FACILITATORS AND IMPEDIMENTS TO MOTHERS’ MONITORING KNOWLEDGE
AND ADOLESCENT ADJUSTMENT IN AFRICAN AMERICAN AND LATINO LOWINCOME FAMILIES
By
Tzu-Fen Chang

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

Family and Child Ecology—Doctor of Philosophy
2014

ABSTRACT
FACILITATORS AND IMPEDIMENTS TO MOTHERS’ MONITORING KNOWLEDGE
AND ADOLESCENT ADJUSTMENT IN AFRICAN AMERICAN AND LATINO LOWINCOME FAMILIES
By
Tzu-Fen Chang
African and Latino American adolescents are disproportionately represented among
adolescents in poverty. Adolescents living in poverty is at high risk for both externalizing and
internalizing behavior problems and poor academic achievement. To investigate protectors for
this specific population, this study examined facilitators and impediments to mothers’ monitoring
knowledge and whether mothers’ monitoring knowledge prevents adolescents’ internalizing and
externalizing behavior problems and delinquency among African American and Latino lowincome families. Four main questions were addressed: 1) how does mothers’ monitoring
knowledge change over time? 2) what factors impede mothers’ knowledge? 3) what factors
facilitate mothers’ knowledge? and 4) how mothers’ long-term knowledge predicts externalizing
and internalizing behavior and school delinquency in late adolescence. All the questions were
addressed by using the Three-City Study. The present study focused on the adolescents being
reared by continuous caregivers and participating in all the three waves of interviews, yielding
319 African American and 354 Latino mother-adolescent dyads.
Analyses involved four steps: 1) to conduct growth mixture modeling to identify
trajectory groups of mothers’ monitoring knowledge; 2) to conduct logistic/multinomial
logistical regression to examine relations of trajectory groups of mothers’ knowledge with
familial, community, and adaptive cultural factors; 3) to conduct multiple regression to predict
adolescents’ externalizing behavior problems and school delinquency with trajectory groups of
mothers’ knowledge; and 4) to conduct ordinal logistic regression to predict adolescents’ clinical

classification for internalizing behavioral problems with trajectory groups of mothers’
knowledge. The findings suggest that for both ethnicities, the majority of mothers maintained
monitoring knowledge in high levels over time. This study identified family routine as a
common facilitator and harsh discipline as a common impediment to mothers’ monitoring
knowledge for both ethnicities. Two facilitators were identified only for the African American
families, i.e., grandmothers’ childrearing involvement and neighborhood cohesion. Adolescent
gender difference in mothers’ monitoring knowledge was only found for the African American
families, with the mothers who rear boys being more likely to experience a sharp decline in
monitoring knowledge than those who rear girls. Positive implications of mothers’ monitoring
knowledge for developmental outcomes were found. Specifically, stable high or slowly declining
monitoring knowledge in adolescence prevents maladjustment, particularly in terms of
internalizing behavior for the African American adolescents and externalizing behavior for the
adolescents in both groups.
The present findings shed light on potential preventions and/or interventions to improve
adolescent developmental outcome and family environment of low-income African Americans
and Latinos. Social workers can advance preventions to educate low-income African American
and Latino mothers about the positive implications of gaining information about their children’s
daily life in adolescence for preventing externalizing and/or internalizing behavior problems. My
findings also indicate the potential benefit of programs that aim to promote low-income mothers
gaining monitoring knowledge, such as establishing family routine. Specifically for the African
American mothers, this study advocates developing programs aiming to strengthen mutual
supports and trust among neighbors and understand positive implications of grandmothers’
childrearing support.

Copyright by
TZU-FEN CHANG
2014

To my beloved father in heaven, mother and the older brother, and God.

v

ACKNOWLEDGEMENTS

Looking back the process of writing the dissertation, I deeply feel that the assistance of
many individuals made this project possible. First and foremost, I need to acknowledge the
contribution of my committee chair, Dr. Desiree Qin. I show my gratitude to Dr. Qin, whose
guidance, support, and patient instruction rendered me an environment to learn conducting
research and thinking problems. I also appreciate the contributions of all my committee members,
including Drs. Brent Donnellan, Steve Gold, Deborah Johnson, and Esther Onaga. Their
guidance and suggestion on research design and framework provided me with a broader view to
structure my dissertation study.
I would also like to express my thanks to a number of my close friends. On this long
journey, they listened to my difficulties, provided comments, and discussed my barriers. I am
grateful to your companionship that transformed this tough processing into a pleasurable one.
I am also greatly indebted to my parents and family members whose immense love and
contribution bolstered each step in my graduate studies. Their support and encouragement will
always give my energy to face challenges in the academic career.
Finally, I would like to deeply express my gratitude to God. You are my rock that
provides me with solid foundation to stand on my feet. You are also my lighthouse that guides
me to walk at every dark night. You give me strength so I can go through the long journey on my
academic pursuit.

vi

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... ix
LIST OF FIGURES ....................................................................................................................... xi
Chapter 1: Introduction ................................................................................................................... 1
Chapter 2: Theoretical Foundations ................................................................................................ 7
2.1 Integrative Ecological Model of Development of Minority Children .................................. 7
2.2 A Process Model of the Determinants of Parenting............................................................ 11
Chapter 3: Background Literature ................................................................................................ 17
3.1 African American Families ................................................................................................. 18
3.1.1 Cultural Values ............................................................................................................ 18
3.1.2 Family Environment .................................................................................................... 19
3.2. Latino American Families.................................................................................................. 25
3.2.1 Cultural Values ............................................................................................................ 25
3.2.2 Family Environment .................................................................................................... 26
3.3 Factors Linked to Parents’ Monitoring Knowledge............................................................ 32
3.3.1 Family Environment .................................................................................................... 32
3.3.2 Neighborhood Context ................................................................................................. 35
3.3.3 Children’s Characteristics ............................................................................................ 37
3.3.4 Adaptive Culture .......................................................................................................... 39
3.4 Parents’ Monitoring Knowledge and Child Outcomes ....................................................... 39
3.5 Chapter Summary ............................................................................................................... 42
Chapter 4: Research Methodology................................................................................................ 44
4.1 Demographics of Low-income Families in Boston, Chicago, and San Antonio ................ 44
4.1.1 Boston .......................................................................................................................... 44
4.1.2 Chicago ........................................................................................................................ 46
4.1.3 San Antonio ................................................................................................................. 47
4.2 The Sampling Procedure of the Three-City Study.............................................................. 49
4.3 The Sample of the Three-City Study .................................................................................. 50
4.4 Analysis Plan ...................................................................................................................... 53
4.4.1 Analyses of Data from the Three-City Study .............................................................. 53
4.4.2 Research Questions ...................................................................................................... 55
4.5 Measures ............................................................................................................................. 59
4.6 Measurement Invariance Tests ........................................................................................... 66
4.7 Reliabilities of the Measures for Family Environment and Neighborhood Context .......... 68
4.8 Reliabilities of the Measures for Adolescents’ Adjustment Outcomes .............................. 73
Chapter 5: Results ......................................................................................................................... 74
vii

5.1 Descriptive Statistics of the Major Variables ..................................................................... 74
5.2 Model Selection and the Specification of the Trajectory Groups for Mothers’ Monitoring
Knowledge ........................................................................................................................... 74
5.3 Description of the Trajectories of Mothers’ Monitoring .................................................... 79
5.4 Correlates of Mothers’ Monitoring Knowledge ................................................................. 82
5.4.1 Family Environment and Trajectories of Mothers’ Monitoring Knowledge ............... 84
5.4.2 Neighborhood Context and Trajectories of Mothers’ Monitoring ............................... 86
5.4.3 Adolescent Gender and Trajectories of Mothers’ Monitoring ..................................... 87
5.4.4 Adaptive Culture and Trajectories of Mothers’ Monitoring Knowledge .................... 88
5.5 Trajectories of Mothers’ Monitoring Knowledge and Adolescent Adjustment Outcomes 90
Chapter 6: Discussion ................................................................................................................... 93
6.1 Trajectories of Mothers’ Monitoring Knowledge ............................................................... 93
6.2 Family Environmental Factors and Mothers’ Monitoring Knowledge .............................. 96
6.3 Neighborhood Context and Mothers’ Monitoring Knowledge ......................................... 100
6.4 Gender and Mothers’ Monitoring Knowledge .................................................................. 103
6.5 Adaptive Culture and Mothers’ Monitoring Knowledge .................................................. 106
6.6 Mothers’ Monitoring Knowledge and Adolescent Outcomes .......................................... 109
6.7 Implications....................................................................................................................... 115
6.8 Limitation .......................................................................................................................... 118
6.9 Future Directions .............................................................................................................. 119
REFERENCES ........................................................................................................................... 121

viii

LIST OF TABLES

Table 1: Demographic Information According to Ethnicity ......................................................... 53
Table 2: Measures of Fit for Mothers’ Monitoring Knowledge and Parenting Styles/Behaviors
for Multiple-Group Confirmatory Factor Analyses Among African Americans and
Latinos Across Times ..................................................................................................... 69
Table 3: Measures of Fit for Family Routine and Neighborhood Context for Multiple-Group
Confirmatory Factor Analyses Among African Americans and Latinos Across Times 70
Table 4: Reliabilities of the Major Variables for African Americans .......................................... 72
Table 5: Reliabilities of the Major Variables for Latinos ............................................................. 72
Table 6: Means and SDs of the Major Variables for African Americans ..................................... 75
Table 7: Means and SDs of the Major Variables for Latinos ....................................................... 76
Table 8: Summary of Information for Selecting Number of Trajectory Classes for African
American Mothers .......................................................................................................... 78
Table 9: Average Latent Class Probabilities for Most Likely Latent Class Membership (Row) by
Latent Class (Column) for African American Mothers .................................................. 78
Table 10: Summary of Information for Selecting Number of Trajectory Classes for Latino
Mothers ........................................................................................................................... 79
Table 11: Average Latent Class Probabilities for Most Likely Latent Class Membership (Row)
by Latent Class (Column) for Latino Mothers ................................................................ 79
Table 12: Intercepts and Slopes of Identified Latent Classes of Mothers’ Monitoring Knowledge
........................................................................................................................................ 80
Table 13: Logits, Standard Errors of Logits, and Odds Ratios From Logistic Regression for The
Relations of Mothers’ Monitoring Knowledge with Family Factors, Neighborhood
Factors, Adaptive Cultural Factor, and Child Gender .................................................... 83

ix

Table 14: Hierarchical Multinomial Logistic Regression Assessing Interaction Between
Adolescent Gender and Types of Family Arrangements for African Americans ........ 89
Table 15: Hierarchical Multinomial Logistic Regression Assessing Interaction between
Adolescent Gender and Types of Family Arrangements Involvement for Latinos ..... 89
Table 16: Regressing Adolescents’ Externalizing Behavior Problems and School Delinquency on
the Trajectory Groups of Mothers’ Monitoring Knowledge ........................................ 92
Table 17: Regressing Adolescents’ Internalizing Behavior Problems on the Trajectory Groups of
Mothers’ Monitoring Knowledge ................................................................................. 92

x

LIST OF FIGURES

Figure 1. Garcia Coll et al.’s Integrative Model for the Study of Developmental Competencies in
Minority Children ............................................................................................................. 8
Figure 2. Belsky’s Model of Influences on Parenting Behaviors ................................................. 12
Figure 3. Concept Map for the Present Study ............................................................................... 16
Figure 4: Trajectories of mothers’ monitoring. A. Frequency for trajectories of mothers’
monitoring for African Americans. B. Frequency for trajectories of mothers’ monitoring
for Latinos. ...................................................................................................................... 81

xi

Chapter 1: Introduction
The United States has been classified as one of the highest-income economies (with
Gross National Income (GNI) per capita of more than $12,196) among many industrialized
nations (World Bank, 2011). Yet, a large proportion of American children under the age of 18
lives in poverty, with 21% living in poor families (where household incomes are less than 100%
of the federal poverty thresholds1) and 22% living in low-income families (where household
incomes are less than 200% of the federal poverty thresholds)(National Center for Children in
Poverty, 2012, June 5). Furthermore, American children are more likely to live in low-income
households compared to children in many other wealthy countries; for example, an American
child is twice to three times more likely to live in a family with income below 125% of the
poverty threshold than a child in Finland, Sweden, or Austria (Child Trends, 2009). In the U.S.,
rates of child poverty vary by race/ethnicity. African and Latino American children are
disproportionately represented among children living in low-income households. Combination of
the two ethnicities comprises of 38% of all children, but represents approximately 53% of
children living in low-income households (Addy & Wight, 2012, February; a report from
National Center for Children in Poverty). Children of the two ethnicities are at higher risk than
European and Asian American children in living in low-income households. In 2010, compared
to only 31% of European American children and 31% of Asian American children, 64% of

1

The federal poverty thresholds are cut-off points of the dollar amounts used to evaluate a person’s or a family’s
poverty status (U.S. Census Bureau, October 26, 2012). Household incomes less than 100% of the federal poverty
thresholds mean that the incomes are exactly below the thresholds; household incomes less than 200% of the federal
poverty thresholds mean that the incomes are below twice the thresholds. The thresholds are determined based on
the size of a household and ages of the members. In 2011, for example, the thresholds are $23,581 for a family of
four members and $18,106 for a family of three members; both cases have one child under the age of 18. Yet,
research indicates that in order to maintain basic living needs, families need at least twice dollar amounts of the
federal poverty thresholds (Cauthen & Fass, 2008). Accordingly, families with incomes below 200% of the federal
poverty thresholds are classified into low-income families.

1

African American children and 63% of Latino American children under the age of 18 lived in
low-income households (Addy & Wight, 2012, February).
Adolescents growing up in poverty are at high risk for both externalized and internalized
problems (e.g., McLoyd & Smith, 2002; Aneshensel & Sucoff, 1996; also see review of Bradley
& Crowyn, 2002) and poor academic achievement (see mata-analytic reviews of Sirin, 2005 and
White, 1982). These adolescents’ disposition to maladjustment is associated with their exposure
to a number of socio-contextual risk factors, such as unsafe neighborhoods and deviant peers in
school (e.g., McLoyd et al., 2009). Among many socio-contextual factors, parents’ monitoring
knowledge (i.e., “knowledge gained about children and their activities”, p.383, Racz &
McMahon, 2011) has been portrayed as one important protective factor that prevents these
adolescents from harm, thereby benefiting their adjustment, and resulting in fewer behavioral
problems and higher rates of high school graduation (see review of Racz & McMahon, 2011).
Although empirical research has noted the protective role of parents’ monitoring
knowledge of adolescents’ whereabouts and activities, a number of gaps remain in contemporary
research on parents’ knowledge about adolescents in low-income households. First, very little is
known about longitudinal associations between parents’ monitoring knowledge and adolescent
adjustment in terms of externalizing and internalizing behavior and school delinquency in lowincome households. Although some studies have examined the longitudinal role of parents’
monitoring knowledge (e.g., Kilgore, Snyder, & Lentz, 2000; Patrick et al., 2005), they focus on
different developmental courses (e.g., from early childhood to early adolescence) and only on
externalized outcomes (e.g., conduct problems). Second, since parents’ monitoring knowledge is
an important protector for adolescents in low-income households, there is a need to understand
what factors facilitate parents’ knowledge of children and their activities. However, there is a

2

significant lack of research examining the impediments and facilitators to parents’ knowledge
among this population from early to late adolescence. Comprehension of the impediments and
facilitators to parents’ monitoring knowledge in this time period will provide low-income parents
with useful strategies to prevent their adolescents from problem behaviors. Third, since Latino
and African Americans have their own cultural traditions distinctive from the majority of the US
(i.e., Whites; see review of Yasui & Dishion, 2007), in the low-income households of the two
ethnic groups there might be unique family and community assets being rooted in their cultural
traditions that can facilitate parents’ monitoring knowledge. However, previous studies on Latino
or African American low-income households primarily focus on barriers to parents’ monitoring
knowledge (Kilgore et al., 2000) or general facilitators to parents’ knowledge across different
ethnicities of low-income households (Patrick et al., 2005). The specific cultural assets that may
benefit parents’ monitoring knowledge have been underexplored. Finally, a body of research and
theories on children and adolescents in poverty mainly focus on socio-contextual risk factors for
family dynamics and child adjustment, including the examples of the family stress model
(Conger & Elder, 1994; Conger & Donnellan, 2007) and the family process model of economic
hardship and adjustment (Conger et al., 1992). Although current scholarly research has also
identified some protective factors (e.g., parents’ social networks and good parent-child
relationship qualities) for this population (e.g., Hill, Bush, & Roosa, 2003; McCabe & Clark,
1999; McLoyd, 1990a), more efforts are needed to explore other positive assets in both family
and other immediate contexts (e.g., neighborhoods) for family dynamics and child development.
To fill the gaps in existing literature, the purpose of the present study is threefold: a) to
examine patterns of trajectories of mothers’ monitoring knowledge; b) to examine the
impediments/facilitators to mothers’ knowledge, including family, neighborhood, and adaptive

3

cultural factors; and c) to examine the associations between the patterns of the trajectories of
mothers’ knowledge and adolescents’ externalizing and internalizing behavior outcomes and
school delinquency for African and Latino American low-income households. For my
dissertation study, I aim to address the following research questions:
Q1: For the first research purpose— to examine patterns of trajectories of parents’
monitoring knowledge, two respective sub-questions were addressed:
1.1) How does mothers’ monitoring knowledge change over time among adolescents in lowincome families for African and Latino Americans?
1.2) Are there different patterns of trajectories of mothers’ knowledge between African and
Latino Americans?
Q2: For the second research purpose— to examine the impediments/facilitators to these
patterns, three respective sub-questions were addressed:
2.1) What socio-contextual factors in neighborhoods and the family context, factors of
adaptive culture, and adolescents’ own characteristics impede mothers’ monitoring
knowledge over time for African and Latino Americans?
2.2) What socio-contextual factors in neighborhoods and the family context, factors of
adaptive culture, and adolescents’ own characteristics facilitate mothers’ monitoring
knowledge over time for African and Latino Americans?
2.3) Are there any differences in impediments and facilitators to mothers’ monitoring
knowledge over time between African and Latino Americans?
Q3: For the third research purpose—to examine the associations between the patterns of
the trajectories of parents’ monitoring knowledge and adolescents’ adjustment, two
respective sub-questions were addressed:

4

3.1) How is trajectory of mothers’ knowledge associated with various domains of
developmental outcomes, including externalizing and internalizing behavior and school
delinquency in late adolescence in low-income households for African and Latino
Americans?
3.2) Are there any differences in patterns of associations between African and Latino
Americans over time?
All the above questions were addressed by using a three-wave longitudinal and largescale survey dataset of the Welfare, Children & Families: A Three-City Study (Cherlin, 1999;
Cherlin et al., 2001; Angel, Burton, Chase-Lansdale, Cherlin, & Moffitt, 2009). This study
focuses on low-income households with at least one child between aged 0-4 or 10-14 in three
cities — Boston, Chicago, and San Antonio. A household was eligible if the household income
was less than 200% of the federal poverty threshold. The study focused on the child and the
child’s female caregivers. Female caregivers are referred to as mothers thereafter, given that over
90% of them were biological or adoptive mothers. At wave 1, 2,402 low-income families
participated in this study. This dataset consists of 1,160 mother-adolescent pairs at wave 1, with
46% African American, 48% Latino, and 6% European American pairs. The majority of
households (about 88%) in this study were headed by single mothers and the focus child’s
biological father did not reside in the household. For the purpose of this study, only African
American and Latino pairs will be focused. The majority of Latino adolescents (93.4%) were
U.S.-born. The original sample of adolescents aged from 10-14 years in 1999; wave 2 data were
collected between 2000 and 2001; and wave 3 data were collected between 2005 and 2006 when
the adolescents were aged from 16-20 years.

5

The present study would contribute to the understanding of family dynamics and
adolescent adjustment in African American and Latino low-income households. Empirical
research based on national data (drawn from the National Longitudinal Study of Youth 1997) has
noted that low-socioeconomic status (SES) parents of African and Latino origins are less likely
to enact effective parenting practices (e.g., failure to monitor and set limits for their children)
than their middle- or high-SES counterparts (e.g., Jones-Sanpei, Day, & Holmes, 2009); and
adolescents in these families are thereby disposed to maladjustment (see reviews of Bradley &
Crowyn, 2002 and Sirin, 2005). The present research seeks to enhance understanding of 1)
whether there are any similar factors that facilitate parents’ monitoring knowledge for both
African and Latino American low-income families over time and further benefit adolescents’
adjustment, and 2) whether African and Latino Americans bear their unique family and
community assets that can facilitate parents’ monitoring knowledge. Exploration of these
questions will provide knowledge for policy-makers and social-workers to develop culturally
sensitive prevention and intervention actions for the purposes of bettering the development and
family environment of Latino and African American adolescents living in poverty.

6

Chapter 2: Theoretical Foundations
The present study explores the potential impediments and facilitators to parents’
monitoring knowledge by investigating the socio-environmental and cultural factors for African
American and Latino adolescents. In understanding the interconnections between adolescents’
cultural and socio-environmental factors and formation of parental knowledge, this study is
grounded in two theoretical foundations: a) García Coll et al.’s integrative model (1996) for the
study of developmental competencies in minority children and b) Belsky’s process model of the
determinants of parenting (Belsky, 1984).
2.1 Integrative Ecological Model of Development of Minority Children
Integrating ecological and cultural perspectives, García Coll and colleagues (1996)
proposed an integrative model that portrays how ethnic minority children’s developmental
courses are shaped by interactions between their distal ecologies and proximal ecologies in
which children have direct contact (including the settings outside the family [communities,
schools, and health care system], adaptive culture of their ethnic groups, and family experiences)
(see Fig. 1). This model highlights that ethnic minority children’s developmental courses are
distinctive from the course of the majority (i.e., White) because they bear unique experiences of
distal and proximal ecologies. In this model, distal ecologies refer to those contexts in which
children do not have direct contact but are influenced indirectly through the mediation of
proximal ecologies. The elements of distal ecologies include social position, segregated
circumstances, and social stratification mechanisms. Social position refers to a constellation of
attributes including race, social class, ethnicity, and gender. For ethnic minority children, their
unique social position (primarily their minority status related to their race and ethnicity) leads to
experiences of social stratification mechanisms (i.e., racism, prejudice, and discrimination) and

7

#6
Child
Characteristics
Age
Temperament
Health Status
Biological Factors
Physical Characteristics

#2
Racism
Prejudice
Discrimination
Oppression

#4
Promoting/
Inhibiting
Environments
Schools
Neighborhoods
Health Care

#1
Social Position
Variables
Race
Social Class
Ethnicity
Gender

#5
Adaptive
Culture
Traditions &
Cultural Legacies
Economic &
Political
Histories
Migration &
Acculturation
Current Contextual

#3
Segregation
Residential
Economic
Social &
Psychological

#8
Developmental
Competencies
Cognitive
Social
Emotional
Linguistic
Biculturalism
Coping with Racism

#7
Family
Structure & Roles
Family Values, Beliefs & Goals

Racial Socialization
Socioeconomic Status

Figure 1. Garcia Coll et al.’s Integrative Model for the Study of Developmental Competencies in Minority Children
8

further contributes to the segregated circumstances. Three common types of segregation
circumstances for ethnic minorities include: a) residential segregation: ethnic minorities likely
concentrate in ethnic enclaves because their social position and the social stratification
mechanisms limit their geographic mobility; b) economic segregation: due to the linguistic
barrier and the cultural unfamiliarity, ethnic minorities likely engage in low-paid, low-skilled,
and labor-intense jobs; this commonly takes place among low-SES immigrants; and c) social and
psychological segregation: this segregation takes place when minorities fail to share common
experiences and values with the majority.
Segregated circumstances and social stratification mechanisms shape ethnic minority
children’s unique experiences of proximal ecologies occurring outside the family, which is
different from the experiences of the majority (i.e. White). Although ethnic majority and
minority children are exposed to similar proximal settings, including communities, schools, and
the health care system, ethnic minority children often encounter different functions and
structures because of their experiences of segregated circumstances and social stratification
mechanisms. This integrative model asserts that these proximal settings (i.e., communities,
schools, and the health care system) can promote or inhibit ethnic minority children’s
development. The extent to which proximal environments promote or inhibit their development
hinges on many factors, including resources and support provided by the environments, the
culture-match between the family and the environment, and safety of the environments. Because
these settings can promote or inhibit minority children’s development, this integrative model
called this body of settings (i.e., communities, schools, and the health care system)
promoting/inhibiting environments (see Fig. 1).

9

Ethnic minority children’s distal ecologies influence adaptive culture of their ethnic
groups and their family experiences through the mediation of promoting/inhibiting environments
(i.e., the proximal ecologies outside the family context). Ethnic minorities develop adaptive
cultures in response to their experiences of segregated circumstances and social stratification
mechanisms, as well as promoting/inhibiting environments. Adaptive culture is characterized by
traditions, economic and political history, migration and acculturation. For example, African
Americans develop one unique tradition by forming strong networks with extended family
members and community members, which helps African American parents overcome economic
hardship and balance both needs of childcare and breadwinning. In terms of family experiences,
ethnic minorities may develop their own definitions of family values, roles, and structures in
response to their experiences of distal ecologies, inhibiting/promoting ecologies, and adaptive
culture. In addition, ethnic minority children’s family experiences are shaped by their own
characteristics (e.g., age, temperament, and gender). Due to formation of adaptive cultures and
unique family experiences, ethnic minority children are likely to experience different
developmental processes from the majority.
García Coll et al.’s (1996) integrative model fills several research gaps not addressed by
previous scholarly models or theories. First, different from other development theories that are
based on cultural perspectives (e.g., genetically deficient model and culturally deficient model),
this integrative model does not view ethnic minorities’ maladaptation as indicative of cultural
deficiency, but instead views it as an adaptive solution in response to their social position and
contexts (see discussion of García Coll et al., 1996). Second, although previous ecology-based
models (e.g., transactional model of Sameroff & Chandler, 1975 and ecological theories of
Bronfenbrenner, 1977, 1979) provide general pictures of the interaction between children,

10

contexts, and development processes, these models fail to consider ethnic minority children’s
unique socio-environmental experiences (e.g., segregated circumstances), as well as the impact
of these experiences on their development. García Coll et al.’s (1996) integrative model fills this
gap. Third, although previous ecology-based models (e.g., ecological theories of Bronfenbrenner)
consider the impact of culture on child development, these models posit that cultural impacts
occur at the macro level, rather than at the micro level (e.g., in family or school context). On the
contrary, García Coll et al. (1996) argue that cultural impacts permeate all levels of social
settings, from the macro- to the micro-level. This perspective provides a more holistic picture of
the interaction between culture, context, and person for ethnic minority children because ethnic
minorities tend to develop their own adaptive cultures that play an important role in shaping their
family experiences and development.
2.2 A Process Model of the Determinants of Parenting
Belsky (1984) proposed a process model portraying what socio-environmental factors
contribute to parenting (see Figure 2). This model is also grounded in ecological perspectives,
underscoring that parenting practices are determined by parents’ socio-environmental
experiences (including marital relations, work circumstances, and social network) as well as
child characteristics. Additionally, this model asserts that parents’ developmental history
influences their parenting through the mediation of personality. In turn, parents’ personality
shapes parenting through the mediation of their marital relations and work. Borrowing the
perspective of the transactional model for child development, Belsky posits that the parent and
the child influence each other’s behaviors. Specifically, this model underscores that children’s
developmental processes are not only the direct product of parenting, but are also shaped by

11

Marital
Relations

Developmental
History

Social
Network
Child
Characteristics

Personality

Parenting

Child
Development

Work

Figure 2. Belsky’s Model of Influences on Parenting Behaviors

12

children’s own characteristics (e.g., temperament) both directly and indirectly (i.e., through the
mediation of parenting).
This model underscores that parents’ socio-environmental experiences are primary
sources of stress or support for parental functioning. In terms of support, parents’ socioenvironmental experiences (i.e., marital relations, work, and social networks) provide support for
parenting through three forms: 1) emotional support, which is exhibited through “love and
interpersonal acceptance” (p. 87, Belsky, 1984); 2) instrumental assistance, which includes
receiving advice on childcare or substantial assistance in child-rearing; and 3) social expectations,
which refer to receiving information about appropriate parenting behaviors. Each type of support
above can produce influence on parenting through direct as well as indirect pathways. Supports
going through the direct pathway bring effects targeting at parental functioning, while those
going through the indirect pathway bring effect via mediators. One example for the direct
pathway is that when parents receive advice on childcare, they directly apply it to childrearing
practices. Regarding to indirect pathway, for example, when a mother feels she is loved by her
husband, her self-esteem tends to increase and this increase in turn benefits efficacy in parenting.
In this example, the husband’s love can be thought as emotional support producing indirect
effect via the mediation of the wife’s psychological adjustment.
This model does not specify the forms of sources of stress, but highlights potential
sources of stress tied with parents’ socio-environmental experiences, including marital
relationship, social network, and work circumstance. From the view of this model, all the abovementioned experiences can bring both support and stress depending on their forms and
environment-person fit. For example, social network is usually beneficial to positive parenting
functioning. However, when the support received from social network does not match the
13

support desired, the support received will bring stress to parenting functioning. Likewise, marital
relationship can provide parents with support in the forms of both emotional support (e.g., love
and affection) and instrumental assistance (e.g., substantial assistance in child-rearing). However,
it can also produce stress for parenting functioning when inter-spousal hostility is high. Work
circumstances can also bring support as well as stress to parenting functioning. For example,
working mothers who feel satisfied with their work tend to have high self-esteem and this
positive psychological adjustment in turn benefits their parenting efficacy. In contrast, a number
of potential stressors are related to work circumstances, including jobs demanding overtime work
and inflexible schedules or jobs limiting opportunities for self-initiation. In addition, maternal
employment often adds parenting stress when the family members do not encourage the mother
to play the dual roles in employment and child-rearing.
By integrating these two theoretical foundations, the present study aims to understand
how formation of parents’ monitoring knowledge is shaped by parents’ promoting/inhibiting
environments (i.e., neighborhood condition), adaptive culture (i.e., grandmothers’ childrearing
involvement), family experiences (i.e., parenting and family relation), and adolescents’
characteristics (i.e., gender). In this study, although links between adolescents’ distal ecologies
and proximal ecologies are not directly examined, the impact of distal ecologies is reflected in
the nature of the present sample. That is, African and Latino Americans tend to experience
economic segregation and thus they are disproportionately represented in the current sample of
low-income families. In examination of family experiences, this study will apply Belsky’s
process model to uncover the interconnections of mothers’ monitoring knowledge with their
socio-environmental experiences (i.e., work circumstances) and their children’s characteristics.
Grandmothers’ childrearing involvement are mentioned in both Garcia Coll et al.’s integrative

14

model and Belsky’s process model but are classified into different categories, with Garcia Coll et
al.’s model placing this childrearing support into adaptive culture, while Belsky’s model
classified this support into a kind of instrumental assistance nested within socio-environment
experiences. Given that grandmothers’ childrearing involvement has been portrayed as reflecting
the importance of family cohesion—one of core cultural values—in African American and
Latino families (Yasui & Division, 2007), the present study adopted the classification of Garcia
Coll et al. for this childrearing support. The concept map for this study is presented in Figure 3.

