Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Intergenerational transmission of values and behaviors over the family life course
(USC Thesis Other)
Intergenerational transmission of values and behaviors over the family life course
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INTERGENERATIONAL TRANSMISSION OF VALUES AND BEHAVIORS
OVER THE FAMILY LIFE COURSE
By
Joohong Min
-------------------------------------------------------------------------------------------------------
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GERONTOLOGY)
May 2014
ii
Dedication
To my loved ones, for always believing in me.
To my family, for their true love and endless support.
To my grandmother, who passed away last year, for inspiring me to study aging.
Without you all, this would not be possible.
iii
Acknowledgements
Special thanks to Dr. Merril Silverstein, my advisor, for allowing me to have an
opportunity to study at the USC Davis School of Gerontology and to Dr. Tara Gruenewald,
co-chair of committee, for her mentorship. I would also like to thank the other member of
my dissertation committee, Dr. Iris Chi as well as my former mentor from Seoul National
University, Gr. Gyeonghae Han, for their continuous support and advice. Thank you to Dr.
Eileen Crimmins for directing me in my final year and for providing guidance regarding my
next steps in academia. I also greatly appreciate my colleagues and friends for their support
across the journey of pursuing a doctoral degree: Jennifer Ailshire, Linda Broder, Roseann
Giarrusso, Elizabeth Hagood, Linda Hall, Seoungyeon Kim, Jessica Lendon, Morgan Levine,
Nick Pisca, and Muriel Rotenberg. Finally, thank you to Danielle Zucker; whose knowledge
of LSOG data along with her endless support and friendship made this possible.
iv
Table of Contents
Dedication……………………………………………………………………………………………….………………….….ii
Acknowledgements………………………………………………………………………………………………………..iii
List of Tables…………………………………………………………………………………………………..…..…….........v
List of Figures………………………………………………………………………………………………………...….....vii
Abstract…………………………………………………………………………………………………………………...….iv
Chapter One: Introduction and Theory……………………………………………………………………….……1
Chapter Two: Overview of Data: The Longitudinal Study of Generations ………………………...17
Chapter Three: Intergenerational Transmission of Values over the Family Life Course…....22
Chapter Four: Intergenerational Similarity in Religious Beliefs over the Life Course………..44
Chapter Five: Intergenerational Transmission of Smoking Behavior…………………….……...….67
Chapter Six: Discussion…………………………………………………………………………………………………88
References……………………………………………………………………………………………………………..……..96
v
List of Tables
Table 3.1. Descriptive Characteristics of Parents and Children………………………………………..33
Table 3.2. Correlations between the latent variables for religious beliefs and
gender role attitudes………………………………………………………………………………………….34
Table 3.3. Measurement models imposing generation and time constraints on
measurement coefficients for latent variables of religious beliefs
and gender role attitudes……………………………………………………………………………………35
Table 3.4. Standardized estimates for structural equation models
predicting transmission religious beliefs……………………………………………………………..38
Table 3.5. Standardized estimates for structural equation models predicting transmission
of gender role attitudes………………………………………………………………………………………39
Table 4.1. Number of children matching to parents in 1971
at each wave of measurement……………………………………………………………………………..50
Table 4.2. Descriptive characteristics of parents and children Table……………………………….51
Table 4.3. Fit indices for growth mixture models of intergenerational similarity
in religious beliefs………………………………………………………………………………………………56
Table 4.4. Representation and growth estimates of three classes derived from growth
mixture model……………………………………………………………………………………………………57
Table 4.5. Multinomial logistic regression estimates predicting membership in
intergenerational religious similarity class…………………………………………………….……58
Table 5.1. Characteristics of the Parents and Children…………………………………………………….77
Table 5.2. Correlation Matrix…………………………………………………………………………………………77
Table 5.3. Multinomial logistic regression estimates predicting child’s tobacco use:
Modeling mechanism only………………………………………………………………………………….78
Table 5.4. Multinomial logistic regression estimates predicting child’s tobacco use:
Anti-smoking socialization only………………………………………………………………………….80
Table 5.5. Multinomial logistic regression estimates of moderating effects
predicting child’s tobacco use: Multivariate mechanisms……………………………………..81
Table 5.6. Moderating effect of anti-smoking socialization (attitude and discouragement)
on intergenerational transmission of child’s tobacco use……………………………………..82
vi
Table 5.7. Parent’s smoking and child’s smoking…………………………………………………………....86
Table 5.8. Parent’s smoking and discouragement…………………………………………………………..86
Table 5.9. Parent’s smoking and attitude_Q1………………………………………………………………….86
Table 5.10. Parent’s smoking and attitude_Q2………………………………………………………………..87
vii
List of Figures
Figure 1.1. Conceptual Framework.………………………………………………………………………………13
Figure 3.1. Measurement and longitudinal structural model of
value transmission between parents and children……………………………………….….….31
Figure 4.1. Trajectories of intergenerational similarity in religious beliefs
over 34 years (1971-2005) (N=1,084)………………………………………………………………..59
viii
Abstract
The intergenerational transmission of family culture has been an important topic of
research in gerontology and the study of aging families. Scholars from both gerontology
and family studies have provided evidence of the impact of parents on their children’s
values and behaviors. Whether intergenerational transmission occurs through
socialization, didactic training, or modeling, it is clear that parental influences are
perpetuated through multiple generations. However, previous studies of intergenerational
transmission have primarily examined the issue using cross-sectional data or retrospective
data, making it impossible to examine whether transmitted values and behaviors endure or
emerge in children as they grow up. Additionally, only a few research investigations have
directly compared mechanisms of intergenerational transmission to explore their relative
importance and potential interaction effects.
The overall goal of this dissertation is to describe the timing, trajectory, and
dynamics of the intergenerational transmission of beliefs and behaviors over a 34 year
period of the family life course. Specifically, three primary questions are investigated. The
first looks at the timing of transmission. Do values and behaviors get transmitted early in
the family life cycle and persist in children over time, or is transmission lagged such that
children show more similarity to their parents’ values and behaviors when they reach
middle adulthood than they do early in life? The second question investigates whether
there are heterogeneous trajectories of intergenerational transmission. Do various patterns
of parent-child similarity exist? If so, does family solidarity affect life course patterns of
intergenerational belief transmission? The third question addresses parents’ influence on
ix
specific behaviors of their children. What is the association between the behaviors, values,
and attitudes of parents and their children’s behavior? Specifically, what are the potential
interactive effects and the relative importance of parents’ behaviors, values, and attitudes
on their children’s smoking behavior in adulthood?
The first question of when beliefs are transmitted was investigated using data
collected over 34 years from the Longitudinal Study of Generations (LSOG). The LSOG is a
four generation longitudinal family study. To explore this question, 775 parent-child dyads
from the 1971 wave and the 2000 wave of the LSOG were analyzed. Time lagged structural
equation modeling and multiple group analysis showed values (religious beliefs and gender
role ideology) were strongly transmitted from parents to their young adult children, and
then remained stable over the life course. This pattern of intergenerational influence was
most robustly observed for religious beliefs. Gender role ideologies had a more lagged
effect than in other values when children reached middle adulthood and strong emotional
closeness within families facilitated value transmission.
The second question about the potential heterogeneous trajectories of
intergenerational transmission of religious beliefs was examined by looking at parent-child
similarity over the life course using data from 8 waves of the LSOG (1971, 1985, 1988,
1991, 1994, 1997, 2000, and 2005). A growth mixture model showed three different types
of trajectories – “stable similar” (80.3%), “stable dissimilar” (10.5%), and “divergent across
time” (9.2%). Most dyads showed a stable similar pattern, providing evidence of the
importance of early relationships with parents. However, stable dissimilar and divergent
groups were also evident. In addition, family-related life transitions (such as becoming a
x
parent) and emotional closeness were factors that predicted the pattern of
intergenerational similarity over the life course.
The final question was addressed by using parent-child dyads from the 2000 wave
of the LSOG study to explore the mediating and moderating effects among modeling and
socialization mechanisms on intergenerational transmission of smoking behavior. Bivariate
models showed that both parents’ smoking behavior (modeling) and parents’ anti-smoking
attitude and discouragement (socialization) were significant mechanisms predicting
parent-child smoking behavior transmission. The effect of parents’ smoking on children’s
smoking was mediated by parents’ anti-smoking attitude, and a conditioned moderating
effect was found. Parental discouragement of smoking had less of an impact on children’s
smoking behavior than these anti-smoking attitudes did, even when strong parent-child
emotional solidarity was present.
This dissertation aims to extend the empirical development of the literature relating
to the intergenerational transmission of values and behaviors. Specifically, it focuses on the
importance of exploring the persistence of intergenerational transmission, investigates the
dynamics of transmission mechanisms, and uses the longitudinal data to more effectively
explain intergenerational transmission over the family life course. These findings have the
potential to inform the development of interventions and prevention efforts that address
values or behaviors that negatively influence psycho-socio-economic outcomes. The final
discussion summarizes the above findings and discusses how they can be explained within
the theoretical context.
1
Chapter One:
Introduction and Theory
A. Introduction
Family social scientists have found that family serves as a key social institution
to maintain continuity in social values over time, with parents serving as the
primary socialization agents for their children (Chodorow, 1978; Parsons & Baltes,
1955). Numerous studies have confirmed the presence of intergenerational
similarity and this similarity has served as evidence of successful intergenerational
transmission (Glass, Bengtson, & Dunham, 1986; Cunningham 2001; Moen, Erickson,
& Dempster-McClain, 1997; Rimal 2003).
Understanding the process of intergenerational transmission and what factors
facilitate it can offer important and valuable insights which will support the
development of more effective interventions for negative behaviors, as well as the
development of policy that better supports individuals’ well-being.
Despite a large volume of research in the field of intergenerational transmission,
few studies have been able to explore the persistence of intergenerational
transmission over the life course. This is partially because of the lack of available
longitudinal data needed to trace the longitudinal trajectory of intergenerational
similarity. In order to optimally investigate intergenerational similarity, data must
be collected from multiple generations of parents and children; in order to optimally
2
investigate the trajectory of this similarity, this dyadic data needs to be collected
over time.
Over the past several decades, society has been rapidly changing and this rapid
change has been greatly influencing individuals’ values and behavior. For example,
important life transitions relating to family, such as marriage and childbearing, have
been deferred (Schoen & Canudas-Romo, 2005). Additionally, there has been a
dramatic increase in divorce rates in the prevalence of historically uncommon
family structures such as cohabitation or single-parent families over the past
decades (Bumpass & Lu, 2000). At a time of rapid societal change, it is both timely
and important to understand whether parents continue to significantly influence
their children’s values and behaviors even beyond young adulthood.
When it comes to examining the longitudinal pattern of intergenerational
similarity, attention also needs to be focused on investigating factors which promote
or hinder the process. Especially within the scope of the family environment, certain
relational factors have been studied in depth and have been shown to facilitate
intergenerational transmission. Children who have emotionally close relationships
with their parents have been found to be more likely to have similar values to their
parents (Bengtson, 2001; Roberts & Bengtson, 1996; Rossi & Rossi, 1990). However,
less research attention has been focused on the long-term effects of these
intergenerational relationships over the life course.
In addition, although research has identified a variety of mechanisms through
which parents contribute to intergenerational similarity (e.g., directly via modeling,
3
discouragement, and attitude), little research has been devoted to figuring out the
relative importance of these different mechanisms and possible moderating effects
between them. Exploring which mechanisms have a greater influence on
intergenerational transmission can provide valuable information for developing
more effective interventions or prevention strategies. For example, smoking
behavior transmission between parent and child is an important topic in prevention
research; determining the relative importance of a given mechanism and how two
mechanisms interact will aid in developing more targeted and efficient
interventions.
Thus, it is important to bridge these research gaps by looking at 1) the
persistence of intergenerational transmission over the life course; 2) the family-
related factors, such as emotional closeness across generations, which contribute to
the pattern of intergenerational transmission trajectories over the family life course;
and, 3) the relative importance of different mechanisms in leading intergenerational
transmission and the possible moderating effects between these mechanisms.
B. Conceptual Framework
B.1. Bioecological Perspectives of human development: How do individuals develop
their own values and behaviors? : Focusing on micro-level
How do individuals develop their own view of the world, establish identity,
and acquire their own values and behaviors? To understand the process,
bioecological perspectives which were initially suggested by Bronfenbrenner
4
(Bronfenbrenner, 1994) provide a valuable theoretical framework. This paradigm
places the human development process within the context of ecological systems
which surround an individual and emphasizes the importance of the individual’s
interactions with this environment over the life span. Bronfenbrenner looks at these
systems or environments as nested features within which individuals and their
environment constantly interplay.
These perspectives suggest five environments: Microsystem, mesosystem,
exosystem, macro system, and chronosystem (Bronfenbrenner, 1994). The
microsystem represents the agents or agencies which most directly influence
individuals’ development, with family being considered a prime example of an
element of the microsystem. The mesosystem refers to the relations between
institutions or groups belonging to the microsystem. The macrosystem is an even
broader concept than mesosystem and it represents the cultural context. Cultural
contexts include a wide range of contexts such as industrialization, law, or socio-
economic status which can greatly influence children’s lives. All
systems/environments influence individuals’ developmental processes, including
how they acquire values and develop behaviors. Among these systems, this
dissertation focuses on the effect of the microsystem, especially the family and
intergenerational relationships within it.
B.2. Family Socialization Theory: How Does Intergenerational Transmission Operate?
Why are family and intergenerational relationships so important? As
individuals acquire their own attitudes, behaviors, and values, various agents
5
ranging from family, peers, neighborhoods, and schools to the public media play a
role in this acquisition process. Among these many agents and agencies, however,
family has been noted as perhaps the most dominant and powerful agent in the
socialization process of children.
