounds or who
are ethnic minorities have been found to express less warmth toward their children than do
parents from higher SES and ethnic majority backgrounds (Hoff-Ginsberg, Laursen, & Tardif,
2002; McLoyd, 1990; Pinderhughes et al., 2000). However, more recent research shows that
while African American and Asian American mothers scored lower on parental warmth than did
European American mothers (Chao & Kanatsu, 2008), low income African American mothers
reported higher displays of warmth toward their children than African Americans mothers with
higher SES (Pinderhughes et al., 2008). In addition, families living in less safe neighborhoods,
despite their ethnic background, have been shown to express less warmth than other families
(Furstenburg et al., 1993). While qualitative research shows that ethnically and
socioeconomically diverse families love and care for their children (e.g., Lareau, 2003), the
expression of these behaviors may differ (e.g., Hill, Bush et al., 2003; Mason et al., 2004).
Differences between males and females. This study found that for males, healthy
marriage had a significant influence on parenting; however, healthy marriage and parenting did
130

not have significant influences on ratings of children‘s social competence. To examine why both
risk and healthy marriage did not have an indirect effect through parenting on children‘s social
competence for males, their involvement with the Head Start child was further explored. First,
the direct effect of parenting on children‘s social competence had only approached significance
for males. Although appraisals of children‘s social competence were not significantly different
between mothers and fathers, (t = -.29, p = .77), mothers‘ appraisals of their positive parenting
behavior (M = 4.05) was significantly higher (t = -3.34, p = .001) than fathers‘ appraisals (M =
3.85). A further look into parenting behaviors between males and females shows a significant
difference (t = - 3.5, p = .001) for ―supportive/engaged parenting‖ between mothers (M = 4.57)
and fathers (M = 4.32), and no significant difference (t = -1.65, p = .10) for ―hostile/coercive‖
parenting (reversed) between mothers (M = 3.52) and fathers (M = 3.39). Thus, the significant
effect of parenting on children‘s social competence for mothers can be due to higher appraisals
of their positive parenting behaviors, namely supportive/engaged parenting behaviors. Another
explanation may be that mothers use different parenting strategies than fathers to influence
children‘s social emotional development.
Fathers in this dataset were quite involved in the parenting of Head Start children, as
fathers are involved in childrearing in numerous ways (Lamb, 2004; Pleck & Masciadrelli,
2004). Fathers who interact with their children in positive ways have significant effects on their
social and emotional development at 2 and 3 years and at pre-kindergarten (Cabrera, Tarkow, &
Shannon, 2006). Although this study did not demonstrate a direct effect of male parenting on
children‘s social competence, other studies have reported similar results (Cabrera, Shannon,
West, & Brooks-Gunn, 2006). First, estimating fathers parenting behaviors and ratings of
children‘s social competence scores may not be appropriate for studying a direct main effect.
131

Second, it is less likely to find an association between fathers‘ reports of parenting behaviors
than when the qualities of fathers‘ parenting behaviors are observed (Ryan, Martin, & BrooksGunn, 2006). Third, it is possible that fathers have different effects on children‘s development
across time (Cabrera et al., 2006). In addition, fathers‘ education and income have been found to
be key predictors of positive father engagement (Cabrera, Shannon, & Tamis-LeMonda, 2007)
and was not included in this current study.
Summary
Although it is quite difficult to disentangle marital and parenting processes and the
effects of one on another because of simultaneous roles fulfilled by both marital and parenting
partners (Fincham & Hall, 2005), this research made several contributions to what is known
about measuring and modeling healthy marriage as well as healthy relationships. This research
demonstrated that the healthy marriage/healthy relationship (HM/HR) constructs were linked to
parenting and child well-being. The measurement fit for low-income subpopulations including
males and females, African Americans and Caucasians, and for married and unmarried couples.
The research compared different approaches for assessing particular constructs. For example,
relationship satisfaction could be measured with a single item of relationship happiness. The
HM/HR measurement consists of several short version scales and the aggregation of constructs
into a summary measure of healthy marriage applies to low-income couples with children. The
unidimensional healthy marriage measurement fit for both males and females and the addition of
two marital virtues, namely forgiveness and teamwork, also loaded well on this measurement.
The factor loadings were significant for both males and females and results of chi-square
difference tests indicated that the factor loadings for the great majority of variables were
invariant across gender. However, the factor loading for relationship stability (one‘s assessment
132

that their relationship is not in trouble) was larger for males where the child commitment (one‘s
assessment of their spouse/partner‘s commitment to their child) was larger for females.
The healthy marriage/healthy relationships measure used in this study can be used to
predict parenting for both mothers and fathers and can be used to predict mothers‘ ratings of
children‘s social competence. Path analysis demonstrated that the Child Trends, Inc. healthy
marriage measurement model fit differed by gender and this was supported by further SEM
analysis. The effect size for the direct path between risk and healthy marriage was significantly
larger for females than for males. The effect size was larger for African American females and
unmarried females. Results using fit indices showed that the fit of the model was good for the
diverse sample. Mediation results showed that for both males and females, healthy marriage is
an important part of the mechanism through which depression and stress effect parenting. For
females, parenting is a mediator between the potentially harmful effects of depression and stress
on their children. In addition, healthy marriage may enhance their children‘s social competence
through positive parenting. Study results may inform marriage and relationship education in
terms of program delivery and evaluation.
Implications
This study demonstrated appropriate use of the healthy marriage/healthy relationships
(HM/HR) measure for diverse low-income couples with children. This short-form measurement
would be helpful to local HMI programs and other marriage relationship education programs for
project evaluation. Specifically, the measure could be used as a baseline at the beginning of the
program as well as a posttest or follow-up survey at the end of the program to analyze outcomes
and report on program impacts related to couple relationships, parenting, and children‘s social
competence. This type of measurement extends the measure of objectives beyond the acquisition
133

of knowledge/skills, changed attitudes, behavioral intentions, and satisfaction with training and
has the potential to demonstrate program impacts more powerfully. An evaluation using this
measurement could also identify areas needed for quality improvement by pinpointing the
HM/HR constructs that do not seem to significantly change after training. This measurement
could help researchers and evaluators understand the effects of personal characteristics
(antecedents) on couple interaction within social and economic environments. As this research
demonstrated, personality characteristics such as depression and stress can hinder positive couple
interaction. Considering the role of such factors is essential in the prevention and treatment of
relationship distress. Understanding the linkages between risk factors and couple interaction is
critical to identifying strategies for helping low-income couples function better in real-life
challenging environments. For example, the contextual influence of stress has strong effects on
marital quality. Both chronic and acute stresses (i.e. job loss, car breakdown, or eviction) can be
destructive to couple interaction, especially for low income people. The inclusion of risk and
even protective factors as antecedents to healthy marriage/healthy relationships is essential when
identifying targets for intervention.
In addition, marital virtues, like forgiveness and teamwork are important in defining and
measuring HM/HR. Marriage and relationship education curriculum would benefit from the
inclusion of lessons on marital virtues. Attention towards marital virtues can help explain not
only effective communication skills, but more so what needs to be communicated. This type of
information can be incorporated into marital therapy and marriage education by helping couples
consider not only their concerns and desires but also the needs of their partners.
The study illustrated the importance of viewing healthy marriage and parenting behavior
together, as they play a role in children‘s development. This is especially applicable when
134

marriages/relationships are weighed down by possible stress and depression because being able
to parent well is made more difficult. Healthy marriage curriculum delivered to low-income
couples with children should most certainly be adapted to include information about parenting.
In addition, because the effects of depression and stress on parenting can partly be explained
through healthy marriage/healthy relationships, the strong parent involvement component of the
Head Start program would benefit from continued inclusion of marriage relationship education
and other supportive services. Especially at a time when HHS, OHS Healthy Marriage Initiative
demonstration programs are approaching a 5-year term limit, it is important to consider
sustainability. The Administration for Children and Families (ACF) is proposing that in 2011,
Congress establish a sizeable fund to competitive three-year grants to states for: (a) responsible
fatherhood initiatives, including those with a marriage component, that rely on strong
partnerships with community-based organizations and for (b) comprehensive family selfsufficiency demonstrations that seek to improve child and family outcomes by addressing the
employment and self-sufficiency needs of parents with serious barriers to self sufficiency. These
efforts will encourage implementation of proven and promising strategies to help fathers and
mothers succeed as working parents. Because of the indirect effect of fathers‘ risk factors
(depression and stress) on fathering through healthy marriage, responsible fatherhood initiatives
would benefit with an emphasis on healthy marriage/healthy relationships.
Limitations
Several limitations of this study should be acknowledged. First, the measurement tool
that was used included healthy marriage construct items that were measured using different
scales. For example, the emotional intimacy variable was constructed by taking the mean from
one item measured on a 4-point Likert agreement scale and the mean from two items measured
135

on a 4-point frequency scale. It would be better for all items measuring the same construct to be
measured using the same scale although the testing of these scale combinations established
satisfactory reliability. Second, the smaller sample size presented limitations on what could be
accomplished using SEM. Instead path analyses used composite scores to test the HM/HR
model by race and marital status. In addition, the smaller sample size restricted the analysis of
healthy marriage antecedents to depression and stress instead of utilizing additional antecedents
in the same model like general health, alcohol and drug dependency, and religiosity. To preserve
sample size, other demographic variables like income were not used because missing values
could not be estimated and those with missing information on income would have to be removed
from analysis. Third, the unfortunate removal of the one-item fidelity variable did not allow for
the true comprehensive definition of healthy marriage. This measured fidelity variable did fit
well in the context of the other variables. Fourth, although this study used an ecological model
as a basis, it was not able to be fully tested. The Child Trends, Inc. measurement model of
healthy marriage is a linear model whereby antecedents lead to healthy marriage which leads to
adult well-being and adult well-being influences child well-being. In the model, marriage is
healthy when it promotes the individual well-being of parents and their children. The purpose of
this research was to test this model. This limited the consideration of bi-directional paths, for
example, how children‘s well-being may influence parenting or healthy marriages. As
hypothesized by Belsky, however, the influence of contextual subsystems of social support (i.e.
healthy marriage/healthy relationship) is greater than the influence of child characteristics on
parental functioning. According to Belsky‘s model, risk characteristics in the child are relatively
easy to overcome, given that either healthy marriage or parenting is not at risk (Belsky, 1984).