15

Family Context
Family Environments:
-Parenting styles/behaviors (authoritative
parenting, harsh discipline, and behavioral
control)
-Establishment of Family routines
-Parents’ socio-environmental experiences:
work circumstances

Distal Ecologies
Parents’ socioenvironmental
experiences:
Household
arrangement

Races/Ethnicities

Adolescent
Characteristics:
Child gender

Social Class
Parents’ monitoring
knowledge

Proximal Ecologies:
-neighborhoods
(problems and
cohesion)

Adolescent
Adjustment in longterm:
-Externalizing
behavior problems

Residential
Segregation

-Internalizing behavior
problems

Adaptive culture:
-Traditions (i.e., grandmothers’
childrearing involvement)

-School Delinquency

Figure 3. Concept Map for the Present Study

16

Chapter 3: Background Literature
According to the process model of parental monitoring (Hayes, Hudson, & Matthews,
2003), parents’ monitoring knowledge of children’s daily whereabouts and activities is the
product of the monitoring process occurring at the third phase. At the first phase, parents set up
rules and expectations on their children’s behavior. Following the previous phase, children
engage in free-time behavior at the second phase. At the third phase, parents gain information
about locations, members, and types of activities of their children’s free-time behavior. Parents’
knowledge is primarily established either through parents’ active solicitation or children’s
voluntary disclosure (Racz & McMahon, 2011). High levels of monitoring knowledge reflect
that parents are able to effectively supervise their children’s daily life as well as children are
willing to discuss and share critical details of their daily behavior.
The traditional perspective on trajectories of parents’ monitoring knowledge asserts that,
on average, parents’ knowledge about children and their activities decreases over time, especially
when the children approach middle and late adolescence (Laird, Pettit, Bates, & Dodge, 2003).
This decline in monitoring knowledge is considered as a result of parents’ gradual release in
controlling and children’s desire for more freedom and privacy as they grow older. In contrast,
some recent studies (Larid, Criss, Pettit, Dodge, & Bates, 2008; Spano et al., 2012) found that
there are variations in patterns of trajectories of parents’ monitoring knowledge, with some
families experiencing consistently high levels, while some experience a moderate or sharp
decline. Their findings suggest that families are diverse in the degrees to which whether
adolescents are allowed to be autonomous managing and determining how and where they spend
their free time.

17

Extending their work, the present study first identified trajectory groups of parents’
knowledge and then explored underlying factors that predict variations in the patterns for African
American and Latino adolescents in low-income families. Previous research in this area points
that levels of parents’ monitoring knowledge are associated to various factors, including family
environment, neighborhood context, and child characteristics. In this following section, previous
findings on each of these factors are reviewed along with discussion on similarities and/or
differences between African Americans and Latinos, which guides research hypotheses in the
present study. Before proceeding to review of previous findings on parents’ monitoring
knowledge and discussion on ethnic differences, the African American and Latino family
contexts are reviewed for purposes of facilitating further discussion on similarities and/or
differences between African Americans and Latinos.
3.1 African American Families
3.1.1 Cultural Values
The family context of African Americans is primarily influenced by two cultural values:
communalism and spirituality. Communalism is defined as “a belief in the importance of group
over the individual and cooperation rather than competition” (Belgrave, Townsend, Cherry, &
Cunningham, 1997, p.423). The cultural value can be traced to the African worldview that
emphasizes every creature in the universe is functionally dependent on each other. The emphasis
on communalism is reflected in the strong communal ties of African Americans. African
Americans tend to form close communal ties with family members, friends, and community
members. Among the early generations of African Americans, they felt that maintaining the
tradition of strong communal ties plays an important role in helping them cope with the slavery
and oppressive experience. Nowadays, having strong communal ties is still an important strength

18

that help African Americans cope with adversities (e.g., economic stress). For example, lowincome African Americans often rely on family members, friends, and community members for
assistance in finance, elder-taking, and childrearing (Yasui & Dishion, 2007).
Another key cultural value, spirituality, is defined as a belief in “the emphasis on
interconnectedness with the natural and the human environments” (Yasui & Division, 2007, p.
152). This cultural value is a trace of the African worldview that stresses interconnections among
all organisms and the importance of harmony (Utsey et al., 2000). The emphasis on spirituality is
reflected in group cooperation and harmony and strong ties with both blood related and fictive
kin. Numerous studies have shown that spirituality plays an important foundation for African
American families and promote intergenerational interdependence and family cohesion, which
benefits their children’s development (e.g., Belgrave et al., 1994; Brody, Stoneman, & Flor,
1996).
3.1.2 Family Environment
Communalism and spirituality bear profound influence on childrearing arrangement for
African Americans. In terms of childrearing arrangement, African Americans tend to feel
obligated to provide childrearing support for blood-related kin or fictive kin such as friends and
community members (McAdoo, 1981). Among sources of kinship support, support from
grandparents is primary since intergenerational connections are highly valued in African
American families (Yasui & Dishion, 2007). Kinship support for childrearing is thought as an
important adaptation to the African American family context. Compared to their European
American counterparts, African American children are more likely to grow up in single mother
headed households. The majority of these single mothers are of working-class, working for lowpaid and labor-intense jobs. The nature of their work condition blocks available time to engage in

19

childrearing tasks. For these mothers, receiving kinship support for childrearing helps share
childrearing responsibilities so that they are able to maintain family function and meanwhile
fulfill economic needs (McAdoo, 2002). Research on African American families has shown that
kinship support is associated to positive family management, including high parental
involvement in schooling and family organization (i.e., clear family routines and schedules) for
both two-parent and single-parent households (Taylor, 1996).
Communalism and spirituality are also reflected in parenting beliefs and behaviors of
African Americans. Since both communalism and spirituality emphasize group interconnection
and harmony with a group, African American parents feel the importance of maintaining family
harmony, parent-child interconnection, and parental authority (Yasui & Dishion, 2007). These
parenting beliefs are distinctive from what European Americans bear in which the priority is to
foster children’s autonomy and independence (Rothbaum & Trommsdorff, 2007). Guided by
parenting beliefs, African and European American parents also engage in distinctive parenting
behaviors. Before proceeding to discussion on comparison in parenting behaviors between
African and European Americans, I reviewed previous research on parenting behaviors below.
One of the most influential dimensions of parenting was proposed by Baumrind (1971).
Baumrind identified two primary dimensions of parenting behavior: parental responsiveness and
parental demandingness. Parental responsiveness refers to warmth, nurturance, and responses to
children’s needs. Parental responsiveness has been portrayed as one crucial parenting behavior
designed to foster children’s autonomy and self-determination (Rothbaum & Trommsdorff,
2007). Parental demandingness is a more complex concept that reflects at least two distinctive
facets of control-oriented parenting behavior. Based on the work of Baumrind, the first facet is
behavioral control, which refers to parents’ attempts to set up clear rules and consistent

20

discipline in order to shape age-appropriate behaviors of children. In the original work of
Baumrind, responsiveness and behavioral control yield four parenting styles: authoritative (high
in both parental responsiveness and behavioral control), authoritarian (low in parental
responsiveness and high in behavioral control), permissive (high in parental responsiveness and
low in behavioral control), and indifferent (low in both parental responsiveness and behavioral
control).
The second facet of demandingness is hostility-coercion, which was not reflected in the
original work of Baumrind. Hostility-coercion is defined as a variety of parenting tactics that aim
to inhibit children’s autonomy and self-regulation, including threats, power assertion,
psychological control, and nagging or harsh physical discipline (Barber, 2002; Bugental, 1992;
Hoffman, 1970; Weis & Toolis, 2010). The crucial difference between the two facets rests on
that behavioral control aims to foster children’s behavioral regulation in order to meet the
constraints of society, while hostility-coercion behaviors are directed to assert parents’ authority
and power over children (Weis & Toolis, 2010). The following discussion focuses on one
specific tactic of hostility-coercion—harsh discipline (including nagging and physical
punishment), which is one of the focus constructs in the present study.
A body of research has shown that compared to European American parents, African
American parents tend to show less responsiveness/acceptance, but exercise more controloriented parenting in terms of both facets, i.e., behavioral control and harsh discipline (e.g.,
Berlin et al., 2009; Bradley, Corwyn, McAdoo, & García Coll, 2001; Jambunathan, Burts, &
Pierce, 2000; Weis & Toolis, 2010; Dearing, 2004). Literature based on Baumrind’s model has
also shown ethnic difference. Specifically, African Americans tend to score lower than European
Americans in authoritative parenting and higher in authoritarian parenting (Dornbusch, Ritter,

21

Leiderman, Roberts, & Fraleigh, 1987; Furstenberg, Cook, Eccles, Elder, & Sameroff, 1999).
The above findings can be explained by the ethnic differences in cultural values, with European
American parents placing more emphasis on fostering children’s autonomy, whereas African
American parents place more emphasis on parental authority and family relatedness.
However, simply making clear-cut differentiations in parenting behavior between
African and European Americans may conceal the variations within these ethnicities. Indeed,
there has been growing attention to diversity of parenting behavior in relation to social contexts,
such as socioeconomic status (SES), within the ethnicities. For example, some studies have noted
that differences between African and European Americans in levels of parental responsiveness,
behavioral control, and harsh discipline are attenuated as SES increases (Bradley et al., 2001;
Weis & Toolis, 2010). Within African Americans, higher SES parents are more likely to exhibit
higher responsiveness as well as lower control and physical discipline than lower SES
counterparts (Dodson, 1997; Kelley, Power, & Wimbush, 1992; McLoyd, 1990b; Smetana,
2000). In line with the above findings, one studying examining typology of African American
family functioning found that the proportions of authoritative and authoritarian type were
comparable as well as the parents of the authoritative type scored higher on the years of parental
education and income than those of authoritarian type (Mandara & Murray, 2002). In addition to
SES, parenting behavior also varies with child gender, with boys perceiving more parental
control than girls (Jones-Sanpei, Day, & Holmes, 2009). This may be due to that African
American parents are aware of boys’ higher risk at behavioral problems.
Literature about associations of parenting with parent-child closeness and parental
warmth for African Americans has shown a trend distinctive from that about European
Americans. For European Americans, parent-child closeness and parental warmth are

22

consistently predicted by authoritative parenting or the parenting tactic aiming to foster
children’s self-regulation, such as behavioral control; but are negatively linked to harsh
discipline or coercive control (Chao, 2001; Jackson-Newsom et al., 2008; Westbrook & Harden,
2010). In contrast, for African Americans, parental warmth can be predicted by all the above
parenting styles or tactics, including authoritative parenting, behavioral control, harsh discipline,
and coercive control for both young children and adolescents (Jackson-Newsom et al., 2008;
Mandara & Murray, 2002; Steele, Nesbitt-Daly, Daniel, & Forehand, 2005; Westbrook &
Harden, 2010). It is noteworthy that the above findings were held after family arrangement and
SES were controlled for, and thus cannot be accounted for by interactions of ethnicity and
SES/family arrangement. Although there also exist some conflicting findings specifically for
harsh discipline and coercive control (i.e., no significant or negative relations between the two
hostility-coercion tactics and parental warmth) (Jackson-Newsom et al., 2008; Mandara &
Murray, 2002), the overall findings suggest that harsh discipline and coercive control may
convey different meanings for European and African Americans. Since parental authority and
power is more highly esteemed in African American families than in their European American
counterparts, African American children may be more likely to interpret harsh discipline and
coercive control as appropriate parenting tactics, and thus tend to perceive such parenting tactics
as representative of parental care and warmth.
Literature about associations between parenting and child adjustment for African
Americans has also shown a trend distinctive from that about European Americans. Similarities
still exist for both ethnicities in that child developmental outcomes are benefitted by parental
responsiveness/warmth (Bradley, Corwyn, Burchinal, McAdoo, & García Coll, 2001; Dearing,
2004) and behavioral control (Barber, Stolz, Olsen, Collins, & Burchinal, 2005; Wang,

23

Pomerantz, & Chen, 2007; Mason et al., 1996). However, a major difference rests on that for
African Americans, harsh discipline or coercive control produces less negative effects on
developmental outcomes or even protective effects, when compared to European Americans. A
body of research has shown that positive associations between harsh discipline/coercive control
and poor outcomes (e.g., internalizing and externalizing behavior problems and low academic
achievement) were weaker or not significant for African American children compared to their
European American peers (Bradley, Corwyn, McAdoo et al., 2001; Hill & Bush, 2001;
Stormshak et al., 2000). One longitudinal study even found that harsh discipline in early
childhood and during early adolescence predicted more externalizing behavior problems reported
at age 16 for European American adolescents, but fewer behavior problems for their African
American peers (Lansford, Deater-Deckard, Dodge, Bates, & Pettit, 2004). Their findings were
held consistent for girls and boys and after family arrangements, SES, and child temperament
were controlled. Together, the above findings suggest that harsh discipline may produce less
negative or even positive implications for African American children’s developmental outcomes
than the outcomes of European Americans. The previous discussion in the present study has
pointed that African American children tend to interpret harsh discipline as an appropriate
parenting tactic and a vehicle that conveys parental care and warmth. Accordingly, African
American children are more likely to be benefitted by such tactic than their European American
counterparts.
Although the above findings suggest ethnic difference in implications of harsh discipline,
there still exist conflicting findings. For example, some studies indicate that the negative effects
of harsh discipline or coercive control on developmental outcomes (e.g., externalizing and
internalizing behavior and academic achievement) are similar for European and African

24

Americans (McLeod & Shanahan, 1993; Shumow, Vandell, & Posner, 1998; Whiteside-Mansell,
Bradley, Owen, Randolph, & Cauce, 2003). The mixed findings may be explained by
developmental processing. Among the above-mentioned findings suggesting ethnic similarity,
children were assessed during preschool or elementary school years, while Lansford et al.’s
(2004) study assessed the effect of harsh discipline in middle to late adolescence. Although it
still calls for further empirical research to document the diversity of implications of harsh
discipline in relation to child age, the above findings suggest that older African American
children are more likely to interpret harsh discipline as an appropriate parenting tactic than
young children given that they have more refined cognitive processing (see the discussion of
Lansford et al., 2004).
3.2. Latino American Families
3.2.1 Cultural Values
Latinos are one of the fastest growing minorities in the US, which is projected to rise to
one-fourth by 2050 (García Coll & Pachter, 2002). Latinos in general refers to people with
origins in Mexico, Central or South America, and the Spanish-speaking Caribbean (Harwood,
Leyendecker, Carlson, Asencio, & Miller, 2002). Although national origins of Latinos are
diverse, Latinos share many cultural heritages. The two primary common cultural heritages are
familismo and respeto, which bear profound influence on the Latino American family context
(Harwood et al., 2002). Familismo refers to “a sense of obligation to and connectedness with
one’s immediate and extended family” (Unger et al., 2002, p.259). According to Ferrari (2002, p.
794), familismo is described as “family unity, with a sense of obligation among family members,
reverence for the elderly, and responsibility to care for all members, especially children.” Overall,
familismo means a strong sense of family cohesion within nuclear family system and among

25

extended family members (e.g., grandparents, aunts, uncles, cousins, and godparents) and it is
maintained through mutual support and placing the importance of family needs over individual
needs. Another central cultural heritage is respeto, referring to “knowing the level of courtesy
and decorum required in a given situation in relation to other people of a particular age, sex, and
social status” (Harwood, Miller, & Lucca Irizarry, 1995, p.98). As respeto is manifested in
family context, it stresses the importance of showing respect and obedience towards parents and
elders (Marin & Vanoss Marin, 1991).
3.2.2 Family Environment
Familismo is reflected in family structure and childrearing arrangement in Latino families.
Compared to European Americans, Latinos are more likely to have strong networks with
extended family members; and multigenerational households are more common among Latinos
(e.g., García Coll, 1993; Miller & Harwood, 2001). Such strong family ties provide Latinos with
great resources of financial support and assistance in childrearing, which has been found as an
important protector for the families experiencing times of crisis, such as financial strain and
single-parent households (Garcia-Preto, 1996; Zea, Mason, & Murguia, 2000). Also, family
support has been found as a predictor of good quality of parenting (e.g., parental acceptance) or a
buffer that decreases negative effects of financial and global life stress on mothers’ nurturance
(i.e., provision of warmth and love in childrearing) (De Leon Siantz & Smith, 1994; Uno,
Florsheim, & Uchino, 1998).
Familismo coupled with respeto bears great influence on parenting beliefs and behaviors
in Latino families. Latino parents tend to place great emphasis on maintaining parental authority,
parent-child interconnection, and family harmony (Harwood et al., 2002). Along with the
previous discussion in the present study for African Americans, parenting beliefs of Latinos are

26

more distinctive from the beliefs of European Americans that stress fostering children’s
autonomy and independence, but are more similar with the beliefs of African Americans that
stress parental authority and harmony. A body of research has found that the degrees to which
Latino parents exercised in both facets of control-oriented parenting (i.e., behavioral control and
harsh discipline) are similar to the degrees of African American parents, and are higher than the
degrees of their European American counterparts (e.g., Blair, Blair, & Madamba, 1999; Bradley,
Corwyn, McAdoo et al., 2001; Dixon et al., 2008; Finkelstein, Donenberg, & Martinovich, 2001;
Hill et al., 2003; Varela et al., 2004; Okagaki & Frensch, 1998). Within Latinos, levels of
control-oriented parenting vary with child gender, with girls perceiving more parental control
than boys (Jones-Sanpei et al., 2009; Perez-Brena, Updegraff, & Umaña-Taylor, 2012). This
might be due to that immigrant parents place different standards in gender role in preserving
cultural heritages. Based on the review of Suarez-Orozco and Qin (2006), immigrant parents
across ethnic backgrounds view daughters as a more important keeper of the culture than boys.
In order to impede daughters from assimilating in the American mainstream culture too quickly,
immigrant parents tend to use strict control over daughters’ daily activities, such as going to
parties or hanging out with friends after school.
Although a number of studies have indicated the ethnic similarity in levels of controloriented parenting for Latinos and African Americans, there also exist conflicting findings.
Specifically, some studies found that the degrees to which African American parents engage in
harsh discipline are higher than the degrees of Latino parents (Jambunathan, Burts, & Pierce,
2000; Weis & Toolis, 2010), while some studies indicated that the degrees are lower for African
Americans in terms of harsh discipline or behavioral control (Dearing, 2004; Florsheim, Tolan,
& Gorman-Smith, 1996). Although it still waits for more empirical research to clarify the above

27

inconsistent findings, such mixed findings may be explained by children’s age. With the
exception of the studies of Jambunathan et al. (2000) and Weis and Toolis (2010), the other
studies assessed the children aged beyond 7 or a wide range of age from 0 to 13. Thus, it is
possible that Latino parents tend to increase degrees of control-oriented parenting as children age
so that the differences in control parenting between Latinos and African Americans are reduced
or even are toward an opposite trend.
The literature on the differences in control-oriented parenting among Latino, African
Americans and European Americans has shown a similar trend, but the literature on the
differences in parental responsiveness/warmth and authoritative parenting has been mixed. Some
studies have indicated that Latino and European American parents display similar degrees of
parental responsiveness/warmth or authoritative parenting and their degrees are higher than the
degrees of African Americans (Berlin et al., 2009; Hill et al., 2003; Jambunathan et al., 2008;
Varela et al., 2004), while some found that Latino and African American parents display similar
degrees of parental responsiveness and their degrees are lower than the degrees of their European
American counterparts (Bradley, Corwyn, McAdoo et al., 2001; Dearing, 2004; Weis & Toolis,
2010).
The above mixed findings may not be explained by any single demographic factors, such
as SES, children’s age, or parents’ educational level, given that both findings suggesting similar
and different levels were based on a wide range of SES, children’s age, and parents’ educational
level. Accordingly, it remains unclear regarding the factors leading to similarities or differences
in parental responsiveness and authoritative parenting among Latinos, African Americans, and
European Americans. Even so, the above findings still point a possible trend in ethnic differences.
That is, Latino parents are more likely to display high degrees of parental responsiveness or

28

authoritative parenting as what European American parents do, and their degrees tend to be
higher than the degrees of their African American parents. This trend is consistent with the view
that family cohesion is more highly esteemed in Latino families than in African American
families (Pachter, Auinger, Palmer, & Weitzman, 2006). The above difference between Latinos
and African Americans may not only be explained by Latinos’ cultural traditions, but also due to
adaptation related to migration. Compared to African Americans, most Latinos are recent
migrants. After migrating to the US, Latino parents may rely on building close relations with
their children and family cohesion to teach their children cultural traditions or reduce the extrafamilial contact that may deteriorate parents’ efficacy in monitoring (Bulcroft, Carmody, &
Bulcroft, 1996; Reese, Kroesen, & Gallimore, 2000).
Literature about associations between parenting and parent-child relationship for Latino
Americans has also shown a trend distinctive from that about European Americans but more
similar to that about African Americans. The previous discussion in the present study points to
that for European Americans, parental warmth is consistently predicted by authoritative
parenting or behavior control, but is negatively linked to harsh discipline and coercive control. In
contrast, like African Americans, parental warmth can be conveyed through all the above
parenting styles or tactics, including authoritative parenting, behavioral control, harsh discipline,
and coercive control for Latinos (Hill et al., 2003; Izzo et al., 2000; Domenech Rodríguez,
Donovick, & Crowley, 2009). These findings suggest that Latino and African American children
are alike in the way they perceive harsh discipline and coercive control. Like African American
children, Latino children tend to interpret harsh discipline and coercive control as a positive
connotation representing parental warmth since parental authority and power is also highly
valued in their traditions. Indeed, Crockett and colleagues (2007) interviewed both first and

29

second generation Mexican American adolescents and found that these adolescents
conceptualized harsh discipline as a form of parents’ caring showing love and warm.
Literature about associations between parenting and child adjustment for Latinos has
shown a trend similar to that about African Americans in that the positive effects of authoritative
parenting, parental responsiveness, and behavioral control for child development are similar, but
the protective role of harsh discipline is more salient, compared with European Americans.
Indeed, harsh discipline in Latino families has been documented to be linked to low levels of
conduct problems and higher cognitive functioning while the same parenting practice was linked
to increased conduct problems and lower cognitive functioning for European Americans (Berlin
et al., 2009; Dearing, 2004). Like African American children, Latino children also tend to
interpret harsh discipline as an appropriate parenting tactic and representative of parental caring
and warmth. Thus, such tactic is more likely to benefit developmental outcomes of Latino
children than the outcomes of their European American counterparts.
However, like African American children, there also exist some conflicting findings in
terms of implications of harsh discipline for Latino children’s development outcomes.
Specifically, a number of research has evidenced that harsh discipline is linked to more
externalizing and internalizing behavior problems for both Latino and European American
children (Bradley, Corwyn, Burchinal et al., 2001; Hill et al., 2003; McLeod, 1993). The mixed
findings may not be simply explained by one single demographic or environmental factor (e.g.,
SES, child age, and neighborhood safety) given that the studies suggesting positive and negative
implications of harsh discipline assessed both a wide range of SES and child age. It needs more
empirical research to clarify the factors accounting for the mixed findings on implications of
harsh discipline. Although there exist the mixed findings, the overall discussion on African

30

Americans and Latinos in the present study suggests that harsh discipline is more optimal to
family relationship characteristics (e.g., warmth and closeness) and children’s overall adjustment
for both ethnicities, compared to European Americans.
In addition to ethnic difference, literature on Latino family environment is often centered
on within-group variations in relation to acculturation. A body of previous findings have
indicated that less and more acculturated Latino parents score similarly in authoritative parenting,
parental responsiveness, and both facets of control-oriented parenting (Berlin et al., 2009; Hill et
al., 2003; Varela et al., 2004). Berlin and colleagues also found that the positive association
between harsh discipline and children’s cognitive functioning is similar for both less and more
acculturated parents. Other research has lent evidence on stability of values in relation to
familismo and respeto during the process of acculturation. For example, Rueschenberg and
Buriel (1989) examined a sample of Mexican American parents with three different generational
statuses (monolingual Spanish-speaking and migrating within five years, monolingual Spanishspeaking or Spanish-dominant and residing at least ten years, and US-born and bilingual or
English-speaking preference). Their findings indicated that the three groups are different in some
values unrelated to the family system (e.g., independence and achievement orientation), but are
alike in their values central to the family system, such as family cohesiveness and authority of
parents.
Although the above findings suggest that Latino family environment appears to be stable
during the process of acculturation, some findings lend conflicting evidence. For example, Hill
and colleagues (2003) found that mothers’ acceptance and coercive control were positively
related for Mexican Americans speaking Spanish, but not for those speaking English. Their
finding suggests that coercive control or harsh discipline may be more likely to be interpreted as

31

parental warmth for less acculturated Latinos than their more acculturated counterparts. In line
with this view, Gonzales and colleagues (2000) found that coercive control was related to fewer
conduct problems only in Spanish-speaking families. All the above-mentioned findings were
based on small community-based samples (sample size ranging between 45 and 177), with the
exception of the study of Berlin et al. based on a national sample. More future large nationalscale research may clarify relations between family environment and acculturation among
Latinos.
3.3 Factors Linked to Parents’ Monitoring Knowledge
In this proceeding section, previous findings on factors associated to levels of parents’
monitoring knowledge are reviewed. These factors include family environment, neighborhood
context, child characteristics, and adaptive culture. Review of each of these factors is along with
discussion on similarities and/or differences between African Americans and Latinos, which
guides research hypotheses in the present study.
3.3.1 Family Environment
Regardless of sources of knowledge from parents’ solicitation or child disclosure, high
degrees of parents’ monitoring knowledge have been linked to indicators of positive parent-child
relationship (e.g., communication full of trust and openness, closeness, and parental warmth) in
White-dominant and Mexican American families based on cross-sectional data (Blocklin,
Crouter, Updegraff, & McHale, 2011; Padilla-Walker, Harper, & Bean, 2011). The previous
discussion in the present study points that for European Americans, the above positive parentchild relationship qualities are likely to be fostered through authoritative parenting or behavioral
control, but are not linked to harsh discipline. In contrast, these positive parent-child relationship
qualities can be linked to both authoritative and both facets of control-oriented parenting (i.e.,

32

behavioral control and harsh discipline) for both African Americans and Latinos. Accordingly,
adolescents of both ethnicities may be more likely to experience consistently high levels of
mothers’ monitoring knowledge when experiencing averagely high levels of both authoritative
and control-oriented parenting of the mothers.
In assessing the low-income African American and Latino mothers’ parenting styles and
tactics, the present study did not apply Steinberg et al.’s approach (1992) in which parenting
styles and/or tactics were classified into static categories (i.e., authoritative and authoritarian
parenting). On the contrary, the study assumes parents may mix their parenting styles and tactics
and thus assessed the relative degrees to which the parents exercise each type of parenting styles
and/or tactics (i.e., authoritative parenting, behavior control, and harsh discipline). The reason
behind is that scholars have noted that parenting practices are not static, but rather fluid and
changing according to situation (Tamis-LeMonda et al, 2008).
In addition to parenting styles and tactics, the present study focused on one indicator of
family relationship quality, i.e., family routine. Family routine is defined as a set of family
regular activities arranged by parents in order to create common family time for family members;
some examples of family routine include the times for playing and talking, mealtimes, and
homework times (Jensen, James, Boyce, & Hartnett, 1983). Based on this definition, high levels
of family routine may provide the mothers with a lot of opportunities to solicit whereabouts and
activities from their children through the regular family time. Meanwhile, their children are also
able to use the family time to share and discuss their free-time activities in or after school. In
other words, high levels of family routine may shape positive dynamics of mother-child
communication in which both the mother and the child can discuss the child’s daily life.