Family socialization theory suggests that early experiences with parents directly
shape the values and behaviors of children. Parents can affect their children’s
values and behaviors through modeling (learning through observation of parents’
behavior and expressed attitudes) and formal training (such as religious instruction
and verbal discouragement that enforces conformity). Classical and operant
conditioning are examples of how parental training of attitudes may be achieved;
value and behavior adoption can occur not only by neutral stimuli and observation,
but also by reward and punishment (De Houwer, Thomas, & Baevens, 2001).
Specifically, children can develop similar behaviors to their parents through
direct modeling. Numerous studies have found intergenerational transmission of
diverse behaviors, from prosocial behavior to behaviors resulting in negative bio-
psycho-social outcomes. These behaviors can include church attendance, smoking
or alcohol consumption, abusive behavior, or health management. This is due to the
direct modeling that children experience through repeated exposure to their
parents’ behaviors; children have many opportunities to observe their parents, so
this modeling effect can be strong. Also, parents dictate the socio-economic context
in which their children are raised which has a strong influence on the formation of
attitudes and behaviors (Alwin, 1990; Steelman and Powell, 1991).
6
Intergenerational behavioral and value similarity between parents and
children can also be the result of learning via parental discouragement or parental
attitudes toward certain behaviors or values (Weistein and Thornton, 1989).
Parents can teach their young children by sharing or showing their opinions, or by
providing strict household rules with an educational purpose. For instance, it has
been widely studied to determine whether an anti-smoking rule or instruction by
parents in the household can decrease the likelihood that a child will start smoking.
Recent research shows that parents’ instructions or discouraging a child have an
influence on their child’s smoking behavior even when one or both of the parents
smoke.
B.3. Life Course Perspectives:
B.3.1. How Does Intergenerational Transmission Change?
As described above, according to the family socialization theory, childhood
and adolescence are considered critical periods for acquiring identity, values, and
experience and parents serve as important agents in socializing their children
during these periods (Alwin, 1984; Smith & Self, 1980; Starrels, 1992; Cunningham,
2001; Fan & Marini, 2000). Socialization theory assumes that parents can influence
children directly by social learning and modeling processes and also indirectly by
passing socio-economic status down.
However, parents compete with other socialization agents (e.g., peers,
teachers, and media) such that adolescence marks a turning point when parental
7
influence begins to wane (Younnis & Smollar, 1985). As they move through the life
course, family relationships—and how much they invest in these relationships—
start to change as well. As a result, adolescents begin to remove themselves from
their parents’ sphere of influence and expand their own identities and values
(Hoffman, 1984; Kroger & Haslett, 1988; Thornton, Orbuch, & Axinn 1995). The
process of individuation accelerates as children experience important events such
as moving out of their parents’ home or marrying and establishing their own family,
leading to greater divergence from the values of their parents (Bucx, Raaijmakers, &
Van Wel, 2010).
In this respect, family life course perspective provides a broad theoretical
framework encompassing early childhood socialization perspectives and
developmental changing perspectives, emphasizing the patterned and systematic
changes as families move through their family life stages and important life
transitions: independent living from parents, marriages, childbearing, widowhood,
and so on. Specifically, family life course perspectives focus on time as one of the
major components and focus on the multidimensional characteristics of time. That is,
the life course perspectives subdivide time into chronological time (age), social
process (cohort), and historical time (event) (Rodgers, 1973). These provide the
keys to why many values and behaviors may or may not be transmitted. Families
also change as a result of these influences. This feature of family life course
perspectives brings to light a unique dimension of the theory as it spotlights family
change.
8
B.3.2. Main Concepts in Life Course Perspectives
The main concepts in life course perspectives are cohorts, transitions,
trajectories, life events, and human agency.
Cohort is the general form for groups of individuals who share the same
birth period and who experience the same specific major social changes. Persons
who are in the same cohort tend to show common life choices and share similar
characteristics. For example, Glen Elder (Elder, 1974), who was one of the pioneers
in applying life course perspectives to empirical research, found that the great
Depression of the 1930s greatly influenced individual and family pathways; he also
found that the cohort who experienced the Great Depression at the same age tended
to develop similar characteristics. Another representative example is what created
the baby boomer cohorts. Many research studies have found that it was one of the
main historical events—World War II—which produced baby boomer cohorts in the
United States and, like Elder’s Depression-era cohort, this cohort displays unique
characteristics (Pearline & Skaff, 1996).
Cohort is an important concept to be considered in the research of
intergenerational transmission. To investigate intergenerational transmission, it is
indispensable to compare values and behaviors across generations. One of the ways
to determine successful transmission is to examine the concordance between parent
and child in the given research topic. Findings of dissimilarity can be attributed to
failed transmission and may be traced to a cohort effect, too. Parents and their
children belong to different birth cohorts and this fact alone can contribute to
9
developing values and behaviors uniquely their own. Thus, the interpretation of
similarity or dissimilarity of values and behaviors between parent and child needs
to consider the possible combination of both cohort and intergenerational
transmission effects.
Transitions represent changes from one stage or role to another. Individuals
within family contexts can experience life transitions by leaving home, getting
married, having children, or experiencing the death of spouse or other significant
family member (Hagestad, 2003). Life course transitions impact not only individuals,
but also the relationship between parent and child. For example, marriage is one of
the representative life transitions in the family context and researchers have found
this transition can have a significant impact on health outcomes (Williams &
Umberson, 2004).
One of the important transitions this dissertation will consider is the
transition to adulthood from adolescence. As individuals move through their life
from adolescence to adulthood, major life events such as leaving home, marrying, or
having children may happen. These life course transitions can greatly influence an
individual’s values, attitudes, beliefs, and behaviors. As an example, leaving home
can trigger parent-child divergence in various aspects, while getting married and
having children can—in their new-found shared experience of the role—provide
leverage toward convergence in their values and behaviors.
Trajectory indicates the pattern over the long-term period which is formed
by influential major life events or life transitions. For example, entering into
10
adulthood from adolescence is a significant transition and frequently prompts major
changes in individual’s lives (Ross, Schoon, Martini, & Sacker, 2009; Schulenberg et
al., 2003). This dissertation pays special attention both to predicting change in the
degree of intergenerational similarity in values and behaviors, and, also, to
identifying what factors contribute to changes over the family life course.
Human Agency is a concept that emphasizes that individuals can make their
own choices and actions, despite the fact that historical and social circumstances
necessarily exert a great influence. This concept naturally leads to one of the main
assumptions of life course perspectives – heterogeneity in life course trajectories.
Trajectory of intergenerational similarity of values and behaviors can be influenced
by factors such as individual developmental stages and life course transitions. A
person can choose the timing of certain life transitions or even choose whether to
make the transition at all. As a result, this can lead to heterogeneity of individuals’
value and behavior trajectories. When it comes to predicting patterns of
intergenerational similarity over the life course, this dissertation will also explore
factors which lead to between-individual variations and different patterns which are
driven by these factors.
B.4. Attachment theory:
What factors facilitate transmission? - Intergenerational Solidarity
Early family socialization theory provides a useful explanatory framework
for understanding mechanisms in intergenerational transmission. Life course
11
perspectives contribute to understanding the changing characteristics and between-
group-variations of the intergenerational similarity over the life course. As a next
step, this paper applies attachment theory to explain factors which may facilitate the
transmission process.
Attachment theory explains the importance of emotional and social
relationships to developmental processes (Bowlby, 1969). Ainsworth (1973)
examined intergenerational relations between infants and mothers and suggested
that attachment is a life-long influencing tie. Attachment with their parents when
children are young is the first and one of the crucial relationships. Merz, Schuengel,
and Schulze (2007) acknowledged that the emotional relationships which are
described from the view of attachment theory provide a valuable opportunity for
enhancing the solidarity concepts. Among the multiple dimensions of
intergenerational solidarity—used to classify intergenerational relationships by
previous researchers (Silverstein & Bengtson, 1997)—affectual solidarity is highly
related to attachment and can be well-explained from the perspective of attachment
theory (Merz et al., 2007).
Family research spanning several decades has demonstrated parent-child
affective solidarity both directly and indirectly affects intergenerational
transmission. Research found that children with strong affectual solidary in early
life were more likely to have similar values and attitudes to their parents (Bengtson,
2001; Roberts & Bengtson, 1996; Rossi & Rossi, 1990). In conclusion, the close
emotional attachment between parent and child which is described as affectual
12
solidarity has been found to have an impact on intergenerational transmission. Thus,
this dissertation explores the facilitating effect of affectual solidarity on
intergenerational transmission over the family life course.
13
Figure 1.1. Conceptual Framework
Conceptual Framework
Transmission
Processes
Transmission
Outcomes
Modeling
Social
Learning
Attachment
Early Affectual Solidarity
Child’s
Values &
Parent’s
Values & Attitudes
Child’s
Behaviors
Parent’s
Behaviors
Change over the Life Course
14
A. Objective of Dissertation
This dissertation addresses three important research gaps in the previous
empirical research on intergenerational transmission. (1) How does
intergenerational similarity change over the family life course? Do the transmitted
values and behaviors persist or degrade over time? (2) What factors shape
heterogeneous patterns of intergenerational transmission among individuals? Does
the quality of family relationships affect intergenerational transmission and inter-
individual variation? (3) What is the relative importance of direct modeling versus
indirect socialization (discouragement, attitude) mechanisms, and do these
mechanisms have a moderating effect?
The main goal of this dissertation is to explore the persistence, factors
contributing to inter-individual variations in transmission patterns, and interaction
among the mechanisms of the intergenerational transmission process. There are
only a few empirical investigations examining intergenerational transmission over
the life course beyond young adulthood using longitudinal data. The majority of past
research on intergenerational transmission focusing on children’s life course is
cross-sectional, and little research has been done utilizing dyadic family data.
Previous research has paid limited attention to potential heterogeneity in
transmission trajectories and factors which may contribute to such inter-individual
variations. This neglect limits knowledge in family studies as to familial
characteristics related to intergenerational transmission patterns. Thus, the current
research will fill some of these knowledge gaps.
15
The specific aims for this dissertation are:
Aim 1 (Chapter Three) examines the timing and persistence of
intergenerational transmission as related to theories of family socialization and the
effect of early family closeness as a facilitator of the transmission process. This
includes an investigation of variation in the degrees and timing of transmission for
social values and beliefs systems. This sub-study will add to the literature because
the findings serve as an impetus to further investigate the longitudinal effects of
intergenerational influence and the power of early family dynamics to shape
children’s values in times of rapid social change.
Aim 2 (Chapter Four) explores how intergenerational similarity in religious
beliefs changes over the family life course. Longitudinal patterns of similarity were
assessed beginning in young adulthood to see what factors—including family life
transitions such as marriage and parenthood—triggered pronounced changes in the
degree of parent-child similarity. A primary focus is on establishing whether early
family dynamics are leading factors in heterogeneous trajectories. Numerous
studies have been limited in their ability to explain the intergenerational
transmission because of their reliance on cross-sectional data. In addition, the few
studies which have used longitudinal data have examined religious transmission
from adolescence to early adulthood only, which limits our knowledge about the
pattern of parent-child similarity as children move beyond early adulthood and
enter midlife. This transition into the midlife period is critical in understanding the
life course trajectories of religious transmission because key life events like
16
marriage and the transition to parenthood often serve as triggers for change in
individual religious beliefs and practices.
Aim 3 (Chapter Five) investigates the intergenerational transmission of
smoking behaviors. The purpose of the third aim is to elucidate the relative
importance of intergenerational transmission mechanisms (modeling, parental
attitude, and parental discouragement) and explore the moderating effect among
mechanisms which have received limited attention in previous research. First,
parental smoking, parental permissive attitude toward smoking, and parental
discouragement are examined as predictors of a child’s tobacco use. Then, the
relative importance and interaction effects among mechanisms are compared. Lastly,
family solidarity is analyzed as a possible facilitating and/or prohibiting factor of
intergenerational smoking behavior.
17
Chapter Two: Overview of Data:
The Longitudinal Study of Generations
A. General Description of Sample
The Longitudinal Study of Generations (LSOG) was used to examine
intergenerational transmission. LSOG began in 1971 with 2,044 original
respondents who were members of three-generation families. Grandparents (G1)
were selected using a multi-stage stratified random sampling procedure from a
population of 840,000 individuals enrolled in an HMO in southern California. Adult
children (G2) and grandchildren (G3) of the G1 grandparents were also invited to
participate in the survey. Follow-up mail surveys were administered to original
respondents and additional spouses in 1985, 1988, 1991, 1994, 1997, 2000, 2005.
The great-grandchildren (G4s) of the G1s who were 16 or older were added in 1991
and in subsequent waves (Bengtson & Schrader, 1982).
LSOG has been widely used in the fields of sociology, family studies, and
gerontology to answer questions about the intergenerational exchange of support,
the transmission of values, and the effect of intergenerational relationships across
multiple generations.
B. Parent-Child Dyad Construction
Unlike most longitudinal surveys, LSOG has followed multiple family
members, including both mother and father and their children. As the survey
progressed over 34 years, multiple generations such as great-grandparents,
18
grandparents, parents, and their children have been followed. The uniqueness of the
LSOG data allow researchers to explore intergenerational similarity and differences
over the life course by constructing parent-child matched dyads. It is especially
important to have this dyadic first-hand information rather than relying on
potentially biased and unreliable second-hand reports when examining
intergenerational similarity, a notable strength of LSOG.