136

Finally, depression and stress were chosen as two risk factors influencing healthy
marriage and this association was quite strong for both genders, but especially for females.
Other risks that may play a different role for males were not considered. Waller and Swisher
(2006) examined how risk factors (i.e. physical abuse, substance abuse, and incarceration) are
related to father involvement and relationship status among unmarried couples using data from
the Fragile Families and Child Wellbeing Study. They found that fathers with risk factors are
less likely to have romantic relationships with mothers. Specifically, there are lower odds of
being married or cohabitating versus having no relationship (controlling for other variables in the
model) with the following associations: physical abuse (51.4% lower odds of being married and
60.6% lower odds of cohabitating verses having no relationship); substance use (66.1% lower
odds of being married and 56.6% lower odds of cohabitating verses having no relationship); and
incarceration (78.3 percent lower odds of being married and 73.5% lower odds of cohabitating
verses having no relationship). Fathers who have been physically abusive and fathers who have
been incarcerated since the child's birth were significantly less involved with their children than
fathers without such risk factors. They also found that parent relationships mediate associations
between risk factors and fathers' involvement. For example, analysis of qualitative data showed
that one reason why physically abusive fathers are not more involved with their children is that
they are no longer in relationships with the mother and that mothers' distrust of physically
abusive fathers mediates fathers' daily activities with children. Although this current study
assessed only depression and stress as risk factors, it did demonstrate that these two risk factors
have significant effects on healthy marriage and parenting for both males and females.
Even with these limitations, this study represents a starting point in testing the healthy
marriage model and identifies future directions of research.
137

Future Research
This research focused on children‘s social development within the Child Trends, Inc.
HM/HR measurement model, but future research can certainly study this model as to how
HM/HR predicts other areas of children‘s well-being like cognitive attainment and academic
achievement. Future studies may also consider how negative marital interactions may influence
children indirectly by decreasing the effectiveness of specific parenting skills, like emotion
coaching. In addition, other risk factors like substance dependency, incarceration, and children
from a previous marriage, can be added when testing the model.
Research on bi-directional family processes is advantageous: the bi-directional
relationship between parent-child and between parenting-marriage/relationships. The influence
of child characteristics on parenting and healthy marriage/healthy relationships, the influence of
parenting behaviors on marriage, and the influence of marital relations on health and mental
health were not considered in this present study but is a direction for future research because this
study focused on the prediction of children‘s well-being. There is support for the influence of a
couple‘s relationship on mental (i.e. Heidemarie, Hyoun, and Capaldi (2009) and physical health
(i.e. Kiecolt-Glaser & Newton, 2001) this dissertation research highlights the influence of mental
health on marital/partner relations. The bi-directional testing of mental health and couple
interaction exist (i.e. Pruchno, Wilson-Genderson, & Cartwright, 2009) and is supported by the
Child Trends, Inc. measurement model. The field would benefit from additional studies using a
bi-directional approach especially for low-income diverse populations.
Studies looking to replicate this dissertation research could further distinguish marital
status into subcategories (i.e. married, remarried, cohabitating, romantic nonresidential, married
but separated). This would allow for an examination of differences in how couple relationship
138

dynamics influence children‘s well-being among parents in different relationship categories. The
dissertation research assessed the HM/HR model between race and between marital status within
gender. Further research may want to explore these demographics across gender. For example,
unmarried women may have different trajectories than married men.
Future research may also want to further analyze healthy marriage measure at the
indicator level to find greater dimensionality for grouping indicators into subsets. This could
also assist with gaining insight into which particular subsets may have greatest impact on
parenting and child well-being for establishing measurement weights.
Research Using Dyadic Datasets
The research results provide some insight into which healthy marriage variables may be
considered ―shared‖ between couples and thus, nonidependent of each other. The seven HM/HR
variables that had correlations of .50 and therefore determined as nonindependent variables were:
emotional intimacy, communication, quality time, teamwork, conflict frequency, stability and
lack of domestic violence (DV). Testing a variety of healthy marriage constructs for nonindependence is a benefit because future research may utilize this information to design
measurements for dyadic analysis. Data were not treated dyadically in this dissertation research
because the focus was on identifying factor structure and the relationships between the constructs
and child outcomes. Future work, though, would likely yield interesting insights as to the ways
in which mothers‘ and fathers‘ own perceptions and their perceptions of the other‘s beliefs about
marriage and parenting are related to child outcomes.
Pursuant to such differences in perceptions, it is interesting to note that some of the
healthy marriage constructs for males and females were independent of each other (in that were
not highly correlated). Interestingly, those variables deemed independent of each other (with
139

correlations less than .50) were couple commitment, child commitment, forgiveness, and
relationship happiness. Specifically, men tended to have slightly more positive ideas about the
future of their relationship than women, were more inclined to rate their partner more positively
in terms of commitment to the Head Start child, and reported higher scores of relationship
happiness than women. These findings are consistent with Lichter and Carmalt (2009).
Although, they compared means across gender and not necessarily within the couple dyad, they
found that the mean scores for husbands on couple commitment was significantly higher than for
their wives, that husbands were significantly more satisfied with their spouse‘s commitment to
the children than were wives and that husbands marital satisfaction score was significantly
higher than for wives. Like this dissertation research, Lichter and Carmalt (2009) did not find
significant differences between males and females on communication, emotional intimacy, nor
domestic violence (they used two of the same DV items as this study and reversed both for a
measure of ―positive marital conflict‖). They did not include quality time, teamwork, conflict
frequency, stability, or forgiveness in their research.
Results from this research showed that women tended to rate men as more ―forgiving‖.
Males‘ higher scores on commitment and relationship happiness may be related to their higher
ratings of forgiveness. Cross-sectional and longitudinal studies have demonstrated that
forgiveness is positively associated with marital longevity, relationship satisfaction and
relationship commitment (Finkel, Rusbult, Kumashiro, & Hannon, 2002; Paleari, Regalia, &
Fincham, 2005; Tsang, McCullough & Fincham, 2006). Although, most studies have found that
gender does not have an effect on forgiveness (Macaskill, Maltby & Day 2002; Toussaint &
Webb, 2005; Friesen, Fletcher, & Overall, 2005), there are a few studies that found gender
differences related to forgiveness (i.e. Orathinkal, Vansteenwegen & Burggraeve, 2008,
140

Kachadourian, Fincham & Davilla, 2004, Shackelford, Buss & Bennett, 2002). For example, a
cross-sectional study found that married women were significantly more likely to forgive than
married men (p < .005, one-tailed, d = .20; Orathinkal, Vansteenwegen, & Burggraeve, 2008).
Further, Kachadourian, and colleagues (2004) explored the effects that the tendency to forgive
and transgression severity have on forgiveness for married couples. Their research found that
wives, who had a greater tendency to forgive, did so regardless of the severity of the offense. In
comparison, husbands, who had a greater tendency to forgive, did so for only severe
transgressions and not minor wrongdoings (Kachadourian et al., 2004). This finding suggests
that the perceived severity of an offense has a greater influence on forgiveness for males than for
females.
Conclusion
In Bronfenbrenner‘s (1978, 1992) ecological model, marital relations are part of a child‘s
microsysytem. Healthy marriage/relationships directly effect children as well as indirectly effect
children through parenting (second-order effects). Healthy marriage/relationships can act as a
social support system for parenting. This is especially true in the context of stress. This study
reinforced the importance of measuring children‘s well-being in the context of
marriage/partnership as well as the parent-child relationship. In addition, it supported the
definition of marriages/intimate relationships as ―healthy‖ because of the established paths
leading to positive parenting and child well-being. Furthermore, the MDRC measure of healthy
marriage was developed for use by married couples. This research made adjustments to the
measure so that it would be inclusive of unmarried couples. In addition, this research made
contributions by studying the association between a comprehensive measure of healthy
marriage/relationships, parenting and child well-being for a diverse sampling of low-income
141

Head Start mothers and fathers at the local level. Interventions, like HMIs, are commendable for
their focus on strengthening family well-being and the new funding with an additional focus on
addressing the employment and self-sufficiency needs of parents will also help to address
families in poverty. Being hopeful for the developing child in poverty requires recognition of
poverty as a national social problem that affects us all and to which we are all collectively called
to respond.

142

REFERENCES

143

REFERENCES
Acs, G. (2007). Can we promote child well-being by promoting marriage? Journal of Marriage
and Family 69(5), 1326-1344.
Administration for Children and Families, Healthy Marriage Initiative. (n.d.). What is HMI.
Retrieved August 8, 2009 from
http://www.acf.hhs.gov/healthymarriage/about/mission.html#background
Administration for Children and Families, Office of Planning Research and Evaluation (n.d.).
Measuring Couple Relationships: Overview. Retrieved August 8, 2009 from
http://www.acf.hhs.gov/programs/opre/strengthen/meas_relation/meas_relation_overview.ht
ml
Afifi, T. O., MacMillan, H., Cox, B. J., Asmundson, G. J. G., Stein, M. B., & Sareen, J. (2009).
Mental health correlates of intimate partner violence in marital relationships in a nationally
representative sample of males and females. Journal of Interpersonal Violence, 24(8), 13981417.
Ahrons, C. (1981). The continuing coparental relationship between divorced spouses. American
Journal of Orthopsychiatry, 51, 415 428.
Amato, P. R. (2000). The consequences of divorce for adults and children. Journal of Marriage
and the Family, 62, 1269-1287.
Amato, P. R. (2005). The impact of family formation change on the cognitive, social, and
emotional well-being of the next generation. The Future of Children, 15(2), 75-96. Retrieved
October 8, 2009 from http://www.jstor.org/pss/3556564
Amato, P. R. (2008). Recent changes in family structure: implications for children, adults, and
society. National Healthy Marriage Resource Center (NHMRC) Brief. Retrieved October
21, 2009 from http://www.healthymarriageinfo.org/docs/May08changefamstructure.pdf
Amato, P. R., & Booth, A. (1997). A generation at risk: Growing up in an era of family
upheaval. Cambridge, MA: Harvard University Press.
Amato, P. R. & Previti, D. (2003). People‘s reasons for divorcing: gender, social class, the life
course, and adjustment. Journal of Family Issues 24, 602–626.
Amato, P. R., & Rogers, S. J. (1999). Do attitudes toward divorce affect marital quality? Journal
of Family Issues, 20(1), 69–86.
Amodio, D. M., & Showers, C. J. (2005). Similarity breeds liking‘ revisited: The moderating role
of communication. Journal of Social and Personal Relationships, 22, 817–836.
Anderson, E. A. (2003). Measurement and family demography. Unpublished memo
commissioned by Child Trends, Inc.
144