33

In addition to the above explanation based on increased opportunities of mother-child
communication, cultural tradition in African Americans and Latinos may also suggest the
positive link between family routine and mothers’ monitoring knowledge for the two ethnicities.
Since family routine is directed to enhance opportunities for family members to engage in
common activities, this family dynamic probably serves as a venue through which family
cohesion is enhanced. As previously discussed in this article, both African Americans and
Latinos place great emphasis on the importance of establishing family cohesion. Given that
family cohesion is a core of their cultural traditions, African American and Latino adolescents
probably enjoy the common family time and take this advantage to share and discuss their daily
life with their mothers. Also, these adolescents may perceive the positive connotation behind
family routine, such as mothers’ care and love, which leads to their intention to voluntarily
disclose their whereabouts and activities. To my knowledge, relation between family routine and
parents’ monitoring knowledge has been underexplored across ethnicities, with exception of the
study of Prelow et al. (2007) finding short-term positive link. The present study aims to address
whether average high levels of family routine over time facilitate mothers’ long-term monitoring
knowledge for the low-income African American and Latino families.
Belsky’s process model posits that parenting behavior is predicted by parents’ work
circumstances. In line with this view, levels of parents’ monitoring knowledge have been found
to be associated with one key family factor, i.e., parents’ work hours. Further, findings on links
between parents’ monitoring knowledge and work hours suggest that the impact of work hours
may depend on family structure and assistance from family networks. Conducting a crosssectional study with in-depth interview of White-dominant samples, Belle and Philips (2010)
found that both mothers and fathers complain that long work hours hinder them from gaining

34

monitoring knowledge of their children. Moreover, this study found that the difficulty of gaining
monitoring knowledge is more common among parents who do not have relatives to rely on for
child-caring. This might be due to that these parents are unable to rely on their relatives to help
monitor their children and gain information about their children. Extending their work, the
present study examined whether there is a link between averagely long work hours of mothers in
adolescence and a decline in their monitoring knowledge over time among low-income African
and Latino American households. Moreover, this study investigated whether this link is buffered
by the support of relatives, specifically grandmothers’ childrearing involvement. The present
study focuses on parenting support from grandmothers because their support has be described as
one primary family support for both African Americans and Latinos (Cox, Brooks, & Valcarcel,
2000; Sotomayor & Applewhite, 1988; Taylor, 1996).
3.3.2 Neighborhood Context
García Coll et al.’s integrative model points that communities can either promote or
inhibit children’s development through the mediation of their family experiences. When
communities are characterized as unsafe and lacking of social cohesion (i.e., trust and mutual
support among neighbors), children’s positive externalizing and internalizing behavior
adjustment is more likely to be inhibited because parents may not effectively monitor their
children or gain information about them. This suggests that parents’ monitoring knowledge is
dependent on the neighborhood context. However, very few studies examine the role of
neighborhoods for parents’ knowledge. One effort was made by Spano and colleagues (2012).
By examining African American youth living in poverty, they found that growing exposure to
violence in neighborhoods increases the likelihood of a decrease in parents’ monitoring
knowledge over time.

35

Social cohesion (characterized as trust and mutual support among neighbors) of
neighborhoods is another critical factor that may influence degrees of parents’ monitoring
knowledge. When living in cohesive communities, parents tend to exert strict control over
children’s activities because misbehavior is not allowed in such communities (Roche, Ensminger,
& Cherlin, 2007). Moreover, cohesive communities are characterized by the mutual supervision
of children’s behaviors and safety. Accordingly, socially organized communities may increase
degrees of parents’ knowledge either through parents’ own efforts or neighbors’ assistance. In
line with this perspective, Lahey et al. (2008) examined a broad (from low to high SES) sample
of Latinos, European Americans, and African Americans, with the finding that adolescents living
in disorganized neighborhoods reported low degrees of parental knowledge. This study only
found this link in a short term. Extending their work, the proposed study examined long-term
associations between neighborhood cohesion and mothers’ monitoring knowledge for lowincome African Americans and Latinos.
The present study surmises the two ethnic groups may be different in the extent to which
consistently disadvantaged neighborhood contexts (i.e., high degrees of neighborhood problems
and low degrees of neighborhood cohesion) are linked to low levels of mothers’ monitoring
knowledge over time. The previous discussion in this article points that both African Americans
and Latinos place emphasis on the importance of family cohesion. However, empirical research,
albeit very limited, has noted that family cohesion may play a more important role in
strengthening family solidarity and interdependence and thus blocking potential harm from
neighborhoods that may threaten positive family processes for Latinos than for African
Americans (Pachter, Auinger, Palmer, & Weitzman, 2006). Using data of female respondents
with children in middle childhood from the National Longitudinal Survey of Youth 79, Pachter

36

et al. (2006) found that disadvantaged neighborhood conditions (i.e., with lack of safety and
social cohesion) are detrimental to positive family processes (e.g., maternal supportive parenting)
for European and African Americans, but such an effect was not found for Latino Americans.
The authors concluded that the strength of family cohesion acts as a stronger firewall that blocks
negative impacts from neighborhoods for Latinos compared to African Americans. Thus,
neighborhood conditions may have little to do with family dynamics (e.g., parenting, parentchild communication, and family relations) for Latino Americans compared to African
Americans. Accordingly, it is surmised that consistently disadvantaged neighborhood conditions
in adolescence are more likely to be linked to low levels of mothers’ monitoring knowledge over
time for African Americans than for Latinos.
3.3.3 Children’s Characteristics
Both García Coll et al.’s integrative model and Belsky’s process model point that
parenting and parents’ family role is impacted by children’s characteristics. In line with both
perspectives, children’s own characteristics account for the extent to which their parents gain
knowledge about their child’s whereabouts and activities. Child gender is one key children’s
characteristic factor that accounts for parents’ monitoring knowledge. Regardless of sources of
knowledge from parents’ solicitation or child disclosure, parents tend to gain more knowledge
about their daughters’ activities and whereabouts than those of their sons across various
countries, such as Hong Kong Chinese (Shek, 2005), European-Americans in the U.S. (Peterson,
Bush, Wilson, & Hennon, 2005; Smetana & Daddis, 2002), and adolescents in Canada, France,
and Italy (Claes, Lacourse, Bouchard, & Perucchini, 2003) and Sweden (Stattin & Kerr, 2000).
This gender difference can be explained by a number of factors, including girls’ high levels of
intention to share information with parents (Stattin & Kerr, 2000), parents’ overprotection of

37

daughters and fear of a daughters’ pregnancy (Veal & Ross, 2006). In addition, parents are better
able to monitor a same-sex child than an opposite-sex child either through active or passive
approaches because parents tend to have more conversations with same-sex children and
encounter fewer barriers in understanding same-sex children’s activities (Vieno, Nation, Pastore,
& Santinello, 2009). Overall, parents are faced with more barriers to gain monitoring knowledge
about boys. Yet, fathers appear to encounter fewer barriers than mothers in gaining knowledge
because of the same-sex parent-child dyad interactions.
The above-mentioned studies offered evidence that the extent to which parents gain
monitoring knowledge varies according to children’s and their own gender. However, their
evidence is based on the samples dominated by two-parent families. It remains unknown whether
single-mother families also encounter more barriers to gain monitoring knowledge about boys.
The living arrangements of single-mother-headed households are diverse, including lone single
mothers and a single mother living with grandparent(s), other kin, a male partner, or grown
siblings (Davidson, 2007). Plausibly, levels of single mothers’ monitoring knowledge about boys
may vary according to their living arrangements. Compared to the single mothers living with a
male partner or other male adult figures in the household, lone single mothers may encounter
more barriers to gain knowledge about boys because of having no male helpers who are able to
help gain information about boys and of the opposite-sex parent-child dyad interaction.
It is also possible that lone single mothers encounter no barriers in gaining knowledge
about their sons because they realize they have no male helpers and thus makes great efforts to
have periodical conversations with the sons. To uncover the aforementioned complexities of
levels of mothers’ monitoring knowledge about boys according to family arrangements, the
present study will examine whether links between child gender and trajectory groups of mothers’

38

monitoring knowledge vary according to household arrangements, specifically among the twoparent family (the mother living with the focus child’s biological father), the lone single mother,
and the single mothers living with a male partner or other male adult figures.
3.3.4 Adaptive Culture
García Coll et al.’s integrative model points that ethnic minorities may develop their
adaptive cultures such as traditions to cope with segregated circumstances (e.g., economic
segregation). In turn, these adaptive cultures influence ethnic minority children’s family
experiences. One salient tradition in African Americans and Latinos is forming strong networks
with extended family members. For both ethnicities, strong ties with extended family members
provide great assistance in childrearing, which has been found as an important protector for the
families experiencing times of crisis, such as poverty or single-parent households. Previous
research has not yet offered evidence on links between patterns of trajectories of parents’
monitoring knowledge and receiving supports in childrearing from relatives. Yet, support from
family members is a possible predictor of consistently high or increasing levels of parents’
monitoring knowledge for African American and Latino families, since the previous discussion
in this article points that kinship support is associated to positive family processes (e.g., high
parental involvement in schooling and clear family routines). Specifically, the present study
focuses on grandmothers’ childrearing involvement because support from grandmothers has been
described as one important part of heritage for both African Americans and Latinos (Cox et al.,
2000; Sotomayor & Applewhite, 1988; Taylor, 1996). It is surmised that averagely high levels of
grandmothers’ childrearing involvement are linked to consistently high or increasing levels of
mothers’ monitoring knowledge over time for both ethnicities.
3.4 Parents’ Monitoring Knowledge and Child Outcomes

39

Both García Coll et al.’s integrative model and Belsky’s process model point that
parenting or parents’ family role is an important predictor of children’s adjustment outcomes. In
line with the above-mentioned theoretical view, current scholarly research has noted the
preventive role of parents’ monitoring knowledge for children’s maladjustment. Children with
parents that possess high degrees of knowledge about their children’s whereabouts and activities
are less likely to develop externalizing behavior problems or school delinquency (e.g., Engels et
al., 2006; Laird et al., 2003). Yet, findings based on longitudinal studies lend caution about
interpreting causal relations between parents’ knowledge and children’s adjustment. Examining
cross-lagged relations, a number of studies (e.g., Bumpus & Hill, 2008; Neumann et al., 2010)
found bidirectional associations between parents’ monitoring knowledge and children’s
externalizing behavior problems or school delinquency. That is, earlier high degrees of parents’
knowledge decrease the likelihood of later externalizing behavior problems and school
delinquency in children, and earlier high degrees of externalizing behavior problems or
delinquency in children decrease the likelihood of later degrees of parents’ knowledge.
Accordingly, parents’ monitoring knowledge and children’s behavioral outcomes/school
delinquency may mutually shape each other over time.
Although previous longitudinal research has demonstrated long-term cross-lagged
relations between parents’ monitoring knowledge and children’s externalizing behavior problems
or school delinquency, it remains unknown about links between trajectory patterns of parents’
knowledge and children’s externalizing behavior problems or school delinquency. Moreover,
despite some recent studies ever examined trajectory patterns of parents’ monitoring knowledge
among children living in working- or middle-class families from early to middle/late adolescence
(Larid et al., 2008; Spano et al., 2012), their work did not investigate links between those

40

trajectory groups of parents’ knowledge and children’s externalizing behavior problems or
school delinquency. Extending their work, the present study examined such links as well as
focuses on internalizing behavior problems. In the present study, internalizing behavior problems
are defined as actions that direct problematic energy toward the self, such as anxious, withdrawn,
and somatic complaints (Achenbach & Rescorla, 2001a). These problems reflect psychological
maladjustment.
To my knowledge, implication of mothers’ monitoring knowledge for adolescents’
internalizing behavior problems has been an underexplored area. Although high levels of
mothers’ monitoring knowledge has been consistently found as brining the positive implication
for externalizing behavior, high levels of this knowledge may not also bring positive implication
for internalizing behavior, based on the view of importance of developing autonomy in
adolescence. While parent-child closeness is important for adolescents’ psychological adjustment,
adolescents with parents who are overprotective may be faced with barriers in establishing
independence, which in turn leads to their disposition to psychological adjustment problems
(Collins & Stenberg, 2006). Indeed, previous findings have pointed that adolescents with parents
who are overprotective tend to report depressive and anxious symptoms (Holmbeck et al. 2000;
MacElhaney & Allen, 2001). Since high levels of mothers’ monitoring knowledge may reflect
their overprotection of children’s daily activities, the high levels of this knowledge may be
associated to internalizing behavior problems.
Alternatively, high levels of mothers’ monitoring knowledge may be a protector based
on the view of importance of parent-child closeness in adolescence. Although adolescence is a
period to develop autonomy, maintaining parent-child closeness remains important for
adolescents’ psychological adjustment (Collins & Stenberg, 2006). Previous studies have found

41

that adolescents who are granted too much autonomy and conceal a lot of secrecy tend to report
psychological adjustment problems (internalizing symptoms) (e.g., Allen et al., 1996; Chen &
Dornbusch, 1998; Engels et al., 2006). These studies further explained that their disposition to
psychological problems may be due to their feeling of being distant and alienated from their
parents. From this view, family dynamics facilitating parent-child closeness—such as parents’
active supervision of their children’s activities or children’s voluntary disclosure—may prevent
children’s internalizing behavior problems. Since both parents’ supervision and children’s
disclosure are important resources of mothers’ monitoring knowledge, high levels of mothers’
monitoring knowledge may be linked to or prevent internalizing behavior problems. Given that
African Americans and Latinos place great emphasis on parent-child connectedness and family
cohesion, high levels of mothers’ long-term monitoring knowledge may be a protector for
adolescents’ internalizing behavior problems.
The present study aims to trace adolescents from early to late adolescence for purposes
of examining how trajectory groups of mothers’ knowledge are linked with the levels of
children’s adjustment in externalizing and internalizing behavior area as well as school
delinquency in late adolescence of African Americans and Latinos living in low-income families
(at the time of Time 3 data collection). The present study surmised that consistently high or
slowly declining levels of mothers’ monitoring knowledge would be linked to low levels of
externalizing and internalizing behavior problems as well as school delinquency in late
adolescence for both ethnicities.
3.5 Chapter Summary
The current scholarly research has noted whether parents’ monitoring knowledge is
impeded or facilitated by socio-environmental factors in family and neighborhood contexts,

42

children’s characteristics, and factors of adaptive culture; and also how parents’ knowledge is
linked to children’s externalizing behavior problems and delinquency. However, more efforts are
needed to investigate longitudinal associations of parents’ knowledge with the aforementioned
impediments, facilitators, and children’s other adjustment (i.e., internalizing behavior).
Specifically, the present study examined the aforementioned factors, including these adolescents’
neighborhoods (i.e., neighborhood problems and social cohesion), family environments (i.e.,
mothers’ parenting styles/behaviors [including authoritative parenting, behavioral control, and
harsh discipline], family routines, and mothers’ work hours), adaptive culture (i.e., grandmothers’
childrearing involvement), and adolescents’ characteristics (i.e., child gender). In terms of
children’s adjustment outcomes, the present study examined adolescents’ externalizing and
internalizing behavior problems and school delinquency. Two potential moderating effects were
also tested. In examining mothers’ work, the interaction of household arrangement by
grandmothers’ childrearing involvement was tested. In examining child gender, the moderating
effect of household arrangement is tested.

43

Chapter 4: Research Methodology
The present study used a three-wave longitudinal survey dataset of the Welfare, Children
& Families: A Three-City Study (Cherlin, 1999; Cherlin et al., 2001; Angel et al., 2009). The
Three-City Study is a longitudinal project that aims to assess the well-being of low-income
families and adjustment of their children in the post-welfare reform era. This study recruited
low-income households in three cities: Boston, Chicago, and San Antonio. In this section,
demographics of low-income families in the focus three cities are first reviewed. Next,
information is provided on the sampling procedure and the sample of the Three-City Study,
analysis plan, and the focus measures.
4.1 Demographics of Low-income Families in Boston, Chicago, and San Antonio
4.1.1 Boston
Boston is the capital of Massachusetts and the largest city in this state (City of Boston.
gov, 2013). During the past two decades (1990-2010), Boston has experienced a great economic
transformation from manufacturing –based industrials to knowledge-based economy and
increasingly economic growth (Kahn & Martin, 2011; a report of Boston Foundation). However,
the official statistics indicate that overall poverty rate remains constant in 1990-2010. Among all
Boston residents, 19.0% were below the federal poverty threshold (the national rate = 13.5%) in
1989, 0.5% increase in 1999, and 0.1% drops in 2005-2009 (the national rate = 14.3% in 20072011) (based on US Census Bureau: Decennial Censuses of 1990 and American Community
Survey: 2009 1-year estimates, calculation conducted by Kahn & Martin, 2011). Among children
under 18, 27.7% lived below the federal poverty threshold in 1989 while almost the same
percentage of children lived in poverty in 2005-2009 (based on US Census Bureau: Decennial
Censuses of 1990 and American Community Survey: 2009 1-year estimates, calculation

44

conducted by Kahn & Martin, 2011). Moreover, the poverty profile shows an increasing
racial/ethnic disparity in family economic levels among children under 18. From 1990 to 2009,
the rate of children living in poverty drops from 18% to 9.5% among European Americans. In
contrast, during the same period, the rates of children living in poverty remain high among ethnic
minorities (Latinos: 45% in 1990 and 40% in 2005-2009; African Americans: 34% in 1990 and
35% in 2005-2009; Asians: 33% in 1990 and 31% in 2005-2009). Further, families with children
in poverty were primarily concentrated in single-mother headed families across ethnic/racial
groups. For example, of households in poverty and with related children under 18, 68%
European American, 82% Latino and 89% African American households were headed by single
mothers in 1999; and 72% European American, 93% Latino and 86% African American
households were headed by single mothers in 1999 (computation based on Census 2000
Summary File 3 and U.S. Census Bureau, 2006 American Community Survey).
Low-income households in Boston are highly concentrated in terms of both geography
and demography (Kahn & Martin, 2011). In regards to family structure, single-mother headed
families are highly clustered in this city and their niches are characterized as poverty, high
unemployment, and dense houses. In regards to racial/ethnic background, most of minority
households, including those of Latino, African, and Asian origins, are highly clustered in highpoverty communities of Boston. In contrast, the majority of European Americans (80% in 20052009) reside in the areas characterized as prosperous economy and new and developing
neighborhoods. This racial residential segregation is reflected in the phenomenon that minority
children are more likely to grow up in low-income households compared to their European
American counterparts in Boston.

45

4.1.2 Chicago
According to the global cities index 2010 of Foreign Policy website (2013), Chicago has
the fourth-largest gross domestic product (GDP) in the world and one of the most important
centers of commerce and trade in the world, which shows that this city is economically
prosperous overall. However, the poverty rate in this city remains higher than the national
average during the past decade. Among all Chicago residents, 19.6% were below the federal
poverty threshold (the national rate = 13.5%) in 2000, while 21.4% were in 2007-2011 (the
national rate = 14.3% in 2007-2011) (US Census Bureau, 2013a, 2013b). The poverty rate
among children underwent fluctuation during the past two decades, with rising to 40 % in the
mid-1990s, dropping to 28% in 2000, and rising again to 32% in 2007-2011. In Chicago,
minority children are more likely to live in the households with incomes below the federal
poverty threshold. In 2000-2011, the rate of children living in poverty remained 8% to 9%
among European Americans. In contrast, during the same period, the rates of children living in
poverty were rising among ethnic minorities (Latinos: 24.8% in 2000 and 36% in 2011; African
Americans: 40.1% in 2000 and 50.7% in 2011; Asian Americans: 21.3% in 2000 and 20.2% in
2011). Among Latino children, the poverty rate is higher among those from Puerto Ricans than
other major groups, i.e., Mexicans and Guatemalans. Like children residing in Boston, children
in poverty are more likely to be reared by single mothers, specifically among African Americans.
For example, of households in poverty and with related children under 18, compared to 43%
European American and 36% Latino households, 81% African American households were
headed by single mothers in 1999; and compared to 60% European American and 43% Latino
households, 83% African American households were headed by single mothers in 2006

46

(computation based on Census 2000 Summary File 3 and U.S. Census Bureau, 2006 American
Community Survey).
In Chicago, although low-income households had been distributed toward this city’s
outer edges, family poverty remained concentrated in a number of communities in the city’s near
west and near south sides between 1990 and 2005 (George, Dilts, Yang, Wasserman, & Clary,
2007; a report of Chapin Hall Center for Children at the University of Chicago). Moreover,
African Americans and Latinos are also highly clustered in these low-income communities. This
is reflected in the phenomenon that African American and Latino children are more likely to be
exposed to neighborhood poverty compared to their European American counterparts in this city.
In 1999, for example, while 5.8% European American children were exposed to neighborhood
poverty, the rate for Latino children (14.5%) was about 2.5 times and the rate for African
American children (22.7%) was about 4 times (based on U.S. Census Bureau, 2000 Census,
Summary File 3, calculation conducted by Diversitydata.org and Harvard School of Public
Health, 2013a).
4.1.3 San Antonio
San Antonio is the second most populous city in Texas and ranked as the seventh most
populous city in the United States (Mildenberg, 2011; a newspaper article of the Washington
Post). Like Boston and Chicago, the poverty rate in this city remains higher than the national
average during the past decade. Among all San Antonio residents, 17.3% were below the federal
poverty threshold (the national rate = 13.5%) in 2000, and 2% increase in 2007-2011 (the
national rate = 14.3% in 2007-2011) (US Census Bureau, 2013c, 2013d). Among children under
18, about one in four children lived in families with household incomes below the federal
poverty threshold (24.3% in 2000 and 27.6% in 2007-2011). The rate of children in poverty is

47

primarily concentrated in ethnic/racial minorities. In 2000, while only 7.3% European Americans
lived in families with household incomes below the federal poverty threshold among children
under 18, 28.5 % Latinos and 26% African Americans lived in the same condition. Based on the
summary of Diversitydata.org and Harvard School of Public Health (2013b), approximately 39%
European American children who attended primary school lived in poverty, while 51% Latino
and 53% African American children attending primary school lived in poverty from 1999 to
2011. The majority of population in this city is composed of Latino origin (approximately 50.4%
in 2000 and 54.1% in 2010) so that Latino children are the largest group in poverty overall. Like
children residing in Boston and Chicago, children in poverty are more likely to be reared by
single mothers. For example, of households in poverty and with related children under 18, 55%
European American, 50% Latino, and 82% African American households were headed by single
mothers in 1999; and 64% European American, 65% Latino, and 71% African American
households were headed by single mothers in 2006 (computation based on Census 2000
Summary File 3 and U.S. Census Bureau, 2006 American Community Survey).
Just like low-income households in Boston and Chicago, low-income households in San
Antonio are demographically and geographically concentrated (Firestone, 2007). This city is in
general viewed as being composed of two contrast worlds, with the west side being characterized
by poverty, unemployment, crime, old houses, and higher rate of school drop-out while the north
side is characterized by prosperous economy, better school, and many developing new
communities. Just as family poverty is strongly concentrated in the west side of this city, Latinos
are also clustered in this district. In contrast, European Americans are highly clustered in the
north side. This geographical and demographical segregation aggravates racial/ethnic

48

socioeconomic gap in San Antonio. Also, this segregation hinders opportunities for Latinos to
gain upward economic mobility and higher educational attainment.
4.2 The Sampling Procedure of the Three-City Study
Using the approach of stratified random sampling, the Three-City Study sampled lowincome families living in low-income neighborhoods in Boston, Chicago, and San Antonio. A
household was eligible if the household income was less than 200% of the federal poverty
threshold. A neighborhood was eligible if 20% households in the neighborhood lived below the
federal poverty threshold in 1990 (Angel et al., 2009; The Inter-University Consortium for
Political and Social Research (ICPSR), 2008, July). The sampling unit of neighborhoods is called
a primary unit (i.e., PU). This study used the PUs defined by US Census blocks in 1990. PUs are
stratified by ethnicity or race. In each focus city, this study selected three large groups (i.e.,
African Americans, Latinos, and European Americans) as the first level of sampling strata (The
Inter-University Consortium for Political and Social Research (ICPSR), 2008, July). Under each
racial group, PUs fitting into the definition of low-income neighborhood were randomly sampled,
with the exception of African American and Latino PUs in Boston and African American PUs in
San Antonio. For the three bodies of PUs, all low-income PUs were selected given that
populations were small. The third level of sampling strata is segments; each of them has 90-120
households. Within each segment, dwelling units were randomly sampled.
Overall, this study randomly sampled 40,000 dwelling units to screen whether the
households met the sampling criteria of this study. The sampling criteria include such as family
income (below 200% of the federal poverty threshold) and children’s age (between 0-4 or 10-14).
Another key criterion was to allow the demographic profile of the final sample to be generalized
to the demographic profile of the low-income population in the three cities. Some examples of

49

demographic indicators include the caregiver’s marital status, the household head’s
race/ethnicity, and family members’ status of social service receipt. In each recruited household,
interviews were administered to one focal child and his/her primary female caregiver (Angel et
al., 2009). These female caregivers are referred to as the mothers given that over 90% were
biological or adoptive mothers. All interviews were face-to-face and conducted by a
computerized device. Interviews of the mothers took approximately two hours, while interviews
of the adolescents took approximately one hour.
4.3 The Sample of the Three-City Study
The sample of the Three-City Study was 2,402 low-income households with a child aged
0-4 or 10-14 years at Time 1 (between March 1999 and December 1999) in low-income
neighborhoods in Boston, Chicago, and San Antonio. Of these participating households, 1,242
households have a child aged 0-4 and 1,160 households have a child aged 10-14. According to
recruiting city, 926 households were from Boston, 762 households from Chicago, and 714
households from San Antonio. The sample was primarily comprised of African Americans and
Latinos, with 47% Latinos, 42% African Americans, 8% European Americans, and 3% of other
races (including American Indians, Asian Americans, and biracial children). Approximately 40%
of the households were receiving cash welfare payments when participating in this study. Only
15.5% of the children were living with their biological father at the time of the first-wave
interview. An average of 16 months after Time 1, the Time 2 data collection occurred between
September, 2000 and June, 2001, with approximately 88% of the original sampled households
participating in this follow-up study. An average of four years after Time 2, the Time 3 data
collection occurred between February, 2005 and January, 2006, with a retention rate of
approximately 80%.