Chapters Three and Four of this dissertation focus on parent-child
relationships and changes in children’s similarity with their parents as they
transition from young adulthood to older adulthood. LSOG data from G2 parents and
their G3 children, selecting by G3 age, are used to examine this. In order to capture
this transition from young adulthood to middle age, LSOG data from 1971 (the first
wave of LSOG)—when the average age of G3s was 19 years old—and data from
2005 (the eighth wave of LSOG)—when the average age of G3s was 52 years old—
are used. These G3s were born between 1945 and 1955, and their G2 parents were
born between 1916 and 1931.
Chapter Five of this dissertation investigates the intergenerational
transmission of smoking behavior which is critical to examine in young adulthood.
LSOG data from G3 parents (born between 1945 and 1955) and their G4 children
(born between 1978 and 1983) are used to evaluate this. In 2000, which is earliest
survey year in which smoking behavior and attitude were measured, the average
age of parents and children is 48.7 years and 22.7 years respectively.
19
In forming these dyads, the same individuals may be present multiple times
since a child may have both a mother and a father, and a parent may have multiple
children. In order to appropriately adjust for family clustering in multivariate
models, robust standard errors were used. Only parent-child dyads which were
present for at least one of the time periods were analyzed.
C. Variable Construction used in Analyses
Intergenerational Transmission. The term “transmission” is used in the literature to
describe a complex process of intergenerational influence that includes direct and
indirect influences. In this paper, the term “intergenerational transmission” is used
to describe congruence or similarity across generations. To measure this
intergenerational similarity, significant beta-coefficients between parents’ values or
behaviors and children’s values or behaviors must be present. In addition, to
explore the trajectories of intergenerational similarity, calculated similarity scores
were used to predict the change of the score over time. This intergenerational
similarity score was calculated as the absolute value of the difference between a
child’s value score at the time of measurement and their parent’s value score at the
baseline. This difference score was reversed to represent similarity between
parents and children.
Values
Religious Beliefs were measured with four items assessing the strength of
beliefs in conservative religious orthodoxy and biblical literalism. These
20
items and their factor loadings for the pooled sample were: Every child should
have religious instruction; God exists as in the Bible; The United States would
be better if religion had more influence; We are all descendents of Adam and
Eve.
Gender Role Ideologies were assessed using a scale comprised of the
following four statements, each rated on a four-point scale from strongly
disagree to strongly agree: (1) Women who want to remove the word “obey”
from the marriage service don’t understand what it means to be a good wife.(2)
Husbands should have the main say in family matters. (3) Women’s liberation
ideas make a lot of sense to me. (4) It goes against nature to put women in
positions of authority over men.
Attitude toward Smoking was measured by asking parents if It’s OK for
people to smoke cigarettes once in a while. Responses ranged on a four point
scale from strongly agree to strongly disagree. Strongly agree and agree were
grouped into a single group since distribution was skewed.
Discouragement of Smoking was assessed using a scale comprised of the
following statement, each rated on a four-point scale from “not at all” to “very
strongly”: When you were growing up, did you mother or father discourage you
from smoking cigarette?
Behaviors
Smoking Behaviors. Participants were asked whether they have ever smoked
cigarettes on a regular basis and whether they still smoke on a regular basis.
21
Then, they were asked when they started and when they quit, if they ever
smoked. Participants were grouped as follows: “never smoked,” “ex-smoker,”
and “current smoker.”
Intergenerational Relationship Quality was measured by using children’s
reported affectual solidarity with each parent in young adulthood. The scale was
comprised of the following five items, measured for mothers and fathers separately:
(1) Taking everything into consideration, how close do you feel is the relationship
between you and your mother/father? (2) How well do you and your mother/father
get along at this point in time? (3) How well do you feel your mother/father
understands you? (4) How well do you feel that you understand your mother/father?
(5) How is communication between you and your mother/father? Responses ranged
on a six point scale from strongly negative to strongly positive.
Relationship type was coded to indicate that the parent in the dyad was a biological
parent (0) or a step parent (1).
Individual Characteristics. Individual characteristics such as age, gender, level of
education, and marital status were included.
22
Chapter Three:
Intergenerational Transmission of Values
Over the Family Life Course
1
A. Introduction
Research on the intergenerational relationships has consistently shown that
the social values of children are strongly associated with those of their parents,
suggesting a process by which parents intentionally and successfully influence the
values of their off-spring (Acock & Bengtson, 1978; Miller & Glass, 1989; Vollebergh,
Iedema & Raaijmakers, 2001; Copen & Silverstein, 2008). Whether
intergenerational similarity, often called transmission, of values occurs directly
through socialization, didactic training, modeling, or indirectly through status
inheritance, the evidence is clear that values are perpetuated through the
generations even accounting for strong cohort differences (Glass, Bengtson &
Dunham, 1986). However, less is known about the persistence and timing of
transmission over the family life course and the role that family dynamics plays in
the process. Almost all studies of intergenerational value transmission have
examined the issue in cross-sectional data where it is not possible to examine
whether values acquired by parents in early life endure or even emerge adulthood
and adopt adult role statuses.
1
This paper has been published in Advances in Life Course Research, Vol. 17 (3), 112-120, 2012. Co-
authors are Merril Silverstein and Jessical Lendon.
Link: http://www.sciencedirect.com/science/article/pii/S1040260812000305
23
This investigation addresses three primary questions. First, this paper asks
whether the timing of transmission matters. Do values get transmitted early in the
family life cycle, and then persist in children over time, or is transmission lagged
such that children fully adopt their parents’ values and beliefs when they reach
middle adulthood? Second, whether values are more powerfully transmitted when
intergenerational cohesion is strong in the early years of family life when sensitivity
to value adoption is considered strong. Third, whether intergenerational
transmission is stronger with respect to (1) belief systems that are privately held,
such as those associated with religiosity, or (2) social attitudes that are guided by
changing public opinion, such as gender role attitudes.
B. Review of Literature
Persistence and the Timing of Intergenerational Transmission
Socialization theory suggests that early experiences with parents directly
shape the values and beliefs of children. Parents can affect their children’s values
through modeling (learning through observation of parents’ behavior and expressed
attitudes) and formal training (such as religious instruction and verbal
discouragement that enforces conformity). Adolescence is considered a critical
period for the development of values and identity, when parents serve as important
agents of socialization of their children (Alwin, 1984; Smith & Self, 1980; Starrels,
1992). However, parents compete with other socialization agents (e.g., peers,
teachers, and media) such that adolescence marks a turning point when parental
24
influence begins to wane (Younnis & Smollar, 1985). Adolescents begin to remove
themselves from their parents’ sphere of influence and expand their own identities
and values (Hoffman, 1984; Kroger & Haslett, 1988). The process of individuation
accelerates as children experience important life transitions such as moving out of
their parents’ home, marrying, or becoming parents, leading to greater divergence
from the values of their parents (Bucx, Raaijmakers, & Van Wel, 2010). Because
timing appears to be a key feature of value socialization, it is important to consider
both pre- and post-adult children when studying the transmission of values
between generations (Acock & Bengtson, 1978; Bengtson, 2001; Hitlin, 2006).
An alternative perspective considers the conditions that produce value
congruence between generations (Cooper, Grotecant, & Gondon, 1983). As
adolescent children transition to early adulthood and beyond, they acquire new
roles that may put their values in alignment with those of their parents. From this
point of view, the success of intergenerational transmission is sensitive to life stage.
Children may need to reach a certain level of maturity or social status before
adopting the values to which they were earlier exposed to by their parents.
Cunningham (2001) found that values become ‘activated’ when children enter a
specific life stage, such as marriage or parenthood, when their susceptibility to early
socialization is maximized. as something akin to a sleeper effect. In this framework,
one might consider that there is a lag between the socialization of children and the
emergence of their values. A similar perspective is offered by proponents of the
status inheritance model, which proposes that parents indirectly transmit their
25
values to children through the transmission of their social status (Acock, 1984).
Adult children are more likely than not to share the socioeconomic status of their
parents and, thus, share the attendant values associated with those social positions.
Therefore, it is possible to argue that parental influence is delayed and becomes
activated when children come to occupy similar social roles and socioeconomic
positions as their parents.
Successful value transmission, whether immediate or delayed, is also likely
enhanced by the quality of early family relationships. Research on family
relationships has provided strong evidence that positive family relationships can
facilitate value transmission between generations (Pearce & Axinn, 1998; Tarvis,
2000; Bengtson, Biblarz, & Roberts, 2002). However, the timing and persistence of
parental influence over the life course are less understood.
The role of type of value orientation: religious beliefs and gender role ideology
The term “values” is used to represent a diffuse set of standards, ideals, and
goals. This paper use the term in the current research to refer to social attitudes and
belief systems, specifically attitudes toward women’s equality and religious beliefs,
respectively. These two types of values were chosen because the transmission of
each is proposed to require varying levels of parental effort (Vedder, Berry, Sabatier,
& Sam, 2009). Research suggests that values are differentially transmitted across
generations (Glass et al., 1986; Rohan & Zanna, 1996, Roest & Dubas, 2010).
Schwartz (1992) proposed that values that are more widely shared within the
community require less effort to successfully transmit to children, whereas values
26
that are more heterogeneous require more effort. The speed of social change has
received little investigation with regard to intergenerational transmission, but it
would be a reasonable to speculate that values undergoing the most rapid
transformation would be less transmissible across generations.
The values of gender egalitarianism diffused most rapidly through the
population during the 1970s and 1980s, but were adopted most vigorously by the
Baby-boom cohort (Cotter, Hermsen, & Vanneman, 2010). As the main factors that
influence gender role attitudes are educational attainment and working status
(Moen et al., 1997), it is likely that parents have a relatively minor direct influence
on their children, but may be prospectively important through status inheritance, the
effects of which would not be observed until education is completed and careers are
developed.
In contrast to social attitudes, religious beliefs are rooted more in family
culture. The family is the primary agent of socialization with regard to religious
beliefs, identity, practice, and salience (Glass et al., 1986; Myers, 1996, 2004;
Bengtson, Copen, Putney, & Silverstein, 2009). As religious values are shaped first
by family members, I expect that these values will be more strongly transmitted
during the younger and more impressionable phase of life and, once established,
exhibit greater stability over time when compared to gender role attitudes. Thus, I
expect parents to more strongly transmit religious beliefs than gender role attitudes
as their children mature into middle adulthood. I also expect gender role attitudes
27
to be transmitted in a lagged manner such that entry into adult roles will trigger
latent predispositions resulting from earlier family relationships.
C. Research Hypothesis
The current study adds to the literature by examining (1) the timing and
persistence of intergenerational influence based on theories of socialization and
status inheritance, (2) the role of early family cohesion as an accelerator of the
transmission process, and (3) variation in the rates and timing of transmission
depending on whether social attitudes or belief systems are studied. This analysis
addresses these issues using longitudinal data from parents and their adolescent
children over a span of almost 30 years.
D. Methodology
D.1. Data and Sample Characteristics
This study used two waves of data from the Longitudinal Study of
Generations (LSOG). The LSOG began in 1971 with 2,044 original respondents who
were members of 341 three-generation families. Grandparents (G1) were selected
via a multi-stage stratified random sampling procedure from a population of
840,000 individuals enrolled in southern California's first large HMO. Adult children
(G2) and adolescent grandchildren (G3) of the grandparents were also invited to
participate in the survey. Follow-up surveys were administered to original
respondents in 1985, 1988, 1991, 1994, 1997, 2000, and 2005. All data have been
28
collected by mail-back surveys (see Bengtson, Biblarz, & Roberts, 2002 for detailed
information).
The data for the current study was taken from the G2 parents and their G3
children in 1971 (W-1) and 2000 (W-7). The sample at baseline consisted of 519 G2
parents and of whom 396 (76.3%) were surveyed at W-7, and 556 G3 children of
whom 454 (81.7%) were surveyed at W-7. When parents and children were
matched to each other within families, the operational sample at baseline consisted
of 775 parent-child dyads (208 mother-daughter dyads, 209 mother-son dyads, 181
father-daughter dyads, and 177 father-son dyads).
Table 4.1 shows characteristics of the parents and children in this sample.
The average age of parents was 44.3 years at W-1 and 73.3 at W-7. The average age
of children was 19.2 years in W-1 and 48.2 years at W-7. Slightly more than half of
both parents and children in the sample were female. Children achieved greater
levels of education by 2000 than their parents.
D.2. Measurement
The two outcomes of interest were religious beliefs and gender role attitudes.
Each was assessed with multiple indicators at both W-1 and W-7 for each
generation. Factor analyses found the measures of each to be unidimensional when
tested by generation and time period, as well as within the pooled sample.
Religious Beliefs: Religious beliefs are measured with four items assessing
the strength of beliefs in conservative religious orthodoxy and biblical literalism.
29
These items and their factor loadings for the pooled sample were: Every child should
have religious instruction (.859); God exists as in the Bible (.944); The United States
would be better if religion had more influence (.908); We are all descendents of Adam
and Eve (.916). Reliability coefficients for these four items were α = .837 at W-1 and
α =.870 at W-7 for G2 parents, and α =.836 at W-1 and α =.888 at W-7 for G3
children.
Gender Role Attitudes: Gender role attitudes were assessed using a scale
comprised of the following four statements, each rated on a four-point scale from
strongly disagree to strongly agree: (1) Women who want to remove the word “obey”
from the marriage service don’t understand what it means to be a good wife.(2)
Husbands should have the main say in family matters. (3) Women’s liberation ideas
make a lot of sense to me. (4) It goes against nature to put women in positions of
authority over men. Reliability coefficients for these four items were α = .679 at W-
1 and α =.725 at W-7 for G2 parents, and α = .788 at W-1 and α = .704 at W-7 for
children.