Anderson, L. S. (2008). Predictors of parenting stress in a diverse sample of parents of early
adolescents in high-risk communities. Nursing Research, 57(5), 340-350.
Arias, I., Lyons, C. M., & Street, A. E. (1997). Individual and marital consequences of
victimization: Moderating effects of relationship efficacy and spouse support. Journal of
Family Violence, 12, 193-210.
Aron, A., Norman, C. C., McKenna, C., & Heyman, R. E. (2000). Couples‘ shared participation
in novel and arousing activities and experienced relationship quality. Journal of Personality
and Social Psychology, 78, 273-284.
Artis, J. E. (2007). Maternal cohabitation and child well-being among kindergarten children.
Journal of Marriage and Family, 69, 222–236.
Atkins, D. C., Baucom, D. H., & Jacobson, N. S. (2001). Understanding infidelity: Correlates in
a national random sample. Journal of Family Psychology, 15, 735-749.
Attala, J. M., Bauza, K., Pratt, H., & Vieira, D. (1995). Integrative review of effects on children
of witnessing domestic violence. Issues in Comprehensive Pediatric Nursing, 18(3), 163-172.
Axinn, W. G., & Thornton, A. (1993). Mothers, children, and cohabitation: The intergenerational
effects of attitudes and behavior. American Sociological Review, 58(2), 233-246.
Axinn, W. G. & Thornton, A. (2000). The transformation in the meaning of marriage. In L. J.
Waite (Ed.), The ties that bind: Perspectives on marriage and cohabitation (pp. 147-165).
New York: Aldine de Gruyter.
Bancroft, L., & Silverman, J. G. (2004). Assessing abusers‘ risks to children. In P. G. Jaffe, L. L.
Baker, & A. J. Cunningham (Eds.), Protecting children from domestic violence: Strategies
for community intervention (pp. 101–119). New York: Guilford.
Baron, R. M., Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173– 1182.
Baumrind, D. (1973). The development of instrumental competence through socialization. In
A.D. Pick (Ed.), Minnesota symposium on child psychology (Vol. 7, pp. 3–46). Minneapolis,
MN: University of Minnesota Press.
Behrman, R. E. & Quinn, L. S. (1994). Children and divorce: Overview and analysis. The
Future of Children 4(1), 4-14.
Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55, 8396.

145

Bender, E. (2005). Depression Treatment in Black Women Must Consider Social Factors.
Psychiatric News, 40(23), 14. Retrieved December 4, 2010 from
http://pn.psychiatryonline.org/content/40/23/14.1.full
Bentler, P M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107,
238-246.
Bentler, P. (1995). Structural equations program manual. Encino: Multivariate Software.
Bentler, P. M., & Bonett, D. G. (1980). Significant tests and goodness of fit in the analysis of
covariance structures. Psychological Bulletin, 88(3)588-606.
Berkowitz, L. (1989). Frustration–aggression hypothesis: Examination and reformulation.
Psychological Bulletin, 106, 59-73.
Blanchard, V. L., Hawkins, A. J., Baldwin, S. A., & Fawcett, E. B. (2009). Investigating the
effects of marriage and relationship education on couples‘ communication skills: A metaanalytic study. Journal of Family Psychology, 23, 203-214.
Blum, J. S., & Mehrabian, A. (1999). Personality and temperament correlates of marital
satisfaction. Journal of Personality, 67, 155–162.
Booth, A., Amato, P., Johnson, D., & Rogers, S. (2003). Conceptualizing and measuring healthy
marriages and positive relationships. Unpublished memo commissioned by Child Trends.
Bornstein, M. H., & Bradley, R. H. (Eds.) (2003). Socioeconomic status, parenting, and child
development. Mahwah, NJ: Erlbaum.
Bradley & Corwyn (2002). Socioeconomic status and child development. Annual Review of
Psychology, 53, 371- 399. Linver, M. R., Brooks-Gunn, J., & Kohen, D. (2002). Family
processes as pathways from income to young children‘s development. Developmental
Psychology, 5, 719–734.
Bradbury, T. N., Fincham, F. D., & Beach, S. R. H. (2000). Research on the nature and
determinants of marital satisfaction: A decade in review. Journal of Marriage and the
Family, 62, 964-980.
Bradbury, T. N., & Karney, B. R. (2004). Understanding and altering the longitudinal course of
marriage. Journal of Marriage and Family, 66, 862–879.
Brandwein, R. A. (1998). Battered women, children, and welfare reform: The ties that bind.
Thousand Oaks, CA: Sage.
Briggs, S. R. & Cheek, J. M. (1986). The role of factor analysis in the development and
evaluation of personality scales. Journal of Personality, 54(1), 106-148.

146

Brody, G. H. & Ge, X. (2001). Linking parenting processes and self-regulation to psychological
functioning and alcohol use during early adolescence. Journal of Family Psychology, 15, 82–
94.
Brody, G. H., Arias, I., & Fincham, F. D. (1996). Linking marital and child attributions to family
processes and parent– child relationships. Journal of Family Psychology, 10, 408–421.
Bronte-Tinkew, J., Guzman, L., Jekielek, S., Moore, K. A., Ryan, S., Redd, Z., Carrano, J., &
Matthews, G. (2004). Conceptualizing and measuring „Healthy Marriages‟ for empirical
research and evaluation studies: A review of the literature and annotated bibliography.
Prepared for the U.S. Department of Health and Human Services, Administration for
Children and Families. Washington, DC: Child Trends.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen
& J. S. Long (Eds.), Testing structural equation models (pp. 136-162). Newsbury Park, CA:
Sage.
Brown, S. L. (2000). Union transitions among cohabitors: The significance of relationship
assessments and expectations. Journal of Marriage and the Family 62, 833-846.
Brown, S. L. (2004). Family structure and child well-being: The significance of parental
cohabitation. Journal of Marriage and Family, 66, 351–367.
Bryant, C. M., & Wickrama, K. A. S. (2005). Marital and intimate relations among Blacks: A
contextual approach. In K. A. Dodge, V. McLoyd, & N. Hill (Eds.), Emerging issues in
Black family life: Context, adaptation, and policy. New York: Guilford Press.
Bryant, C. M., Wickrama, K. A. S., Bolland, J., Bryant, B. M., Cutrona, C. E., & Stanik, C. E.
(2010). Race matters, even in marriage: identifying factors linked to marital outcomes for
African Americans. Journal of Family Theory and Review, 2, 157-174. DOI:10.1111/j.17562589.2010.00051.x
Buehler, C., & Gerard, J.M. (2002). Marital conflict, ineffective parenting, and adolescents‘
maladjustment. Journal of Marriage and the Family, 64, 78–92.
Buehlman, K. T., Gottman, J. M., & Katz, L. F. (1992). How a couple views their past predicts
their future: Predicting divorce from an oral history interview. Journal of Family Psychology,
5, 295-318.
Bumpass, L. L. & Lu, H. H. (2000). Trends in cohabitation and implications for children's family
contexts in the United States. Population Studies 54, 29-41.
Burstein, N. R., Lindberg, L. D., Fein, D. J., Page, L. & LaRock, R. J. (2003, March). Guide to
data sources on the determinants of marriage and cohabitation. Bethesda, Maryland: Abt
Associates, Inc
147

Cabrera, N. J., Shannon, J. D., & Tamis-LeMonda, C. (2007). Fathers' Influence on Their
Children's Cognitive and Emotional Development: From Toddlers to Pre-K. Applied
Developmental Science, 11(4), 208-213.
Cabrera, N., Shannon, J. D., West, J., & Brooks-Gunn, J. (2006). Parental interactions with
Latino infants: Variation by country of origin and English proficiency. Special Issue on Race,
Ethnicity, and Culture in Child Development, 74, 1190–1207
Cabrera, N., Tarkow, A., & Shannon, J. (2006). Fathers‟ and mothers‟ contribution to their
preschoolers‟ emotional regulation. Presented at World Association of the Infant Mental
Health, July, Paris, France.
Caldera, Y. M., & Lindsey, E. W. (2006). Coparenting, mother-infant interaction, and infantparent attachment relationships in two-parent families. Journal of Family Psychology, 20(2),
275-283.
Campbell, J. C., & Lewandowski, L. A. (1997). Mental and physical health effects of intimate
partner violence on women and children. Psychiatry Clinic North America, 20(2), 353-374.
Cano, A., & O‘Leary, K.D. (2000). Infidelity and separations precipitate major depressive
episodes and symptoms of nonspecific depression and anxiety. Journal of Consulting and
Clinical Psychology, 68, 774–781.
Carasso, A., & Steuerle, C. E. (2005). The hefty penalty on marriage facing many households
with children. The Future of Children, 15(2), 157–175.
Carlson, M. J. & Högnäs, R. S. (2009). Coparenting in fragile families. Princeton University,
Woodrow Wilson School of Public and International Affairs, Center for Research on Child
Wellbeing. Working Paper # 1188. Retrieved January 13, 2010 from
http://crcw.princeton.edu/workingpapers/WP09-13-FF.pdf
Carlson, M. J., & McLanahan, S. S. (2006). Strengthening unmarried families: Could enhancing
couple relationships also improve parenting? Social Service Review, 80, 297-321.
Carlson, M., McLanahan, S., & England, P. (2004). Union formation in fragile families.
Demography 41, 237-261.
Carrere, S., Buehlman, K. T., Gottman, J. M., Coan, J. A., & Ruckstuhl, L. (2000). Predicting
marital stability and divorce in newlywed couples. Journal of Family Psychology, 14, 42-58.
Carroll, J. S. & Doherty, W. J. (2003). Evaluating the effectiveness of premarital prevention
programs: A meta-analytic review of outcome research. Family Relations, 52, 105-118.
Catalano, S. (2007). Intimate partner violence in the United States. U.S. Department of Justice,
Bureau of Justice Statistics. Retrieved January 20, 2010 from
http://www.ojp.usdoj.gov/bjs/intimate/ipv.htm.
148