50

The proposed research focuses on the subsample of mother-adolescent dyads at Time 1
(N = 1160; the age of the adolescents: M =11.95, SD = 1.43; the age of the mothers: M = 38.38,
SD = 8.46). The Three-City Study followed up both the adolescents who were reared by
continuous and new caregivers. The present study only focuses on the adolescents who were
reared by continuous caregivers and participated at all three time points. This criterion yields 741
eligible mother-adolescent dyads, with the adolescents aged 10-14 (M =12.26, SD = 1.54) at
Time 1; these adolescents consist of 58 European Americans, 325 African Americans, 361
Latinos, and 10 adolescents of other races (including American Indians and those not reporting
race). Based on the purpose of this study, analyses only involve the mother-adolescent dyads
with African- and Latino-origins. Further examination indicates that six African American and
seven Latino mother-adolescent dyads did not report levels of mothers’ monitoring knowledge
across all the three time points or had missing values over half of the variables focused in the
present study. Excluding these dyads yields 319 African American and 354 Latino motheradolescent dyads for further analyses. Of the focus households, 20.2% were receiving cash
welfare payments when participating. The majority (82.6%) of the focus Latino adolescents were
second generation or beyond, given that they were U.S.-born and their parents were either
foreign- or U.S.-born. Among the Latino mothers, 165 (46.6%) were second generation or
beyond and 189 (53.4%) were first generation (60 born in Mexico, 56 in Puerto Rico, 39 in
Dominican Republic, and 34 in other countries).
The average years the participant mothers received formal education were about twelve
years for both ethnicities, meaning that on average both African American and Latino mothers
had completed high school. However, African American mothers received longer formal
education than Latino mothers (t(673) = 4.04, p < .001). Of the focus participants, 74.1%

51

reported the monthly household income. The ranges of monthly household income were wide for
both ethnicities, with $0 to $4,989 for African Americans and $0 to $4,450 for Latinos. The two
ethnicities were similar in monthly household income (t(502) = .41, p = .684). The household
arrangements were reported at Time 1 and Time 2. Among the Latinos, the stability of household
arrangements was high across types (with a range between 52% and 69%) from Time 1 to Time
2, with the exception of the single-mothers living with kin (29% of this type at Time 1 staying
the same at Time 2). Among the African Americans, the stability of household arrangements was
high for the lone single mothers, the single mothers living with a male partner, and the single
mothers living with grown child (with a range between 52% and 71%) from Time 1 to Time 2,
but was low for the household with the focal child’s biological father (36% of this type at Time 1
staying the same at Time 2) and the single-mothers living with kin (44% of this type at Time 1
staying the same at Time 2). Considering the types of household arrangements at Time 2 were
reported at the middle point of the Three-City Study, the present study regarded the household
arraignments at Time 2 as representative of the average types across times and used for further
analyses. A chi-square goodness-of-fit test was conducted to determine if African Americans and
Latinos were comparable in the types of household arrangements at Time 2. The result indicated
that there was a significant difference in the proportion of types of household arrangements by
ethnicity (χ2(4, N = 673) = 14.07, p =. 007). Follow-up tests were conducted by examining the
standardized residuals. The results indicated that the difference between African Americans and
Latinos was explained by the proportion of two-parent families (the focal child’s biological
father lives in the household). Specifically, the standardized residual for two-parent families was
-2.10 for African Americans and 2.00 for Latinos, which suggests that the observed percentage
for African Americans was lower than the expected percentage while the observed percentage

52

for Latinos was higher than the expected percentage. The percentages for the types of household
arrangements and other demographic information are shown in Table 1 by ethnicities.
Table 1: Demographic Information According to Ethnicity
African American

Latino

(n = 319)

(n = 354)

Variable
%
M
SD
%
M
SD
Female
48
48
Youth Age
12.24
1.37
12.28
1.43
Parental Education (Years)
11.08
2.10
10.38
2.33
Monthly Household Income
1145.21 926.27
1176.70
807.39
Household Arrangements
Two Parents (The Focal Child’s
9.4
17.8
Biological Father Lives in the
Household)
Two Parents (The Mother Lives with
15.6
19.2
a New Partner)
Mother Only
53.9
44.6
Mother and Grown Child
14.4
14.1
Mother and Kin
6.6
4.2
Sample City
San Antonio
23
31
Boston
35
38
Chicago
42
31
Note. Youth age, parental education, and monthly household income were reported at Time 1; and
household arraignments was reported at Time 2.

4.4 Analysis Plan
The proceeding section detailed how the data from the Three-City Study are analyzed, as
well as how the analyses addressed each of the research questions.
4.4.1 Analyses of Data from the Three-City Study
Analyses of data from the Three-City Study involved four domains: 1) to identify
trajectory groups of mothers’ monitoring knowledge for both ethnicities; 2) to identify correlates
of trajectory groups of mothers’ monitoring knowledge from potential antecedents, including
53

adolescents’ neighborhood condition (neighborhood problems and social cohesion), family
environment (family routines, mothers’ parenting styles/behaviors [including harsh discipline,
behavioral control, authoritative parenting], and mother-child relationship), and mothers’ socioenvironmental experiences (mothers’ work hours), adaptive culture (grandmother’s involvement
in childrearing), and adolescents’ gender; 3) to examine interactions between mothers’ work
hours and grandmothers’ childrearing as well as between gender and family arrangements for
predicting trajectory groups of mothers’ monitoring knowledge; and 4) to predict adolescents’
adjustment outcomes (including externalizing and internalizing behavior problems and school
delinquency) by trajectory groups of mothers’ monitoring knowledge.
The domain (1) was conducted by using growth mixture modeling (GMM; Muthén,
2004). GMM is an extension of conventional growth modeling. Conventional growth modeling
assumes that “all individuals are drawn from a single population with common population
parameters” (Muthén, 2004, p. 348). In other words, the individuals share a common intercept
and slope (or shape) of the growth curve. In contrast, GMM relaxes this assumption by allowing
heterogeneity of a population. That is, a population consists of subpopulations that are different
in terms of parameters (e.g., the mean intercept and slope). Using this analysis approach, the
present study identified latent classes of trajectories of mothers’ monitoring knowledge in
African Americans and Latinos as well as compare whether the two groups are different in types
of latent classes and prevalence of each class (if the two groups do not differ in the types of
latent classes). GMM was conducted by Mplus 7.0 (Muthén & Muthén, 1998-2012). Adolescents
participated at Time 1 interview at different times, ranging from age 10 to 14. To allow
adolescents with different assessment time, RANDOM was selected to estimate a growth model
with random slopes across participants (Muthén & Muthén, 1998-2012).

54

The domain (2) to (4) were conducted by logistic or multinomial logistic regression
analysis in which latent classes that are identified in the domain (1) were regressed on the
predictors mentioned in domain (2) and examining moderating effects mentioned in domain (4).
Logistic regression analysis is conducted if the number of identified latent classes is two, while
multinomial logistic regression analysis is conducted if the number of identified latent classes is
more than three. For domain (3), ordinary least square (OLS) regression analysis was conducted
for the outcome variables—externalizing behavior problems and school delinquency, given that
both the outcome variables were based on continuous scales. For the model predicting
internalizing behavior problem, ordered logistic regression analysis was conducted given that
this outcome variable was based on an ordinal scale.
4.4.2 Research Questions
1.1) How does mothers’ monitoring knowledge change over time among adolescents in lowincome families for African Americans and Latinos?
1.2) Are there different patterns of trajectories of mothers’ knowledge between African and
Latino Americans?
GMM was conducted to identify trajectory groups of mothers’ monitoring knowledge for
African Americans and Latinos separately. The two ethnicities were compared in terms of types
of latent classes and prevalence of each class (if the two groups do not differ in the types of
latent classes) qualitatively.
2.1) What socio-contextual factors in neighborhoods and the family context, factors of adaptive
culture, and adolescents’ own characteristics facilitate mothers’ monitoring knowledge over
time for African Americans and Latinos?

55

2.2) What socio-contextual factors in neighborhoods and the family context, factors of adaptive
culture, and adolescents’ own characteristics impede mothers’ monitoring knowledge over
time for African Americans and Latinos?
2.3) Are there any differences in facilitators and impediments to mothers’ monitoring knowledge
over time between African Americans and Latinos?
To identify the correlates that facilitate and impede mothers’ monitoring knowledge,
logistic regression analysis was conducted to regress trajectory groups of mothers’ knowledge on
average levels of each of these factors from Time 1 to Time 3 separately. These factors include
neighborhood conditions (neighborhood problems and social cohesion), family environments
(family routines, mothers’ parenting styles/behaviors, and mother-child relationship), mothers’
socio-environmental experiences (mothers’ work hours), adaptive culture factors (grandmothers’
childrearing involvement), and adolescents’ characteristics (gender). For both ethnicities, the
controlled variables for each factor included types of family arrangements and mothers’
educational level. Only for the Latinos, mothers’ acculturation was also controlled for given that
the previous discussion in the present study pointed links between mothers’ acculturation and
parenting styles or behaviors. Four dummy codes were used for family arrangements (lone single
mother as the reference group): two parents (the focal child’s biological father lives in the
household), two parents (the mother lives with a new partner), single mother living with grown
child, and single mother with living kin. One single dummy code was used for gender (1 =
female). The way to create dummy codes for acculturation was discussed in the following
section of measures. Results based on logistic or multinomial logistic regression analysis would
show what factors facilitate mothers’ knowledge over time (i.e., constantly high, increasing, or
slowly declining) and what factors impede mothers’ knowledge over time (i.e., declining or

56

constantly low). The impediments and facilitators for the two ethnicities were compared
qualitatively.
Two moderating effects were focused: a) the moderating effect of grandmothers’
childrearing involvement on associations between trajectory groups of mothers’ knowledge and
levels of their work hours, and b) two-way interactions of household arrangement with male
adult by child gender for the outcome variable of mothers’ monitoring knowledge. For a), the
present study examined the moderating effect of average levels of grandmothers’ childrearing
involvement from Time 1 to Time 3. For b), the household arrangement with male adult was
coded into lone single mothers, two parents (the focal child’s biological father lives in the
household), and the single mothers living with a male partner or other male adult figures (i.e.,
grown children and relatives). The single mothers living with a male partner and those with other
male adult figures were combined into one category since a small number of mothers living with
grown male children and male relatives.
Six hypotheses were tested:
H1: In terms of family environment, it is expected that for both ethnicities, mothers maintain
high levels of knowledge over time for the adolescents perceiving the following characteristics
of family environment in averagely high levels from Time 1 to Time 3: 1.a) mothers’
authoritative parenting, 1.b) mothers’ harsh discipline, 1.c) mothers’ behavioral control, and 1.d)
well-established family routines. One hypothesis is proposed for mothers’ work hours: 1.f)
mothers’ averagely long work hours from Time 1 to Time 3 are linked to consistently low or a
decline in their monitoring knowledge over time.
H2: In terms of neighborhoods, ethnic differenced are expected. Specifically, 3.a) the averagely
high levels of neighborhood problems and 3.b) the averagely low levels of neighborhood

57

cohesion are linked to consistently low or decreasing levels of mothers’ monitoring knowledge
for African Americans, but not for Latinos.
H3: In terms of adaptive culture factors, it is expected that for both ethnicities, mothers will
maintain high levels of knowledge over time when mothers receive averagely high levels of
grandmothers’ childrearing involvement from Time 1 to Time 3.
H4: In terms of adolescents’ characteristics, it is expected that for both ethnicities, mothers’
knowledge will remain higher when adolescents are girls.
H5: For both ethnicities, associations between average low levels of mothers’ knowledge and
averagely long work hours from Time 1 to Time 3 are weaker for mothers receiving averagely
high levels of grandmothers’ childrearing involvement from Wave 1 to Wave 3.
H6: For both ethnicities, lone single mothers will gain less knowledge about boys than the
married mother (living with the focal child’s biological father) or the single mothers living with a
male partner or other male adult figures (i.e., grown child and kin) in the household.
3.1) How is mothers’ longitudinal monitoring knowledge associated with adolescent adjustment
in terms of various domains, including externalizing and internalizing behavior problems,
and school delinquency in late adolescence among adolescents in low-income households
for African Americans and Latinos?
3.2) Are there any differences in patterns of associations between African and Latino Americans
over time?
Associations between trajectory groups of mothers’ monitoring knowledge and
adolescent adjustment at Time 3 were examined for both ethnicities. This analysis addresses how
change in mothers’ knowledge over time predicts adolescents’ adjustment in late adolescence.
Second, the present study compared whether the two ethnicities were different in the types of the

58

adolescent adjustment that is significantly predicted by mothers’ monitoring knowledge
qualitatively.
One hypothesis was tested:
H7: For both ethnicities, it is surmised that consistently high levels of mothers’ knowledge
predict the following adolescent outcomes at Time 3: 7.a) low levels of externalizing behavior
problems, 7.b) low levels of internalizing behavior problems, and 7.c) low levels of school
delinquency. The present study did not surmise ethnic differences since the previous study has
not yet evidenced.
4.5 Measures
The Three-City Study used many existing measures and created measures for assessing
low-income households’ family environment, neighborhood conditions, and children’s
adjustment. The focus measures reported by the mothers included: behavioral control,
authoritative parenting, family routine, their works’ hours per week, mothers’ linguistic
acculturation, grandmothers’ childrearing involvement, neighborhood problems, neighborhood
cohesion, and internalizing behavior problems; the focus measures reported by the adolescents
included: mothers’ monitoring knowledge, harsh discipline, externalizing behavior problems,
and school delinquency. The focused measures were discussed as follows. All the mean of the
following measures at each time were averaged to create the average levels of the measures from
Time 1 to Time 3 for further analyses, with the exception of maternal monitoring knowledge,
mothers’ work hours, and adolescents’ internalizing and externalizing behavioral problems and
delinquency. The ways of transformation of the above five variables are discussed in the
respective sections.

59

Maternal Monitoring Knowledge. The Three-City Study assessed maternal monitoring
knowledge by using the measure of parental monitoring knowledge developed by Steinberg,
Mounts, Lamborn, & Dornbusch (1991). Five items assessed adolescents’ perception of how
much the mother knows about their friends, activities after school, activities at night, the use of
free time, and money use. Responses were given along a 3-point Likert-type scale, including: 0 =
doesn’t know, 1 = knows a little, and 2 = knows a lot. The final factor solution resulting from the
exploratory factor analyses (EFAs) and confirmatory factor analyses (CFAs) indicates that all
items were loaded on one latent factor and the loadings of all the items were equivalent across
ethnicities. All the five items were averaged to create the mean maternal monitoring knowledge
at each time for both ethnicities. The mean maternal monitoring knowledge at each time was
used for further analyses of growth mixture modeling.
Harsh Discipline. The Three-City Study developed the measure of maternal punishment based on
the guidance of psychological experts in family functioning (Winston, 1999). Across all the time
points, five maternal-reported items on punitive parenting were used to assess maternal
punishment. The five items assessed youth perception of whether the mother scolded, spanked,
threatened to spank, punish her child if the child didn’t behave, threatened to put the child out of
the house during the past 12 months. Responses were given along a 6-point Likert-type scale,
including: 0 = never in past 12 months, 1 = several times in the past 12 months, 2 = once a month,
3 = several times a month, 4 = once a week, and 5 = more than once a week. The final factor
solution resulting from the EFAs and CFAs excluded two items that had equivalent loadings on
the maternal punishment factor across times within each ethnicity but had different loadings
between the African Americans and the Latinos. The two excluded items were “How often
caregiver spanks/hits the child” and “How often caregiver threatened to kick the child out”. The

60

rest three items were averaged to create the mean maternal punishment at each time for both
ethnicities.
Authoritative Parenting. The Three-City Study developed the measure of authoritative parenting
based on the guidance of psychological experts in family functioning (Winston, 1999). The
present measure of authoritative parenting assessed two primary elements of authoritative
parenting style, i.e., parental warmth and control, which is in line with the definition of
Baumrind (1971). At Time 1 and 2, authoritative parenting was assessed with five items, with
two items measuring parental warmth (including praising and understanding the child’s feeling)
and three items measuring parental control (including making clear rules, explaining the reason
of setting rules, and having strict punishment while the child was misbehaved). At Time 3, the
items assessing praising and making clear rules were dropped in the Three-City Study. Responses
were given along a 4-point Likert-type scale, including: 0 = definitely false, 1 = sort of false, 2 =
sort of true, and 3 = definitely true. In order to have the same measurement construct across
times, only the three items administrated at all the time points were used for further analyses.
The final factor solution resulting from the EFAs and CFAs indicates that all the three items
were loaded on one latent factor and the loadings of all the items were equivalent across items
for both ethnicities. The three items were averaged to create the mean authoritative parenting at
each time for both ethnicities.
Behavioral Control. The Three-City Study did not directly assess mothers’ perception of
restrictive parenting, but assessed their perception of permissive parenting. The measure of
permissive parenting was developed based on the guidance of psychological experts in family
functioning (Winston, 1999). This study used the reversed score of permissive parenting as the
index of the levels of mothers’ restrictive parenting. The measure of permissive parenting was

61

only administrated at Time 1 and 2. Across the two time points, permissive parenting was
assessed with five items; some examples included: “I avoid giving the child chores” and “I drop
a rule if the child objects to it.” Responses were given along a 4-point Likert-type scale,
including: 0 = definitely false, 1 = sort of false, 2 = sort of true, and 3 = definitely true. The final
factor solution resulting from the EFAs and CFAs indicates that all the five items were loaded on
one latent factor and the loadings of all the items were equivalent across items for both
ethnicities. Thus, the five items were reversed and averaged to create the mean behavioral
control at each time for both ethnicities.
Family Routine. The Three-City Study developed the measure of family routine based on the
family routine inventory (Jensen et al., 1983). Across all the time points, five items were
administrated to assess mothers’ perception of how often their family has family time and eat
breakfast and dinner at the same time, her child and she have homework/nighty rituals, and the
child has the same bedtime. Responses were given along a 4-point Likert-type scale, including: 0
= almost never, 1 = sometimes, 2 = usually, and 3 = always. The final factor solution resulting
from the EFAs and CFAs indicates that all the five items were loaded on one latent factor and the
loadings of all the items were equivalent across items for both ethnicities. Thus, the five items
were averaged to create the mean mother-child anger and alienation at each time for both
ethnicities.
Mothers’ Overtime Work. The Three-City Study asked the mother how many hours she usually
worked per week at main job at each time point. Since the present study focused on whether the
mother worked overtime, the raw work hours per week was recoded into a dummy variable—the
status of overtime work—by coding 0 for work hours below 40 and 1 for work hours larger than
40 at each time point. Forty-hour was chosen as the threshold because working over 40 hours per

62

week is defined as overtime work (United States Department of Labor, n.d.). The status of
overtime work across times were summed and used as an index for the degrees of overtime work
from Time 1 to Time 3.
Mothers’ Linguistic Acculturation. Because the Latino sample in the present study was diverse in
linguistic use and fluency in English, an index of English fluency and preference was served as a
proxy for acculturation. English fluency and preference has often been used as an indicator of
acculturation in many acculturation measures (e.g., Cuellar, Harris, & Jasso, 1980; Marin &
Gamba, 1996) or a proxy for acculturation. This is because immigrants tend to prefer English to
Spanish if they are highly acculturated to the US (Griffith & Villavicencio, 1985; Krause,
Bennett, & Van Tran, 1989). Based on the suggestion of Loukas and colleagues (2007), the
present study used a two-step process to infer levels (high versus low) of mothers’ linguistic
acculturation. First, the mothers reporting English as their native language were classified into
high-acculturation group. The mothers reporting English as not their native language (N = 230)
were further inferred to levels of linguistic acculturation based on their fluency in English.
Mothers were asked how well they could speak, read, and write along a 4-point Likert-type scale
ranging from 0 (“not at all”) to 3 (“very well”). The present study used the criterion of Loukas et
al. (2007) in which mothers who scored 2 or higher in average were classified into highacculturation group. Among those with English as not their native language, 81 were classified
into high acculturation. Overall, 208 Latino mothers were classified into high-acculturation
group and 146 were classified into low-acculturation group. The high-acculturation mothers tend
to be US-born, with 150 US-born of the 208 high-acculturation mothers. Similarly, the lowacculturation mothers tend to be foreign-born, with 131 foreign-born of the 146 lowacculturation mothers. The levels of linguistic acculturation for the African American sample

63

was not used for further analyses because there was too little variability in acculturation among
the African American participants, with 1.6% classified as low-acculturation group.
Grandmothers’ Childrearing Involvement. In the Three-City Study, one item was used to assess
whether the child’s grandmother was involved in childrearing. That item asked the mother “How
much responsibility does the grandmother take in the child’s daily care, such as preparing food
for the child or helping him/her get dressed?”. The response was given along a 3-point Likerttype scale, including: 0 = none, 1 = some responsibilities, and 2 = complete responsibilities. The
raw value at each time was averaged to produce the average levels of grandmothers’ childrearing
involvement from Time 1 to Time 3 for further analyses.
Neighborhood Problems. The Three-City Study developed a 11-items measure of neighborhood
problems. Neighborhood problems were measured by the mother’s ratings of problems (e.g.,
unsafe street, unemployment, burglaries and thefts, gangs, and unsupervised children) in her
neighborhood. Responses were given along a 3-point Likert-type scale, including: 0 = not a
problem, 1 = somewhat a problem, and 2 = a big problem. The final factor solution resulting
from the EFAs and CFAs indicates that all the eleven items were loaded on one latent factor and
the loadings of all the items were equivalent across items for both ethnicities. Thus, all the items
were averaged to create the mean neighborhood problems at each time for both ethnicities.
Neighborhood Cohesion. Robert Sampson’s Collective Efficacy Scale (Sampson, Raudenbush,
& Earls, 1997) was administrated to the mother to assess social cohesion of neighborhood. At
each time, four items of this scale measured levels of perceived social cohesion (e.g., “People
around here are willing to help neighbors,” and “people in this neighborhood can be trusted”).
Responses were recorded along a 4-point Likert-type scale ranging from 0 (“strongly disagree”)
to 3 (“very likely”). The final factor solution resulting from the EFAs and CFAs indicates that all

64

the four items were loaded on one latent factor and the loadings of all the items were equivalent
across items in both groups. Thus, all the items were averaged to create the mean neighborhood
cohesion at each time across the entire sample.
Internalizing Behavioral Problems. At Time 3, three subscales of the Child Behavior Checklist
(CBCL; Achenbach & Rescorla, 2001a) for ages 6-18 were used to assess mothers’ perception of
levels of their children’s internalizing behavioral problems. The three subscales were anxious,
withdrawn, and somatic complaints. For the adolescents aged above 18, the same three subscales
of the Adult Behavior Checklist (AC; Achenbach & Rescorla, 2001b) were administrated to the
adolescents to assess their perception of anxious and withdrawn problems and somatic
complaints. The investigators of the Three-City Study summed the three subscales, created the tscore of internalizing behavioral problems, and transformed the t-score to create clinical
classification for internalizing behavioral problems. Scores below 95th percentiles are considered
in the normal range, scores between the 95th and 98th percentiles are considered in the
borderline clinical range, and scores in the 98th-100th percentile are considered in the clinical
range. The clinical classification for internalizing behavioral problems was used for further
analyses.
Externalizing Behavioral Problems. The Three-City Study assessed adolescents’ externalizing
behavioral problems by adapting the items from the National Longitudinal Study of Youth
(Borus, Carpenter, Crowley, & Daymont, 1982) and the Youth Deviance Scale (Gold, 1970;
Steinberg, Mounts, Lamborn, & Dornbusch, 1991). At each time, the adolescents were
administered six items that assessed deviant behavior including stealing, getting in trouble with
the police, damaging property, carrying a weapon, attacking someone, and fighting during the
past year. The adolescents were also administered another five items that assessed the use of

65

cigarettes, alcohol, marijuana, and hard drug during the past year. Responses were given along a
4-point Likert-type scale, including: 0 = never, 1 = once or twice, 2 = several times, and 3 =
often. The Three-City study did not provide information on each single item at Time 3. Thus, the
EFAs and CFAs were not allowed to be conducted. Yet, the investigators of the Three-City Study
conducted factor analysis, which indicates that the deviant behavior and the drug use subscales
were loaded into one factor. Accordingly, they averaged the z-scores of all the eleven items to
create one average z-score as an index for externalizing behavioral problems. The present study
used this average z-score for further analyses.
School Delinquency. The Three-City Study assessed adolescents’ school delinquency by
adapting the items from the National Longitudinal Study of Youth (Borus, Carpenter, Crowley,
& Daymont, 1982) and the Youth Deviance Scale (Gold, 1970; Steinberg, Mounts, Lamborn, &
Dornbusch, 1991). At each time, the adolescents were administrated with five items that assessed
school delinquency including the following behaviors: skipping school, copying homework,
being suspended or expelled, getting detentions, and cheating during the past year. Responses
were given along a 4-point Likert-type scale, including: 0 = never, 1 = once or twice, 2 = several
times, and 3 = often. Like the measure of externalizing behavioral problems, the Three-City
study did not provide information on each single item at Time 3. Thus, the EFAs and CFAs were
not allowed to be conducted. The investigators of the Three-City Study provided the mean of zscores of all the five items, which was used in the present study.
4.6 Measurement Invariance Tests
Exploratory factor analyses (EFAs) and multiple-group confirmatory factor analyses
(CFAs) were conducted to determine whether measurement constructs were similar or invariant
between African Americans and Latinos for a series of measures for family environment (i.e.,

66

maternal monitoring, parenting styles/behaviors, and family routines) and neighborhood context
(i.e., neighborhood problems and cohesion). The analyses were conducted by Mplus 7. A number
of prior research has noted that the linear SEM (structural equation modeling) approach can
produce biased estimation of fit indices, factor loadings, and spurious factors when an ordered
categorical scale has small number of scale points (Bauer, 2005; Yang & Green, 2011). The
research also suggests the nonlinear SEM approach should be chosen for ordered categorical
scales. Accordingly, the nonlinear SEM approach was used to conduct the EFAs and the CFAs.
The EFAs were conducted for African Americans and Latinos separately to explore the
measurement constructs. The results of EFAs showed that all measures for family environment
and neighborhood context have one similar overarching factor among both groups. Next,
multiple-group CFAs were conducted to determine equivalence in item loadings and thresholds.
A series of nonlinear SEMs with different levels of invariance was conducted to determine
whether the focus measures are configural invariance (with the same number of factors), weak
invariance (with the same units of loading), and strong invariance (with the same thresholds of
scale points in indicators). A series of chi-square difference tests were computed to determine
whether two adjacent models are similar. Non-significant chi-square differences suggest that the
two adjacent models are similar and higher level of invariance is present. If strong invariance is
violated, partial invariance (i.e., item loadings are equivalent but thresholds of scale points in
some items are not equivalent) is tested and chi-square difference test is also conducted to
determine equivalence between configural invariance and partial invariance models. The results
of CFAs showed that all the measures are partial invariance (i.e., item loadings are equivalent but
thresholds of scale points in some items are not equivalent), with the exception of neighborhood
cohesion that is strong invariance. Partial invariance was acceptable for all the measures given

67

that comparing ethnicities in factor means for the measures was not the focus of the present study
(Bauer, 2005; Ployhart & Oswald, 2004). The exceptional measure was mothers’ monitoring
knowledge. Given that the present study estimated the trajectory of change in mothers’
monitoring knowledge across three time points for each ethnicity, comparing the relative degrees
of mothers’ monitoring knowledge provides important information for estimating the initial level
(the intercept) and the rate of change (the slope). Although chi-square difference test showed that
the strong invariance and weak invariance models were significantly different, the results of the
fit indices indicated that strong invariance model is good fit (RMSEA ≤ .07 and CFI ≥ .95; Hu &
Bentler, 1999; Steiger, 2007). This suggests that strong invariance model for mothers’
monitoring knowledge was well specified and able to represent the factorial structure. The fit
indices are shown in Table 2 for the measures of mothers’ monitoring knowledge and parenting
styles/behaviors, and in Table 3 for family routine and neighborhood factors.
4.7 Reliabilities of the Measures for Family Environment and Neighborhood Context
Yang and Green (2011) point out that coefficient alpha can underestimate scale reliability
because the three underlying assumptions are likely to be violated. These assumptions are: 1)
classical item-score assumption: “Item scores are a simple sum of item true and error scores.”; 2)
tau equivalency assumption: “The same true scores underlie all items and equally contribute to
all item scores. Within the framework of factor analysis, items have equal loadings on a single
underlying factor.”; and 3) uncorrelated errors assumption: “Item error scores between any pair
of items are uncorrelated.”(p.379; Yang & Green, 2011). Yang and Green also indicated that the
SEM approach provides more precise estimation of scale reliability because it allows the above-

68

Table 2: Measures of Fit for Mothers’ Monitoring Knowledge and Parenting Styles/Behaviors
for Multiple-Group Confirmatory Factor Analyses Among African Americans and Latinos
Across Times
Variable
Mothers’ monitoring
knowledge
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value
Harsh Discipline
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value
Authoritative
parenting
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value
Behavioral Control
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value

Configural
Invariance

Weak
Invariance

Strong
Invariance

Partial
Invariance

193.44
144
.99
.03

206.59
154
.99
.03
13.15
10
.215

346.20
198
.97
.05
139.61
44
.001

224.12
168
.99
.03
30.68
24
.163

64.87
30
.98
.06

70.59
36
.98
.05
5.72
6
.455

541.69
91
.77
.12
471.10
55
.001

89.74
51
.98
.05
24.87
21
.253

47.51
30
.98
.03

64.19
36
.97
.05
16.68
6
.011

155.83
76
.92
.06
91.65
40
.001

81.89
51
.97
.04
34.38
21
.033

146.14
59
.96
.07

146.94
63
.96
.06
.80
4
.938

219.50
105
.95
.06
72.56
42
.002

159.08
75
.96
.06
12.14
16
.734

69

Table 3: Measures of Fit for Family Routine and Neighborhood Context for Multiple-Group
Confirmatory Factor Analyses Among African Americans and Latinos Across Times
Variable
Family routine
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value
Neighborhood
problems
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value
Neighborhood
cohesion
Chi-square
df
CFI
RMSEA
ΔChi-square
Δ df
p value

Configural
Invariance

Weak
Invariance

Strong
Invariance

Partial
Invariance

272.09
144
.96
.05

278.80
154
.96
.05
6.71
10
.753

523.49
224
.91
.06
244.69
80
.001

309.58
165
.96
.05
30.78
21
.077

1287.70
918
.98
.03

1273.52
940
.99
.03
14.18 a
22
.895

1590.83
1045
.98
.04
317.31
105
.001

1296.12
965
.99
.03
8.42
47
1.000

161.1
78
.99
.06

165.9
86
.99
.05
4.8
8
.779

261.94
141
.99
.05
96.04
55
.001

189.26
111
.99
.05
23.36
33
.893

70

mentioned assumptions to be violated. Accordingly, the SEM approach was conducted to
compute the reliabilities of the measures for all the family environment and neighborhood
context.
The SEM approach in estimating reliability involves two steps: 1) assessment of CFA
models: “a series of CFA models are evaluated to determine the most appropriate model based
on a researcher’s understanding of the scale of interest and the empirical fit of the models to the
item data.”; and 2) estimation of the reliability coefficient: “using the parameter estimates from
the model chosen in Step 1.” (p. 384, Yang & Green, 2011). Following the above steps, a series
of CFAs were conducted to produce parameter estimates for all the measures across times for
African Americans and Latinos. In line with the previous discussion in the section of the
measurement invariance tests, the nonlinear SEM approach was used to conduct CFAs. Since the
formula for computing the reliability coefficient based on nonlinear SEM approach is complex,
Green and Yang’s (2009) SAS-program-based syntax was used to calculate the coefficient. All
the coefficients are shown in Table 4 and 5 for African Americans and Latinos, respectively. The
coefficient alphas were also reported. The reliabilities of all the measures were considered
adequate for both ethnicities (above .70 for the measures with many items and .65 for the
measures with few items; Kaplan & Saccuzzo, 2009; Schmitt, 1996), with the exception of
authoritative parenting across all time points and harsh discipline at Time 1 that were considered
low.