Intergenerational Solidarity: To measure the strength of emotional cohesion
between generations, children’s reported of affectual solidarity with each parent at
W-1 was used (Bengtson & Roberts, 1991). The scale comprised the following five
items, measured for mothers and fathers separately: (1) Taking everything into
consideration, how close do you feel is the relationship between you and your
mother/father? (2) How well do you and your mother/father get along at this point in
time? (3) How well do you feel your mother/father understands you? (4) How well do
30
you feel that you understand your mother/father? (5) How is communication between
you and your mother/father? Responses ranged on a six point scale from strongly
negative to strongly positive. Reliability for these five items was α =.887. After a
composite score was calculated, it was then dichotomized into those having high
solidarity with a parent, as indicated by being above the average score total (N =
409), and those having low solidarity with a parent, as indicated as being below the
average score (N = 366).
Control variables: Following variables were included as control variables for
each generation: age (in years), gender, (0=male, 1=female), and education (1-8).
Because education of children continued to increase over time, an updated
education variable was included for this generation at W-7.
D.3. Analytical Strategy
To address our research questions, a cross-lagged structural equation model
representing religious beliefs and gender role attitudes was tested as latent
constructs. The basic model (Figure 3.1) shows the latent constructs with their
manifest indicators, measurement errors, and the proposed inter-relationships
among them. This model proposes that parents influence the values of their
children at W-1 and at W-7. The influence at W-7 can occur directly as a lagged or
sleeper effect on children and indirectly through stability of earlier formed values.
While it is assumed that values are transmitted downward from parents to children
at W-1, the possibility was opened that the possibility that children, as they mature,
31
may influence their parents over time, so the relationship between values of parents
and children at W-7 are estimated as bidirectional.
Structural models testing pathways of intergenerational value transmission
were estimated for the entire sample of dyads, for each value type. Then
intergenerational solidarity was stratified by high and low solidarity in a multiple
group model to test whether early emotional closeness moderated transmission
effects.
Figure 3.1. Measurement and longitudinal structural model of value transmission
between parents and children.
32
Attrition diagnostics revealed that respondents missing in W-7 had less
education than those who were retained and fathers and sons were more likely to
be lost to follow-up than mothers and daughters. To account for biases associated
with systematic attrition we estimated our models using all available data such that
respondents with only W-1 data were included in the analysis, as were follow-up
data from parents and/or children whose cross-generational partners were not
followed-up at W-7. Full Information Maximum Likelihood provided unbiased
estimates by simultaneously incorporating missing data patterns and their
predictors into the estimation process (Graham, 2003).
Estimated models are assessed by goodness of fit statistics, including CFI,
RMSEA, and chi-square difference testing in nested models, where CFI >.9 and
RMSEA <.05 generally indicate a good model fit (Maruyama, 1998). Because the
same children and parents could appear multiple times across dyads in the same
families, robust standard errors were used to adjust for family clustering in
multivariate models.
E. Results
For descriptive purposes, mean scores for each of the values were calculated
for both generations at W-1 and W-7. These means (Table 3.1.), show that at both
time periods, children were less religious and more had more egalitarian gender
role attitudes than their parents. Cross-generational and cross-time correlations for
the scales are shown in Table 3.2. Correlations between generations were stronger
33
at W-1 than at W-7 suggesting greater correspondence earlier than later in the
family life course. Stability in both generations was stronger for children than their
parents, showing that children are more prone to change their values over time.
Correlations of religious beliefs, both across generations and over time, were about
twice that of gender role attitudes, suggesting that the religious beliefs have a
stronger familial basis and are more enduring.
Table 3.1. Descriptive Characteristics of Parents (N=519) and Children (N=556)
Variables M SD Range
Number
of Valid
Cases
Parents
Mother (vs. father) .54 -- 0-1 519
Age (W-1) 44.25 5.20 30-67 519
Education (W-1) 4.86 1.41 1-8 514
Religious beliefs (W-1) 3.07 .85 1-4 480
Religious beliefs (W-7) 2.79 .75 1-4 255
Gender role attitudes (W-1) 2.51 .74 1-4 395
Gender role attitudes (W-7) 2.97 .54 1-4 277
Children
Daughter (vs. son) .51 -- 0-1 556
Age (W-1) 19.23 2.67 15-26 556
Education (W-1) 4.19 1.13 1-8 552
Education (W-7) 5.60 1.32 1-8 312
Solidarity with parent (W-1) 4.25 1.06 1-5 553
Religious beliefs (W-1) 2.71 .92 1-4 501
Religious beliefs (W-7) 2.63 .86 1-4 262
Gender role attitudes (W-1) 2.57 .84 1-4 404
Gender role attitudes (W-7) 3.21 .55 1-4 267
34
Table 3.2. Correlations between the latent variables for religious beliefs and gender
role attitudes
Parent's Value Child's Value Parent's Value Child's Value
W-1 W-1 W-7 W-7
Parent's Value W-1
0.524** 0.790** 0.455**
Child's Value W-1 0.261**
0.453** 0.635**
Parent's Value W-7 0.657** 0.244**
0.423**
Child's Value W-7 0.212** 0.420** 0.203**
Note. N = 723. Correlations for religious beliefs appear above the diagonal; those for
gender role attitudes appear below the diagonal. W-1 = 1971; W-7 = 2000.
* p< .05. ** p<.01. *** p<.001.
Before estimating the structural model, a test was run to determine if the
measurement model for each latent construct was consistent between generations
and over time. Table 3.3 shows the results for comparisons of models that imposed
equality constraints on measurement coefficients to one that allowed none or fewer
constraints. The first test showed that for each set of values, imposing equality
constraints between generations did not produce a significant decrement in model
fit. The second test, adding equality constraints over time showed no decrement in
model fit for both values. Thus, the fully constrained model became the basis for the
estimated structural models examining transmission effects for religious beliefs and
gender role attitudes. Measurement coefficients for this model are shown in the
bottom half of Table 3.3.
The results of the structural equation models are found in Table 3.4 for
religious beliefs and Table 3.5 for gender role attitudes. In each set of models, all
exogenous variables were allowed to correlate. I note that models for both sets of
35
values met the criteria for an adequate model fit based on CFI and RMSEA goodness-
of-fit indexes.
Table 3.3. Measurement models imposing generation and time constraints on
measurement coefficients for latent variables of religious beliefs and gender role
attitudes
Religious Beliefs Gender Role Attitudes
Model
Comparisons
Model Fit Increment Model Fit Increment
Generation-
constrained vs.
unconstrained
∆ χ²[df] = 7.214 [3], p > .05 ∆ χ²[df] = 2.043 [3], p > .05
Generation- &
time-
constrained vs.
generation-
constrained
model
∆ χ²[df] = 12.537[6], p > .05 ∆ χ²[df] = 3.964 [6], p > .05
Generation- &
time-
constrained
model
Measurement Structure Measurement Structure
Item Unstandardized
Measurement
Coefficient
Item Unstandardized
Measurement
Coefficient
We are all
descendents of
Adam and Eve
1.000 Wives should
obey their
husbands
1.000
Every child should
have religious
instruction
.496 Husbands
should have the
main say in
marriage
.959
God exists as in
the Bible
.965 Women’s
liberation ideas
make a lot of
sense to me
.755
The United States
would be better if
religion had more
influence
.673 Women should
not have
authority over
men
.758
Note. N=775. Generation equality constraints force measurement coefficients to be
equal across parents and children within each wave of measurement. Generation
and time constraint force measurement coefficients to be equal for parents and
children within and across waves of measurement.
36
Transmission of Religious Beliefs
Turning first to transmission effects in religious beliefs, Table 3.4 shows that
religiosity is transmitted downward from parents to children at W-1 when the
children are adolescents and young adults and then remains stable over time as the
children matured. The indirect transmission effect, an indicator of transmission
persistence (derived by multiplying the early transmission effect by the children’s
stability coefficient) was .316. There were no cross-lagged effects between
generations in either direction, nor was there a cross-generational correlation at W-
7. Since a significant zero-order correlation between generations at W-7 was found
(see Table .2), it can be inferred that this similarity resulted from an earlier
correspondence in religiosity that then remained stable over time. When the model
was stratified by high and low solidarity, there were no differences in estimated
coefficients that would indicate differential rates of transmission between the two
groups.
Transmission of Gender Role Attitudes
Comparable models for gender role attitudes are shown in Table 3.5. Gender
role ideology was transmitted from parent to child at W-1 and then remained stable
in children over time. No cross-lagged effects are observed, nor was there a cross-
generational correlation at W-7. The indirect transmission effect of parents on
children was .176, substantially smaller than the transmission persistence of
religious beliefs. Multiple group analysis based on solidarity revealed that in the
high solidarity group there was a lagged effect, whereby early parental values
37
produced an effect on gender role egalitarianism more later in the lives of their
offspring. In the low solidarity group, no such effect was observed. A test
comparing this model to one with equality constraints imposed on the lagged
coefficient produced a significant decrement in model fit ( ∆ χ²[df] = 9.175 [1], p
> .01), confirming the model with differential coefficients between groups is
preferable.
38
Table 3.4. Standardized estimates for structural equation models predicting
transmission religious beliefs.
Full Sample
(N=775)
Multi-group
Low Solidarity
(N=366)
High
Solidarity
(N=409)
Stability Effects
Parents [W-1]-[W-7] .873*** .882*** .791***
Children [W-1]-[W-7] .591*** .487*** .688***
Synchronous Effects
Parent [W-1] - Child [W-1] .534*** .461*** .652***
Parent [W-7] - Child [W-7] .014 -.079 .084
Cross-lagged Effects
Parent [W-1]-Child [W-7] .112 .123 .050
Child [W-1]-Parent [W-7] .027 .032 .069
Exogenous Variables
Parent [W-1]
Age - .057 -.012 -.008
Mother -.139 -.111 -.151
Education -.191*** -.179** -.191**
Parent [W-7]
Age .053 .002 .015
Mother -.033 -.185 .063
Daughter -.140 -.325** .074
Child [W-1]
Age -.011 -.060 -.007
Mother -.021 .016 -.033
Daughter .315** .164* .254*
Child [W-7]
Age .028 .071** -.006
Mother -.041 .126 -.091*
Daughter .258* .211 .299*
Education -.114* -.186** -.058
R
2
Parent [W-1] .070 .071 .074
R
2
Child [W-1] .351 .261 .446
R
2
Parent [W-7] .750 .831 .700
R
2
Child [W-7] .523 .440 .603
χ² (df) 847.15(405)*** 803.17(386)***
CFI .895 .901
RMSEA .053 .053
Note. W-1 = 1971; W-7 = 2000. * p< .05. ** p<.01. *** p<.001.
39
Table 3.5. Standardized estimates for structural equation models predicting
transmission of gender role attitudes.
Full Sample
(N=775)
Multi-Group
Low Solidarity
(N=366)
High
Solidarity
(N=409)
Stability Effects
Parents [W-1]-[W-7] .890*** .890*** 1.027***
Children [W-1]-[W-7] .472*** .445*** .464***
Synchronous Effects
Parent [W-1] - Child [W-1] .368*** .461*** .419***
Parent [W-7] - Child [W-7] .014 -.036 .017
Cross-lagged Effects
Parent [W-1]-Child [W-7] .095 -.130 .346*
Child [W-1]-Parent [W-7] -.058 -.090 -.045
Exogenous Effects
Parent [W-1]
Age .052 -.020 .023
Mother .269*** .436** .305***
Education .369*** .385*** .235**
Parent [W-7]
Age -.005 .023 - .068
Mother .011 .084 - .059
Daughter .072 .082 - .003
Child [W-1]
Age .057** .120 .179**
Mother -.091 -.141** .028
Daughter .033 -.037 .080
Child [W-7]
Age -.144 -.081 -.021
Mother -.048 .038 -.106*
Daughter .080* .171 .000
Education .122 .298** -.072
R
2
Parent [W-1] .155 .171 .164
R
2
Child [W-1] .155 125 .218
R
2
Parent [W-7] .864 .788 .985
R
2
Child [W-7] .277 .284 .447
Chi-square fit (df) 577.421(405)*** 537.807(386)***
CFI .903 .914
RMSEA .033 .032
Note. W-1 = 1971; W-7 = 2000. * p< .05. ** p<.01. *** p<.001.
40
F. Discussion
The purpose of this paper was to examine the transmission of values from
parents to children in order to identify the timing, persistence, and family context of
intergenerational transmission of religious beliefs and attitudes toward gender
equality. Using a unique longitudinal dyadic design it was able to be traced whether
transmission of values (1) occurred early and was persistent over three decades in
the lives of children, (2) became evident later in the life course of children when
they achieved mature adult statuses, and (3) was more powerful when
intergenerational relationships were emotional close. It was also proposed that
personal values of religiosity would be more strongly communicated than social
values of gender role ideology.
Overall, I was found that values were transmitted by parents at an early stage
in the lives of their children and subsequently remained stable in those children
over almost three decades. The correspondence of values between parents and
their mature children was due to transmission that occurred at an earlier life stage.
This supports socialization theory’s basic tenet that individuals permanently
acquire value orientations from important social agents at critical stages of
development.
Transmission was also manifest in a lagged manner for gender role attitudes,
but only under the condition of strong intergenerational solidarity. That is,
transmission of gender role egalitarianism did not emerge until the child reached
midlife and only for those who earlier had very close relationships with their
41
parents. It is possible that this sleeper effect is more likely to be found among
mother-daughter dyads than among other gender configurations, as gender role
attitudes have been found to become more salient when women enter work and
family roles (Cunningham, 2001).