Centers for Disease Control and Prevention. (2008, February). Adverse health conditions and
health risk behaviors associated with intimate partner violence, morbidity and mortality
weekly report. Retrieved January 20, 2010 from
www.cdc.gov/mmwr/preview/mmwrhtml/mm5705a1.htm.
Chao, R., & Kanatsu, A. (2008). Beyond socio-economics: Explaining ethnic group differences
in parenting through cultural and immigration processes. Applied Developmental Science,
12(4), 181–187.
Cherlin, A. J., Burton, L. M., Hurt, T. R., & Purvin, D. M. (2004). The influence of physical and
sexual abuse on marriage and cohabitation. American Sociological Review 69, 769-789.
Cherlin, A. J., Cross-Barnet, C., Burton, L. M., & Garrett-Peters, R. (2008). Promises they can
keep: Low-income women's attitudes toward motherhood, marriage, and divorce. Journal of
Marriage and Family, 70, 919-933.
Christensen, S. A., & Miller, R. B. (2006). Areas of desired change among married midlife
individuals. Journal of Couple & Relationship Therapy, 5, 35-57.
Ciabattari, T. (2004). Cohabitation and housework: The effects of marital intentions. Journal of
Marriage and Family, 66(1), 118–125.
Clark, M. S. (1988). Interpersonal processes in close relationships. Annual Review of
Psychology, 39, 609-672.
Clarkberg, M., Stolzenberg, R. M., & Waite, L. J. (1995). Attitudes, values, and the entrance into
cohabitational unions. Social Forces, 74, 609-634.
Clements, M. L., Stanley, S. M., & Markman, H. J. (2004). Before they said ―I do‖:
Discriminating among marital outcomes over 13 years based on premarital data. Journal of
Marriage and Family, 66, 613-626.
Comrey, A. L. (1988). Factor-analytic methods of scale development in personality and clinical
psychology. Journal of Consulting and Clinical Psychology, 56, 754-761.
Conduct Problem Prevention Research Group (CPPRG) (1995). Psychometric properties of the
social competence scale teacher and parent ratings (Fast Track Project technical report).
University Park, PA: Pennsylvania State University.
Cordova, J. V., Gee, C. G., and Warren, L. Z. (2005). Emotional skillfulness in marriage:
intimacy as a mediator of the relationship between emotional skillfulness and marital
satisfaction. Journal of Social and Clinical Psychology, 24, 218-235.
Cowan, C. P., & Cowan, P. A. (2003). The conceptualization and measurement of „Healthy
Marriages/Couple Relationships‟. Unpublished memo commissioned by Child Trends.

149

Cowan, P. A., Cowan, C. P., Pruett, M. K., & Pruett, K. D. (2009). Promoting fathers‘
engagement with children: Preventative interventions for low-income families. Journal of
Marriage and Family, 71, 663 – 679.
Cowan, P. A., Cowan, C. P., Schulz, M. S., & Heming, G. (1994). Prebirth to preschool family
factors in children's adaptation to kindergarten. In R. D. Parke & S. G. Kellam (Eds.),
Exploring family relationships with other social contexts. Family research consortium:
advances in family research (Vol. 4, pp. 75-114). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Crnic K, & Low C. (2002). Everyday stresses and parenting. In Bornstein M (ed.), vol. 5,
Handbook of Parenting: Practical Issues in Parenting (2nd edn) (pp. 243 – 267). Mahwah,
NJ: Lawrence Erlbaum Associates:
Cummings, E. M., & Davies, P. T. (2002). Effects of marital conflict on children: Recent
advances and emerging themes in process-oriented research. Journal of Child Psychology
and Psychiatry, 43, 31–63.
Cutrona, C. E., Russell, D., Abraham, W., Gardner, K., Melby, J., Bryant, C., & Conger, R.
(2003). Neighborhood context and financial strain predictors of marital interaction and
marital quality in African American couples. Journal of Personal Relationships, 10, 389 –
409.
Deal, J. E., Wampler, K. S., & Halverson. (1992). The importance of similarity in the marital
relationship. Family Process, 31, 369–382.
DeKlyen M, Brooks-Gunn J, McLanahan S, & Knab J. (2006). The mental health of married,
cohabiting, and non-coresident parents with infants. American Journal of Public Health, 96,
1836-1841.
Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977). Maximum likelihood from incomplete
data via the EM algorithm (with discussion). Journal of the Royal Statistical Society Serial B
39, 1-38.
Diamond, L. M. (2003). Love matters: Romantic relationships among sexual-minority
adolescents. In P. Florsheim (Ed.), Adolescent romantic relations and sexual behavior:
Theory, research, and practical implications (pp. 85–107). Mahwah, N.J.: Lawrence
Erlbaum Associates.
Dindia, K. & Timmerman, L. (2003). Accomplishing romantic relationships. In Brant R.
Burleson & John O. Greene (Eds.) Handbook of Communication and Social Interaction Skills
(pp. 685–721). Mahwah, NJ: Lawrence Erlbaum.
Dion, R. M. (2005). Healthy marriage programs: learning what works. The Future of Children
15, 139–156.

150

Doss, B. D., Rhoades, G. K., Stanley, S. M., & Markman, H. J. (2009). The effect of the
transition to parenthood on relationship quality: an 8-year prospective study. Journal of
Personality and Social Psychology, 96(3), 601-619.
Duck, S., & Barnes, M. (1992). Disagreeing about agreement: Reconciling differences about
similarity. Communication Monographs, 59, 199-208.
Duncan, G. J.,& Magnuson, K. A. (2005).Can family socioeconomic resources account for racial
and ethnic test score gaps. Future of Children, 15, 35 – 54.
Dunifon, R., & Kowaleski-Jones, L. (2002). Who‘s in the house? Race differences in
cohabitation, single parenthood, and child development. Child Development, 73, 1249–1264.
Early Childhood Learning and Knowledge Center. OHS Healthy Marriage Initiative Grants
retrieved from:
http://eclkc.ohs.acf.hhs.gov/hslc/Family%20and%20Community%20Partnerships/New%20P
arental%20Involvement/Healthy%20Marriage/famcom_ann_00315_100307.html
Eckstein, D., & Goldman, A. (2001). The Couple‘s Gender-Based Communication Questionnaire
(CGCQ).The Family Journal: Counseling and Therapy for Couples and Families, 16, 83-86.
Edin, K. (2000). Few good men: Why poor women don‘t remarry. The American Prospect, 11,
112-133.
Edin, K. (2000). What do low-income mothers say about marriage? Social Problems, 47(1), 112133.
Edin, K. & Kefalas, M. J. (2005). Promises I can keep: Why poor women put motherhood before
marriage. Berkeley: University of California Press.
Edin, K. & Reed, J. M. (2005). Why don't they just get married? Barriers to marriage among the
disadvantaged. Future of Children, 15(2), 117–37.
Edin, K., England, P., & Linnenberg, K. (2003, September). Love and distrust among unmarried
parents. Paper presented at National Poverty Center Conference: Marriage and Family
Formation Among Low Income Couples: What do we Know From Research? Ann Arbor, MI.
Ellis, B. J., & Malamuth, N. M. (2000). Love and anger in romantic relationships: A discrete
systems model. Journal of Personality, 68, 525–556.
Ellwood, D. T. & Jencks, C. (2004). The uneven spread of single-parent families: What do we
know? Where do we look for answers? In K. Neckerman (Ed.), Social Inequality (pp. 3-77).
New York: Russell Sage.
Erel, E., & Burman, B. (1995). Interrelatedness of marital relations and parent–child relations: A
meta-analytic review. Psychological Bulletin, 118, 108–132.
151

Ewing, J.A. (1984). Detecting alcoholism: The CAGE questionnaire. Journal of the American
Medical Association, 252, 1905–1907.
Fagan, P. F. (2003). Healthy marriage measures. Unpublished memo commissioned by Child
Trends.
Fein, D. (2003). Determinants of marriage and cohabitation among disadvantaged Americans:
Two themes. Presented at the Marriage and Family Formation among Low-Income Couples:
What Do We Know from Research? Georgetown University: National Poverty Center.
Fein, D. J. (2004). Married and poor: Basic characteristics of economically disadvantaged
married couples in the U .S. Working Paper SHM-01, Supporting Healthy Marriage Project,
New York: MDRC.
Fein, D. J., Burstein, N. R., Fein, G. G., & Lindberg, L. D. (2003). The determinants of marriage
and cohabitation among disadvantaged Americans: Research findings and needs. Final
Report, Marriage and Family Formation Data Analysis Project. Abt Associates, Cambridge,
MA.
Fincham, F. D., & Beach, S. R. H. (2002). Forgiveness in marriage: Implications for
psychological aggression and constructive communication. Personal Relationships, 9, 239–
251.
Fincham, F. D., Bradbury, T. N., Arias, I., Byrne, C. A., & Karney, B. R. (1997). Marital
violence, marital distress, and attributions. Journal of Family Psychology, 11(3), 367-372.
Fincham, F. D., & Hall, J. H. (2005). Parenting and the marital relationships. In: Luster T,
Okagaki L, editors. Parenting: An ecological perspective (pp. 205–233). Mahwah, NJ:
Erlbaum.
Fincham, F., Stanley, S., & Beach, S. (2007). Transformative processes in marriage: An analysis
of emerging trends. Journal of Marriage and Family, 69, 275 - 292.
Finkel, E. J., Rusbult, C. E., Kumashiro, M., & Hannon, P. A. (2002). Dealing with betrayal in
close relationships: Does commitment promote forgiveness? Journal of Personality and
Social Psychology, 82, 956–974.
Firestone, R. W., & Firestone, L. (2004). Methods for overcoming the fear of intimacy. In D.
J.Mashek & A. P. Aron (Eds.), Handbook of closeness and intimacy (pp. 375–395). Mahwah,
NJ: Lawrence Erlbaum Associates.
Fishman, E. A., & Meyers, S. A. (2000). Marital satisfaction and child adjustment: Direct and
mediated pathways. Contemporary Family Therapy, 22(4), 437-452.
Floyd, F. J., Gilliom, L. A., & Costigan, C. L. (1998). Marriage and parenting alliance:
Longitudinal prediction of change in parenting perceptions and behaviors. Child
Development, 69(5), 1461-1479.
152