71

Table 4: Reliabilities of the Major Variables for African Americans

Variable
Mothers’ monitoring
knowledge
Harsh discipline
Authoritative parenting
Behavioral control
Family routines
Neighborhood problems
Neighborhood cohesion

African Americans
Time1
Nonlinear
Alpha SEM
.71
.71
.59
.48
.51
.64
.89
.85

.61
.54
.66
.66
.91
.86

Time2
Alpha
.71

Time3
Nonlinear
SEM
.73

.66
.61
.59
.66
.91
.82

.69
.63
.69
.69
.93
.83

Alpha
.82
.66
.50
N/A
.70
.91
.87

Nonlinear
SEM
.85
.67
.60
N/A
.72
.93
.87

Table 5: Reliabilities of the Major Variables for Latinos
Latinos
Time1
Variable
Mothers’ monitoring
knowledge
Maternal punishment
Authoritative parenting
Behavioral control
Family routines
Neighborhood problems
Neighborhood cohesion

Alpha
.69
.63
.55
.62
.69
.91
.86

Time2
Nonlinear
SEM
.71

Alpha
.77

.63
.63
.73
.72
.93
.86

.65
.60
.60
.64
.91
.86

72

Time3
Nonlinear
SEM
.80
.65
.62
.67
.67
.93
.86

Alpha
.83
.63
.61
N/A
.71
.91
.87

Nonlinear
SEM
.85
.65
.62
N/A
.74
.93
.87

4.8 Reliabilities of the Measures for Adolescents’ Adjustment Outcomes
Both open accessible and restricted datasets of the Three-City Study did not provide raw
data for each item of internalizing and externalizing behavioral problems and school delinquency.
Thus, the scale reliability was not allowed to be computed. Yet, the open accessible documents
of the Three-City Study provided information on the Cronbach’s Alphas for the scales of
externalizing behavioral problems (α = .80) and school delinquency (α = .67) at Time 3 for all
the adolescent participants. The documents did not provide information on the Cronbach’s Alpha
for the scale of the sum of internalizing behavioral problems at Time 3, but offered the Alphas
for the subscales of internalizing behavioral problems at Time 3, which range from .71 (for
somatic complaint) to .79 (for anxiety) for CBCL internalizing behavioral problem and from .76
(for withdrawn) to .85 (anxiety) for ABCL internalizing behavioral problem. The reliabilities of
all the scales were considered adequate.

73

Chapter 5: Results
5.1 Descriptive Statistics of the Major Variables
The descriptive statistics of the major variables are shown in Table 6 and 7 for the
African Americans and the Latinos, respectively. The mean and SD for internalizing behavior
problems were not reported for both ethnicities because the present study used its clinical
classification. For the African American adolescents, 78.4% were classified into the normal
range, 9.3% were into the borderline range, and 12.3% were into the clinical range. For the
Latino adolescents, 71.8% were classified into the normal range, 7.4% were into the borderline
range, and 20.8% were into the clinical range. With the exception of the adolescents’ outcome
variables, all the variables were averaged across the three time points for further analyses. Two
out of the three time points needed to be valid, non-missing to create the average score for all the
variables. The percentages of missing data for all the average score of the focus variables ranged
between 0% and 29.7% for the African Americans and 0% and 33.1% for the Latinos. Full
Information Maximum Likelihood (FIML; Allison, 2001) estimation was used for all the
following analyses in order to include as many participants as possible in the study.
5.2 Model Selection and the Specification of the Trajectory Groups for Mothers’ Monitoring
Knowledge
Muthén and Muthén (2002) suggest four criteria for choosing the number of latent
trajectory classes. First is the Bayesian Information Criteria (BIC), a fit index used to compare
relative model fit across models. BIC is defined as follows:
BIC = −2 log L+p ln(n), (1)
Where L is the loglikelihood value of the model, p is the number of parameters and n is the
sample size. Lower BIC suggests better model fit. The second criterion is the Lo–Mendell–Rubin

74

Table 6: Means and SDs of the Major Variables for African Americans

Variable
Mothers’ monitoring knowledge
Harsh Discipline
Authoritative parenting
Behavioral Control
Family routines
Mothers’ work hours/per week
Grandmothers’ Childrearing
Involvement
Neighborhood problems
Neighborhood cohesion
Internalizing Behavior Problems
Externalizing Behavior Problems
School Delinquency

Time1
M
1.63
1.18
2.60
2.24
1.75
16.85
.52

SD
.39
.86
.47
.58
.65
19.13
.63

.90
1.31
NA
NA
NA

.56
.86
NA
NA
NA

n
318
312
319
319
319
299
245

Time2
M
1.60
1.21
2.60
2.22
1.62
19.50
.39

SD
.42
.94
.45
.59
.67
19.08
.57

319
315
NA
NA
NA

.85
1.32
NA
NA
NA

.57
.83
NA
NA
NA

75

n
301
309
303
303
303
298
229

Time3
M
1.51
.78
2.56
NA
1.52
19.83
.32

SD
.50
.74
.48
NA
.70
18.42
.55

n
318
319
305
NA
296
311
197

303
292
NA
NA
NA

.80
1.41
NA
-.14
-.22

.59
.87
NA
.42
.62

304
288
319
319
319

Table 7: Means and SDs of the Major Variables for Latinos

Variable
Mothers’ monitoring knowledge
Harsh Discipline
Authoritative parenting
Behavioral Control
Family routines
Mothers’ work hours/per week
Grandmothers’ Childrearing
Involvement
Neighborhood problems
Neighborhood cohesion
Internalizing Behavior Problems
Externalizing Behavior Problems
School Delinquency

Time1
M
1.62
1.12
2.62
1.89
1.81
16.85
.52

SD
.39
.87
.47
.65
.69
19.13
.63

.78
1.34
NA
NA
NA

.57
.85
NA
NA
NA

n
354
348
354
354
354
348
241

Time2
M
1.64
1.03
2.55
1.83
1.74
19.51
.39

SD
.40
.81
.48
.62
.65
19.08
.57

352
340
NA
NA
NA

.73
1.36
NA
NA
NA

.56
.87
NA
NA
NA

76

n
327
328
327
327
327
322
228

Time3
M
1.56
.71
2.49
NA
1.57
19.83
.32

SD
.47
.67
.57
NA
.70
18.42
.55

n
354
354
343
NA
339
352
210

326
312
NA
NA
NA

.69
1.57
NA
-.12
-.19

.58
.81
NA
.41
.65

341
316
354
354
354

likelihood ratio test (LMR-LRT). LMR-LRT is a likelihood ratio–based method for testing k − 1
classes against k classes. LMR-LRT does not require the premise that models are nested, which
is a necessary condition in the classic chi-square testing. The application of LMR-LRT
has been criticized (Jeffries, 2003), but it is still unclear about what factors affect its use in
practice. The third criterion is entropy. Entropy is an index to summarize the posterior
probability of membership in the most likely latent class for one (Collins & Lanza, 2010).
Entropy ranges between 0 and 1. There is no clear criterion about to which extent entropy is high
enough. The higher entropy indicates better latent class separation (Celeux & Soromenho, 1996).
The fourth criterion is to consider the usefulness and interpretability of latent trajectory classes.
Some considerations include whether there are distinct qualitative differences across identified
latent classes.
Table 8 shows indexes of model fit for the African American mothers. For the African
American mothers, only the indexes for 1-class, 2-class, and 3-class model were reported
because models more than three classes did not converge, indicating poor model-fit. For the
African American mothers, the values of BIC for 3-class model were smaller than the values of
BIC for 1-class and 2-class model. Entropy of 3-class model (=.78) were acceptable. LMR-LRT
was not reported since Mplus software cannot compute this index when RANDOM was chosen
as the type of analysis. Additionally, the posterior probabilities (diagonal values) of 3-class
model are all reasonably high, .90 for the smaller-size group, .77 for the middle-size group,
and .93 for the larger-size group (Table 9; the proportions of class 1, class 2, and class3
= .191, .732, and .075). Although the entropy of 2-class model (=.86) was higher than the
entropy of 2-class model, 3-class model was selected given that its value of BIC was smaller and

77

there were distinct qualitative differences across the identified three latent classes for the African
American mothers.
Table 8: Summary of Information for Selecting Number of Trajectory Classes for African
American Mothers
Number of
Classes
1
2
3
4

Loglikelihood AIC

BIC

-454.32
920.65
-424.93
867.86
-411.67
847.33
Not well identified

943.22
901.72
892.48

Sample-size
Adjusted BIC
924.19
873.17
854.42

Entropy
N/A
.86
.78

Table 9: Average Latent Class Probabilities for Most Likely Latent Class Membership (Row)
by Latent Class (Column) for African American Mothers

1
2
3

1
.769
.060
.057

2
.173
.934
.041

3
.057
.006
.902

Table 10 shows indexes of model fit for the Latino mothers. For the Latino mothers, the
values of BIC for 2-class model were smaller than the values of BIC for 1-class and 3-class
model. Like the African American mothers, LMR-LRT was not reported. Entropy of 2-class
model (=.78) was acceptable for the Latino mothers. Additionally, the posterior probabilities
(diagonal values) of 2-class model are all reasonably high, .84 for the smaller-size group and .95
for the larger-size group (see Table 11; the proportions of class 1 and class 2 = .112 and .887).
Although the entropy of 3-class model (=1.00) was higher than the entropy of 2-class model, two
of the posterior probabilities of 3-class model are zero, indicating 3-class model does not fit the
data of the Latino mothers. The combined information suggests good model fit for 2-class model
for the Latino mothers.

78

Table 10: Summary of Information for Selecting Number of Trajectory Classes for Latino
Mothers
Number of
Classes
1
2
3

Loglikelihood AIC

BIC

-507.75
-478.86
-515.20

1050.72
1010.54
1100.83

1027.51
975.72
1054.40

Sample-size
Adjusted BIC
1031.69
981.99
1062.76

Entropy
N/A
.78
1.00

Table 11: Average Latent Class Probabilities for Most Likely Latent Class Membership (Row)
by Latent Class (Column) for Latino Mothers
1
.842
.053

1
2

2
.158
.947

5.3 Description of the Trajectories of Mothers’ Monitoring
Figure 4 shows distinct patterns of the level of mothers’ monitoring knowledge from
age 10 to 20 for both the African American and Latino mothers (see Panel A and B for African
Americans and Latinos, respectively). Among the African American mothers, 73.3% were
classified into stable high monitoring group given that the level of mothers’ monitoring started
high and maintained high over time; 19.2% were classified into high declining monitoring group
given that the level of mothers’ monitoring started high but declined quickly over time; and 7.5%
were classified into stable low monitoring group given that the level of mothers’ monitoring
starts low and maintained low over time. The estimated values of intercepts and slopes are shown
in Table 12.

79

Table 12: Intercepts and Slopes of Identified Latent Classes of Mothers’ Monitoring Knowledge

African Americans
Stable High Monitoring Knowledge Class
High Declining Monitoring Knowledge Class
Stable Low Monitoring Knowledge Class
Latinos
Slowly Declining Monitoring Knowledge Class
Low Class Monitoring Knowledge Class
*** p< .001; + p < .10.

Intercept
M

SE

Slope
M

SE

1.719***
2.063***
.815***

.037
.072
.154

.001
-.130***
.011

.001
.016
.026

1.832***
.853

.155
.603

-.028+
.054

.015
.052

Among the Latino mothers, 88.8% were classified into slowly declining monitoring
group given that the level of mothers’ monitoring started high and declined slightly over time
(the point estimate of slope = -.028, p < .072, in a marginal significance level). The rest 11.2%
were classified into low monitoring group given that the level of mothers’ monitoring started low
and maintained low over time (the point estimate of slope = .054, p < .293, which was not
significant). Together, there exist both similarities and differences in the trajectory patterns of
mothers’ monitoring between the African American and Latino mothers. Specifically, the
majority of the mothers maintained high levels of monitoring, while very few mothers had low
levels of monitoring over time in both ethnicities. However, the two ethnicities were also
different in that about one-fifth of the African American mothers’ monitoring declined quickly
over time, while this group was not identified among the Latino mothers.

80

Mothers' Monitoring

A

Stable High
Monitoring
(73.3%)

2.2
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0

High
Declining
Monitoring
(19.2%)
Stable Low
Monitoring
(7.5%)

Age
10 11

12 13 14

15 16 17 18

19 20

Mothers' Monitoring

B

2.2
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0

Low
Monitoring
(11.2%)
Slowly
Declining
Monitoring
(88.8%)

Age
10 11 12 13 14 15 16 17 18 19 20

Figure 4: Trajectories of mothers’ monitoring. A. Frequency for trajectories of mothers’
monitoring for African Americans. B. Frequency for trajectories of mothers’ monitoring for
Latinos.

81

5.4 Correlates of Mothers’ Monitoring Knowledge
For both ethnicities, logistic or multinomial logistic regression analyses were conducted
to identify correlates of mothers’ monitoring knowledge from the family environmental factors
(i.e., harsh discipline, authoritative parenting, behavioral control, family routine, and mothers’
working hours), neighborhood factors (i.e., neighborhood problems and neighborhood cohesion),
adaptive cultural factor (i.e., grandmothers’ childrearing involvement), and adolescent gender.
For the African American families, since three trajectory groups of mothers’ knowledge were
identified, multinomial logistic regression analysis was utilized to examine relations between the
likelihood of belonging to the stable high monitoring knowledge group versus the stable low
group or the high declining group and average levels of all the above-mentioned factors
simultaneously. For the Latino families, given that only two trajectory groups of mothers’
knowledge were identified, logistic regression analysis was utilized to examine relations between
the likelihood of belonging to the slowly declining monitoring knowledge class versus the low
group and the average levels of all the above-mentioned factors simultaneously. For both
ethnicities, mothers’ education level and types of family arrangement were controlled. Only for
the Latinos, mothers’ acculturation was also controlled. The results are shown in Table 13.

82

Table 13: Logits, Standard Errors of Logits, and Odds Ratios From Logistic Regression for The Relations of Mothers’ Monitoring
Knowledge with Family Factors, Neighborhood Factors, Adaptive Cultural Factor, and Child Gender

Predictors
Family Factors
Harsh Discipline
Authoritative
Parenting
Behavioral
Control
Family Routine
Mothers’
Overtime Work
Neighborhood
Factors
Neighborhood
Problems
Neighborhood
Cohesion
Adaptive Cultural
Factor
Grandmothers’
Childrearing
Involvement
Adolescent Gender
(ref. male)

Latinos
Slowly Declining Class
(ref. Low Class)
Logit
OR
95% CI

Wald
Statistic

Stable High Class
(ref. High Declining Class)
Logit
OR
95% CI

Wald
Statistic

African Americans
Stable High Class
(ref. Stable Low Class)
Logit
OR
95% CI

Wald
Statistic

-.59*
.62

.56
1.87

.33, .94
.72, 4.87

4.79*
1.62

-.55*
-.40

.58
.67

.35, .95
.28, 1.63

4.66*
.78

-.83**
.50

.44
1.66

.26, .75
.59, 4.67

9.25**
.92

-.14

.87

.42, 1.79

.15

.27

1.32

.67, 2.48

.72

.71

2.02

.78, 5.29

2.09

.81*
-.14

2.25
.87

1.17, 4.32
.47, 1.63

5.89*
.19

.60*
-.65

1.82
.52

1.05, 3.15
.24,1.15

4.51*
2.63

.60
-.02

1.81
.98

.96, 5.65
.24, 4.08

1.47
.00

.03

1.03

.43, 2.48

.01

.30

1.35

.62, 2.95

.58

.59

1.80

.50, 6.49

.80

-.08

.93

.53, 1.62

.07

-.37

.69

.40, 1.20

1.73

.81*

2.27

1.02,
5.00

4.02*

-.20

.82

.45, 1.50

.40

.67+

1.22

.88, 4.27

2.73+

.09

1.10

.45, 2.67

.05

.26

1.30

.66, 2.57

.58

.52+

1.68

.93, 3.05

2.95+

.48

1.62

.73, 3.61

1.39

Note. 1. Mothers’ acculturation and education and household arrangement were controlled for the Latinos; mothers’ education and
family arrangement were controlled for the African Americans; and the estimates of the controlled variables were not shown in the
table. 2. *p < .05; ** p < .01; + p < .10.

83

5.4.1 Family Environment and Trajectories of Mothers’ Monitoring Knowledge
For the African American families, the results indicate that average high levels of harsh
discipline significantly predicted less likelihood of belonging to the stable high monitoring
knowledge group versus the stable low group or the high declining group. Specifically, with each
one unit increase in harsh discipline, the odds of the stable high monitoring knowledge versus the
stable low group decreased by 56% (OR = .44, p = .003) and the odds of the stable high
monitoring knowledge versus the high declining group decreased by 42% (OR = .58, p = .031).
These results suggest that harsh discipline was an impediment to mothers’ monitoring knowledge
over time. In contrast, another family factor—family routine—predicted higher likelihood of
belonging to the stable high monitoring knowledge group versus the high declining group.
Specifically, the odd of the stable high monitoring knowledge versus the high declining
monitoring knowledge increased by 82% with each one unit increase in family routine (OR =
1.82, p = .034). This result suggests that family routine was a facilitator to the African American
mothers’ monitoring knowledge over time. Authoritative parenting, behavioral control, and
mothers’ overtime work were not associated with the likelihood of belonging to the stable high
monitoring knowledge group versus the stable low group or the high declining group, suggesting
that all the above factors were neither facilitators nor impediments to mothers’ monitoring
knowledge.
A number of findings on the Latino families paralleled to the findings on the African
American families. First, earlier harsh discipline was also an impediment to the Latino mothers’
monitoring knowledge over time. Specifically, the odds of the slowly declining monitoring
knowledge versus the stable low group decreased by 40% with each one unit increase in harsh
discipline (OR = .56, p = .029), indicating that the average high levels of harsh discipline

84

predicted less likelihood of belonging to the slowly declining monitoring knowledge group
versus the low group. Second, family routine was also a facilitator to the Latino mothers’
monitoring. Specifically, the odds of the slowly declining monitoring knowledge versus the
stable low group increased by 125% with each one unit increase in family routines, indicating
that the average high levels of family routines predicted higher likelihood of belonging to the
slowly declining monitoring knowledge group versus the low group. In addition, similar to the
African American families, authoritative parenting, behavioral control, and mothers’ overtime
work were also not significantly associated with the likelihood of belonging to the slowly
declining monitoring knowledge group versus the low group for the Latino families. Accordingly,
the three factors were neither facilitators nor impediments for the Latino families.
A further analysis was conducted to test whether the effects of mothers’ overtime work
on their monitoring knowledge over time were moderated by grandmothers’ childrearing
involvement. This logistic/multiple logistic regression model included all the predictors and
controlled variables in Table 13 as well as the interaction of mothers’ work hours and
grandmothers’ childrearing involvement for both ethnicities. For both ethnicities, the interaction
of mothers’ work hours and grandmothers’ childrearing involvement was not significant. For the
African American families, OR of the stable high monitoring knowledge class versus the low
stable class = .67, 95% CI of OR [.04, 10.87], p = .775; and OR of the stable high monitoring
knowledge class versus the sharply declining class = .69, 95% CI of OR [.06, 7.69], p = .764. For
the Latino families, OR of the slowly declining monitoring knowledge class versus the low class
= .99, 95% CI of OR [.98, 1.00], p = .162. To evaluate whether the interactions of the mothers’
overtime work variable and grandmothers’ childrearing involvement adds predictability over and
above the main effects, further chi-square difference tests were conducted between the relax

85

model (in which interactions were freely estimated) and the constrained model (in which
interactions were constrained to be zero). Further chi-square difference test indicates that the
interaction of mothers’ overtime work and grandmothers’ childrearing involvement was not
significant for either the African Americans (∆χ2(∆df=2) = .17, p = .919) or the Latinos
(∆χ2(∆df=1) = 2.55, p = .110). This suggests that the effects of mothers’ overtime work on their
monitoring knowledge over time did not vary with average levels of grandmothers’ childrearing
involvement over time for both ethnicities.
5.4.2 Neighborhood Context and Trajectories of Mothers’ Monitoring
For the African American families, the results indicate that average high levels of
neighborhood cohesion significantly predicted more likelihood of belonging to the stable high
monitoring knowledge group versus the stable low group. The odds of the stable high monitoring
knowledge versus the stable low group increase by 127% with each one unit increase in
neighborhood cohesion (OR = 2.27, p = .046). In contrast, the likelihood of belonging to the
stable high monitoring knowledge versus the high declining group was not predicted by
neighborhood context. The present study did not find trajectory groups of mothers’ monitoring
knowledge were related to average levels of neighborhood problems. Together, these results
suggest that average high levels of neighborhood cohesion facilitated mothers’ monitoring
knowledge over time, but the average levels of neighborhood problems were neither facilitators
nor impediments. For the Latino families, the likelihood of belonging to the slowly declining
monitoring knowledge group versus the low group was not significantly associated with any
neighborhood context factors. This suggests that neighborhood context did not predict the levels
of mothers’ monitoring knowledge over time for the Latino families.

86

5.4.3 Adolescent Gender and Trajectories of Mothers’ Monitoring
For the African American mothers, the results indicate that the likelihood of belonging to
the stable high monitoring knowledge group versus the low stable group was not significantly
predicted by adolescent gender (OR = 1.62, p =.238). In contrast, the likelihood of belonging to
the stable high monitoring knowledge group versus the sharply declining group was predicted by
adolescent gender. Converting logit value of adolescent gender (.52) into odds ratio (1.68)
showed that being a girl increases the odds of having mothers with the stable high monitoring
knowledge versus the sharply declining knowledge by 68% (p = .086) with a marginal
significance level. In other words, the African American mothers who rear boys were more likely
to experience high declining monitoring knowledge over time. These findings suggest that
African American mothers may not be faced with challenges in gaining information about their
sons in early adolescence but are less likely to effectively supervise them between middle and
late adolescence. In contrast, for the Latino mothers, the likelihood of belonging to the slowly
declining monitoring knowledge group versus the low group was not significantly predicted by
adolescent gender (OR = 1.30, p = .447). The findings suggest that compared to those rearing
boys, the Latino mothers rearing girls experienced similar levels of monitoring knowledge over
time.
A further analysis was conducted to test whether the relation between child gender and
mothers’ monitoring knowledge over time were moderated by types of household arrangement
with male adult (i.e., comparing the mother living with the child’s biological father or male
adults versus the single lone mother). This logistic/multiple logistic regression model included
all the predictors and controlled variables shown in Table 13 and the interaction of adolescent
gender and the household arrangement with male adult for both ethnicities. The only difference

87

is that the family arrangement was replaced by the household arrangement with male coded with
two dummy variables, i.e., biological father (versus reference group—lone single mother) and
male adult (versus reference group— lone single mother). As shown in Table 14 and 15, the
interaction of adolescent gender and family arrangements was not significant for both ethnicities.
Further chi-square difference test indicates that the interaction of adolescent gender and family
arrangements was not significant for either the African Americans (∆χ2(∆df=4) = .72, p = .949)
or the Latinos (∆χ2(∆df=2) = 2.39, p = .303). The overall findings suggests that mothers’
monitoring knowledge about girls and boys were not different between the household headed by
the single lone mother, the mother living with the child’s biological father or male adults, for
both ethnicities.
5.4.4 Adaptive Culture and Trajectories of Mothers’ Monitoring Knowledge
For the African American mothers, the results indicate that the likelihood of belonging to
the stable high monitoring knowledge group versus the low stable group was not significantly
predicted by average levels of grandmothers’ childrearing involvement over time (OR of the
stable high monitoring knowledge class versus the low stable class = 1.64, 95% CI of OR [.58,
4.55], p = .354). Yet, the likelihood of belonging to the stable high monitoring knowledge group
versus the declining group marginally and significantly increased by 22% with each one unit
increase in average levels of grandmothers’ childrearing involvement (OR = 1.22, p = .098). This
suggests that among the mothers having initially high levels of monitoring knowledge, some of
them (those who were classified into the high declining group) were more likely to experience a
sharp decline in monitoring knowledge when they received low levels of childrearing support
from the grandmothers. For the Latino mothers, the likelihood of belonging to the slowly
declining monitoring knowledge group versus the low group was not significantly predicted by

88

Table 14: Hierarchical Multinomial Logistic Regression Assessing Interaction Between Adolescent Gender and Types of Family
Arrangements for African Americans

Models
Step1

Step 2

Predictors
Gender (ref. Male)
Biological Father (BF)
(ref. Lone Single
Mother)
Male Adult (MA) (ref.
Lone Single Mother)
Gender ×BF
Gender ×MA

Stable High Class
(ref. High Declining Class)
Logit
Odds Ratio 95% CI
+
.53
1.70
.93, 3.08
.07
1.07
.39, 2.97

Stable High Class
(ref. Stable Low Class)
Logit Odds Ratio 95% CI
.56
1.76
.78, 3.94
.35
1.42
.27, 7.52

.13

1.05

.56, 2.31

.89

2.43

.89, 7.25

-.84
-.22

.43
.80

.05, 3.70
.19, 3.36

-.09
-.37

.91
.69

.03, 29.41
.09, 5.52

∆χ2(∆df=4)

.72

Note. Family factors, neighborhood factors, grandmothers’ childrearing involvement, and mothers’ education were controlled for each model but the estimated
logits and odds ratios were not reported in this table.

Table 15: Hierarchical Multinomial Logistic Regression Assessing Interaction between Adolescent Gender and Types of Family
Arrangements Involvement for Latinos

Models
Step1

Step 2

Predictors
Gender (ref. Male)
Biological Father (BF) (ref.
Lone Single Mother)
Male Adult (MA) (ref. Lone
Single Mother)
Gender ×BF
Gender ×MA

Slowly Declining Class
(ref. Low Class)
Logit
Odds Ratio 95% CI
.22
1.25
.63, 2.48
-.02
.98
.40, 2.39
-.26

.77

.35, 1.69

1.47
.25

4.34
1.29

.70, 26.95
.36, 4.61

∆χ2(∆df=2)

2.39

Note. Family factors, neighborhood factors, grandmothers’ childrearing involvement, and mothers’ education and acculturation were controlled for each model
but the estimated logits and odds ratios were not reported in this table.