Both the strength of early transmission and the extent of stability were
noticeably weaker for gender role attitudes than they were for religious beliefs,
suggesting that social attitudes are less family-based than are religious beliefs, as
well as more malleable in younger generations. This result supports a separation-
individuation perspective emphasizing the decline of parents’ influence when their
children achieve residential, financial, and emotional independence (Hoffman, 1984;
Kroger & Haslett, 1988). Turbulent social transformations in values, such as that
occurring with regard to women’s rising educational attainment and participation in
the labor market beginning in the 1970s, eroded the influence of parents whose
values were adopted at an earlier time. Here it is important to note that even when
attitudes change substantially on average, stability can still be strong within
individuals over time. Our findings support the notion that Baby-boomers were
swayed by an emerging Zeitgeist of gender egalitarianism rendering their earlier
expressed social values less deterministic of later outlooks.
On the other hand, religious beliefs are core values that are more powerfully
transmitted early in the lives of children, and, despite change in the religious
landscape of society, remain the province of training and socialization within the
family unit. These values are less susceptible to outside influences and
42
consequently more stable across generations and over one’s lifetime than are social
attitudes.
F.1. Contributions
In conclusion, this study showed values were robustly transmitted from
parents to their adolescent and young adult children, and then tended to remain
stable over the life course. This pattern of intergenerational influence and intra-
individual stability was most strongly observed for religious beliefs. Close family
relationships served as facilitators of value transmissions but only in the lagged
transmission of gender role attitudes. These findings will contribute to our
knowledge as an impetus to consider a wider assortment of values along the
personal-social continuum when investigating how parents and early family
dynamics, in conjunction with social change, shape children’s values into adulthood.
F.2. Limitations
This study had several limitations that deserve mention. First, gender
differences in intergenerational transmission should be considered for future
research. We know from other research that intergenerational value transmission is
stronger between same sex relationships than between opposite sex relationships.
Although we controlled for gender in our analyses, we did not examine how gender
interacts with value transmission and intergenerational solidarity. Second, the
analysis was relied on the reports of children to assess emotional closeness, yet
there can be differences between parents’ and children’s perspectives in evaluating
43
intergenerational relationships, with parents generally reporting stronger and
children generally reporting weaker affectual solidarity. Third, future research in
this area would benefit from including a wider assortment of socioeconomic factors
such as income occupation, and employment status to examine more precisely the
relative contribution of direct transmission vs. status inheritance in value similarity
between generations.
44
Chapter Four:
Intergenerational Similarity in Religious Beliefs
over the Family Life Course
2
A. Introduction
Religious ideology is one of the most enduring belief systems in society—
particularly in the United States—influencing life decisions and behaviors across the
lifespan (Booth, Johnson, Branaman, & Sica, 1995; Edgell, 2006; Lehrer, 1999; Levin,
1996; Miller & Thoresen, 2003; Waite & Lehrer, 2004). Parents have long been
recognized as the primary agent socializing children to religious beliefs and other
fundamental doctrines (Alwin, 1984; Erikson, 1950; Glass, Bengtson, & Dunham,
1986), a process that takes place through formal training and/or modeling of
desired behaviors (Acock & Bengtson, 1978; Bengtson, et al., 2009; Miller & Glass,
1989). Religious beliefs, in particular, appear to be shaped by family culture because
such beliefs emerge from formal religious instruction and rituals orchestrated by
parents and are reinforced by the commitment of parents to insure
intergenerational religious continuity in their children (Glass et al., 1986; Min,
Silverstein, & Lendon, 2012).
Yet the ways that individuals acquire and then maintain their religious
beliefs, and the factors that facilitate this process within families are not well
2
This paper has been submitted in Journal of Marriage and Family. Co-authors are Merril Silverstein and
Tara Gruenewald.
45
established in the empirical literature. The ability of parents to inculcate their
religious beliefs in their children is likely to be contingent based on the quality of
the parent-child relationship, whether the relationship is step or biological, and at
what point in the family life-stage religious beliefs are assessed. In this investigation
the temporal patterns or trajectories by which adult children diverge and/or
converge with their parents’ religious beliefs is described and relational factors and
family events of marriage and parenthood that may give rise to those trajectories is
examined.
B. Review of Literature
Research shows relatively robust correlations in religious beliefs between
parents and children, a correspondence that is usually taken as strong evidence of
intergenerational religious transmission (Bengtson et al., 2009; Erickson, 1992;
Francis & Brown, 1991; Francis & Gibson, 1993). However, high correlations are not
inconsistent with the existence of absolute differences or incongruence between
generations in their religious orientations that are driven by factors inside and
outside the family system. Cohort and historical factors may create religious gaps
between generations, as society becomes more secularized or religious; Research
suggests that the current generation of young adults is less religious than their
parents’ generation (Flannelly, Galek, Kytle, & Silton, 2010; Uecker, Regnerus, &
Vaaler, 2007). In addition, as children become independent they seek identities
separate from their parents and are increasingly influenced by their peers causing a
46
distancing from their families of orientation (Petts, 2009). On the other hand,
children may show a convergence with parental beliefs as they adopt similar roles
(e.g., parent) and environments. This paper suggests that intergenerational religious
congruence may not be an immutable property of families, but may ebb and flow
through the adult life span of children as children assume adult roles.
The ability to discern religious belief similarity between parents and their
children has been limited by reliance on cross-sectional data (Acock & Bengtson,
1980; Francis & Gibson, 1993; Hoge, Petrillo, & Smith, 1982; Milevsky, Szuchman, &
Milevsky, 2008). The few investigations using longitudinal data to address this issue
have examined religious transmission to children from adolescence to early
adulthood (Myers, 1996; Petts, 2009). Thus, little has been known about the pattern
of parent-child congruence in religiosity as children move beyond early adulthood
into periods of the life course when conservatizing events of marriage and family
formation might result in a closing of the generational gap.
The life course approach provides a conceptual framework for the current
investigation as it is concerned with the synchrony of individual, family, and
historical time in individual development (Elder, 1978; Hareven, 1994).
Intergenerational congruence should be manifest most strongly during the critical
period of adolescence (Alwin, 1984; Cunningham, 2001; Smith & Self, 1980) when
parents influence the religious orientations of their children by instituting formal
religious training, celebrating religious holidays, attending religious services
together, and stressing the importance of religious values in the home. This
47
congruence may become crystallized such that children maintain their received
beliefs over time. On the other hand, it is possible that the intergenerational
congruence degrades over time as the importance of other social agents (e.g., peers,
teachers, and media) grows and children become independent from their parents.
(Baltes, 1979; Bengtson & Kuypers, 1971; Hess & Warning, 1978; Rossi, 1980).
While some studies suggest that intergenerational transmission of religiosity and
other values is not dependent on life stage of children (Myers, 1996), others suggest
otherwise (Francis & Brown,1991; Willits & Crider, 1989).
As children move through the life course, family relationships and level of
investment in these relationships also change (Grusec, Goodnow, & Kuczynski,
2003; Younnis & Smollar, 1985). As a result, young adults may begin to remove
themselves from their parents’ sphere of influence and adopt their own identities
and values (Kroger & Haslett, 1988; Thornton, Orbuch, & Axinn, 1995). Thus, it is
possible that parent-child congruence may decrease as they mature into young
adulthood, but then increase again as those children mature into older phases of
adulthood and take on roles and responsibilities similar to those of their parents.
Intergenerational congruence in religious beliefs may also be influenced by
the nature of the parent-child relationship such as its quality or emotional closeness
and whether it is a biological or a step relationship. Emotional closeness leads to
more successful modeling of beliefs and attitudes, reinforcing intergenerational
continuity through stronger attachment styles (Bengtson, Biblarz, & Roberts, 2002;
Carlson & Knoester, 2011; Dudley, 1978; Hoge et al., 1982; Min et al., 2012; Myers,
48
1996). Step-relationships are apt to be less committed and of shorter duration than
biological ones, factors that impede intergenerational transmission of ideologies
(Carlson & Knoester, 2011). Previous research suggests that parents in step-parent
relationships have relatively little influence over their step-children (Ganong &
Coleman, 2004; Thomson, McLanahan, & Curtin, 1992) including the transmissions
of values (Carlson & Knoester, 2011; Myers, 1996).
C. Research Hypothesis
To summarize, this study examined how intergenerational congruence in
religious beliefs changes over time by capitalizing on dyadic longitudinal data
collected over 34 years. Trajectories in congruence were assessed beginning in
adolescence to determine if changes in congruence with parents are sensitive to
prospective family events of marriage and parenthood, as well as to early family
conditions as represented by emotional closeness with parents and whether the
relationship was biological or step. This paper hypothesizes that religious beliefs are
more consistent between generations in emotionally close and biological
relationships compared to less close and step-relationships, respectfully, and that
this higher degree of consistency will persist across the family life course. This
paper also anticipate that the emergence of a trajectory in which religious similarity
between generations in families will follow a life course pattern of early period
divergence followed by later convergence as children transition into marriage and
parenthood.
49
D. Methodology
D.1. Data and Sample Characteristics
Dyadic data from eight waves of the Longitudinal Study of Generations
(LSOG) was used to examine life span patterns in religious divergence. The LSOG
began in 1971 with 2,044 respondents from 341 three-generation families.
Grandparents (G1) participants were selected through multi-stage stratified random
sampling of 840,000 individuals enrolled in southern California’s first large HMO.
Enrollment into the survey was restricted to grandparents who had at least one
grandchild between the ages of 16 and 26. All adult children (G2) and grandchildren
(G3) in the designated age range descending from the G1s were recruited to
participate in the survey. The sample was resurveyed in 1985, 1988, 1991, 1994,
1997, 2000, and 2005. Follow-up rates of eligible participants were 65% between
1971-1985 (with an additional 10% recapture rate in subsequent surveys) and
averaged 85% for panels after 1985 (see Bengtson et al., 2002, for detailed
information). All surveys were mail-back questionnaires except in 2005 when about
half the G3s participated via a web-based survey.
Given that this study aims to follow religious beliefs in children from
adolescence to midlife, this paper relied on data from G3 children (N=554) and their
G2 parents (N=556) who participated in 1971. After matching children to their
parents, the analytic sample consisted of 1,084 G2 parent-G3 child dyads consisting
of 473 mother-daughter dyads (43.6%), 196 mother- son dyads (18.1%), 117
50
father-daughter dyads (10.8%), and 298 father-son dyads (27.5%). Table 4.1 shows
participation and matching rates in each wave. Because children’s religious beliefs
at each wave were contrasted with their parents’ religious beliefs at baseline, only
the religious beliefs of children were time-varying. This strategy was used to
maximize the sample size (G2 parents had accelerating mortality and non-
participation rates over the study interval) and localized the parents’ beliefs at the
time they would have been most effectively communicated to their children. It is
noted that a robustness check in which we estimated models incorporating time-
varying data from parents yielded results substantively similar to the strategy used
herein.
Table 4.1. Number of children matching to parents in 1971 at each wave of
measurement (N=1,084)
Number of Children
1971 1985 1988 1991 1994 1997 2000 2005
Parents in 1971 1084 705 702 665 672 625 620 602
51
Table 4.2 shows characteristics of parents and children in this study. The
average ages of parents and children in 1971 were 43.8 years and 19.2 years,
respectively, and the average age of children in 2005 was 53.0 years. Slightly more
than half of parents in the sample were mothers (54%) and a similar percentage of
children were daughters (55%). The large majority of children became parents
(80%) and parents (73%) by 1985.
Table 4.2. Descriptive characteristics of parents and children
Variables M or % SD Range
Number
of Valid
Cases
Parents
Mother (vs. father) 54 -- 701
Age (W-1) 43.82 5.26 30-67 701
Education (W-1) 4.82 1.37 1-8 696
Religious beliefs (W-1) 3.03 .86 1-4 697
Children
Daughter (vs. son) 55 -- 758
Age (W-1) 19.09 2.59 16-26 758
Education (W-2) 5.21 1.40 2-8 519
Religious beliefs (W-1) 2.69 .91 1-4 758
Married (by W-2) 80.04 -- 521
Became parent (by W-2) 72.73 -- 484
Solidarity with parent (W-1) 4.22 1.06 1-6 742
Note: W-1 = 1971; W-2 = 1985; W-7 = 2005
52
D.2. Measurement
Religious beliefs were measured by the strength agreement with four statements
representing conservative religious orthodoxy in terms of biblical literalism and
religious universalism: (1) God exists as in the Bible. (2) We are all descendants of
Adam and Eve. (3) The United States would be better if religion had more influence.
(4) Every child should receive religious instruction. Statements were evaluated (and
scored) with the following response options: strongly disagree (1), disagree (2),
agree (3), strongly agree (4). An additive scale was calculated from the four items
ranging from 4 to 16. Reliability coefficients across generations and waves of
measurement ranged from .84 to .89. The average scores of religious beliefs for
parents and children in 1971 were 3.03 and 2.69, respectively, showing that parents
had stronger religious beliefs than their children at baseline (see Table 4.2).
Religious belief congruence was operationalized as the absolute value of
differencing children’s religious belief scale scores at each wave of measurement
(between 1971 and 2005) from their parent’s religious belief scale scores in 1971.
The resulting difference score was reversed so that it represented the strength of
child-parent similarity ranging from 0-12, with 0=no similarity and 12=completely
similar.
Relationship quality and type: To measure relationship quality between
generations, we used children’s report of affectual solidarity with each parent in
1971 (Bengtson & Roberts, 1991). The scale comprised the following five items,
measured for mothers and fathers separately: (1) Taking everything into
53
consideration, how close do you feel is the relationship between you and your
mother/father? (2) How well do you and your mother/father get along at this point in
time? (3) How well do you feel your mother/father understands you? (4) How well do
you feel that you understand your mother/father? (5) How is communication between
you and your mother/father? Responses ranged on a six point scale from strongly
negative to strongly positive. Reliability for these five items was α =.89.