Fowers, B. J. (2000). Beyond the myth of marital happiness: How embracing the virtues of
loyalty, generosity, justice, and courage can strengthen your relationship. San Francisco:
Jossey-Bass.
Fowers, B. J. (2001). The limits of a technical concept of a good marriage: exploring the role of
virtue in communication skills. Journal of Marital and Family therapy, 27(3), 327-340.
Fowers, B. J. (2003). Conceptualizing and measuring healthy marriages and positive
relationships. Unpublished memo commissioned by Child Trends. Retrieved August 14,
2009 from http://www.childtrends.org/Files/CT_HealthyMarriage.pdf
Fowers, B. J., Bucker, J., Calbeck, K. B., & Harrigan, P. (2003). How do social scientists define
a good marriage? Unpublished manuscript.
Friesen, M. D., & Fletcher, G. J. O. (2007). Evidence for a distinct forgiveness prototype:
Convergent and discriminant validity. Personal Relationships, 14, 209-223 from
http://www.shatterdmen.com/DV%20homicides%20cut%20in%20half.htm
Funk J. L., & Rogge R. D. (2007). Testing the ruler with item response theory: Increasing
precision of measurement for relationship satisfaction with the couples satisfaction index.
Journal of Family Psychology, 21, 572–583.
Furstenberg, F. F., Jr. (2001). The fading dream: Prospects for marriage in the inner city. In E.
Anderson & D. Massey (Eds.), Problem of the Century: Racial Stratification in the U.S. (pp.
222-247). New York: Russell Sage.
Furstenberg, F. F., Belzer, A., Davis, C., Levine, J. A., Morrow, K., Washington, M., & Wilson,
W. J. (ed) (1993) Sociology and the public agenda (pp. 231-258). Sage Publications,
Newbury Park.
Garcia-Moreno, C., Jansen, H. A., Ellsberg, M., Heise, L., & Watts, C. H. (2006). Prevalence of
intimate partner violence: Findings from the WHO multi-country study on women‘s health
and domestic violence. Lancet, 368, 1260-1269.
Garfinkle, I., & McLanahan, S. (2000). Fragile families and child well-being: a survey of new
parents. Focus 21(1), 9-11.
Gaunt, R. (2006). Couple similarity and marital satisfaction: Are similar spouses happier?
Journal of Personality, 74(5), 1–20.
Gibson-Davis, C. (2009). Money, marriage, and children: Testing the financial expectations and
family formations theory. Journal of Marriage and Family, 71(1), 146-160.
Gibson-Davis, C. M., Edin, K., & McLanahan, S. (2005). High hopes but even higher
expectations: The retreat from marriage among low-income couples. Journal of Marriage
and Family, 67, 1301-1312.
153

Glass, S. P. (2002). Couple therapy after the trauma of infidelity. In A.S. Gurman & N.S.
Jacobson (Eds.), Clinical handbook of couple therapy (3rd ed., pp. 488–507). New York:
Guilford.
Goodwin, P., McGill, B., Chandra, A. Who marries and when? Age at first marriage in the
United States, 2002. NCHS data brief, no 19. Hyattsville, MD: National Center for Health
Statistics. 2009. Retrieved November 1, 2009 from
http://www.cdc.gov/nchs/data/databriefs/db19.htm
Goodwin, R. (1992). Overall. just how happy are you? The magical question 31 of the Spanier
Dyadic Adjustment Scale. Family Therapy, 19. 273-275.
Gordon, K. C., & Baucom, D. H. (2003). Forgiveness and marriage: Preliminary support for a
synthesized model of recovery from a marital betrayal. American Journal of Family Therapy,
31, 179–199.
Gordon, K. C., Hughes, F. M., Tomcik, N. D., Dixon, L. J., & Litzinger, S. C. (2009). Widening
spheres of impact: The role of forgiveness in marital and family functioning. Journal of
Family Psychology, 23(1), 1-13.
Gottman, J. M. (1994). What predicts divorce? The relationship between marital processes and
marital outcomes. Hillsdale, NJ: Erlbaum.
Gottman, J. M., Coan, J., Carrerre, S., & Swanson, C. (1998). Predicting marital happiness and
stability from newlywed interactions. Journal of Marriage and the Family, 60, 5-22.
Gottman, J. M., & Notarius, C. I. (2000). Decade review: Observing marital interaction. Journal
of Marriage and the Family, 62, 927-947.
Graefe, D.R. & Lichter D.T. (2002). Marriage among unwed mothers: whites, blacks, and
Hispanics compared. Perspectives on Sexual and Reproductive Health 34, 286-293.
Graefe, D. R., & Lichter, D. T. (2008). Marriage patterns among unwed mothers: Before and
after PRWORA. Journal of Policy Analysis and Management, 27(3), 479-497.
Graham-Bermann, S. A., & Edleson, J. L. (Eds.). (2001). Domestic violence in the lives of
children: The future of research, intervention, and social policy. Washington, DC: American
Psychological Association.
Grych, J. H., & Fincham, F. D. (2001). Interparental conflict and children‘s adjustment: An
overview. In J. H. Grych & F. Fincham (Eds.), Child Development and Interparental
Conflict. New York: Cambridge University Press.
Guzzo, K. B. (2009). Marital intentions and the stability of first cohabitations. Journal of Family
Issues 30(2), 179-205.

154

Halberstadt, A. G., Crisp, V. W., & Eaton, K. L. (1999). Family expressiveness: A retrospective
and new directions for research. In P. Philippot, R. S. Feldman, & E. Coats (Eds.), The social
context of nonverbal behavior (pp. 109–155). New York: Cambridge University Press.
Hall, J. H. & Fincham, F. D. (2006). Relationship dissolution following infidelity: the roles of
forgiveness and attributions. Journal of Social and Clinical Psychology, 25(5), 508-522.
Retrieved December 1, 2009, from Psychology Module. (Document ID: 1060410471).
Hargrave, T. D. (2000). What does it mean to grow strong families? Journal of the Texas
Association for Marriage and Family Therapy. 5, 13-19.
Harknett, K. (2008). Mate availability and unmarried parent relationships. Demography, 45(3),
555-571.
Harknett, K. & McLanahan, S. S. (2004). Racial and ethnic differences in marriage after the birth
of a child. American Sociological Review, 69, 790-811.
Hassebrauck, M., & Fehr, B. (2002). Dimensions of relationship quality. Personal Relationships,
9, 253–270.
Hawkins, A. J., Blanchard, V. L., Baldwin, S. A., & Fawcett, E. B. (2008). Does marriage and
relationship education work? A meta-analytic study. Journal of Consulting and Clinical
Psychology.
Hawkins, A. J., & Fackrell, T. A. (2010). Does couple education for low-income couples work?
A meta-analytic study of emerging research. Journal of Couple & Relationship Therapy, 9,
181-191.
Hawkins, A. J., Fowers, B. J., Carroll, J. S., & Yang, C. (2007). Conceptualizing and measuring
marital virtues. In S. L. Hofferth, & L. M. Casper (Eds.), Handbook of measurement issues in
family research (pp. 67-83). Mahwah, NJ: Lawrence Erlbaum Associates.
Hawkins, A & Ooms, T. (2010). What works in marriage and relationship education? A review
of lessons learned with a focus on low-income couples. National Healthy Marriage
Resource Center. Retrieved on October 29, 2010 from:
http://www.acf.hhs.gov/healthymarriage/pdf/whatworks_edae.pdf
Heller, P., & Wood, B. (1998). The process of intimacy: Similarity, understanding and gender.
Journal of Marital and Family Therapy, 24, 273-288.
Henry, R. G., Miller, R. B., & Giarrusso, R. (2005). Difficulties, disagreements, and
disappointments in late-life marriages. International Journal of Aging and Human
Development, 61, 243-264.
Hill, H. D. (2007). Steppin' out: Infidelity and sexual jealousy among unmarried parents. In P.
England & K. Edin (Eds.), Unmarried Couples with Children (pp. 104-132). New York:
Russell Sage Foundation.
155

Hill, N. E. , Bush, K. R. and Roosa, M. W. (2003) Relations between parenting and family
socialization strategies and children's mental health: Low income, Mexican American and
Euro-American mothers's and children's perspectives. Child Development, 74, 189-204.
Hoff, E., Laursen, B., & Tardif, T. (2002). Socioeconomic status and parenting. In: Bornstein
MH, editor. Handbook of parenting 2nd ed. (pp. 231–252). Mahwah, NJ: Lawrence Erlbaum
Association.
Hofferth, S. L. & Casper, L. M. (2007). Handbook of measurement issues in family research.
Mahwah, NJ: Lawrence Erlbaum Associates Publishers.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.
Hu, L. T., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to
underparameterized model misspecification. Psychological Methods, 3, 424-453.
Hunter, W. M., Cox, C. E., Teagle, S., Johnson, R. M., Mathew, R., Knight, E. D., & Leeb, R. T.
(2003). Measures for assessment of functioning and outcomes in longitudinal research on
child abuse. Volume 1: Early Childhood. Retrieved December 12, 2009 from the
LONGSCAN web site http://www.iprc.unc.edu/longscan/
Huth-Bocks, A. C., & Hughes, H. M. (2008). Parenting stress, parenting behavior, and children‘s
adjustment in families experiencing intimate partner violence. Journal of Family Violence,
23, 243-251.
Jaffe, P. G., Baker, L. L., & Cunningham, A. J. (2004). Protecting children from domestic
violence: Strategies for community intervention. New York: Guilford.
Jekielek, S. M., Moore, K. A., Carrano, J., Matthews, G. (2003). Conceptualizing and measuring
„healthy marriages‟ for empirical research and evaluation studies: recommendation memos
from experts in the field (Task Two). Washington, DC : Child Trends.
Jekielek, S. J., Moore, K. A., Carrano, J., & Matthews, G. (2004). Conceptualizing and
measuring „healthy marriages‟ for empirical research and evaluation studies: executive
summary. Washington, D.C.: Child Trends, Inc. Retrieved November 1, 2009 from
http://www.childtrends.org/PDF/HealthyMarriage/RecommendationMemos.pdf
Johnson, S. M. (1996). The practice of emotionally focused marital therapy, creating connection.
New York, NY: Brunner/Mazel
Johnson, M. P. (1999, November). Two types of violence against women in the American family:
Identifying patriarchal terrorism and common couple violence. Paper presented at the
National Council on Family Relations annual meetings, Irvine, CA.