89

average levels of grandmothers’ childrearing involvement (OR = .82, p = .527). The overall
findings suggest that the average levels of grandmothers’ childrearing involvement was neither a
facilitator nor an impediment to mothers’ monitoring knowledge over time for the Latino
families, but was a facilitator for the African American families.
5.5 Trajectories of Mothers’ Monitoring Knowledge and Adolescent Adjustment Outcomes
For both ethnicities, ordinary least square (OLS) regression analyses were conducted to
test relations between trajectory types of mothers’ monitoring knowledge and adolescents’
externalizing behavior problems and school delinquency (see Table 16); and ordinal logistic
regression analyses were conducted to test relations between trajectory types of mothers’
monitoring knowledge and adolescents’ internalizing behavior problems (see Table 17). One
similar result was found for both groups. Compared to those whose mothers had high levels of
monitoring knowledge over time (i.e., belonging to the stable high or slowly declining class), the
adolescents whose mothers had low levels of monitoring knowledge (i.e., belonging to the high
declining or the low class) were more likely to display externalizing behavioral problems in late
adolescence for both the African American families (βs = .17 and .17, ps = .002 and .001, for the
high declining class versus the stable high class and the stable low class versus the stable high
class, respectively) and Latino families (β = .15, p < .001). This suggests that high levels of
mothers’ monitoring knowledge prevented adolescents from externalizing problems.
Ethnic differences were also found. First, high levels of mothers’ monitoring knowledge
predicted lower levels of internalizing problems in late adolescence for the African American
adolescents, but not for the Latino adolescents. Specifically, among the African Americans,
converting the logit (=1.04) into the odds ratio (=2.82) for the stable low class versus the stable
high class indicates that the odds of displaying higher levels of internalizing behavioral problems

90

versus lower levels in late adolescence were 2.82 times greater for the adolescents whose
mothers had stable low levels of monitoring knowledge than those whose mothers had stable
high levels of monitoring knowledge over time (see Table 17). In contrast, the trajectory type of
mothers’ monitoring knowledge was not significantly related to the Latino adolescents’
internalizing behavioral problems. Together, this suggests that high levels of mothers’
monitoring knowledge prevented adolescents from internalizing behavioral problems in late
adolescence only in the African American families, but not in the Latino families.
Second, high levels of mothers’ monitoring knowledge predicted less school delinquent
behavior in late adolescence for the Latino adolescents, but not for the African American
adolescents. As shown in Table 16, among the Latinos, compared to those whose mothers had
high levels of monitoring knowledge over time (i.e., belonging to the slowly declining class), the
adolescents whose mothers had low levels of monitoring knowledge (i.e., belonging to the low
class) had more school delinquent behavior (β = .10, p = .074) in late adolescence with a
marginal significance level. In contrast, the trajectory type of mothers’ monitoring knowledge
was not significantly related to the African American adolescents’ delinquent behavior. Together,
this suggests that high levels of mothers’ monitoring knowledge prevented adolescents from
school delinquent behavior in late adolescence only in the Latino families, but not in the African
American families.

91

Table 16: Regressing Adolescents’ Externalizing Behavior Problems and School Delinquency on
the Trajectory Groups of Mothers’ Monitoring Knowledge

B
African Americans
High Declining
Class (ref. Stable
High Class)
Stable Low Class
(ref. Stable High
Class)
Latinos
Low Class (ref.
Slowly Declining
Class )

Externalizing Behavior Problems
SE
β
R2

.18**

.06

.17

.28***

.09

.17

.20***

.07

.15

.10

.06

B

School Delinquency
SE
β
R2

.11

.10

.07

.06

.17

.02

.23+

.14

.10

.05

.05

Note. 1. Adolescents’ gender, mothers’ educational level, and family arrangement types were controlled
for each outcome variable; only for the Latinos, mothers’ acculturation was also controlled. 2. ** p < .01;
*** p < .001; + p < .10.

Table 17: Regressing Adolescents’ Internalizing Behavior Problems on the Trajectory Groups of
Mothers’ Monitoring Knowledge
Internalizing Behavior Problems
Logit
Odds Ratio
95% CI

Wald
Statistic

African Americans
High Declining Class (ref. Stable
.04
1.04
.55, 1.97
.02
High Class)
Stable Low Class (ref. Stable High 1.04*
2.82
1.23, 6.48
5.95*
Class)
Latinos
Low Class (ref. Slowly Declining
.30
1.34
.69, 2.61
.75
Class )
Note. 1. Adolescents’ gender, mothers’ educational level, and family arrangement types were controlled
for each outcome variable; only for the Latinos, mothers’ acculturation was also controlled. 2. * p < .05.

92

Chapter 6: Discussion
The present study shed light on mothers’ monitoring knowledge in terms of its
trajectories, facilitators and impediments, and implications for adolescents’ developmental
outcomes among low-income African American and Latino families. In both the low-income
African American and Latino families in this study, the majority of mothers maintain monitoring
knowledge with high levels across the adolescent years. In examining family environment and
mothers’ monitoring knowledge, the present study identified family routine as a common
facilitator and harsh discipline as a common impediment to mothers’ monitoring knowledge for
both ethnic groups. Some unique facilitators were identified only in the low-income African
American families, i.e., grandmothers’ childrearing involvement and neighborhood cohesion.
Adolescent gender difference in mothers’ monitoring knowledge was also only found in African
American families. Specifically, the African American mothers who rear boys were more likely
to experience a sharp decline in monitoring knowledge than those who rear girls. Regarding
implication of mothers’ monitoring knowledge, the present finding suggests that stable high or
slowly declining monitoring knowledge across the adolescent years prevents adolescents from
maladjustment, particularly in terms of internalizing behavior adjustment for the African
American adolescents and externalizing behavior adjustment for the adolescents in both groups.
The findings, implications, and limitations are discussed below.
6.1 Trajectories of Mothers’ Monitoring Knowledge
My findings indicate that there were variations in patterns of trajectories of the primary
caregivers’ (i.e., mothers’) monitoring knowledge in the low-income African American and
Latino families in this study, with the majority of the mothers having high levels of monitoring
knowledge over time (stable high for the African American mothers and slowly declining for the

93

Latino mothers), while a small portion of the mothers had stable low or declining levels in both
groups. High levels of mothers’ monitoring knowledge reflect their active engagement in
tracking of or children’s willingness to disclose whereabouts and daily activities. Accordingly,
most of the low-income African American and Latino mothers may maintain high levels of
control over their children and solicit whereabouts and activities from them. Another possible
explanation is that the parent-child dyad is characterized by high levels of trust and closeness so
that most children are willing to actively disclose their daily activities. Regardless of either
possibility, my findings suggest that parent-child connectedness and parental authority are
probably highly valued in these low-income households. The family dynamic that the present
study identifies is inconsistent with the previous findings, which suggest that although there are
cross-ethnic variations in the onset of developing autonomy, parents tend to gradually release
their control and children are provided with more autonomy as they enter into middle or late
adolescence across both groups (see the review of Smetana, 2002). This contradiction may be
explained by both the cultural values and social context of low-income African American and
Latino families. Both African American and Latino families emphasize the importance of parentchild connectedness and parental authority (Harwood et al., 2002; Yasui & Dishion, 2007). Thus,
although developing autonomy is an universal developmental need for adolescents, parents’ high
control and monitoring as well as children’s voluntary disclosure can be very common in
adolescence and even as children enter into late adolescence in both the low-income African
American and Latino families.
This contradiction may also be explained by the social context of the African American
and Latino families in the study. In the present sample, all the African American and Latino
families were classified as low-income households and located in the low-income neighborhoods

94

(i.e., 20% households in the neighborhood lived below the federal poverty threshold in 1990).
Low-income neighborhoods are in general characterized as having a lot of risk factors, such as
gun use, drug and alcohol abuse, and gang activity. In order to prevent children from those risk
factors, these parents may make great efforts to gain monitoring knowledge either through active
tracking of their children’s behavior, or being consistently informed by the children. In line with
this view, the previous findings on parental monitoring of low-income African and European
American families found that over half of the parents maintained stable high or slightly declining
levels of monitoring knowledge over time (Laird et al., 2008; Spano et al., 2012). Extending the
studies of Laird et al. and Spano et al., the present findings suggest that the low-income Latino
mothers may also rely on high levels of monitoring knowledge to protect their children from the
risk factors in relation to low-income neighborhoods during the overall adolescent years. In sum,
maintaining intense monitoring knowledge may be a strategy to buffer negative impact of the
social context surrounding low-income families across ethnic groups.
Since the most of the African American and Latino mothers in the study were single
mothers, maintaining intense monitoring knowledge among the majority of them may also be a
strategy to overcome barriers to supervising children’s behavior that are due to multiple and high
density of time obligations. Low-income single mothers tend to take on multiple responsibilities,
including full-time caregiving and full-time employment (Roy, Tubbs, & Burton, 2004).
Furthermore, Roy et al. found that their full-time employment is often composed of multiple
part-time jobs that require transitioning from one work place to another. The findings of Roy et
al. suggest that intense obligations compress the low-income single mothers’ available time to
get along with and physically track of their children. Acknowledging the limited time in caring

95

and supervising children, the mothers in the current sample may make great efforts to improve
the quality of monitoring in order to efficiently protect their children from behavioral problems.
Together, to clarify whether the present findings are attributed to the mothers’ cultural values,
social context and position, or the synergic effects of all the factors, more efforts are needed to
examine patterns of trajectories of parents’ monitoring knowledge across diverse ethnic,
socioeconomic, and marriage status groups.
6.2 Family Environmental Factors and Mothers’ Monitoring Knowledge
The present study identified one common facilitator and one common impediment to
mothers’ monitoring knowledge from family environmental factors in both the low-income
African American and Latino families. In terms of the specific common facilitator, average high
levels of family routine during adolescence were identified as predicting high likelihood of
slowly declining levels of mothers’ monitoring knowledge for the Latino families and stable high
levels of knowledge for the African American families. Families with high levels of family
routine are characterized by a set of family regular activities arranged by parents for the purposes
of creating common family time, including such as mealtimes, homework times, and the times
for playing and talking (Jensen et al., 1983). A lot of common family time probably provide both
the mother and the child with many opportunities to discuss the child’s whereabouts and
activities, and thus facilitate the mother to gain information about their children. In other words,
high levels of family routine may shape positive dynamics of mother-child communication in
which both the mother and the child can discuss the child’s daily life.
The positive implication of family routine for formation of mothers’ long-term
monitoring knowledge also affirms the general view on the cultural traditions in the African
American and Latino families. Both ethnicities have been described as valuing the importance of

96

family cohesion (e.g., Yasui & Dishion, 2007; Harwood et al., 2002). Family routine possibly
functions as an important venue through which family solidarity and cohesion is strengthened
because this routine creates opportunities for family members to engage in common activities.
Given that family cohesion is highly valued in the culture, the African American and Latino
adolescents in the study were likely to appreciate family routine, and thus used this opportunity
to discuss their daily issues with their mothers. These adolescents may also interpret routine as
conveying positive meaning, such as mothers’ care and involvement, which possibly enhance
mother-child trust and the children’s voluntary disclosure.
In addition to family environment, this study also shed light on the long-term relations
between parenting styles and tactics and mothers’ monitoring knowledge in the low-income
African American and Latino families. This study did not assume mothers’ parenting styles as
static typologies and classified mothers into specific categories such as authoritative versus
authoritarian, which is the measurement approach conducted by Steinberg and his colleagues
(Steinber, Lamborn, Dornbusch, & Darling, 1992; Steinberg, Lamborn, Darling, Mounts, &
Dornbusch, 1994). Rather, this study assumed that mothers could mix their styles of parenting,
which is based on the recent view of Tamis-LeMonda et al. (2008) suggesting that the nature of
parenting practices is fluid rather than static typology. Adopting this approach, the present
findings present how the relative degrees to which the low-income parents in both ethnic groups
exercise each type of parenting styles and/or tactics (i.e., authoritative parenting, behavior
control, and harsh discipline) predict their monitoring knowledge in a long term. In consistent to
the research hypotheses, the study did not identify authoritative parenting and behavioral control
as facilitators to mothers’ monitoring knowledge for either the African American or Latino
families. Although both authoritative parenting and behavioral control can convey mothers’ care

97

and warmth as associated with parent-child closeness (Jackson-Newsom et al., 2008; Westbrook
& Harden, 2010), their links to parent-child closeness may not be as strong as other family
environmental factors—such as family routine—that more directly reflects the efforts that
parents make to enhance venues of parent-child communication and consolidate parent-child
closeness. This suggests that mothers’ monitoring knowledge is more likely to be fostered
through the explicit establishment of positive family dynamics.
By contrast, mothers’ harsh discipline was identified as one common impediment to their
monitoring knowledge in both the low-income African American and Latino families in the
study. This finding is not in line with the study hypothesis inferred from the previous research.
The previous research suggests that harsh discipline is likely interpreted as a positive connotation
representing parents’ care and warmth among African American and Latino children (e.g., Hill et
al., 2003; Steele et al., 2005). Furthermore, coercive control-oriented parenting tactics, such as
harsh discipline, has been found as protection against the negative influence of the risk factors in
low-income neighborhoods on children (Bradley et al., 2001). Thus, those children living in lowincome families may be more likely to perceive parents’ care and warmth from this kind of
parenting behavior. Based on the above view, mothers’ harsh discipline is less likely to lead to
mother-child alienation and may even predict mother-child closeness in both the low-income
African American and Latino families. Given that mothers are more likely to gain monitoring
knowledge in a context full of mother-child closeness (Blocklin et al., 2011; Padilla-Walker et al.,
2011), mothers’ harsh discipline is probably able to facilitate mothers to gain information about
their children’s whereabouts and activities.
The present finding may be explained by the method of measuring mothers’ monitoring
knowledge. This study assesses mothers’ monitoring knowledge by adolescent report.

98

Accordingly, from the eyes of both the African American and Latino adolescents in the study,
mothers’ harsh discipline may be perceived as conveying negative messages, such as mothers’
hostility and anger. When those adolescents feel that the relation with the mother is lack of
closeness and love, they tend not to allow their mothers to gain information about their daily
activities.
Interpreting the finding also needs to take two points into consideration. First, the finding
was only based on the three items from the measure of mothers’ harsh discipline, including the
items assessing whether the mother scolded, threatened to spank, and punish her child if the child
didn’t behave. Caution needs to be kept in mind regarding whether the few items can really
capture a comprehensive construct of harsh discipline and unpack the relations between mothers’
harsh discipline and monitoring knowledge. Second, the present sample was primarily comprised
of low-income single mothers. Thus, the present finding on the negative link between mothers’
harsh discipline and monitoring knowledge can only be generalized to a particular set of African
American and Latino families. The finding, however, cannot present the complex picture of
mothers’ harsh discipline and monitoring knowledge across a variety of family composition (e.g.,
two-parent family) and socioeconomic context (e.g., middle or upper class) for African
American and Latino families.
Indicators of both parenting and family relation were found to be associated with mothers’
long-term monitoring knowledge. Yet, another family environmental factor—mothers’ overtime
work—was neither a facilitator nor an impediment to their monitoring knowledge in both the
low-income African American and Latino families. This finding is inconsistent with the finding
of the previous qualitative study indicating that both mothers and fathers perceived long work
hours as a barrier hampering them from understanding their children’s daily activities (Belle &

99

Philips, 2010). One possible reason behind this contradiction may be due to distinction between
parents’ perception and real situation. Although parents tend to perceive overtime work as an
obstacle for supervising their children, it does not necessarily mean that they inevitably fail to
gain information about the children’s daily life. Being aware of this parenting difficulty, the
African American and Latino mothers in the study may make great efforts to overcome through
establishing well-organized family routines as previously discussed in this chapter. For example,
these mothers may arrange mealtimes to enhance mother-child communication and take this
advantage of understanding the child’s school life. The above-mentioned possibility may be
clarified by further research investigating mothers’ strategies for coping with impact of overtime
work on monitoring.
6.3 Neighborhood Context and Mothers’ Monitoring Knowledge
The present study found ethnic difference in whether long-term levels of mother’s
monitoring knowledge are associated with neighborhood context. In the Latino families, the
long-term trajectory patterns of mothers’ monitoring knowledge were not linked either with
average levels of neighborhood problems or with average levels of neighborhood cohesion. In
contrast, in the African American families, the long-term trajectory patterns of mothers’
monitoring knowledge were associated with average levels of neighborhood cohesion. The above
distinction is in line with the study hypothesis that consistently disadvantaged neighborhood
conditions in adolescence are less likely to be linked to low levels of mothers’ monitoring
knowledge over time for the Latino families than for the African American families. The
previous research has pointed that compared to the family environment of African Americans,
Latinos’ family environment is less likely to be impacted by neighborhood context because
Latinos have stronger family cohesion, including strong connections between siblings (Pachter et

100

al., 2006). Pachter et al. assert that the strong family cohesion plays as an important firewall that
blocks risks from low-income neighborhoods and reduces their impacts on positive family
dynamics of Latinos. Another possible reason is that African American families place great
emphasis on communalism (i.e., cooperation among community members) and thus tend to view
fictive kin (e.g., community members) as a part of their family system and feel obligated to
provide childrearing support for fictive kin (McAdoo, 1981; Yasui & Dishion, 2007). According
to the discussion of Yasui and Dishion (2007), although Latino families also value cooperation
within group, they primarily rely on family members rather than fictive kin in childrearing
support. On the one hand, Latino families may lack the advantage of having close-knit neighbors
to help with child-rearing (e.g., monitoring each other’s children). On the other hand, however,
the family dynamics of Latinos is less susceptible to the impact of neighborhood condition.
Based on either of the above possibilities, neighborhood conditions may have little to do with
indicators of good quality of family dynamics, such as slowly declining mothers’ monitoring
knowledge, in the Latino families in the study.
In contrast, the African American mothers in the study were more likely to gain
information about their children in a stable low level if living in low quality neighborhoods,
specifically characterized as low cohesion. In this study, neighborhood cohesion was assessed
with the degrees to which mothers perceived their neighborhood as close-knit and a good place
to rear children as well as neighbors are willing to help each other and can be trusted. The
previous discussion in this chapter has pointed to that African American families value the
importance of mutual support among community members. Thus, highly cohesive neighborhood
may provide the low-income African American mothers in this study with the advantage to get
access to neighbors’ assistance in child-rearing, such as supervising their children or gathering

101

the information about the children. However, since neighborhood connection is highly valued,
the African American mothers’ efficacy in gaining monitoring knowledge is more vulnerable to
the impact of disadvantaged neighborhood condition. Also, the previous discussion in this
chapter indicates that African American families lack stronger family cohesion as a firewall
blocking risks associated with low quality neighborhoods compared to Latino families. In sum,
both the neighborhood and family conditions provide possible explanations for the link between
low levels of monitoring knowledge and low neighborhood cohesion for the low-income African
American mothers in the study.
Although the average levels of neighborhood cohesion account for the likelihood of stable
high versus stable low monitoring knowledge, those levels cannot distinguish the likelihood of
experiencing stable high versus high declining monitoring knowledge for the African American
mothers. In contrast, the likelihood of stable high versus high declining knowledge is primarily
explained by the within-familial factors, including harsh discipline, family routine, grandmothers’
childrearing support, and child gender. In addition, further examination of the average levels of
neighborhood cohesion across the three trajectory groups indicated that the mothers with high
declining monitoring knowledge (M = 1.48) and those with stable high knowledge (M = 1.33)
lived in the communities with similarly high levels of neighborhood cohesion (t(289) = 1.66, p
< .2932), while those with stable low knowledge had lowest levels of neighborhood cohesion (M
= 1.07).
The possible explanation behind the above findings is that the long-term levels of African
American mothers’ monitoring knowledge may be determined by two stages—with
neighborhood condition being the first determinant and family environment the next determinant.
When living in the neighborhoods are characterized as low cohesion, such as with lack of mutual
2

The p-value was adjusted based on pairwise multiple comparisons with Bonferroni adjustment.

102

support in supervising children’s behavior, the African American mothers tend to gain stable low
levels of information about their children over time and within-familial factors have very little to
do with formation of mothers’ monitoring knowledge (with the exception of harsh discipline).
Next, among the mothers living in highly cohesive neighborhood and gaining high levels of
information about their children in early adolescence, within-familial factors play the major role
in determining why some mothers maintain stable high knowledge versus some starts high levels
but experience a sharp decline across adolescent years. This two-stage process implies that
within-family factors become influential for the formation of mothers’ monitoring knowledge
specifically only when the neighborhood quality turns good. This process again possibly points
to the quiet susceptibility of African Americans’ family dynamics is to the influence of
neighborhoods, specifically those characterized as disadvantaged.
6.4 Gender and Mothers’ Monitoring Knowledge
My findings indicate ethnic difference in relations between mothers’ monitoring
knowledge and child gender. In the African American families, mothers’ monitoring knowledge
varied according to child gender. Specifically, African American mothers who reared sons were
more likely to be classified in the group of sharp decline in monitoring knowledge rather than
stable high monitoring knowledge. In contrast to the findings of the African Americans, the
present findings of the Latino families indicate that mothers’ monitoring knowledge did not vary
according to child gender, with the majority of both Latino boys and girls with mothers having
slowly declining monitoring knowledge (88.6% and 88.8%, respectively). This may be explained
by ethnic difference in the mothers’ fear of gendered risk behaviors. Both the African American
and Latina girls in the study were at even higher risk of teenage pregnancy, given that their
proportion of having ever had sexual intercourse (63% and 55% for the African American and

103

the Latina girls, respectively) is above the national average (47.4%)(Centers for Disease Control
and Prevention, n.d.). The previous research has pointed that parents across races and SES in
general are in fear of their daughter’s pregnancy and thus monitor girls closely (Jones-Sanpei et
al., 2009). More specifically, the research on the low-income African American and Latino
single mothers indicates that the majority hold conservative attitudes toward teenage sexual
activities for their daughters (Kotchick, Dorsey, Miller, & Forehand, 1999). For example,
daughters should not have sex until she is married and teenage pregnancy will devastate their
daughters’ future education and career. Thus, both the African American and Latino mothers in
the study may actively gain information about their daughters’ whereabouts and activities in
order to prevent the daughters from engagement in sexual activities and decrease their risk of
pregnancy.
Yet, different from the African American mothers, the Latino mothers may have dual
fears. That is, they may not only be in fear of the daughter’s pregnancy but also of the son’s
engagement in externalizing behavior problems, and thus may make great efforts in supervising
their sons and daughters to a similar level. Indeed, further examination of gender difference in
externalizing behavior problems indicates the degrees of such problems for the Latino boys were
higher than the degrees for their female counterparts across adolescent years (measured at all
three waves)3. Given that the Latino boys displayed higher levels of tendency to engage in
externalizing behavior problems than the Latina girls since very early stage of adolescence, their
mothers may tend to be protective of the sons in order to decrease such gendered risk behaviors
specifically prevalent among their sons. Furthermore, the following examination of interaction

3

For the Latino adolescents, a series of t-tests indicated that the Latino boys displayed higher levels of externalizing
behavior problems than their female counterparts at all the waves, with Ms = -.07 and -.16 (at Time1), -.11 and -.19
(at Time 2), and -.05 and -.19 (at Time 3) for the boys and girls, respectively; and ts(352) = 2.35, 2.30, and 3.21, ps
< .019, .022, and .001at Time 1, Time2, and Time3, respectively.

104

between child gender and the household arrangement with male adult indicates that the Latino
lone single mothers’ monitoring knowledge about boys and girls were at a similar level.
Theoretically, lone single mothers may be faced with more barriers in understanding free-time
activities of their sons than of their daughters due to the opposite-sex parent-child interaction and
the lack of childrearing assistance from male adults (Vieno et al., 2009). Inconsistent with the
previous perspective, the present findings suggest the possibility that the Latino lone single
mothers may encounter no barriers or contribute all their energy in supervising their sons
because of no male adults in the household and their strong desires to protect their sons from
externalizing behavior problems.
In contrast, although the African American boys displayed more externalizing behavior
problems than their female counterparts in the late adolescence (measured at Time 3), no gender
difference was found in early and middle adolescence (aged from 10-16, measured at Time 1 and
2)4. Accordingly, the African American mothers may not think it is necessary to be protective of
sons as they are of daughters, because at least in early and middle adolescence, their sons were
not at higher risk of engaging in externalizing behavior problems, which is in general viewed as
common risk behaviors among adolescent boys (Leadbeater, Kuperminc, Blatt, & Hertzog,
1999). Moreover, it is noteworthy that the present findings do not suggest African American
mothers living with their spouse or other male adults (e.g., a male partner or grown male child)
gain more information about their sons than do the African American lone single mothers. Based
on the previous research on parental monitoring (Vieno et al., 2009), male adult figures may be
able to assist the African American mothers in gaining supplement information about their sons
4

For the African American adolescents, a series of t-tests indicated that the African American boys displayed higher
levels of externalizing behavior problems than their female counterparts only at Time 3, with Ms = -.03 and -.22 for
the boys and girls, respectively; and t(317) = 4.03, p < .001. The African American boys and girls were not different
in the levels at Time 1 and Time 2, with Ms = -.04 and -.10 (Time 1) and -.05 and -.09 (Time 2) for the boys and
girls, respectively; and ts(317) = 1.61 and .81, ps = .109 and .420 at Time 1 and Time2, respectively.

105

due to being faced with fewer obstacles discussing and understanding same-sex children’s
activities. Yet, given that the gendered risk (i.e., the risk of teenage pregnancy) among the
African American girls are more salient, the African American mothers may still concentrate
energy on monitoring their daughters and not deliberately rely on male adults to gain supplement
information about their sons. The above-mentioned possibilities about the mothers’ gendered
strategies for monitoring for both ethnicities may be clarified by further qualitative research
investigating mothers’ attitudes toward monitoring sons versus daughters.
6.5 Adaptive Culture and Mothers’ Monitoring Knowledge
The present study focused on the long-term association between mothers’ monitoring
knowledge and one specific adaptive culture factor—grandmothers’ childrearing involvement.
Grandmothers’ assistance in childrearing is one key indicator of strong family cohesion, which is
nested within cultural traditions of both African American and Latino families. Childrearing
support from kin (e.g., grandparents) has been thought as an important factor that is associated to
good quality of parenting (such as parental warmth) or buffers the negative effects of life stress
on quality of parenting (Cox et al., 2000; Sotomayor & Applewhite, 1988; Taylor, 1996). The
above view is supported by the present finding of the African American families. Specifically,
compared to those experiencing sharply declining monitoring knowledge over time, the African
American mothers who experience stable high monitoring knowledge were more likely to
receive average high levels of grandmothers’ childrearing involvement. This suggests that high
levels of grandmothers’ childrearing involvement facilitated mothers’ monitoring knowledge in
long term.
In the present study, both the African American mothers with stable high monitoring
knowledge and those with high declining monitoring knowledge gained high levels of

106

information about their children in early adolescence. However, the major difference between the
two groups of mothers is that the former’s knowledge about their children was consistently high
over time, while the latter experienced a sharp decline in the monitoring knowledge. Based on
this difference, grandmothers’ childrearing involvement may play an important role in assisting
the mothers in gaining information about their children’s daily activities in middle or late
adolescence in the African American families. Middle and late adolescence has been portrayed
as the period when children have more chances to explore the extra-familial world and their
needs in independence and autonomy also accelerate (Collins & Steinberg, 2006). Such
expansion of extra-familial experiences and their growing needs in being independent probably
impede the African American mothers from gaining information about their children either
through their own active solicitation or the children’s self-disclosure. Yet, if the mothers receive
strong childrearing support from the grandmother during this period, they may rely on the
grandmother to help monitor their children so that they are able to be more efficient in
supervising their children’s free-time activities.
By contrast, the findings of the Latino families indicate that the grandmothers’
childrearing involvement was neither a facilitator nor an impediment to mothers’ monitoring
knowledge. The findings are inconsistent with the previous research on Latinos suggesting that
grandmothers’ childrearing involvement is related to good quality of parenting (Contreras, López,
Rivera-Mosquera, Raymond-Smith, & Rothstein, 1999). The reasons behind the no effect of the
grandmothers’ assistance are not clear. One possible explanation is that extensive grandmother
support in childrearing may result in the dispersion of responsibility for supervising children’s
free-time activities and the decrease in obligation for the mothers to actively solicit or understand
their children’s daily life. Indeed, Contreras’ (2004) study of Puerto Rican mothers also suggests

107

that grandmothers’ excessive childrearing assistance may confuse the mothers about assignment
of responsibility for the parenting and discourage them to establish their own parenting skills
(e.g., showing positive affect). Contreras’ study and the present study focused on different
populations in relation to the age of the caregiver and the focus child, with the former focusing
on the adolescent mothers with infants or toddlers (in age from 12 to 42 months) while the
present study consisted of adult mothers with adolescents. Albeit existence of this difference,
both studies suggest that indicators of family cohesion (e.g., grandmothers’ childrearing support)
may not necessarily lead to good quality of parenting.
Another possibility is that the grandmother-grandchild intergenerational gap may be too
large so that the grandmother may not be able to efficiently help the mother gain information
about the child’s free-time activities. Although grandparents who are involved in childrearing are
able to absorb knowledge from grandchildren, they are also faced with barriers in understanding
the grandchildren’s life due to being exposed to different cohort experiences (Hirshorn, 1998).
Moreover, many immigrant families experience high degrees of intergenerational gaps because
of different rates of assimilation into the mainstream culture across generations (Kwak, 2003)
and intergenerational language gap (Portes & Rumbaut, 1996; 2001). In general, older
generations (e.g., parents or grandparents) tend to identify their cultures of origin and assimilate
into the new culture at a slower rate because they grew up in their countries of origin during their
formative years. Also, older generations like speaking their native language and may not be
fluent in English. In contrast, younger generations (e.g., grandchildren) are more likely to
identify with the new culture and speak in English fluently because of exposure to schooling and
media during their formative years. Accordingly, the Latino grandmothers may be faced with
many challenges in parenting their grandchildren, such as tracking of their free-time activities,

108

because they do not understand the American mainstream culture as their grandchildren do as
well as have language barrier in understanding English. Likewise, the Latino grandchildren may
think that their grandmothers are not able to understand their daily life due to the gap in cultural
values and language, and thus conceal their personal information from their grandmothers. Such
stagnation in grandmother-grandchild information flow perhaps explains why the grandmothers’
childrearing involvement does not facilitate mothers’ monitoring knowledge.
Although both the dispersion of responsibility for supervising children and
intergenerational gap provide possible explanations for no effect of the grandmothers’ assistance
on mothers’ monitoring knowledge for the Latinos, it is unclear regarding which one does
portray the Latino family dynamics. In addition, if both obstacles do exist in the Latino families,
it is unknown why these appear not influential for the African Americans. The grandmothergrandchild intergenerational gap for the African Americans may not be as large as for the
Latinos, since almost all the African American families in the present study were identified as
migrating to the US at least beyond second generations. Yet, future research is called for to
clarify why grandmothers’ childrearing support seems not leading to the dispersion of
responsibility for supervising children in the African American families. Future qualitative
approach may provide more information on differences in implication of grandmother support
for family dynamics between African American and Latino families.
6.6 Mothers’ Monitoring Knowledge and Adolescent Outcomes
Examining associations between trajectory groups of mothers’ monitoring knowledge
and developmental outcomes in late adolescence indicates that mothers’ stable high or slowly
declining monitoring knowledge may prevent adolescents’ maladjustment, primarily in terms of
externalizing behavior area. Both the African American and Latino adolescents were least likely

109

to have externalizing behavior problems when their mothers maintained their monitoring
knowledge in a high level (stable high for the African Americans and slowly declining for the
Latinos) rather than experiencing stable low or high declining knowledge. The present findings
are in line with the previous view suggesting that adolescents’ externalizing behavior problems
can be prevented by high levels of parents’ monitoring knowledge, regardless of being formed
either through their active solicitation or children’s self-disclosure (e.g., Engels et al., 2006;
Laird et al., 2003). The findings of Engels et al. and Laird et al. were based on cross-sectional
data or two-adjacent time points. Extending their work, the present study provided evidence on
the preventive role of mothers’ consistently high monitoring knowledge for externalizing
behavior problems of African American and Latino adolescents in low-income families.
It remains unclear whether mothers’ consistently high monitoring knowledge also plays
significant role in preventing adolescents’ externalizing behavior problems among middle to
upper classes. Yet, the present study suggests that at least among the adolescents in the lowincome African American and Latino families, such stable high knowledge across the adolescent
years is an important protector for their adjustment in externalizing behavior. This finding is not
in line with some previous view asserting that parents may not need to be involved in their
children’s daily life too much during middle and late adolescence, since teenagers during this
period are much like adults and this period is the timing for them to gradually establish their own
boundary from the family system (see discussion of Collins & Stenberg, 2006 and Smetana,
2002). On the contrary, the present finding suggests that parents’ (at least mothers’) protection
and supervision over children’s daily activities remains critical and beneficial to the children’s
development, at least in terms of externalizing behavior adjustment of the African American and
Latino adolescents in the low-income families.