Relationship type: was coded as whether the parent in the dyad was a
biological parent (0) or a step parent (1).
Control variables: included gender of children (0=son; 1=daughter) and
gender of parents (0=father; 1=mother). Although there is evidence of gender
matching in religious transmission across generations, the interaction between
gender of children and gender of parents could not be tested because including this
term caused the model not to converge. Also controlled were age of child, and the
highest education achieved by children and parents, ranging from1-8, with 1=less
than 8 years of education and 8=post graduate education. Finally, parent’s baseline
religious beliefs were controlled as it has been found that more extreme religious
orientations of parents result in lower correspondence with children’s religiosity
over time (Petts, 2009).
54
D.3. Analytical Strategy
To identify discernible types of trajectories of intergenerational similarity in
religious beliefs, Growth Mixture Modeling (GMM), a variant of random effects
modeling, was employed using Mplus 5.2 software (Jung & Wickrama, 2008). GMM
groups heterogeneous patterns of individual change into common classes, with each
subject belonging to a specific growth pattern, and incorporates the ability to
estimate individual predictors of class membership.
In order to identify the best functional form of the growth patterns, a no
growth model (intercept only), a linear growth model (intercept and linear term),
and a non-linear growth model (intercept, linear term, quadratic term) were
successively examined (A model with a cubic term was excluded as it failed to
converge). The non-linear model yielded the best fit across a variety of class
specifications and was selected for further analysis.
To determine the number of classes that best describe the growth patterns in
the data, model fit indices were compared across increasingly complex non-linear
models ranging from 1-4 classes. Typically, the most parsimonious model with an
acceptable fit is considered optimal. There are a number of options for
comparatively assessing fit, including the Akaike information criterion (AIC), the
Bayesian information criterion (BIC), sample size-adjusted Bayesian (SSBIC) index;
entropy; and the Lo-Mendell-Rubin likelihood ratio test (LRT). Models with lower
values for the criterion indices and higher entropy values are considered better
fitting models. The LRT statistically tests the improvement in fit between sequential
55
models such that a significant p-value is interpreted as a statistically significant
improvement in fit for the inclusion of one more class (Nylund, Asparouhov, &
Muthen, 2007).
Models were estimated using a robust full information maximum-likelihood
(FIML) estimation procedure to treat missing data based on the MAR assumption
that attrition is random conditional on observed variables in the model (Enders,
2001). Attrition of G3s was non-random with respect to gender and education, with
greater attrition of men and lower educated individuals; both variables were
included as predictors in the estimated models. In another diagnostic we compared
a model that included respondents with incomplete data to a model that considered
only respondents who participated at all waves. There were no statistically
significant differences in the estimates or in the general trends of trajectories.
E. Results
GMM goodness of fit statistics is provided for multiple models in Table 4.3.
These statistics suggest that the optimal number of trajectory classes was either
three or four. Although the four class model had slightly smaller AIC, BIC, SSBIC
values than the three-class model, its LRT was not statically significant suggesting
that there is no detectable improvement in fit from the three-class model. Thus, the
three-class model was selected as the preferred grouping of the individual growth
estimates.
56
Table 4.3. Fit indices for growth mixture models of intergenerational similarity in
religious beliefs(N=1,084)
Growth Mixture Model
Fit index 1 classes 2 classes 3 classes 4 classes
AIC 22522.78 22290.33 22177.85 22103.53
BIC 22607.58 22395.08 22302.56 22248.19
SSBIC 22553.59 22328.38 22223.16 22156.08
Entropy - .79 .84 .78
LRT p value - .003 .033 0.39
Note. AIC = Akaike’s information criterion; BIC = Bayesian information criterion;
SSBIC = sample size-adjusted Bayesian information criterion; LRT = Lo-Mendell-
Rubin test
Table 4.4 shows the intercept and linear and quadratic slopes for three class
model. These estimates were used to illustrate predicted values in Figure 4.1 that
display growth patterns in religious belief similarity between parents and children.
The most common class, capturing 80.3% (n=870) of the sample, was characterized
by a stable-similar pattern that shows high similarity with parents in adolescence
that persists across time. Although linear and quadratic terms are statistically
significant for this class (see Table 4.4), the rate of change is small and because the
positive linear coefficient represents the instantaneous slope in 1971 (when
time=0), the negative quadratic term simply moderates this early increase in
similarity such that the overall pattern is one of stability.
57
Table 4.4. Representation and growth estimates of three classes derived from growth
mixture model (N=1,084)
Intercept Slope Quadratic
Class Membership
Probability
Est. (SE) p≤ Est. (SE) p≤ Est. (SE) p≤
Stable-similar 80.3% 9.66 (.11) .000 0.02 (.10) .030 -0.00 (.03) .045
Stable-dissimilar 10.5% 4.17 (.42) .000 0.06 (.59) .358 -0.00(.17) .534
Divergent 9.2% 9.57 (.34) .000 -0.40 (.40) .000 0.01 (.11) .000
The next largest cluster representing 10.5 % of the sample (n= 114) was
characterized as stable-dissimilar, characterized by low similarity at baseline and no
significant pattern of change over time. A third class capturing 9.2% of the sample
(n=100) exhibited a pattern of divergence across time. This group had similar
religious beliefs as their parents at baseline but decreased in similarity at an
accelerating rate though young adulthood before reaching a plateau of dissimilarity
thereafter.
58
Table 4.5. Multinomial logistic regression estimates predicting membership in
intergenerational religious similarity class (N=1,084)
Stable-dissimilar
(vs. Stable-similar)
Divergent
(vs. Stable-similar)
Predictor
Est. SE OR
Est. SE OR
Child’s age
.05 .05 1.05
-.06 .06 .94
Mother (vs. father)
-.59 .25 .55*
-.34 .30 .71
Daughter (vs. son)
-.54 .34 .58
-.85 .34 .43*
Parent’s education
.05 .10 1.05
-.03 .11 .97
Child’s education
.25 .13 1.38
-.02 .13 1.05
Parent’s religious
beliefs at baseline
.25 .10 1.28*
.01 .06 1.01
Solidarity with
parent at baseline
-.30 .12 .74*
.02 .12 1.02
Step relationship
(vs. biological
relationship)
.27 .51 1.31
1.16 .46 3.19*
Child married by
1985
-.11 .34 .90
.52 .50 1.68
Child became parent
by 1985
-.70 .41 .50
-.84 .41 .43*
* p ≤ .05; ** p ≤ .01; *** p≤.001.
Multinomial logistic regression examined factors that predicted membership
in the stable-dissimilar and divergent growth classes relative to the stable-similar
class that served as the reference group. Estimates shown in Table 4.5 reveal that
children were less likely to be stable-dissimilar with their mothers than their fathers,
59
and daughters were less likely than sons to be divergent with their parents over
time. Children were more likely to be stable-dissimilar with parents who held
stronger religious beliefs.
Figure 4.1. Trajectories of intergenerational similarity in religious beliefs over 34
years (1971-2005) (N=1,084)
Turning to the theoretically informative variables, it was found that of quality
and type of relationship with parents predicted class membership. Children who
reported weaker intergenerational solidarity at baseline were more likely to be in
the stable-dissimilar class than those expressing stronger solidarity, and children
were more likely to be divergent with step-parents than with biological parents.
Examining coefficients for the adoption family roles related to marriage and
60
parenthood, only parenthood predicted class membership. Children who became
parents by 1985 were less likely to be a member the divergent class relative to the
stable-similar class, suggesting a drift of the childless from the religious beliefs of
their parents as they grew older. However, marrying during the interval considered
did not predict class membership.
F. Discussion
The goal of this analysis was to examine the extent of religious agreement
between parents and children over more than three decades in the family life
course. Using longitudinal data between 1971 and 2005, intergenerational
differences in religious beliefs as children aged from adolescence and young
adulthood to early middle age were identified. Using a growth mixture model
approach, three distinct patterns of intergenerational similarity were identified.
Perhaps the most striking finding is that the large majority of children remained
fairly consistent in the extent to which they expressed similar beliefs to those of
their parents.
Further, the dominant group of children manifested high degrees of
intergenerational similarity, even as they established independent identities and
families of their own indicating that religious beliefs are reproduced across
generations with little degradation over time. This finding goes against the
dominant narrative of religious rebellion of the younger generation and is more
61
consistent family socialization and impressionable youth theories which suggest
that early home life experiences not only influence the formation of beliefs, values
and attitudes but their maintenance as well. However, it should be noted that
findings of religious weakening in the literature typically relies on evidence of
declining religious service attendance, religious switching, or de-identification with
a religious denomination rather than beliefs.
The boomerang type of trajectory characterized by early period divergence
and later period convergence was expected but did not materialize. However, one in
ten children experienced a distancing from the religious beliefs of their parents, a
group while small, might be in the vanguard as the driving agent of social change
observed in diverging religious orientations across generations. Nevertheless, this
group exhibited a slightly uptick in similarity as they entered midlife, a reversal
towards greater similarity occurring roughly at the same age as when their parents
expressed their religious beliefs.
F.1. Contributions
Our findings provide support for the power of early relationships with
parents and later family events to the degree to which the religious beliefs of
children are tethered to those of their families of orientation. The finding that
weaker relationships were related to larger intergenerational gaps however may be
symptomatic of families where disagreements in fundamental beliefs result in
conflict and tension in relationships. That large religious gaps persist over three
62
decades suggest the possibility of continued emotional distancing as well, as it is not
unusual for religious rebellion to cause long-term relational rifts between
generations in the family (Bengtson, Putney, & Harris, 2013). On the other hand, it is
not difficult to imagine that children with close and warm intergenerational
relationships at this critical time in late adolescence and early adulthood would be
more likely to identify with their parents and that this emotional closeness serves as
an accelerant to religious transmission. This interpretation is consistent with earlier
research showing that children who had close relationships with their parents were
more likely to adopt their parents’ religious orientations (Dudley, 1978; Hoge et al.,
1982; Myers, 1996).
The finding that stepchildren tend to diverge from the religious beliefs of
their step-parents was expected in that step-parents, on average, have less early
exposure to their step-children than do biological parents. However, this finding
should be interpreted in light of the fact that information about when step-
relationships began was lacked. Future research investigating the connection
between the timing of remarriage and step-family formation and intergenerational
similarity in attitudes will help us better understand the process of
intergenerational transmission in non-traditional family structures.
Evidence that becoming a parent is associated with greater similarity in
religious beliefs with one’s own parents over time is consistent with the idea that
families have strong desires to maintain intergenerational continuity in religious
orientations. However, the childless may seek an alternative child-free life-style due
63
to their adoption of attitudes that diverge from those of their upbringing. Whether
attitudes proceed, or are outgrowths of, family transitions, remains an open
question. Nevertheless findings speak to the potential interdependence of
ideological change and parental role adoption.
F.2. Limitations
Several limitations of the present research deserve mention. First, this paper
explored intergenerational agreement but not the directionality of disagreement.
Thus, the current findings do not speak to whether children were more or less
religious than their parents, though most of the disagreement trended in a negative
direction toward weaker religious beliefs of children. This paper focused on
absolute differences so as to inclusively consider all generation gaps as meaningful.
The relatively small group who diverged from their parents in a more religious
direction did not form a discernible trajectory group in supplemental analyses,
however future research will be benefited by focusing on this interesting group.
Second, measurement of parents’ beliefs was restricted to the baseline
period for practical reasons. This paper recognized that parents may still influence
their children well into adulthood and it is possible that parental influence on
children changes across family life course (and may even be bidirectional). To
account for parental change a supplemental analysis was estimated using parents’
religious beliefs that were time-varying and found as the preferred model the same
64
three trajectories reported in this paper. However, low power from the small
sample size caused the fit statistics to be unstable and the multinomial regression to
not converge. Thus, this paper elected to use models incorporating parental beliefs
from the initial wave of measurement only.
Third, it was not able to examine the gender composition of parent-child
dyads with an interaction term due to sample size restrictions. Others have found
belief transmission similarity is stronger among same sex parent-child dyads (Axinn
& Thornton, 1993). Larger samples would aid in the examination of these cross-
gender patterns.
Forth, this paper examined only one cohort of dyads, and it is possible that
the younger generation is unique in being baby-boomers who were first measured
just following the societal upheavals and social changes of the 1960s. To the extent
that this generational pairing is unique, caution should be exercised in generalizing
the study’s findings to other cohorts.
The weaknesses noted above should be understood in light of unique dyadic
data and the 34 years over which children were followed. These features allowed for
an assessment of children at both temporal ends of the family life cycle to examine
the roles played by early exposure to families of orientation and later adoption of
family roles in full adulthood.
In conclusion, this study showed more stability than change and more
agreement than disagreement in the religious beliefs of children relative to their
parents. Perhaps this reflects the deep role that religion plays in American families
65
and the persistence of religious belief systems that may survive the waning
behavioral and participatory manifestations of religiosity. This paper hopefully
contributes to a better understanding of religious beliefs over the life course from
an intergenerational perspective and the role that family dynamics play in the
process.