156

Johnson, M. P. (2001). Conflict and control: Symmetry and asymmetry in domestic violence. In
A. Booth, A. C. Crouter & M. Clements (Eds.), Couples in Conflict (pp. 95-104). Mahwah,
NJ: Lawrence Erlbaum Associates.
Johnson, W. E. (2003). Conceptualization and measurement of positive couple relationships.
Unpublished memo commissioned by Child Trends.
Johnson, M. P. (2006). Violence and Control in Intimate Relationships: Intimate Terrorism and
Other Types of Domestic Violence. Boston: Northeastern University Press.
Johnson, M. P. (in press). Conflict and control: Gender, symmetry, and asymmetry in domestic
violence. Violence Against Women, 12(11), 1-16.
Johnson, M. P., Conklin, V., & Menon, N. (2002, November). The effects of different types of
domestic violence on women: Intimate terrorism vs. situational couple
Johnson, M. P., & Leone, J. M. (2005). The differential effects of intimate terrorism and
situational couple violence: Findings from the national violence against women survey.
Journal of Family Issues, 26, 322–349.
Johnson, W. E., Shears, J. & Perry, A. R. (in press). Paternal involvement among African
American fathers: Evidence from the early head start and fragile families and child wellbeing
study. Research on Social Work Practice.
Joreskog, K. G., Sorbom, D. (1984), LISREL VI User's Guide, Scientific Software, Mooresville,
IN.
Kachadourian, L. K., Fincham, F. D., & Davila, J. (2005). Attitudinal ambivalence, rumination
and forgiveness of partner transgressions in marriage. Personality and Social Psychology
Bulletin, 31, 334-342.
Kamp Dush, C. M. (2009). The association between family of origin structure and instability and
mental health across the transition to adulthood. In H. E. Peters & C. M. Kamp Dush (Eds.)
Marriage and family: Complexities and perspectives. New York City, NY: Columbia
University Press.
Karney, B. R., & Bradbury, T. N. (1995). The longitudinal course of marital quality and stability:
A review of theory, methods, and research. Psychological Bulletin, 118, 3-34.
Karney, G., Garvan, C., & Thomas, M. (2003). Family formation in Florida: 2003 baseline
survey of attitudes, beliefs, and demographics relating to marriage and family formation.
University of Florida.
Katz, L. F., & Low, S. M. (2004). Marital violence, co-parenting, and family-level processes in
relation to children's adjustment. Journal of Family Psychology, 18(2), 372-382.

157

Kelly, A. B., Fincham, F. D., & Beach, S. R. H. (2003). Communication skills in couples: A
review and discussion of emerging perspectives. In J. O. Green & B. R. Burleson (Eds.)
Handbook of communication and social interaction skills (pp. 723-751). Mahwah, New
Jersey: Lawrence Erlbaum.
Kenny, D. A. (2008). Reflections on mediation. Organizational Research Methods, 11, 353–358.
Kenny, D. A., Kashy, D. A., & Cook W. L. (2007). Dyadic data analysis. New York NY:
Guilford Press.
Kiecolt-Glaser, J. K., & Newton, T. L. (2001). Marriage and health: His and hers. Psychological
Bulletin, 127, 472–503.
Kitzmann, K. M. (2000). Effects of marital conflict on subsequent triadic family interactions and
parenting. Development and Psychopathology, 36, 3–13.
Kline R. B. (1998). Principles and practice of structural equation modeling. Guilford Press, New
York, NY.
Knox, V. & Fein, D. (2008). Designing a marriage education demonstration and evaluation for
low-income married couples. MDRC.
Koball H. L., Moiduddin E., Henderson J., Goesling B., Besculides M. (2010). What do we
know about the link between marriage and health? Journal of Family Issues, 31(8), 10191040
Kolbo, J. R., Blakely, E. H., & Engleman, D. (1996). Children who witness domestic violence: A
review of empirical literature. Journal of Interpersonal Violence, 11(2), 281-293.
Krishnakumar, A. & Buehler, C. (2000). Interparental conflict and parenting behaviors: A metaanalytic review. Family Relations, 49, 25–44.
Lamb, M. E. (2004). The Role of the Father in Child Development. New York: John Wiley &
Sons, Inc.
Lareau, A. (2003). Invisible Inequality: Social Class and Child Rearing in Black and White
Families. American Sociological Review, 67 (October), 747-776.
Larson, J. H., Hammond, C. H., & Harper, J. M. (1998). Perceived equity and intimacy in
marriage. Journal of Marital and Family Therapy, 24(4), 487-506.
Laurenceau, J. P., Feldman Barrett, L., & Rovine, M. J. (2005). The interpersonal process model
of intimacy in marriage: A daily-diary and multilevel modeling approach. Journal of Family
Psychology, 19, 314–323.

158

Laurenceau, J. P., Stanley, S. M., Olmos-Gallo, A., Baucom, B. & Markman, H. J. (2004).
Community-based prevention of marital dysfunction: Multilevel modeling of a randomized
effectiveness study. Journal of Consulting and Clinical Psychology, 72, 933–943.
Lawrence, E., & Bradbury, T. N. (2001). Physical aggression and marital dysfunction: A
longitudinal analysis. Journal of Family Psychology, 15(1), 135-154.
Leone, J. M., Johnson, M. P., & Cohan, C. L. (2003, November). Help-seeking among women in
violent relationships: Factors associated with formal and informal help utilization. Paper
presented at the National Council on Family Relations annual meetings, Vancouver, British
Columbia.
Leone, J. M., Johnson, M. P., Cohan, C. M., & Lloyd, S. (2001, June). Consequences of different
types of domestic violence for low-income, ethnic women: A control based typology of malepartner violence. Paper presented at the International Network on Personal Relationships,
Prescott, Arizona.
Lerman, R. I., Acs, G., & Bir, A. (2007). An economic framework and selected proposals for
demonstrations aimed at strengthening marriage, employment, and family functioning
outcomes. Urban Institute. Retrieved December 3, 2009 from
http://www.urban.org/publications/411655.html
Levinger, G. (1979). A social exchange view on the dissolution of pair relationships. In R. L.
Burgess & T. L. Huston (Eds.), Social exchange in developing relationships (pp. 169 – 193).
New York: Academic Press.
Lewis, J. M. & Gossett, J. T. (1999). Disarming the past: How an intimate relationship can heal
old wounds. Redding, CT: Zeig, Tucker & Theisen, Inc.
Lewis, R., & Spanier, G. (1979). Theorizing about the quality and stability of marriage. In W. R.
Burr, R. Hill, F. I. Nye, & I. Reiss (Eds.), Contemporary theories about the family (vol. 1, pp.
268–294). New York: Free Press.
Lichter, D. T., & Carmalt, J. H. (2009). Religion and marital quality among low-income couples.
Social Science Research, 38, 168–187.
Lichter, D. T. & Graefe, D. R. (2001). Finding a mate? The marital and cohabitation histories of
unwed mothers. In L.L. Wu & B. Wolfe (Eds.), Out of Wedlock: Causes and Consequences
of Nonmarital Fertility (pp. 317-344). New York: Russell Sage Foundation.
Lichter, D. T., Graefe, D. R., & Brown, R. B. (2003). Is marriage a panacea? Union formation
among economically disadvantaged unwed mothers. Social Problems 50, 60-86.
Lichter, D. T., Qian, Z., & Mellott, L. (2006). Marriage or dissolution? Union transitions among
poor cohabiting women. Demography 43(2), 223-240.

159

Lincoln, K. D., & Chae, D. H. (2001). Stress, marital satisfaction, and psychological distress
among African Americans. Journal of Family Issues, 31(8), 1081-1105.
Locke, H. J., Wallace, K. M. (1959). Short marital-adjustment and prediction tests: Their
reliability and validity. Marriage and Family Living, 21, 251–255.
Lovejoy, M. C., Weis, R., O‘Hare, E., & Rubin, E. C. (1999). Development and initial
validation of the Parent Behavior Inventory. Psychological Assessment, 11, 534-545.
Luthar, S. (1999). Poverty and children's adjustment. Los Angeles: Sage.
Macaskill, A., Maltby, J., & Day, L. (2002). Forgiveness and Emotional Empathy. The Journal
of Social Psychology, 142, 663-665.
MacKinnon D. P., Lockwood CM, Hoffman JM, West SG, Sheets V. (2002). A comparison of
methods to test mediation and other intervening variables. Psychological Methods, 7, 83–
104.
MacKinnon D. P., Warsi G, Dwyer J. H. (1995). A simulation study of mediated effect
measures. Multivariate Behavioral Research, 30, 41–62.
Magnusson, K. A., & Duncan, G. J. (2002). Parents in poverty. In M. H.Bornstein (Ed.),
Handbook of parenting (2nd ed., pp. 95–121). Mahwah, NJ: Erlbaum.
Manning, W. D., & Lamb, K. A. (2003). Adolescent well-being in cohabiting, married and
single-parent families. Journal of Marriage and Family, 65, 876–893.
Margolin, G., Gordis, E. B., & John, R. S. (2001). Coparenting: A link between marital conflict
and parenting in two-parent families. Journal of Family Psychology, 15(1), 3-21.
Marshall, L. L., Weston, R., & Honeycutt, T. C. (2000). Does men's positivity moderate or
mediate the effects of their abuse on women's relationship quality? Journal of Social &
Personal Relationships, 17(4-5), 660-675.
Martin, S. P. (2006). Trends in marital dissolution by women‗s education in the United States.
Demographic Research 15, 537-560.
Mason, C., Walker-Barnes, C., Tu, S., Simons, J. & Martinez-Arrue, R. (2004). Ethnic
differences in the affective meaning of parental control behaviors. Journal of Primary
Prevention 25, 59-79.
Mauldon, J., London, R., Fein, D. and Bliss, S. (2002) What Do They Think? Welfare Recipients‟
Attitudes Towards Marriage and Childbearing, Research Brief Welfare Reform and Family
Formation Project No. 2, Bethesda, MD: Abt Associates.
McCall, M. (1995). Orientation, outcome, and other-serving attributions. Basic and Applied
Social Psychology, 17, 49-64.
160