110

Two possible reasons may explain the above implication. The first is regarding social
circumstance of the current sample. In the present study, all participating households were of
low-income and drawn from low-income neighborhoods. As discussed previously in this article,
low-income neighborhoods are in general characterized as having a lot of risks, such as gang
activities and drug abuse. Being exposed to this disadvantaged condition, these adolescents
probably are at high risk at engaging in problem behaviors. In face of such condition, the
adolescents may be free from the impact of risk factors when their mothers continuously take an
active action of gathering information about their daily whereabouts and activities either through
active soliciting or building unimpeded communication channels that facilitate the children’s
self-disclosure, even in middle and late adolescence.
The second reason is in relation to the cultural values of African Americans and Latinos.
As previously discussed in this article, both African American and Latino families place great
emphasis on the importance of parental authority and parent-child connectedness. Thus, although
development of autonomy is a universal developmental need, both the African American and
Latino adolescents in the present sample may interpret indicators of mother-child connectedness
and maternal authority (i.e., mothers’ stable high monitoring knowledge) as mothers’
demonstration of care and love. When perceiving such protective connotation behind this family
dynamics, they may be more likely to appreciate mothers’ supervision and overprotection, and
thus tend to stay away from externalizing behavior problems.
By contrast to the findings on externalizing behavior problems, the findings on school
delinquency do not support that mothers’ stable high or slowly declining monitoring knowledge
plays a significant preventive role. The African American adolescents whose mothers maintained
stable high monitoring knowledge did not have less of a tendency to engage in school delinquent

111

behavior than their peers whose mothers experienced stable low or high declining knowledge.
For the Latino adolescents, although those whose mothers with slowly declining monitoring
knowledge scored lower in school delinquency than those whose mothers with low monitoring
knowledge over time, this difference was with a statistically marginal level (p = .100).
One possible reason that explains why the preventive role of mothers’ monitoring
knowledge for school delinquency is not salient may be due to that the focus areas of the
measure for mothers’ monitoring knowledge are not quite related to children’s school life. In the
present study, mothers’ monitoring knowledge was assessed with their knowledge about
children’s whereabouts after school and at night as well as activities with regards to friends, freetime activities, and money usage. Albeit the indicator about children’s whereabouts after school
is in relation to school area, the other four indicators do not explicitly focus on the issues about
children’s school life—such as copying homework, being suspended or expelled, and cheating—
which are the focus of the current measure for school delinquency. To examine the above
possibility calls for the further research distinguishing mothers’ monitoring knowledge on the
regular activities from the school area. Doing so will help clarify whether preventive roles of
mothers’ monitoring knowledge are generally effective across different areas of maladjustment,
or only occur when the focus areas of mothers’ monitoring knowledge correspond to those of
children’s daily issues.
Little is known about whether high levels of mothers’ monitoring knowledge are a
protector or a risk for adolescents’ internalizing behavior problems. The present study shed light
on implication of mothers’ long-term monitoring knowledge for internalizing behavior
adjustment for the African American and Latino adolescents living in the low-income families.
In this study, the African American adolescents whose mothers with stable high monitoring

112

knowledge were less likely to have internalizing behavior problems than those whose mothers
with stable low knowledge. This finding is in line with the perspective that high levels of
mothers’ monitoring knowledge may protect adolescents from internalizing behavior problems
since high levels of mothers’ monitoring knowledge reflect parent-child closeness and this
closeness still plays an important role in benefitting psychological adjustment in adolescence
(Collins & Stenberg, 2006). This also corresponds to the possibility that the culture of African
Americans emphasizes importance of parent-child connectedness so that consistently high levels
of mothers’ monitoring knowledge are more likely to bring positive implication for children’s
developmental outcomes.
Interestingly, regardless of their mothers with stable high or high declining monitoring
knowledge, the tendency of the African American adolescents in the study to have internalizing
behavior problems was similar. The overall finding indicates that the present sample of African
American adolescents’ susceptibility to internalizing behavior problem is growing when their
mothers experience consistently low levels since early adolescence; in contrast, their tendency
decreases when their mothers experienced high levels of monitoring knowledge at least in early
adolescence. In other words, establishing family dynamics facilitating parent-child closeness in
early adolescence may bring the most profound positive effect on adolescents’ internalizing
behavior adjustment, while a decline in such establishment in middle or late adolescence appears
not very influential, for the African American adolescents living in the low-income families.
The reason behind the difference in levels of importance of mothers’ monitoring
knowledge between early and middle/late adolescence for the African American adolescents in
the study may be explained by children’s social context across different adolescent stages.
Although children across races/ethnicities may have different onsets to develop autonomy, there

113

is still universal commonality. Compared to those in early adolescence, teenagers in middle or
late adolescence have more opportunities to explore the extra-familial environments (Grotevant,
1998), and thus may view keeping some distance from their parents or concealing secrecy as a
demonstration of being independent. Also, as children age, their need for autonomy and freedom
is growing (Collins & Stenberg, 2006) so that they are more likely to take parents’ insufficient
involvement in monitoring and supervision for granted. Accordingly, the African American
teenagers in this study may not view a decline in mothers’ monitoring knowledge in middle or
late adolescence as an indicator for their alienation from the mothers, and thus their internalizing
behavior (psychological) adjustment is not influenced.
By contrast to the findings on the African American families, the findings on the Latino
families indicate that the adolescents whose mothers with slowly declining monitoring
knowledge and those whose mothers with low monitoring knowledge were not different in their
tendency to have internalizing behavior problems. This suggests that high levels of mothers’
monitoring knowledge over time is neither a protector nor a risk for internalizing behavior
problems of Latino adolescents living in the low-income families. The reasons behind such no
link may be explained by acculturation of Latino adolescents. High- and low-acculturation
Latino adolescents interpret the family dynamics (e.g., mothers’ monitoring knowledge) that
facilitates parent-child closeness are different. High-acculturation Latino adolescents may view
establishment of this family dynamics as a threat to development of independence, because they
tend to assimilate the American mainstream culture that emphasizes the importance of
developing autonomy and independence in adolescence (Kwak, 2003). In contrast, lowacculturation adolescents may view this dynamics as a way to maintain closeness with their
mothers since they identify the family dynamics aligned with their cultural tradition, such as

114

family cohesion and parent-child connectedness. Compared to the low-acculturation counterparts,
the high-acculturation Latino adolescents are more likely to develop internalizing behavior
problems since the family dynamics behind high levels of mothers’ monitoring knowledge does
not match with their need. The present study only had information on adolescents’ generational
status, but did not assess their acculturation orientation with a variety of indicators—including
behavior, ethnic identity, and values—which are more able to reflect Latino adolescents’
acculturation. To examine the previously-mentioned possible reason, thus, future research is
called for to investigate a variety of the above-mentioned acculturation indicators and examine
relations between mothers’ monitoring knowledge and adolescents’ internalizing behavior
problems across different acculturation levels.
6.7 Implications
By examining mothers’ long-term monitoring knowledge and its relation with family and
neighborhood context as well as developmental outcomes in late adolescence, the present study
shed light to potential prevention and/or intervention actions for purposes of improving
adolescent developmental outcome and family environment of low-income and urban African
Americans and Latinos. First, examining relations between mothers’ long-term monitoring
knowledge and adolescent developmental outcome suggests that maintaining monitoring
knowledge with a stable high or slowly declining level across the overall adolescent years can
prevent maladjustment, at least in internalizing behavior adjustment for the African American
adolescents and externalizing behavior adjustment for the adolescents in both ethnicities.
Accordingly, social workers who serve youth and families may advance prevention programs to
educate low-income African American and Latino mothers, particularly those living in urban and
low-income neighborhoods, about the positive implications of gaining information about their

115

children’s whereabouts and activities in adolescence for the prevention of externalizing and/or
internalizing behavior problems. The programs can also be targeted at addressing these
populations about the importance of sustaining high levels of knowledge about their children’s
daily issues. As children enter into middle and late adolescence, mothers may think supervision
and monitoring are no longer necessary because their children are much like adults and it is a
period for the children to establish personal boundaries. On the contrary to the above idea, the
present findings suggest that mothers’ active involvement remain a critical parenting tactic that is
able to protect their children from risks in relation to low-income neighborhoods and prevent
their behavior problems.
Second, the present findings on the facilitators and impediments to mothers’ monitoring
knowledge provide useful information for social workers to think about in designing preventions
and/or interventions aiming to help mothers maintain high levels of monitoring knowledge
across adolescent years. In terms of family environment, family routine was identified as a
common facilitator to mothers’ monitoring knowledge for low-income families in both
ethnicities. Accordingly, social workers may design the prevention programs including the
workshops or curriculums that aim to guide low-income African American and Latino mothers
to learn strategies for establishing common family times to enhance positive parent-child
communication.
The workshops or curriculums may also include discussion on implication of harsh
discipline for low-income African Americans and Latinos. There remains unclear whether harsh
discipline brings positive or negative effects on parent-child relationship in the African American
and Latino families. A number of previous research indicates that harsh discipline is linked to
parent-child closeness or parental warmth in the African American and Latino families (e.g., Hill

116

et al., 2003; Steele et al., 2005), suggesting this parenting tactic probably is interpreted as
conveying positive connotation, such as parents’ care, for the children in both ethnicities. In
contrast, the present finding suggests that harsh discipline may still deliver negative meanings
(e.g., parents’ hostility) for the children in both ethnicities and impede parent-child closeness,
and thus average high levels of harsh discipline over time may hinder opportunities for mothers
to maintain monitoring knowledge with a stable high or slowly declining level. Since there
remain debates over implication of harsh discipline, social workers may present scholarly
findings to show complexity in this area. Also, they can encourage mothers to share their
personal experience on use of harsh discipline and its effects on quality of parent-child
relationship and adolescents’ developmental outcomes. These discussions probably can help the
low-income African American and Latino mothers integrate scholarly research and practical
experience to think about the potential pros and cons of harsh discipline for family relation and
adolescent development.
Specifically for urban and low-income African American families, a number of
suggestions are provided for facilitating mothers’ monitoring knowledge. In terms of
neighborhood context, the present finding suggests that these mothers’ monitoring knowledge is
quiet susceptible to impact of neighborhood cohesion and this impact may be more dominant
than the impact of within-familial factors (e.g., family cohesion and parenting behavior). Hence,
the present study advocates that social workers can develop prevention and/or intervention
programs that are targeted at strengthening mutual supports and trust among neighbors. Doing so
can enhance the overall cohesion in communities and in turn promote positive dynamics of the
low-income African American families, particularly in terms of mothers’ supervision and
monitoring on their children’s whereabouts and activities.

117

In addition to the above implication for social workers, the present finding also provides
the implication to policy-makers in terms of how budget-decision may play an important role in
enhancing neighborhood cohesion in the low-income African American families. In reality, the
budget cuts have occurred across 46 states in the US in all major services for community
development during the past decade (Center on Budget and Policy Priorities, 2011). The budget
reductions can halt the programs for supporting community connection (e.g., the operation of
community centers) or deteriorate the quality of the programs. Since mutual support among
community members plays an important role in benefiting the low-income African American
mothers’ parenting efficacy, the study advocates that policy-makers deliberate how to
appropriately allocate budgets to support the services for strengthening community cohesion.
Another suggestion is to address the importance of grandmothers’ childrearing support.
The present study identified grandmothers’ childrearing involvement as a facilitator to positive
family dynamics, particularly in terms of high levels of mothers’ monitoring knowledge. Also,
grandmothers’ childrearing involvement has been described as reflecting emphasis on family
connectedness for the African American families. Accordingly, social workers can design the
workshops on addressing potential positive implications of ethnic heritages for shaping family
dynamics and use grandmothers’ childrearing support as an example to discuss its positive
implication for mothers’ efficacy in supervising adolescents’ free-time activities in the urban and
low-income African American families.
6.8 Limitation
Four limitations have to be kept in mind for interpretation of the present findings. For one,
the present urban and low-income populations were recruited in three cities, i.e., Chicago,
Boston, and San Antonio. Thus, the present findings cannot be generalized to the low-income

118

populations in other urban areas in the U.S. Future research on national low-income populations
is called for. Second, the sample size of each Latino subgroup was too small to be analyzed
separately, and thus the present study examined the Latino sample as a whole. Recent research
has cautioned that investigating Latinos as a whole can conceal the within-group heterogeneity
(Harwood et al., 2002). Yet, it is noteworthy that the present Latino sample is still homogeneous
to some extent, at least in terms of the low-income background. Also, the three major Latino
subgroups in this study have comparable poverty rates (25.0%, 26.3%, and 25.6% for Mexicans,
Dominicans, and Puerto Ricans, respectively) between 2007 and 2011 based on the US Census
(2013). Accordingly, the internal heterogeneity may not be so salient in the present sample.
Third, although both the mother and the child were surveyed, the present study did not
assess the same constructs (e.g., mothers’ monitoring knowledge) from both adolescent and
maternal report. Whereas some research has indicated levels of consistency between adolescent
and parent report in family dynamics is high (Park et al., 2008), future research will benefit from
comparing constructs based on both adolescent and maternal report. Fourth, the present study
only assessed mothers’ monitoring knowledge at three time points. Future research measuring
more time points will provide more precise estimation of trajectories of monitoring knowledge.
6.9 Future Directions
A number of future directions are suggested for enhancing our understanding of parents’
monitoring knowledge. First, while mothers are the primary caregivers in the low-income
households, examining fathers’ monitoring knowledge will provide rich information on family
dynamics and its relation to adolescent development in the low-income populations. Second,
future research is called for to distinguish mothers’ monitoring knowledge on regular activities
from school area. Doing so will benefit understanding of whether preventive roles of mothers’

119

monitoring knowledge are only effective when the focus areas of mothers’ monitoring
knowledge match with those of children’s daily issues. Third, assessing a variety of information
on acculturation (i.e., behavior, racial/ethnic identity, cultural values) will address whether
associations between mothers’ monitoring knowledge and adolescents’ internalizing behavior
problems vary according to adolescents’ acculturation orientation.
Fourth, this study used only three items to conceptualize harsh discipline. Future
research using more items may capture more reliable construct of harsh discipline and help
clarify relations between mothers’ harsh discipline and monitoring knowledge. Next, future
research will benefit from examining issues on mothers’ monitoring knowledge for other lowincome populations (e.g., European and Asian Americans) and middle- or upper-class families
across races/ethnicities. Doing so will enhance understanding of fluidity and complexity of
family dynamics and how its relationship to adolescent development varies according to
socioeconomic status and races/ethnicities. The last, the present study has identified a number of
facilitators and impediments to mothers’ monitoring knowledge for both the low-income African
American and Latino families. Future qualitative research (e.g., interviewing both parents and
adolescents and observation of family dynamics) is called for to unpack the underlying
mechanism by which the identified factors facilitate or impede mothers’ monitoring knowledge.
Qualitative research will also help clarify why mothers’ monitoring knowledge is unrelated to
some factors, such as their overtime work for both ethnicities and grandmothers’ childrearing
involvement for the Latino families.

120

REFERENCES

121

REFERENCES

Achenbach, T. M., & Rescorla, L. A. (2001a). ASEBA School-Age Forms & Profiles. Burlington,
VT: University of Vermont, Research Center for Children, Youth & Families.
Achenbach, T. M., & Rescorla, L. A. (2001b). Manual for the ASEBA adult forms &
profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth &
Families.
Addy, S., & Wright, V. R. (2012, February). Basic facts about low-income children, 2010children under age 18. National Center for Children in Poverty. Retrieved from
http://www.nccp.org/publications/pub_1049.html
Allen, J. P., Hauser, S. T., O'Connor, T. G., Bell, K. L., & Eickholt, C. (1996). The connection of
observed hostile family conflict to adolescents' developing autonomy and relatedness
with parents. Development and Psychopathology,8(02), 425-442.
Allison, P. D. (2001). Missing data. Thousand Oaks, CA: Sage.
Armsden, G. C., & Greenberg, M. T. (1987). The Inventory of Parent and Peer Attachment:
Individual differences and their relationship to psychological well-being in adolescence.
Journal of Youth and Adolescence, 16, 427–454.
Aneshensel, C. S., & Sucoff, C. A. (1996). The Neighborhood Context of Adolescent Mental
Health. Journal of Health and Social Behavior, 37(4), 293-310.
Angel, R., Burton, L., Chase-Lansdale, P. L., Cherlin, A., & Moffitt, Robert. (2009). Welfare,
Children, and Families: A Three-City Study [Computer file]. ICPSR04701-v7. Ann
Arbor, MI: Inter-university Consortium for Political and Social Research [distributor],
2009-02-10.
Barber, B. K. (1996). Parental psychological control: revisiting a neglected construct. Child
Development, 67(6), 3296–3319.
Barber, B. K., Stolz, H. E., Olsen, J. A., Collins, W. A., & Burchinal, M. (2005). Parental
support, psychological control, and behavioral control: Assessing relevance across time,
culture, and method. Monographs of the Society for Research in Child Development,
70(4), 1–147.
Bauer, D. J. (2005). The role of nonlinear factor-to-indicator relationships in tests of
measurement equivalence. Psychological methods, 10(3), 305.
Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology
Monographs, 2(4), 1-103.
122

Belgrave, F. Z., Cherry, V. R., Cunningham, D., Walwyn, W., Letlaka-Rennert, K., & Phillips, F.
(1994). The influence of Africentric values, self-esteem, and Black identity on drug
attitudes among African American fifth graders: A preliminary study. Journal of Black
Psychology, 20, 143–156.
Belgrave, F. Z., Townsend, T., Cherry, V. R., & Cunningham, D. (1997). The influence of an
Africentric worldview and demographic variables on drug knowledge, attitudes, and use
among African American youth. Journal of Community Psychology, 25, 421-433.
Belle, D., & Phillips, B. (2010). Impediments to parental monitoring in the after-school hours: A
qualitative analysis. In V. Guilamo-Ramos, J. Jaccard, & P. Dittus (Eds.), Parental
monitoring of adolescents. New York, NY: Columbia University Press.
Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55, 8396.
Berlin, L. J., Ispa, J. M., Fine, M. A., Malone, P. S., Brooks-gunn, J., & Brady-smith, C. (2009).
Correlates and consequences of spanking and verbal punishment for low-income White,
African American, and Mexican American toddlers. Child Development, 80(5), 1403–1420.
Berry, J. W. (1980). Acculturation as varieties of adaptation. In A. M. Padilla (Ed.),
Acculturation: Theory, models and some new findings (pp. 9-25). Boulder, CO:
Westview Press.
Blair, S. L., Blair, M. C. L., & Madamba, A. B. (1999). Racial/ethnic differences in high school
students' academic performance: Understanding the interweave of social class and
ethnicity in the family context. Journal of Comparative Family Studies, 539-555.
Blocklin, M. K., Crouter, A. C., Updegraff, K. A., McHale, S. M. (2011). Sources of parental
knowledge in Mexican American families. Family Relations, 60(1), 30-44.
Borus, M. E., Carpenter, S. W., Crowley, J. E., & Daymont, T. N. (1982). Pathways to the future,
Volume II: A final report on the National Survey of Youth Labor Market Experience in
1980. Columbus: Center for Human Resource Research, Ohio State University.
Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomic status and child development. Annual
Review Psychology, 53, 371–399.
Bradley, R. H., Corwyn, R. F., Burchinal, M., Mcadoo, H. P., & Coll García, C. (2001). The
home environments of children in the United States Part II : Relations with behavioral
development through age thirteen. Child Development, 72(6),1868-1886.
Brody, G. H., Stoneman, Z., & Flor, D. (1996). Family wages, family processes, and youth
competence in rural married African American families. In E. M. Hetherington, & E. A.
Blechman (Eds.), Stress, coping, and resiliency in children and families (pp. 173–188).
Hillsdale, NJ, England: Lawrence Erlbaum Associates.
123

Bronfenbrenner, U. (1977). Toward an experimental ecology of human development. American
Psychologist, 32, 513–531.
Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard
University Press.
Bugental, D.B. (1992). Affective and cognitive processes within threat-oriented family systems.
In I.E. Sigel, A.V. McGillicuddy-DeLisi, & J.J. Goodnow (Eds.), Parental belief systems:
The psychological consequences for children (pp. 219–248). Hillsdale, NJ: Erlbaum.
Bulcroft, R. A., Carmody, D. C., and Bulcroft, K. A. (1996). Patterns of parental independence
giving to adolescents: Variations by race, age, and gender of child. Journal of Marriage
& Family. 58, 866–883.
Bumpus, M. F., & Hill, L. G. (2008). Secrecy and parent-child communication during middle
childhood: Associations with parental knowledge and child adjustment. Parenting, 8(2),
93–116.
Cauthen, N. K., & Fass, S. (2008). Measuring income and poverty in the United States. New
York, NY: National Center for Children in Poverty, Columbia University, Mailman
School of Public Health.
Celeux, G., & Soromenho, G. (1996). An entropy criterion for assessing the number of clusters
in a mixture model. Journal of Classification, 13, 195-212.
Centers for Disease Control and Prevention (n.d.). Sexual risk behavior: HIV, STD, & teen
pregnancy prevention. Retrived from the Centers for Disease Control and Prevention
website: http://www.cdc.gov/healthyyouth/Sexualbehaviors/index.htm.
Claes, M., Lacourse, E., Bouchard, C., & Perucchini, P. (2003). Parental practices in late
adolescence, a comparison of three countries: Canada, France and Italy. Journal of
Adolescence, 26, 387-399.
Clark, R., Novak, J. D., & Dupree, D. (2002). Perceptions of parenting practices as predictors of
coping strategies in African American adolescents. Journal of Adolescence, 25, 373–384.

Chao, R. K. (2001). Extending research on the consequences of parenting style for Chinese
Americans and European Americans. Child Development, 72(6), 1832–43.
Chen, Z. Y., & Dornbusch, S. M. (1998). Relating aspects of adolescent emotional autonomy to
academic achievement and deviant behavior. Journal of Adolescent Research, 13(3), 293319.

124

Cherlin, A. J. (1999). Welfare, children and families: A three-city study, wave 1 (Data Set 01–04,
Ruiz RO, McKean EA, and Peterson, JL)Machine-readable data file and documentation.
John Hopkins University (Producer). Los Altos, CA: Sociometrics Corporation, Child
Well-being and Poverty Data Archive (Producer and Distributor).
Cherlin, A.J., Angel, R., Burton, L., Chase-Lansdale, P. L., Moffitt, R., & Wilson, J. (2001).
Welfare, children, and families: A three-city study (Boston, Chicago, and San
Antonio),Wave 2, September 2000–June 2001 (Data Set 24–29, Mince J, Ruiz RO,
McKean EA, and Peter- son, JL) [Machine-readable data file and documentation]. Johns
Hopkins University, University of Texas, Pennsylvania State University, Northwestern
University, Harvard University (Producer). Los Altos, CA: Sociometrics Corporation,
Child Well-being and Poverty Data Archive (Producer and Distributor).
Child Trends (2009). Child research brief, 2009-11. Washington, DC: Author.
City of Boston. gov (2013). About Boston. Retrieved from
http://www.cityofboston.gov/visitors/about/
Collins, L. M., & Lanza, S. T. (2010). Latent class and latent transition analysis: With
applications in the social, behavioral, and health sciences (Vol. 718). John Wiley & Sons.
Collins, W. A., & Steinberg, L. (2006). Adolescent devel- opment in interpersonal context. In N.
Eisenberg (Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and
personality development (6th ed., pp. 1003–1067). Hoboken, NJ: Wiley.
Conger, R. D., & Donnellan, M. B. (2007). An interactionist perspective on the socioeconomic
context of human development. Annual Review of Psychology, 58, 175–199.
Conger, R. D., & Elder Jr., G. H. (1994). Families in troubled times: Adapting to change in rural
America. New York, NY: Aldine De Gruyter.
Contreras, J. M. (2004). Parenting Behaviors Among Mainland Puerto Rican Adolescent
Mothers: The Role of Grandmother and Partner Involvement. Journal of Research on
Adolescence, 14(3), 341–368.
Contreras, J. M., Lopez, I. R., Rivera-Mosquera, E. T., Raymong- Smith, L., & Rothstein, K.
(1999). Social support and adjustment among Puerto Rican adolescent mothers: The
moderating effect of acculturation. Journal of Family Psychology, 13, 228–243.
Cortés, D. E. (1995). Variations in familism in two generations of Puerto Ricans. Hispanic
Journal of Behavioral Sciences, 17, 249–255.
Cortés, D. E., Rogler, L. H., & Malgady, R. G. (1994). Biculturality among Puerto Rican adults
in the United States. American Journal of Community Psychology, 22, 707–721.