66
APPENDIX 1
Table 4.6. Number of Biological and Step relationships in Each Wave
Wave Dyads Number of
Cases
Biological
Relationships
(%)
Step
relationships
(%)
1971
Parent in 1971 –child in 1971
1084 878 (81.0) 129 (11.9)
1988
Parent in 1971 –child in 1988
705 635 (90.1) 70 (9.9)
1991
Parent in 1971 –child in 1991
665 590 (88.7) 75 (11.3)
1994
Parent in 1971 –child in 1994
672 600 (89.3) 72 (10.7)
1997
Parent in 1971 –child in 1997
625 550 (88.0) 75 (12.0)
2000
Parent in 1971 –child in 2000
620 549 (88.5) 71 (11.5)
2005
Parent in 1971 –child in 2005
602 530 (88.0) 72 (12.0)
67
Chapter Five:
Intergenerational Transmission of Smoking Behavior
A. Introduction
Studies examining intergenerational transmission of tobacco use have shown
that children whose parents smoke are also more likely to smoke, themselves
(Jackson & Henriksen, 1997; Chassin, et al., 1988; Kandel & Wu, 1995; Sargent &
Dalton, 2001). A number of possible within-family mechanism have been proposed
for parent-child smoking transmission, including: genetic factors (Boardman,
Blalock, & Pampel, 2010; Bock & Goode, 2006), exposure to parental smoking, or
modeling behaviors, parental socialization through attitudes, and discouragement
(Jackson & Dickinson, 2003; Jackson & Henriksen, 1997; Ennett, Bauman, Foshee,
Pemberto, & Hicks, 2004; Tinsley, 1992; Chassin et al., 2010).
Nevertheless, there is little consensus regarding which components of
parental socialization are most important for influencing child’s smoking
behavior—modeling, implicit, or explicit—and whether the parent-child
relationship moderates the link between socialization and child’s smoking behavior.
As a result, this study examines which mechanisms (e.g. parental smoking behavior
(modeling), parental attitude (implicit anti-socialization), or parental
discouragement (explicit anti-socialization)) most strongly predict child’s smoking
behavior and whether intergenerational emotional closeness moderate the
association between parental anti-socialization and child’s smoking.
68
Comparing mechanisms’ relative significance may shed light on which factors
are most important for smoking behavior, and thus facilitate the building of better
targeted anti-smoking interventions. Furthermore, if the mechanisms act
competitively, understanding their relative significance may help identify what
circumstances are more likely to foster intergenerational transmission of smoking.
Given that most smokers begin smoking around age 16—a time when they are
typically still living with, and most likely influenced by their parents—knowing
which family circumstances increase smoking uptake may allow us to intervene
before smoking behavior ever begins.
B. Review of Literature
Research has consistently shown that tobacco use is transmitted from one
generation to another (Avenevoli & Merikangas, 2003; Darling & Cumsille, 2003;
West, Sweeting & Ecob, 2002; Jackson & Henriksen, 1997; Chassin et al., 1984;
Kandel & Wu, 1995; Sargent & Dalton, 2001) and has identified several possible
mechanisms for this phenomenon. Twin studies have suggested that smoking
behavior may be 50% heritable (Avenevolie & Merikangas, 2003; Boardman et al.,
2010; Bock & Goode, 2006; Li, 2003). However, there is evidence that the remaining
variability in intergenerational transmission of smoking may partly depend on
behavior modeling, internalization of parents’ attitude toward smoking, and direct
socialization.
Behavior modeling, explicit socialization, and implicit socialization
69
Extensive research has gone into understanding how behavior modeling—
regardless of parental smoking—influences tobacco use among adolescence.
Avenevoli and Merikangs reviewed 87 studies assessing the influence of parent and
sibling smoking on adolescent smoking and found modest associations between
parent—child concordance of smoking. Nevertheless, it is not clear whether this
transmission resulted from modeling, exposure to permissive smoke family
environment, or variations in the levels of parental discouragement.
In an effort to explain the occurrence of intergenerational transmission of
smoking behavior, various types of parental influence has been examined. For
instance, one mechanism that has received a significant amount of research
attention is parental anti-smoking socialization. There is evidence that positive
parenting practices such as parental support (Simons-Morton, 2004; Wills et al.,
2004) and parental monitoring (Dishion & McMahon, 1998) may play a protective
role in child’s smoking behavior. Also, research found that children living in smoke-
free family environments were significantly less likely to smoke than children living
in households with no restrictions on smoking (Farkas et al., 2000; Proescholdbell,
Chassin, & MacKinnon, 2000; Wakefield, et al., 1998).
In addition to explicit socialization, such as parental monitoring, smoke-free
rules, or discouragement of smoking, implicit socialization such as knowledge or
attitudes towards smoking have also been hypothesized to influence behavior.
Henricksen & Jackson examined the effect of three different types of anti-smoking
socialization (consisting of knowledge, attitude, and discouragement of smoking)
70
and found that these implicit socialization also impacted a child’s likelihood to
smoke (Henriksen, & Jackson, 1998; Proescholdbell et al., 2000).
Relative importance of mechanisms
Over the years, research has pointed to numerous different parental
mechanisms for explaining the likelihood of tobacco use among children, ranging
from behavior modeling to discouraging smoking and instilling negative attitude
toward smoking. However, such studies often examine the impact of various
mechanisms in a single model—forcing them to compete with each other—and then
report which mechanisms remained significant (Bailey, Ennett, & Ringwalt, 1993;
Chassin et al., 1998; Hill, Hawkins, Catalano, Abbott, & Guo, 2005; Kodl, &
Mermelstein, 1993; Jackson, & Henriksen, 1997). On the other hand, only a few
studies have focused on the relative importance of mechanisms (Nolte, Smith,
O’Rourke, 1983; Newman & Ward, 1989; Bailey et al., 1993). As a result, there is a
need to further explore the effect of predictors independently and then compare this
to findings from integrated models.
Interaction between behavior modeling and implicit/explicit socialization
Additionally, few studies have considered the interaction between parent’s
smoking behaviors and parent’s anti-smoking socialization. While several papers
examined the moderating effect of family bonds (Doherty & Allen, 1994; Wen, Tsai,
Cheng, Hsu, Chen & Lin, 2005) and gender (Chassin et al., 2003) on
intergenerational smoking transmission, little is known about the interaction effect
71
of parents’ anti-socialization attitude or discouragement (Bailey et al., 1993) and
whether this influences intergenerational smoking transmission.
While parent’s smoking behavior has been known to increase the likelihood
of their child smoking, in contrast, parent’s anti-smoking socialization has been
shown to decrease the likelihood of their child smoking. Unfortunately, little is
known regarding whether anti-smoking socialization has the potential to buffer the
intergenerational transmission of smoking, even under circumstances where one or
both of the parents are smokers, or whether anti-smoking socialization would
increase the transmission process, if children’s’ tendencies are to disobey their
parent’s discouragement as a result of observing parent’s discordance between
attitudes and behavior. Therefore, the investigation of such interactions may answer
in what way parents’ socialization influences their child’s smoking even when
parents smoke.
According to the literature, the extent and rate of parental influence on
child’s socialization to smoking highly depends upon the quality of intergenerational
relationships (Tilson, McBride, Lipkus, & Catalano, 2004; Hill et al., 2005). Given that
an understanding of whether attitudes or discouragement moderate the
intergenerational transmission of smoking and if so, what family related factor
contribute to this effect, has the potential to provide practical information that may
help build more targeted intervention and possibly prevent smoking uptake in
adolescence and young-adults, this study aims to examine whether the moderating
72
effect of parent’s anti-socialization on intergenerational smoking transmission
varies by family solidarity.
C. Research Purpose
Using dyadic data of parents and their children, the current study adds to the
literature by examining (1) the relative importance of mechanisms between
parental modeling and direct/indirect anti-socialization; (2) the moderating effect
of anti-socialization, to see whether parent’s anti-socialization reinforces or buffers
the influence of parental smoking behavior on child’s smoking; and (3) the role of
family cohesion as an activator of the moderating effect. We hypothesized that
parent’s smoking behavior is positively associated with child’s smoking behavior,
whereas parent’s anti-smoking socialization is negatively associated with child’s
smoking behavior. Furthermore, we hypothesize that the interaction between
parent’s smoking behavior and parent’s anti-smoking socialization will influence
child’s smoking, conditional on parent-child’s emotional closeness.
D. Methodology
D.1. Data and Sample Characteristics
Our analyses took advantage of parent-child dyadic data from the
Longitudinal Study of Generations (LSOG). The LSOG began in 1971 with 2,044
respondents from 341 three-generation families. Grandparents (G1) participants
were selected through multi-stage stratified random sampling of 840,000
73
individuals enrolled in southern California’s large HMO. Enrollment in to the survey
was restricted to grandparents who had at least one grandchild between the ages of
16 and 26. All adult children (G2) and grandchildren (G3) in the designated age
range descending from the G1s were recruited to participate in the survey. The
sample was resurveyed in 1985, 1988, 1991, 1994, 1997, 2000, and 2005. Follow-up
rates of eligible participants were 65% between 1971 and 1985 (with an additional
10% recapture rate in subsequent surveys) and averaged 85% for panels after 1985
(see Bengtson et al., 2002, for detailed information). All survey were mailed-back
questionnaires except in 2005 when about half the G3s participated via a web-based
survey.
We utilized data from G3 parents (N=435) and their G4 children (N=469)
who participated in the survey in 2000, which is the first year tobacco-use was
assessed. The total analytic sample included 190 father-son dyads (26.7%), 142
mother-son dyads (19.9%), 219 father-daughter dyads (30.7%), and 162 mother-
daughter dyads (22.7%).
D.2. Measurement
The outcome of interest was child’s tobacco use and predictors included
parents’ tobacco use, attitude toward smoking, and discouragement of smoking.
Affectual solidarity was included as a moderating variable. Age, gender, education of
both parent and child were controlled.
74
Smoking Behavior: We classified parents and children as either having “never
smoked,” being an “ex-smoker,” and being a “current smoker” on the basis of the
following two questions: “Have you ever smoked cigarettes on a regular basis?” and
“Do you still smoke in a regular basis?”
Explicit anti-smoking socialization: To measure parents’ active anti-smoking
socialization, , children were asked When you were growing up, did your mother or
father discourage your from smoking cigarettes? Responses ranged from 1 (very
strongly) to 4 (not at all) This item was reverse coded. Higher scores mean that
children perceived their parents more strongly discouraged them from smoking
cigarettes.
Implicit Anti-Smoking Socialization: To measure parents’ attitude toward
smoking, parents responded to the following two statements:“It’s okay for people to
smoke cigarettes once in a while” and “Smoking should be banned in all public places.”
Responses ranging from 1 (strongly agree) to 4 (strongly disagree). The second
statement was reverse coded and two items were averaged. A greater score
indicated stronger anti-smoking attitude.
Intergenerational Affectual Solidarity: To measure the strength of affection
between generations, we used children’s reported of affectual solidarity. The scale
comprised the following five items: (1) Taking everything into consideration, how
close do you feel is the relationship between you and your mother/father? (2) How
well do you and your mother/father get along at this point in time? (3) How well do
you feel your mother/father understands you? (4) How well do you feel that you
75
understand your mother/father? (5) How is communication between you and your
mother/father? Responses ranged from 1 (strongly negative) to 6 (strongly positive)
and were averaged to create the composite scale. Scores above the median were
classified as “high solidarity” and scores below the median indicated “low solidarity.”
Control variables: We included as control variables the following
characteristics for each generation: age (in years), gender, (0=male, 1=female), and
education (1=8
th
grade or less, 2=some high school, 3=high school, 4=training after
high school, 5=some college, 6=college or university graduate, 7=more academic
years beyond college, and 8=post-graduate degree).
D.3. Analytic Strategy
Univariate descriptive analyses were performed to describe parent and child
socio-demographic and smoking-related characteristics and a correlation matrix
assessed the associations between smoking behavior and smoking related attitudes
and discouragement.
Two bivariate multinomial regression models investigated the effect of parent’s
smoking behavior and parent’s anti-smoking socialization on children’s smoking
behavior (ex-smoker vs. non-smoker and current smoker vs. non-smoker). Then we
examined these two components simultaneously in a multivariate model to test the
relative importance of behavior, attitude, and discouragement (variance inflation
factor (VIF) indices were less than 1.5, indicating there was no multicollinearity).
To examine the potential interaction between parents’ smoking behavior and
parent’s anti-smoking socialization interaction terms of the three measures were
76
entered into the model. Then finally intergenerational solidarity was stratified by
high and low solidarity in a multiple group model to test whether emotional
closeness is a significant moderator of parental socialization and the interactions
between the socialization mechanisms.
E. Results
Table 5.1 shows descriptive characteristics of parent-child dyads. The
average ages of parents and children were 48.7 years and 22.7 years respectively.
43 percent of the parents were mothers and 53 percent were daughters. Total of 42
percent of participating parents have reported that they have used tobacco and 58
percent have reported as never having smoked. Among the 42 percent of parents
with experience in tobacco use, 37.3 percent have responded that they are still
smoking regularly when the survey was conducted. Also, 31.7 percent of children
have reported experiencing tobacco use and of the 31.7 percent, 49.1 have reported
as no longer smoking, while 50.9 percent still smoke regularly.
77
Table 5.1. Characteristics of the Parents and Children (N=713)
Variable Parent Child
Range Mean (SD) % Mean (SD) %
Age
P: 34-69
C: 16-40
48.72 (.17) 22.66 (.19)
Female
42.64 53.44
Education
1-8 5.42 (.05) 4.37 (.06)
Having Child
100 20.90
Never Smoke
57.92 68.30
Ex-Smoke
26.37 15.57
Current Smoke
15.71 16.13
Smoking Attitude
1-4 3.12 (.03)
Discouragement
3.38 (.04)
Affectual Solidarity
4.30 (.04)
Table 5.2 is a correlation matrix showing the smoking practices and
behaviors of parents and children and the relationships among parents’ tobacco
usage, attitudes towards smoking, and the degrees of discouragement in their child’s
tobacco use. First, parents’ usage of tobacco was positively associated with their
child’s tobacco use. Similarly, Anti-smoking attitude of the parents was positively
related to the parents’ stricter discouragement in their child’s tobacco use. Parents
who smoke were more likely to have a more lenient behavior towards discouraging
their children who smoke (For details, refer to Appendix 2).