McClintock-Comeaux, M., Anderson, E. A., & Kuvalanka, K. A. (2004). TANF and marriage
education: Utilizing marriage legislation to design a new educational curriculum that meets
the needs of low-income families. Vision 2004: What is the Future of Marriage? (pp. 77-83).
Minneapolis, MN: National Council on Family Relations.
McGinnis, S. L. (2003). Cohabiting, dating, and perceived costs of marriage: A model of
marriage entry. Journal of Marriage and Family, 65, 105-116.
McGroder, S. M. (2000). Parenting among low-income African American single mothers with
preschool-age children: Patterns, predictors, and developmental correlates. Child
Development, 71(3) 752-771.
McLanahan, S. (2002). Life without father: What happens to children? Context. 1(1), 35-44.
McLanahan, S., & Sandefur, G. (1996). Growing up with a single parent: What hurts, what
helps. Cambridge, MA: Harvard University Press.
McLoyd, V. C. (1990). The impact of economic hardship on Black families and children:
Psychological distress, parenting, and socioemotional development. Child Development, 61,
311-346.
McLoyd, V. C. (1998). Socioeconomic disadvantage and child development. American
Psychologist, 53, 185–204.
McNeal, C., & Amato, P. (1998). Parents‘ marital violence: Long-term consequences for
children. Journal of Family Issues, 19, 123–139.
Meyers, L. S., Gamst, G., & Guarino, A. J. (2006). Applied multivariate research design and
interpretation. Thousand Oaks, CA: Sage.
Mills, T. L., & Combs, Y. J. (2002). Environmental factors, income inequality, and health
disparity: Emerging research and policy implications for Black Americans. African American
research perspectives: an occasional report of the Program for Research on Black
Americans, 8(2), 29 – 39.
Mincy, R. (2002). Who should marry whom? Multiple partner fertility among new parents.
Center for Research on Child Well-being Working Paper No. 2002-03. Princeton, NJ:
Princeton University. Retrieved November 1, 2009 from
http://crcw.princeton.edu/fragilefamilies
Mincy, R., Pouncy, H., Reichert, D., & Richardson, P (2004). Fragile Families in Focus: A
Look at How Never Married Low-Income Parents Perceive Marriage and Relationships.
Baton Rouge, LA, TANF Executive Office of Oversight and Evaluation.
Monte, L. M. (2007). Blended but not the Bradys: Navigating unmarried multiple partner
fertility. In P. England & K. Edin (Eds.), Unmarried couples with children (pp. 183- 203).
New York: Russell Sage.
161

Moore, K., Jekielek, S., Bronte-Tenkew, J., Guzman, L., Ryan, S., & Redd, Z. (2004). What is
healthy marriage? Defining the concept. Washington, DC: Child Trends. Retrieved July 24,
2009 from http://158.71.31.30/healthymarriage/pdf/Child_Trends-2004.pdf
Moore, K. A., Bronte-Tinkew, J., Jekielek, S., Guzman, L., Ryan, S., Redd, Z., Carrano, J.,
Matthews, G. (2007). Developing measures of healthy marriages in relationships. In S. L.
Hofferth, & L. M. Casper (Eds.), Handbook of measurement issues in family research.
Mahwah, NJ: Lawrence Erlbaum Associates.
Morrill, M. I., Hines, D. A., Mahmood, S., Cordova, J. (2010). Pathways between marriage and
parenting for wives and husbands: The role of coparenting. Family Process, 49(1), 59-73.
Moss, B. F., & Schwebel, A. I. (1993). Marriage and romantic relationships. Family Relations,
42, 31-37.
Mulaik, A., & Millsap, R. E. (2000). Doing the four-step right. Structural equation modeling: A
Multidisciplinary Journal, 7, 36–73.
Murry, V. M., Brown, P. A., Body, G. H., Cutrona, C. E., & Simons, R. L. (2001). Racial
discrimination as a moderator of the links among stress, maternal psychological functioning,
and family relationships. Journal of Marriage and the Family, 63, 915 – 926.
National Healthy Marriage Resource Center. (2010). Administration for Children and Families
Healthy Marriage Initiative, 2002-2009: An Introductory Guide. January 2010, available at:
http://www.healthymarriageinfo.org/docs/ACFGuideto09.pdf
Nock, S. L. (1995). A comparison of marriages and cohabiting relationships. Journal of Family
Issues, 16, 53-76.
Ooms, T. (2002). Strengthening couples and marriage in low income communities. In A. J.
Hawkins, L. D. Wardle, & D. O. Coolidge (Eds.), Revitalizing the institution of marriage for
the twenty-first century: An agenda for strengthening marriage (pp. 79–100). Westport, CT:
Praeger.
Ooms, T. (2007). Adapting healthy marriage programs for disadvantaged and culturally diverse
populations: what are the issues? Couples and marriage series, No. 10. Center for Law and
Social Policy.
Ooms, T., & Wilson, P. (2004). The challenges of offering marriage and relationship education
to low-income populations. Family Relations, 53, 440–447.
Orathinkal, J., Vansteenwegen, A., & Burggraeve, R. (2008). Are demographics important for
forgiveness? The Family Journal, 16, 20-27.
Osborne, C., & McLanahan, S. (2007). Partnership instability and child well-being. Journal of
Marriage and Family, 69(4), 1065-1083. Retrieved December 1, 2009, from Research
Library Core. (Document ID: 1386963771).
162

Owens, E. & Shaw, D. (2003). Poverty and early childhood adjustment. In S.S. Luthar (Ed.),
Resilience and vulnerabilities: Adaptation in the context of childhood adversities (pp.243266). New York: Cambridge University Press.
Paleari, F. G., Regalia, C., & Fincham, F. D. (2005). Marital quality, forgiveness, empathy, and
rumination: A longitudinal analysis. Personality and Social Psychology Bulletin, 31, 368–
378.
Paleari, G., Regalia, C., & Fincham, F.D. (2005). Marital quality, forgiveness, empathy, and
rumination: A longitudinal analysis. Personality and Social Psychology Bulletin, 31, 368378.
Patrick, S. & Beckenbach, J. (2009). Male perceptions of intimacy: Qualitative study. Journal of
Men's Studies, 17(1), 47.
Patrick, S., Sells, J. N., Giordano, F. G., & Tollerud, T. R. (2007). Intimacy, differentiation, and
personality variables as predictors of marital satisfaction. The Family Journal: Counseling
and Therapy for Couples and Families, 15, 359-367.
Pickhardt, C. (2003). Strengths of the Single Parent. Adoption.com Library. Retrieved on
November 1, 2010 from: http://library.adoption.com/articles/strengths-of-the-singleparent.html
Pinderhughes, E. E. , Dodge, K. A. , Bates, J. E. , Petitt, G. S. & Zelli, A. (2000). Discipline
responses: Influences of parents' socioeconomic status, ethnicity, beliefs about parenting,
stress, and cognitive-emotional processes. Journal of Family Psychology, 14, 380-400.
Pinderhughes, E. E., Hurley, S. & the Conduct Problems Prevention Research Group. (2008).
Disentangling ethnic and contextual influences among parents raising youth in high-risk
communities. Applied Developmental Science, 12(4), 211–219.
Pleck, J. H., & Masciadrelli, B. (2004). Paternal Involvement by U.S. Residential Fathers:
Levels, Sources, and Consequences. In M. E. Lamb (Ed.), The Role of the Father in Child
Development (pp. 222-271). Hoboken, NJ: Wiley & Sons.
Power TG. (2004). Stress and coping in childhood: The parents‘ role. Parenting, 4, 271–317.
Preacher, K. J., & Hayes, A. F. SPSS Macro For Multiple Mediation. Written by Andrew F.
Hayes, The Ohio State University. Retrieved on November 4, 2010 from:
http://www.comm.ohio-state.edu/ahayes/
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and
comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879
891.
Presser, H. B. (2000). Nonstandard work schedules and marital instability. Journal of Marriage
and the Family, 62, 93-110.
163

Previti, D., & Amato, P. R. (2004). Is infidelity a cause or a consequence of poor marital quality?
Journal of Social and Personal Relationships, 21, 217-230.
Pruchno, R., Wilson-Genderson, M., and Cartwright, F.P. (2009). Depressive symptoms and
marital satisfaction in the context of chronic disease: a longitudinal dyadic analysis. Journal
of Family Psycholgy, 23(4), 573-584.
Qian, Z., Lichter, D. T., & Mellott, L. (2005). Out-of-wedlock childbearing, marital prospects
and mate selection. Social Forces, 84, 473-491.
Ragan, E. P., Einhorn, L. A., Rhoades, G. K., Markman, H. J., & Stanley, S. M. (2009).
Relationship education programs: Current trends and future directions. In J. H. Bray & M.
Stanton (Eds.), Handbook of family psychology. Hoboken, NJ: Wiley-Blackwell.
Raley, R. K. & Bumpass, L. L. (2003). The topography of the plateau in divorce: Levels and
trends in union stability after 1980. Demographic Research, 8, 246-258.
Raver, C. C., Gershoff, E. T., & Aber, J. L. (2007) Testing Equivalence of Mediating Models of
Income, Parenting, and School Readiness for White, Black, and Hispanic Children in a
National Sample. Child Development 78:1, 96–115.
Rector, R. E., Pardue M. G., & Noyes, L. R. (2003). „Marriage Plus‟: Sabotaging the president's
efforts to promote healthy marriage. Washington, D.C.: The Heritage Foundation. Retrieved
September 8, 2009 from http://www.heritage.org/Research/Welfare/BG1677.cfm.
Reed, J. (2005, August). How and why do unmarried parents break-up? Paper presented at the
annual meeting of the American Sociological Association, Philadelphia, PA.
Reis, H. T., & Gable, S. L. (2003). Toward a positive psychology of relationships. In C. L. M.
Keyes, J. Haidt, C. L. M. Keyes & J. Haidt (Eds.), Flourishing: Positive psychology and the
life well-lived. (pp. 129-159). Washington, DC US: American Psychological Association.
Rhoades, G. K., Stanley, S. M. & Markman, H. J. (2009). Couples‘ reasons for cohabitation:
Associations with individual well-being and relationship quality. Journal of Family Issues,
30, 233–46.
Roberts, P. (2007). Out of order? Factors influencing the sequence of marriage and childbirth
among disadvantaged americans, Center for Law and Social Policy, http://www.clasp.org.
Rogge, R. D., & Bradbury, T. N. (1999). Till violence does us part: The differing roles of
communication and aggression in predicting adverse marital outcomes. Journal of
Consulting and Clinical Psychology, 67, 340-351.
Rusbult, C. E., Bissonnette, V. I., Arriaga, X. B., Cox C. L. (1998). Accommodation processes
across the early years of marriage. In T. N. Bradbury (Ed.) The Developmental Course of
Marital Dysfunction. (pp. 74-113). New York: Cambridge University Press.
164