125

Cox, C., Brooks, L. R., & Valcarcel, C. (2000). Culture and caregiving: A study of Latino
grandparents. To grandmother’s house we go and stay: Perspectives on custodial
grandparents, 218-232.
Crockett, L. J., Brown, J., Russell, S. T., & Shen, Y.-L. (2007). The meaning of good parentchild relationships for Mexican American adolescents. Journal of Research on
Adolescence, 17(4), 639–668.
Crouter, A. C., Helms-Erikson, H., Updegraff, K., & McHale, S. M. (1999). Conditions
underlying parents’ knowledge about children’s daily lives in middle childhood:
Between- and within- family comparisons. Child Development, 70, 246-259.
Cuellar, I., Harris, L. C., & Jasso, R. (1980). An acculturation scale for Mexican American
normal and clinical populations. Hispanic Journal of Behavioral Sciences, 2(3), 199-217.
Davidson, P. R. (2007). Diversity in living arrangements and children’s economic well-being in
single-mother households. In B. A. Arrighi, & D. J. Maume (Eds.), Child poverty in
America Today, Volume 1: Families and children. Westport, CT: Praeger Publishers.
De Leon Siantz, M. L., & Smith, M. S. (1994). Parental factors correlated with developmental
outcome in the migrant Head Start child. Early Childhood Research Quarterly, 9, 481–
503.
Dearing, E. (2004). The developmental implications of restrictive and supportive parenting
across neighborhoods and ethnicities: Exceptions are the rule. Journal of Applied
Developmental Psychology, 25(5), 555–575.
Delgado-Gaitan, C. (1993). Parenting in two generations of Mexican American families.
International Journal of Behavioral Development, 16, 409–427.
Derogatis, L. R. (2000). BSI 18: The brief symptom inventory 18: Administration,
scoring and procedures manual. Minneapolis: National Computer Systems, Inc.
Diversitydata.org and Harvard School of Public Health (2013a). Profile summary of ChicagoNaperville-Joliet, IL-IN-WI. Retrieved from
http://diversitydata.sph.harvard.edu/Data/Profiles/Show.aspx?loc=308
Diversitydata.org and Harvard School of Public Health (2013b). Profile summary of San Antonio,
Texas. Retrieved from:
http://diversitydata.sph.harvard.edu/Data/Profiles/Show.aspx?loc=1224.
Dixon, S. V., Graber, J. A., & Brooks-gunn, J. (2008). The Roles of Respect for Parental
Authority and Parenting Practices in Parent–Child Conflict Among African American,
Latino, and European American Families. Journal of Family Psychology, 22(1), 1–10.

126

Domenech Rodriguez, M. M., Donovick, M. R., & Crowley, S. L. (2009). Parenting styles in a
cultural context: Observations of “protective parenting” in first‐generation Latinos. Family
process, 48(2), 195-210.
Dornbusch, S. M., Ritter, P., Liederman, P., Roberts, D., & Fraleigh, M. (1987). The rela- tion of
parenting style to adolescent school performance. Child Development, 58, 1244–1257.
Dodson, J. E. (1997). Conceptualizations of African American families. In H. P. McAdoo (Ed.),
Black Families, 3rd edition (pp. 67–93). Thousand Oaks, CA: SAGE Publications.
Ebaugh, H. R., & Curry, M. (2000). Socioeconomic position, poverty, and family research.
International Review of Research in Mental Retardation, 37, 97–129.
Engels, R. C. M. E., Finkenauer, C., & van Kooten, D. C. (2006). Lying behavior, family
functioning, adjustment in early adolescence. Journal of Youth and Adolescence, 35, 949958.
Ferrari, A. M. (2002). The impact of culture upon child rearing practices and definitions of
maltreatment. Child Abuse and Neglect, 26, 793–813.
Finkelstein, J. S., Donenberg, G. R., & Martinovich, Z. (2001). Maternal control and adolescent
depression : Ethnic differences among clinically referred girls. Journal of Youth and
Adolescence, 30(2), 155–171.
Fite, P. J., Colder, C. R., Lochman, J. E., & Wells, K. C. (2006). The mutual influence of
parenting and boys’ externalizing behavior problems. Journal of Applied Developmental
Psychology, 27(2), 151–164.
Firestone, J. M. (2007). Family context, income adequacy, and child-care needs: The bottom line
for San Antonio’s future. In B. A. Arrighi, & D. J. Maume (Eds.), Child poverty in
America Today, Volume 1: Families and children. Westport, CT: Praeger Publishers.
Florsheim, P., Tolan, P. H., Gorman-Smith, D. (1996). Family processes and risk for
externalizing behavior problems among African American and Hispanic boys. Journal of
Consulting and Clinical Psychology, 64(6), 1222–1230.
Fomby, P. (2006). Time use scale in documentation for created variables continuing/new and
separated caregiver data files. Baltimore, MD: Johns Hopkins University.
Foreign Policy (2013). The Global Cities Index 2010. Retrieved from:
http://www.foreignpolicy.com/node/373401?wp_login_redirect=0.
Furstenberg, F. F., Cook, T. E., Eccles, J., Elder, G., & Sameroff, A. (1999). Managing to make
it: Urban families and adolescent success. Chicago, IL: Univ. Chicago Press.

127

García Coll, C. (1993). What do we mean by extended family? A closer look at Hispanic
multigenerational families. Journal of Cross-Cultural Gerontology, 8, 137–146.
García Coll, C., Crnic, K., Lamberty, G., Wasik, B.H.., Jenkins, R., Garcia, H. V., & McAdoo, H.
P. (1996). An integrative model for the study of developmental competencies in minority
children. Child Development, 67, 1891-1914.
García Coll, C., & Pachter, L. M. (2002). Ethnic and minority parenting. In M. H. Bornstein
(Ed.), Handbook of parenting: Social conditions and applied parenting (2nd ed.,Vol. 4,
pp. 1-20). Mahwah, NJ: Lawrence Erlbaum Associates.
Garcia-Preto, N. (1996). Latino families: An overview. In M. McGoldrick, J. Giordano, & J. K.
Pearce (Eds.), Ethnicity and family therapy (pp. 141–154). New York: Guilford Press.
George, R., Dilts, J., Yang, D-H, Wasserman, M., & Clary, A. (2007). Chicago children and
youth 1990-2010: Changing population trends and their implications for services.
Chicago, IL: Chapin Hall Center for Children at the University of Chicago.
Gold, M. (1970). Delinquent behavior in an American city. Belmont CA: Brooks/Cole.
Gonzales, N. A., Pitts, S. C., Hill, N. E., & Roosa, M. W. (2000). A mediational model of the
impact of interparental conflict on child adjustment in a multiethnic low-income sample.
Journal of Family Psychology, 14, 365–379.
Green, S. B., & Yang, Y. (2009). Reliability of summed item scores using structural equation
modeling: An alternative to coefficient alpha. Psychometrika,74(1), 155-167.
Griffith, J. E., & Villavicencio, S. (1985). Relationships among acculturation, sociodemographic
characteristics and social supports in Mexican American adults. Hispanic Journal of
Behavioral Sciences, 7(1), 75-92.
Grotevant, H. D. (1998). Adolescent development in family contexts.
Hardway, C, & Fuligni, A. J. (2006). Dimensions of family connectedness among adolescents
with Mexican, Chinese, and European backgrounds. Developmental Psychology, 42(6),
1246–1258.
Harwood, R., Leyendecker, B., Carlson, V., Asencio, M., Miller, A. (2002). Parenting among
Latino families in the U.S. In M. H. Bornstein (Ed). Handbook of parenting: Vol. 4:
Social conditions and applied parenting (2nd ed.) (pp. 21-46). Mahwah, NJ, US:
Lawrence Erlbaum Associates Publishers.
Harwood, R. L., Miller, J. G., & Irizarry, N. L. (1995). Culture and attachment: Perceptions of
the child in context. NewYork, NY: Guilford Press.

128

Hayes, L., Hudson, A., & Matthews, J. (2003). Parental monitoring: A process model of parentadolescent interaction. Behaviour Change, 20(01), 13-24.
Hill, N. E., Bush, K. R., & Roosa, M. W. (2003). Parenting and family socialization strategies
and children’s mental health: Low-income Mexican-American and Euro-American
mothers and children. Child Development, 74(1), 189-204.
Hirshorn, B. A. (1998). Grandparents as caregivers. Handbook on grandparenthood, 200-214.
Hoffman, M. L. (1970) Moral development, In: Mussen, P. H. (Ed.), Carmichael’s Manual of
Child Psychology (vol. 2) (pp. 260–359). New York: Wiley.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal, 6(1), 1-55.
Izzo, C., Weiss, L., Shanahan, T., & Rodrugruez-Brown, F. (2000). Parental self-efficacy and
social support as predictors of parenting practices and children’s socioemotional
adjustment in Mexican immigrant families. Diverse Families, Competent Families, 20,
197–213.
Jackson-Newsom, J., Buchanan, C. M., & McDonald, R. M. (2008). Parenting and perceived
maternal warmth in European American and African American adolescents. Journal of
Marriage and Family, 70, 62–75.
Jambunathan, S., Burts, D. C., & Pierce, S. (2000). Comparisons of parenting attitudes among
five ethnic groups in the United States. Journal of Comparative Family Studies, 31, 395–
406.
Jeffries, N. (2003). A note on ‘Testing the number of components in a normal mixture.’
Biometrika, 90, 991–994.
Jensen, E.W., James, S. A., Boyce, W.T., & Hartnett, S. A. (1983). The family routines
inventory: development and validation. Social Science & Medicine, 17(4), 201–211.
Jones-Sanpei, H. A., Day, R. D., & Holmes, E. K. (2009). Core family process measures in the
NLSY97: Variation by gender, race, income, and family structure. Marriage & Family
Review, 45(2-3), 140-167.
Kahn, C. B., & Martin, J. K. (2011). The measure of poverty: A Boston indicators of project
special report. Boston, MA: Boston Foundation.
Kaplan, R. M., & Saccuzzo, D. P. (2009). Psychological testing: Principles, applications, and
issues (7th ed.). Belmont, CA: Thompson Wadsworth.

129

Kelley, M. L., Power, T. G., & Wimbush, D. D. (1992). Determinants of disciplinary practices in
low-income Black mothers. Child Development, 63, 573-582.
Kilgore, K., Snyder, J., & Lentz, C. (2000). The contribution of parental discipline, parental
monitoring, and school risk to early- onset conduct problems in African American boys
and girls. Developmental Psychology, 36, 835-845.
Kline, R. B. (2005). Principles and practice of structural equation modeling. New York, NY:
Guilford Press.
Kotchick, B. A., Dorsey, S., Miller, K. S., & Forehand, R. (1999). Adolescent sexual risk-taking
behavior in single-parent ethnic minority families. Journal of Family Psychology, 13(1),
93.
Krause, N., Bennett, J., & Van Tran, T. (1989). Age differences in the acculturation process.
Psychology and Aging, 4(3), 321-332.
Kwak, K. (2003). Adolescents and their parents : A review of intergenerational family relations
for immigrant and non-immigrant families. Human Development, 46, 115-136.
Lahey, B. B., Van Hulle, C. A., D’Onofrio, B. M., Rodgers, J. L., & Waldman, I. D. (2008). Is
parental knowledge of their adolescent offspring’s whereabouts and peer associations
spuriously associated with offspring delinquency? Journal of Abnormal Child Psychology,
36(6), 807-823.
Laird, R. D., Criss, M. M , Pettit, G. S., Dodge, K. A., & Bates, J. E. (2008). Parents’ monitoring
knowledge attenuates the link between antisocial friends and adolescent delinquent
behavior. Journal of Abnormal Child Psychology, 36(3), 299–310.
Laird, R. D., Pettit, G. S., Bates, J. E., & Dodge, K. A. (2003). Parents’ monitoring-relevant
knowledge and adolescents’ delinquent behavior: evidence of correlated developmental
changes and reciprocal influences. Child Development, 74(3), 752–768.
Lamborn, S. D., Dornbusch, S. M., & Steinberg, L. (1996). Ethnicity and community context as
moderators of the relations between family decision making and adolescent adjustment.
Child Development, 67, 283–301.
Lansford, J. E., Deater-deckard, K., Dodge, K. A., Bates, J. E., & Pettit, G. S. (2004). Ethnic
differences in the link between physical discipline and later adolescent externalizing
behaviors. Journal of Child Psychology and Psychiatry, 45(4), 801–812.
Leidy, M. S., Guerra, N. G., & Toro, R. I. (2010). Positive parenting, family cohesion, and child
social competence among immigrant Latino families. Journal of family psychology, 24(3),
252–260.

130

Lindahl, K. M., & Malik, N. M. (1999). Marital conflict, family processes, and boys’
externalizing behavior in Hispanic American and European American families. Journal
of Clinical Child Psychology, 28, 12 – 24.
Loukas, A., Suizzo, M-A., & Prelow, H. M. (2007). Examining resource and protective factors in
the adjustment of Latino youth in low income families: What role does maternal
acculturation play? Journal of Youth and Adolescence, 36(4), 489–501.
McElhaney, K. B., & Allen, J. P. (2001). Autonomy and adolescent social functioning: The
moderating effect of risk. Child Development, 72, 220–231.
Maguin, E., & Loeber, R. (1996). Academic Performance and Delinquency. Crime and Justice,
20, 145-264.
Marín, G., & Gamba, R. J. (1996). A new measurement of acculturation for Hispanics: The
Bidimensional Acculturation Scale for Hispanics (BAS).Hispanic Journal of Behavioral
Sciences, 18(3), 297-316.
Marin, G., &Vanoss Marin, B. (1991). Research with Hispanic populations. Newbury Park, CA:
Sage Publications.
Mandara, J., & Murray, C. B. (2002). Development of an empirical typology of African
American family functioning. Journal of Family Psychology, 16(3), 318–337.
Mason, C. A., Cauce, A. M., Gonzales, N., & Hiraga, Y. (1996). Neither too sweet nor too sour:
problem peers, maternal control, and problem behavior in African American adolescents.
Child Development, 67(5), 2115–2130.
McAdoo, H. P. (1981). Black families, Beverly Hills, CA: Sage.
McAdoo, H. P. (2002). African American parenting. In M. H. Bornstein (Ed). Handbook of
parenting: Vol. 4: Social conditions and applied parenting (2nd ed.) (pp. 47-58).
Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers.
McCabe, K. M., & Clark, R. (2010). Family protective factors among urban African American
youth. Journal of Clinical Child Psychology, 28(2), 137–150.
McLeod, J. D., & Shanahan, M. J. (1993). Poverty, parenting, and children’s mental health.
American Sociological Review, 58(3), 351–366.
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.

131

McLoyd, V. C. (1990). Minority children: Introduction to the special issue. Child Development,
61, 263–266.
McLoyd, V. C., Kaplan, R., Purtell, K. M., Bagley, E., Haraway, C. R., & Smalls, C. (2009).
Poverty and socioeconomic disadvantage in adolescence. In R. M. Lerner & L. Steinberg
(Eds.), Handbook of Adolescent Psychology (3rd ed.), Vol. 2 (pp. 444-491). Hoboken, NJ:
Wiley.
McLoyd, V. C., & Smith, J. (2002). Physical discipline and behavior problems in African
American, European American, and Hispanic children: Emotional support as a moderator.
Journal of Marriage and the Family, 64, 40-53.
Mildenberg, D.(2011). Population growth in Texas eclipses national rate. The Washington Post.
Retrieved from
http://www.washingtonpost.com/wpdyn/content/article/2011/02/17/AR2011021707179.h
tml?nav=hcmoduletmv.
Miller, A. M., & Harwood, R. L. (2001). Long-term socialization goals and the construction of
infants’ social net- works among middle-class Anglo and Puerto Rican mothers.
International Journal of Behavioral Development, 25, 450– 457.
Muthén, B. O. (2004). Latent variable analysis: Growth mixture modeling and related
techniques for longitudinal data. In D. Kaplan (ed.), The Sage Handbook of Quantitative
Methodology for the Social Sciences, pp. 345-368. Thousand Oakes: Sage Publications.
Muthén, L. & Muthén, B. O. (1998-2012). Version 7. Los Angeles, CA: Muthen & Muthen.
National Center for Children in Poverty (2012, June 5). United States demographics of lowincome children. Retrieved from http://www.nccp.org/profiles/US_profile_6.html.
Neumann, A., Barker, E. D., Koot, H. M., & Maughan, B. (2010). The role of contextual risk,
impulsivity, and parental knowledge in the development of adolescent antisocial behavior.
Journal of Abnormal Psychology, 119(3), 534–545.
Okagaki, L., & Frensch, P. A. (1998). Parenting and Children ’ s School Achievement : A
Multiethnic Perspective, 35(1), 123–144.
Pachter, L. M., Auinger, P., Palmer, R., & Weitzman, M. (2006). Do parenting and the home
environment, maternal depression, neighborhood, and chronic poverty affect child
behavioral problems differently in different racial-ethnic groups? Pediatrics, 117(4),
1329–1338.
Padilla-Walker, L. M., Harper, J. M., & Bean, R. A. (2011). Pathways to parental knowledge:
The role of family process and family structure. The Journal of Early Adolescence, 31(4),
604-627.

132

Patrick, M. R., Snyder, J., Schrepferman, L. M., & Snyder, J. (2005). The joint contribution of
early parental warmth, communication and tracking, and early child conduct problems on
monitoring in late childhood. Child Development, 76, 999-1014.
Perez-Brena, N. J., Updegraff, K. A., & Umaña-Taylor, A. J. (2012). Father- and motheradolescent decision-making in Mexican-origin families. Journal of Youth and
Adolescence, 41(4), 460–473.
Peterson, G. W., Bush, K. F., Wilson, S. M. & Hennon, C. B. (2005), Family relationship
predictors of parent-adolescent conflict: Cross-cultural similarities and differences.
International Family Strengths Consortium, Cuernavaca, Mexico. Retrieved Dec 13, 2011,
from http://www. ciesas.edu.mx/proyectos/mifs2005/papers/04/peterson_bush_wilson
_hennon.pdf.
Phinney, J. S., Ong, A., and Madden, T. (2000). Cultural values and intergenerational value
discrepancies in immigrant and non-immigrant families. Child Development, 71, 528–539.
Ployhart, R. E., & Oswald, F. L. (2004). Applications of mean and covariance structure analysis:
Integrating correlational and experimental approaches. Organizational Research Methods,
7, 27–65.
Portes, A., & Rumbaut, R. G. (1996). Immigrant America: A portrait. Los Angeles, CA:
University of California Press.
Portes, A., & Rumbaut, R. G. (2001). Legacies: The Story of the Immigrant Second Generation.
Berkeley, CA: University of California Press.
Prelow, H. M., Loukas, A., & Jordan-Green, L. (2007). Socioenvironmental risk and adjustment
in Latino youth: The mediating effects of family processes and social competence.
Journal of Youth and Adolescence, 36(4), 465–476.
Racz, S. J., & McMahon, R. J. (2011). The relationship between parental knowledge and
monitoring and child and adolescent conduct problems: a 10-year update. Clinical Child
and Family Psychology Review, 14(4), 377–98.
Rankin, B. H., & Quane, J. M. (2002). Social contexts and urban adolescent outcomes: The
interrelated effects of neighborhoods, families, and peers on African-American youth.
Social Problems, 49, 79–100.
Reese, L., Kroesen, K., & Gallimore, R. (2000). Agency and school performance among urban
Latino youth. In R. Taylor & M. Wang (Eds.), Resilience across contexts: Family, work,
culture, and community (pp. 295- 332). Mahwah, NJ: Erlbaum.
Roche, K. M., Ensminger, M. E., & Cherlin, A. J. (2007). Variations in parenting and adolescent
outcomes among African American and Latino families living in low-income, urban areas.
Journal of Family Issues, 28(7), 882–909.
133

Rothbaum, F., & Trommsdorff, G. (2007). Do roots and wings complement or oppose one
another: The socialization of relatedness and autonomy in cultural context. In J. Grusec,
& P. Hastings (Eds.), Handbook of socialization (2nd ed., pp. 461–489). New York:
Guilford Press.
Rueschenberg, E., and Buriel, R. (1989). Mexican American family functioning and
acculturation: A family systems perspective. Hispanic Journal of Behavioral Sciences, 11,
232–244.
Sameroff,A. J., & Chandler, M. (1975). Reproductive risk and the continuum of caretaking
casualty. In F. D. Horowitz, E. M. Hetherington, & S. Scarr-Salapatek (Eds.), Review of
child development research (Vol. 4, pp. 187-244). Chicago: University of Chicago Press.
Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A
multilevel study of collective efficacy. Science, 277, 918–924.
Schmitt, N. (1996). Uses and abuses of coefficient alpha. Psychological Assessment, 8, 350-353.
Shek, D. T. L. (2005). Perceived parental control processes, parent-child relational qualities, and
psychological well-being in Chinese adolescents with and without economic
disadvantage. The Journal of Genetic Psychology, 166(2), 171-188.
Shumow, L., Vandell, D. L., & Posner, J. K. (1998). Harsh, firm, and permissive parenting in
low-income families: Relations to children’s academic achievement and behavioral
adjustment. Journal of Family Issues, 19, 483–507.
Smetana, J. G. (2000). Middle-class African American adolescents’ and parents’ conceptions of
parental authority and parenting practices: A longitudinal investigation. Child
Development, 76, 1672 – 1686.
Smetana, J. G. (2002). Culture, autonomy, and personal jurisdiction in adolescent-parent
relationships. Advances in child development and behavior,29, 52-89.
Smetana, J. G., & Daddis, C. (2002). Domain-specific antecedents of parental psychological
control and monitoring: The role of parenting beliefs and practices. Child Development,
73, 563 - 580.
Smokowski, P. R., Rose, R., & Bacallao, M. L. (2008). Acculturation and Latino family
processes: How cultural involvement, biculturalism, and acculturation gaps influence
family dynamics. Family Relations, 57, 295–308.
Spano, R., Rivera, C., Vazsonyi, A. T., & Bolland, J. M. (2012).Specifying the interrelationship
between exposure to violence and parental monitoring for younger versus older
adolescents: a five year longitudinal test. American Journal of Community Psychology,
49, 127–141.
134

Spencer, M. B., Dobbs, B., & Phillips Swanson, D. (1988). African American adolescents:
Adaptational processes and socioeconomic diversity in behavioural outcomes. Journal of
Adolescence, 11(2), 117-137.
Spencer, M. B., Hartmann, T., & Dupree, D. (1997). A phenomenological variant of ecological
systems theory (PVEST): A self-organization perspective in context. Development and
Psychopathology, 9(4), 817–833.
Sirin, S. R. (2005). Socioeconomic status and academic achievement: A meta-analytic review of
research. Review of Educational Research, 75(3), 417–453.
Sotomayor, M., & Applewhite, S. R. (1988). The Hispanic elderly and the extended
multigenerational family. Hispanic elderly in transition: Theory, research, policy and
practice, 121-133.
Stattin, H., & Kerr, M. (2000). Parental monitoring: a reinterpretation. Child Development, 71(4),
1072-1085.
Steele, R. G., Nesbitt-Daly, J. S., Daniel, R. C., & Forehand, R. (2005). Factor structure of the
parenting scale in a low-income African American sample. Journal of Child and Family
Studies, 14(4), 535–549.
Steiger, J. H. (2007). Understanding the limitations of global fit assessment in structural equation
modeling. Personality and Individual Differences, 42(5), 893-898.
Steinberg, L., Mounts, N. S., Lamborn, S. D., & Dornbusch, S. M. (1991). Authoritative
parenting and adolescent adjustment across varied ecological niches. Journal of Research
on Adolescence, 1(1), 19-36.
Suárez-Orozco, C., & Qin, D. B. (2006). Gendered perspectives in psychology: Immigrant origin
youth. International Migration Review, 40(1), 165–198.
Tamis‐LeMonda, C. S., Way, N., Hughes, D., Yoshikawa, H., Kalman, R. K., & Niwa, E. Y.
(2008). Parents' goals for children: The dynamic coexistence of individualism and
collectivism in cultures and individuals. Social Development,17(1), 183-209.
Taylor, R. D. (1996). Kinship support, family management, and adolescent adjustment and
competence in African American families. Developmental Psychology, 32, 687–695.
The Inter-University Consortium for Political and Social Research (ICPSR) (2008, July). The
three-city study data workshop presented at The Inter-University Consortium for Political
and Social Research (ICPSR), Ann Arbor, MI. Presentation materials retrieved from:
http://www.icpsr.umich.edu/icpsrweb/3CITIES/index.jsp.

135

Unger, J. B., Ritt-Olson, A., Teran, L., Huang, T., Hoffman, B. R., & Palmer, P. (2002). Cultural
values and substance use in a multiethnic sample of California adolescents. Addiction
Research and Theory, 10, 257–279.
Uno, D., Florsheim, P., & Uchino, B. N. (1998). Psychosocial mechanisms underlying quality of
parenting among Mexican- American and white adolescent mothers. Journal of Youth
and Adolescence, 27, 585–605.
US Census Bureau (2000). Census 2000 Summary File 3. Retrieved from US Census Bureau
American FactFinder website:
http://factfinder2.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t
US Census Bureau (2006). U.S. Census Bureau, 2006 American Community Survey. Retrieved
from US Census Bureau American FactFinder website:
http://factfinder2.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t
US Census Bureau, Social, Economic, and Housing Statistics Division: Poverty (October 26,
2012). How the Census Bureau Measures Poverty. Retrieved from:
http://www.census.gov/hhes/www/poverty/about/overview/measure.html
US Census Bureau (2013a). State and country quickfacts 2000 of Chicago. Retrieved from US
Census Bureau achievement website:
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk.
US Census Bureau (2013b). 2007-2011 American Community Survey 5-Year Estimates of
Chicago. Retrieved from US Census Bureau achievement website:
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk.
US Census Bureau (2013c). 2007-2011 American Community Survey 5-Year Estimates of San
Antonio. Retrieved from US Census Bureau achievement website:
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk.
US Census Bureau (2013d). State and country quickfacts 2000 of San Antonio. Retrieved from
US Census Bureau achievement website:
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk.
United States Department of Labor (n.d.). Work hours. Retrieved from
http://www.dol.gov/dol/topic/workhours/overtime.htm
Utsey, S. O., Adams., E. P., & Bolden, M. (2000). Development and initial validation of the
Africultural coping systems inventory. Journal of Black Psychology, 26, 194–215.

Varela, R. E., Vernberg, E. M., Sanchez-Sosa, J. J., Riveros, A., Mitchell, M., & Mashunkashey,
J. (2004). Parenting style of Mexican, Mexican American, and Caucasian-non-Hispanic

136

families: social context and cultural influences. Journal of Family Psychology, 18(4),
651–657.
Vieno, A., Nation, M., Pastore, M., & Santinello, M. (2009). Parenting and antisocial behavior:
A model of the relationship between adolescent self-disclosure, parental closeness,
parental control, and adolescent antisocial behavior. Developmental Psychology, 45,
1509-1519.
Wang, Q., Pomerantz, E. M., & Chen, H. (2007). The role of parents’ control in early
adolescents’ psychological functioning : A longitudinal investigation in the United States
and China. Child Development, 78(5), 1592–1610.
Weis, R., & Toolis, E. E. (2010). Parenting across cultural contexts in the USA: assessing
parenting behaviour in an ethnically and socioeconomically diverse sample. Early Child
Development and Care, 180(7), 849–867.
Westbrook, T. R., & Harden, B. J. (2010). Pathways among exposure to violence, maternal
depression, family structure, and child outcomes through parenting: A multigroup
analysis. The American Journal of Orthopsychiatry, 80(3), 386–400.
White, K. R. (1982). The relation between socioeconomic status and academic achievement.
Psychological Bulletin, 91(3), 461-481.
Whiteside-Mansell, L., Bradley, R. H., Owen, M. T., Randolph, S. M., & Cauce, A. M. (2003).
parenting and children behavior at 36 months: Equivalence between African American
and European American mother-child dyads. Parenting: Science and Practice, 3(3), 197–
234.
Winston, P. (1999). Welfare, Children and Families: A Three-City Study: Overview and design.
Retrieved May 6, 2006, from http://www.jhu.edu/~welfare/overviewanddesign.pdf.
Winston, P., Angel, R. J., Burton, L. M., Chase-Landsdale, P. L., Cherlin, A. J., Moffit, R. A., &
Wison, W. J. (1999). Welfare, children and families: A three-city study: Overview and
design. Retrieved August 7, 2002, from http://www.jhu.edu/∼welfare/overviewanddesign.
pdf.
World Bank (2011). The world bank indicators database. Washington, DC: Author.
Yang, Y., & Green, S. B. (2011). Coefficient alpha: A reliability coefficient for the 21st
century?. Journal of Psychoeducational Assessment, 29(4), 377-392.
Yasui, M., & Dishion, T. J. (2007). The ethnic context of child and adolescent problem behavior :
Implications for child and family interventions. Clinical Child and Family Psychology,
10(2), 137-179.

137

Zea, M. C., Mason, M. A., & Murguia, A. (2000). Latino/Latina religious traditions: Implications
for psychotherapy. In A. E. Bergin, & P. S. Richards (Eds.), Psychotherapy and religious
diversity: A guide for mental health professionals (pp. 397–419). Washington, DC:
American Psychological Association.

138