Rusbult, C. E., Kumashiro, M., Finkel, E. J., & Wildschut, T. (2002). The war of the roses: An
interdependence analysis of betrayal and forgiveness. In P. Noller & Feeney, J. A. (Eds.),
Understanding marriage: Developments in the study of couple interaction (pp. 251–281).
New York: Cambridge University Press.
Rusbult, C. E., Verette J., Whitney, G. A., Slovik, L. F., & Lipkus I. (1991). Accommodation
processes in close relationships: theory and preliminary empirical evidence. Journal of
Personality and Social Psychology, 60, 53-78.
Ryan, R., Martin, A., & Brooks-Gunn, J. (2006). Is one good parent good enough? Patterns of
mother and father parenting and child cognitive outcomes at 24 and 36 months. Parenting, 6,
211–228.
Sassler, S. (2004). The process of entry into cohabiting unions. Journal of Marriage and Family
66, 491-505.
Schneller, D. P., & Arditti, J. A. (2004). After the break-up: Interpreting divorce and rethinking
intimacy. Journal of Divorce and Remarriage, 42, 1–37.
Schoen, R., Rogers, S. J., & Amato, P. R. (2006). Wives‘ employment and spouses‘ marital
happiness: Assessing the direction of influence using longitudinal couple data. Journal of
Family Issues, 27, 506-528.
Schoppe, S. J., Mangelsdorf, S. C., & Frosch, C. A. (2001). Coparenting, family process, and
family structure: Implications for preschoolers' externalizing behavior problems. Journal of
Family Psychology, 15(3), 526-545.
Schoppe-Sullivan, S. J., Mangelsdorf, S. C., Frosch, C. A., & McHale, J. L. (2004). Associations
Between Coparenting and Marital Behavior From Infancy to the Preschool Years. Journal of
Family Psychology, 18(1), 194-207.
Schroder, K. E. E., Carey M. P., Vanable P. A. (2003). Methodological challenges in research on
sexual risk behavior: II. Accuracy of self-reports. Annals of Behavioral Medicine, 26, 104–
123.
Shackelford, T. K., Buss, D. M., & Bennett, K. (2002). Forgiveness or breakup: Sex differences
in responses to a partner‘s infidelity. Cognition and Emotion, 16, 299–307.
Shapiro, S. E., Lasarev,M. R. & McCauley, L. (2002) Factor analysis of Gulf War illness: what
does it add to our understanding of possible health effects of deployment? American Journal
of Epidemiology, 156, 578-585.
Smock, P. J., & Manning, W. D. (2003). The conceptualization and measurement of relationship
quality: Insights from a qualitative study of cohabiting young adults. Unpublished memo
commissioned by Child Trends.

165

Smock, P. J., Manning, W. D., & Porter, M. (2005). Everything's there except money.' how
money shapes decisions to marry among cohabitors. Journal of Marriage and Family 67,
680-696.
Snyder, D.K., & Schenider, W.J. (2002) Affective reconstruction: A pluralistic, developmental
approach. In A. Gurman & N. Jacobson (Eds.), Clinical Handbook of Couple Therapy (151179). New York: Guilford Press.
Sobel ME. (1982). Asymptotic confidence intervals for indirect effects in structural equation
models. In: S. Leinhardt (Ed.) Sociological Methodology. Washington, DC: American
Sociological Association.
Squires, J. (2004). Parent completed screening for social emotional problems in young children:
The effects of risk/disability status and gender on performance. Infant Mental Health
Journal, 25(1), 62-73.
Stanley, S. (2004). What is a Healthy Marriage? Excerpt from testimony given before the Senate
Finance Committee, May 5, 2004. Retrieved October 8, 2009 from www.clasp.org
Stanley, S. M. (2007). Assessing couple and marital relationships: beyond form and toward a
deeper knowledge of function. In S. L. Hofferth & L. M. Casper (Eds.), Handbook of
Measurement Issues in Family Research (pp. 85–99). Mahwah, NJ: Lawrence Erlbaum
Associates.
Stanley, S. M., Amato, P. R., Johnson, C. A., & Markman, H. J. (2006). Premarital education,
marital quality, and marital stability: Findings from a large, random, household survey.
Journal of Family Psychology, 20, 117-126.
Stanley, S. M., Markman, H. J. & Whitton, S.W. (2002). Communication, conflict and
commitment: Insights on the foundations of relationship success from a national survey.
Family Process, 41(4), 659-675.
Stanley, S. M., Whitton, S. W., & Markman, H. J. (2004). Maybe I do: Interpersonal
commitment and premarital or nonmarital cohabitation. Journal of Family Issues, 25, 496–
519.
Stanley, S. M., & Markman, H. J. (1992). Assessing commitment in personal relationships.
Journal of Marriage and the Family, 54, 595-608.
Strauss, M. A. (1992). Children as witnesses to marital violence: A risk factor for lifelong
problems among a nationally representative sample of American men and women. Report of
the Twenty-Third Ross Roundtable. Columbus, OH: Ross Laboratories
Strauss, M. A., & Gelles, R. J. (1990). Physical violence in American families. New Brunswick,
NJ: Transaction Publishers.

166

Sweeney, M. M., Hongbo, W., & Videon, T. (2009). Reconsidering the association between
stepfamilies and adolescent well-being. In H.E. Peters & C. M. Kamp Dush (Eds.), Marriage
and family: Perspectives and complexities (pp. 177-225). New York: Columbia University
Press.
Sweeney, M., & Phillips, J. (2004). Understanding racial differences in marital disruption:
Recent trends and explanations. Journal of Marriage and Family, 66, 639 – 650.
Teitler, Julien O. & N. E. Reichman. (2008). Mental Illness as a Barrier to Marriage Among
Mothers with an Out-of-Wedlock Birth. Journal of Marriage and Family 70(4), 770-782.
Testa, M., & Leonard, K. E. (2001). The impact of marital aggression on women's psychological
and marital functioning in a newlywed sample. Journal of Family Violence, 16(2), 115-130.
Thomas, A., & Sawhill. (2005). For love and money? The impact of family structure on family
income. Future of Children, 15, 57-74.
Tomarken, A. J. & Waller, N. G. (2005). Structural equation modeling: Strengths, limitations,
and misconceptions. Annual Review of Clinical Psychology 1, 31-65.
Toussaint, L. & Webb, J. R. (2005). Gender differences in the relationship between empathy and
forgiveness. Journal of Social Psychology, 145, 673-685.
Toussaint, L. L., & Webb, J. R. (2005). Theoretical and empirical connections between
forgiveness, mental health, and well–being. In E. L. Worthington (Ed.), Handbook of
forgiveness (pp. 340–362). New York: Routledge.
Toussaint, L. L., Williams, D. R., Musick, M. A., & Everson, S. A. (2001). Forgiveness and
health: Age differences in a U.S. probability sample. Journal of Adult Development, 8, 249–
257.
Tsang, J., McCullough, M. E., & Fincham, F. D. (2006). The longitudinal association between
forgiveness and relationship closeness and commitment. Journal of Social and Clinical
Psychology, 25, 448-472.
Upchurch, D. M., L. A. Lillard & C. W. A. Panis. (2002). ―Nonmarital Childbearing: Influences
of Education, Marriage, and Fertility.‖ Demography, 39, 311-329.
Ventura, S. J. & Bachrach, C. A. (2000). Nonmarital childbearing in the United States, 1940-99.
National Vital Statistics Reports, 48(16). Hyattsville, MD: National Center for Health
Statistics.
Vinokur, A. D., Price, R. H., & Caplan, R. D. (1996). Hard times and hurtful partners: How
financial strain affects depression and relationship satisfaction of unemployed persons and
their spouses. Journal of Personality and Social Psychology, 71, 166 – 179.

167

Waite, L. J. & Gallagher, M. (2000). The case for marriage: Why married people are happier,
healthier, and better off financially. New York: Doubleday.
Waller M. R., McLanahan, S. S. (2005). ‗His‘ and ‗her‘ marriage expectations: determinants and
consequences. Journal of Marriage and Family, 67, 53–67.
Waller M. R., Swisher, R. (2006). Fathers‘ risk factors in fragile families: implications for
‗healthy‘ relationships and father involvement. Social Problems, 53(3), 392–420.
Waring, E. M. (1988). Enhancing marital intimacy through facilitating cognitive self-disclosure.
New York: Brunner/Mazel.
Weigle, D. J. (2003). A communication approach to the construction of commitment in the early
years of marriage: A qualitative study. The Journal of Family Communication, 3, 1–19.
Whitfield, C. L., Anda, R. F., Dube, S. R., Felittle, V. J. (2003). Violent childhood experiences
and the risk of intimate partner violence in adults: assessment in a large health maintenance
organization. Journal of Interpersonal Violence, 18(2), 166-185.
Whitton, S. W., Stanley, S. M., & Markman, H. J. (2002). Sacrifice in romantic relationships: An
exploration of relevant research and theory. In H. T. Reiss, M. A. Fitzpatrick, & A. L.
Vangelisti (Eds), Stability and change in relationship behavior across the lifespan (pp. 156181). Cambridge, England: Cambridge University Press.
Widom, C. S. (1989). Does violence beget violence? A critical reexamination of the literature.
Psychological Bulletin, 106, 3-28.
Wilcox, B. W., & Nock, S. L. (2006). What‘s love got to do with it? Equality, equity,
commitment and women‘s marital quality. Social Forces, 84, 1321–1345.
Wilcox, W. B., & Wolfinger, N. H. (2007). Then comes marriage? Religion, race, and marriage
in urban America. Social Science Research, 36, 569-589.
Williams, D. R. & Collins, C. (1995). U.S. socioeconomic and racial differences in health:
patterns and explanations. Annual Review of Sociology, 21, 349–86.
Williams, D. R., Takeuchi, D. T., & Adair, R.K. (1992). Socioeconomic status and psychiatric
disorder among blacks and whites. Social Forces, 71, 179–94.
Witvliet, C. V., Ludwig, T. E., & Vander Laan, K. L. (2001). Granting forgiveness or harboring
grudges: Implications for emotion, physiology, and health. Psychological Science, 12, 117–
123.
Wolak, J., & Finkelhor, D. (1998). Children exposed to partner violence. In J. L. Jasinski & L.
M. Williams (Eds.), Partner violence: A comprehensive review of 20 years of research (pp.
73-112). Thousand Oaks, CA: Sage.
168

Yeung, W. J., Linver, M, & Brooks-Gunn, J. (2002). How money matters for young children‘s
development: parental investment and family processes. Child Development, 73(6): 18611879.

169