Table 5.2. Correlation Matrix (N=713)
Child’s
Smoke
Parent’s
Smoke
Discourag
ement
Smoking
Attitude
Affectual
Solidarity
Child’s Smoke
1.000
Parent’s Smoke
.208** 1.000
Parent’s Discouragement
-.055 -.290*** 1.000
Parent’s Smoking Attitude
-.181*** -.516** .275** 1.000
Affectual Solidarity
-.029 -.128*** .308*** .068 1.000
78
Bivariate relationships between parent’s smoking and child’s smoking
Parents’ tobacco usage was related to child’s tobacco use (see Table 5.3). A
child with a smoking parent was significantly more likely to have an experience with
smoking. When a parent has ever smoked, the child was more than twice as likely to
smoke (OR=2.42, p < .001), or still smoke regularly (OR=2.12, p < .01). Parents who
still smoke regularly are more likely to have a child who has smoked (OR=2.17, p
< .05) and who still regularly smoke (OR=2.29, p <.05). Generally, parents’ use of
tobacco was positively associated with child’s tobacco use and such relationship
was not varied by whether parent ‘has smoked but not anymore’ or ‘still smoke
regularly’.
Table 5.3. Multinomial logistic regression estimates predicting child’s tobacco use:
Modeling mechanism only (N=713)
Child: Ex Smoker
(vs. Never Smoke)
Current Smoker
(vs. Never Smoke)
Predictor Est. SE OR Est. SE OR
Socio-economic variables
Child’s Age .11 .04 1.11*** .09 .03 1.10***
Daughter (ref. son) .56 .47 1.75* .36 .46 1.44
Child’s Education -.03 .13 .97 -.39 .08 .68***
Parents Age -.03 .03 .97 -.03 .03 .97
Mother (ref. father) .17 .20 1.18 .04 .18 1.04
Parent’s Education -.19 .07 .83* -.07 .09 .93
Affectual Solidarity .01 .11 1.01 -.04 .12 .97
Mechanisms
Parent’s Smoking (ref. Never)
Ex-Smoker .88 .69 2.42*** .75 .58 2.12**
Current-Smoker .78 .75 2.17* .83 .79 2.29*
* p ≤ .05; ** p ≤ .01; *** p≤.001.
79
Bivariate relationships between parent’s anti-smoking socialization and child’s
smoking
Next I have examined the association between parents’ anti-smoking
attitude/discouragement and child’s smoking without including parent’s smoking
behavior (Table 5.4). After controlling the socio-economic variables, parents’
negative attitudes towards smoking was negatively related to child’s smoking
experience. Regardless of whether the child was still smoking or not, the child
whose parents portray higher level of anti-smoking attitude were less likely to be an
ex-smoker (OR=.63, p < .001), or to be a current smoker (OR=.61, p <.001). Unlike
the parents’ attitude toward smoking, parents’ degree of discouragement seem to
have no direct association with child’s tobacco use.
80
Table 5.4. Multinomial logistic regression estimates predicting child’s tobacco use:
Anti-smoking socialization only (N=713)
Child: Ex Smoker
(vs. Never Smoke)
Current Smoker
(vs. Never Smoke)
Predictor Est. SE OR Est. SE OR
Socio-economic variables
Child’s Age .11 .04 1.11**
*
.10 .04 1.10**
Daughter (ref. son) .50 .43 1.65* .37 .46 1.45
Child’s Education -.03 .12 .97 -.37 .08 .69***
Parents Age -.02 .02 .98 -.02 .03 .98
Mother (ref. father) .15 .20 1.16 .06 .20 1.07
Parent’s Education -.24 .07 .78** -.11 .09 .90
Affectual Solidarity -.03 .14 .97 -.09 .13 .92
Mechanisms
Parent’s Attitude (Anti-Smoking) -.47 .08 .63*** -.49 .10 .61***
Parent’s Discouragement .10 .17 1.10 .15 .20 1.16
* p ≤ .05; ** p ≤ .01; *** p≤.001.
Multivariate model of Intergenerational transmission of smoking behavior
In the multivariate model, all three mechanisms - parent’s tobacco usage,
anti-smoking attitude, and discouragement of smoking were introduced (see Table
5.5). Both parental smoking and attitude were significantly associated with child’s
smoking but not discouragement of smoking. Interestingly, although the parent’s
previous smoking experience was significant in predicting a child’s smoking,
whether a parent currently smokes or not was not related to the child’s smoking.
This result is different from the findings of univariate model (see Table 5.3). This
result indicates that parent’s anti-socialization mechanism partially mediated the
relationships between parent’s current smoking and child’s smoking. Parents who
81
still regularly smoked tended to have a more permissive attitude which was related
to child’s smoking.
Table 5.5. Multinomial logistic regression estimates of moderating effects predicting child’s
tobacco use: Multivariate mechanisms (N=713)
Child: Ex Smoker
(vs. Never Smoke)
Current Smoker
(vs. Never Smoke)
Predictors Est. SE OR Est. SE OR
Socio-economic variables
Child’s Age .11 .04 1.11** .10 .04 1.10**
Daughter (ref. son) .52 .44 1.68 .38 .47 1.47
Child’s Education -.02 .13 .98 -.36** .09 .70
Parents Age -.03 .03 .97 -.02 .04 .98
Mother (ref. father) .20* .26 1.22 .08 .20 1.09
Parent’s Education -.22 08 .80 -.09 .09 .91
Affectual Solidarity -.04 .13 .96 -.09 .12 .92
Mechanisms
Parent’s Smoking (ref. Never)
Ex-Smoker .84** .68 2.32 .70* .58 2.02
Current-Smoker .32 .53 1.37 .35 .56 1.42
Parent’s Attitude (Anti-
Smoking)
-.44** .09 .64 -.44* .12 .64
Parent’s Discouragement .16 .18 1.18 .20 .21 1.23
* p ≤ .05; ** p ≤ .01; *** p≤.001.
Moderating effects involving parent’s anti-smoking socialization on child’s smoking
Lastly, two multinomial regressions were analyzed to see the interaction
effects of 1) parent’s attitude and 2) parent’s discouragement on parent-child
transmission of smoking. Both effects were not significant (not shown) not until the
analyses were performed separately according to the level of intergenerational
emotional closeness (Table 5.6). This result shows that the emotional closeness
between parent and child is an important condition which activates the
82
complementary effect of parents’ discouragement toward child’s smoking on
intergenerational transmission of smoking behavior (OR=-1.47, p < .05). When a
child had closer relationships with their parent, parent’s smoking was less likely to
be transmitted. Also, the parents’ degree of discouragement was significantly lower
and the likelihood of child’s smoking even when regularly smokes.
Table 5.6. Moderating effect of anti-smoking socialization (attitude and
discouragement) on intergenerational transmission of child’s tobacco use (N=713)
Low solidarity High Solidarity
Child: Ex-Smoke
(vs. Never)
Current-Smoke
(vs. Never)
Child: Ex-Smoke
(vs. Never)
Current-Smoke
(vs. Never)
Interaction Terms
Est. SE OR Est. SE OR Est. SE OR Est. SE OR
Ex-Smoke x
Attitude
-.79 .21 .45 .50 .87 1.65 -.44 .57 1.20 -.40 .33 .66
Current-Smoke x
Attitude
.68 .96 1.97 -.17 .52 .84 .16 .35 .36 -1.24 .26 .08
Ex-Smoke x
Discouragement
-.47 .21 .62 .16 .50 1.18 -.60 .45 .94 -.99 .31 .37
Current-Smoke x
Discouragement
-.31 .32 .74 .12 .34 1.13 -1.47 .17 .23* -1.75 .16 .17
Note. age, sex, and education of parent and child, affetual solidarity, parental
smoking, attitude, discouragement were included in the model
* p ≤ .05; ** p ≤ .01; *** p≤.001.
F. Discussion
The goal of this paper was to examine how the use of tobacco is transmitted
intergenerationally, and how parents' smoking practice related characteristics—
such as: tobacco use experience, anti-smoking attitudes, and prohibiting
discouragement—are associated with the likelihood of child smoking behavior.
83
Additionally, we also examined parents' anti-smoking attitudes/strict
discouragement moderated the association between parent and child smoking
practice. Finally, looked at whether family solidarity played a role in mitigating or
enhancing this interaction.
Findings based on the 713 cases of parent-child dyads from the Longitudinal
Study of Generations suggest that both modeling and socialization play a significant
role in predicting a child’s tobacco use. These researches confirm the findings of
previous studies examining parental influence on child’s smoking practices (Chassin
et al., 1984; Darling & Cumsille, 2003; Kandel & Wu, 1995; Sargent & Dalton,
2001Farkas et al., 2000; Proescholdbell et al., 2000).
Our results also showed that parent’s attitude mediates the relationship
between parents’ regular smoking behaviors and child’s smoking. This is
noteworthy given that it suggests that while parents’ tobacco usage is important to
child’s smoking practices, much of this association has to do with parental attitudes
about smoking. Moreover, we found that parents who still regularly smoke tended
to have more permissive attitude toward smoking which accounted for a significant
amount of the increase in smoking prevalence among their children.
This study also found a conditioned interactional effect between parents’
smoking practice and discouragement on tobacco use of their offspring. While
initially, we did not find a significant interaction between modeling and anti-
socialization on child’s smoking behavior—a result which is consistent with
previous findings (Bailey, Ennett, & Ringwald, 1993), this interaction became
84
significant when intergenerational emotional closeness was also considered. Results
showed that among children who had close relationships with their parent, parents’
increased discouragement was associated with a significantly lower likelihood of
the child smoking, even when the parent was a regular smoker. This suggests that
only families who reported having close emotional relationships experienced the
beneficial effect of parents’ discouragement on intergenerational transmission of
smoking behavior. It is noteworthy that intergenerational emotional closeness was
essential to provoke the facilitating effect of anti-smoking socialization on
intergenerational smoking transmission.
F.1. Contributions
Previous work has documented the influence of parental behavior on
children’s smoking behaviors, and has suggested various familial mechanisms to
explain this, including: 1) direct modeling of children from the exposure to parents’
smoking, 2) indirect socialization through parent’s attitude towards smoking, and 3)
degree of discouragement toward children’s smoking. However, limited attention
has been given to distinguishing the relative importance and possible correlation
among these various factor.
In response to this, our paper aims to contribute to the better understanding
of the contribution of both direct modeling and anti-socialization to smoking
transmission between generations. Using available dyadic data, we were able to
measure more directly the influence of parents’ smoking. In addition, this paper
85
used information of parent-child emotional closeness to explore whether
relationship influenced the association between the modeling and anti-socialization
interaction and intergenerational transmission of smoking behavior. The findings of
the current study about socialization mechanisms and the activating role of
emotional closeness have the potential to facilitate the development of more
focused studies on smoking-related intervention, operating within family context.
F.2. Limitations
Interpretation of the study findings should be made in the context of the
following considerations. Although the current study was able to make predictions
regarding smoking behavior of parents and their children, the information on when
each person started and quit smoking were asked retrospectively. Longitudinal data
with more detailed information needs to be analyzed to further explore the possible
change of smoking behavior trajectories. Also, this paper used the parents’ current
attitudes towards smoking, which may not represent their attitude when their
adolescent children were experiencing the critical period of smoking transmission.
Finally, this paper did not include the factor regarding the amount of tobacco use of
the participants which may benefit the further research. Nevertheless, this paper is
strengthened by exploring contributions of various mechanisms using dyadic data
and by investigating the role of intergenerational emotional closeness in provoking
the facilitating effect of anti-smoking socialization on intergenerational smoking
transmission.
86
APPENDIX 2
Table 5.7. Parent’s smoking and child’s smoking
Child’s Smoking
N (%)
Parent’s Smoking Never Ex Current Total
Never
320 (44.9) 46 (6.5) 47 (6.6) 413 (57.9)
Ex
107 (15.0) 41 (5.8) 40 (5.6) 188 (26.4)
Current
60 (8.4) 24 (3.4) 28 (3.9) 112 (15.7)
Total
487 (68.3) 111 (15.6) 115 (16.1) 713 (100.0)
Table 5.8. Parent’s smoking and discouragement
Parent discourage smoking
N (%)
Parent’s smoking Not at all Not very
strongly
Somewhat
strongly
Very
strongly
Total
Never
13 (1.9) 28 (4.0) 63 (9.1) 298 (42.9) 402 (57.9)
Ex
21 (3.0) 26 (3.8) 37 (5.3) 98 (14.1) 182 (26.2)
Current
17 (2.5) 23 (3.3) 23 (3.3) 47 (6.8) 110 (15.9)
Total
51 (7.4) 77 (11.1) 123 (17.7) 443 (63.8) 694 (100.0)
Table 5.9. Parent’s smoking and attitude_Q1
OK for people to smoke once in a while
N (%)
Parent’s Smoking Strongly
Agree
Agree Disagree Strongly
Disagree
Total
Never
6 (0.9) 53 (7.7) 123 (17.9) 216 (31.4) 398 (57.9)
Ex
2 (0.3) 44 (6.4) 61 (8.9) 77 (11.2) 184 (26.7)
Current
15 (2.2) 64 (9.3) 18 (2.6) 9 (1.3) 106 (15.4)
Total
23 (3.3) 161 (23.4) 202 (29.4) 302 (43.9) 688 (100.0)
87
Table 5.10. Parent’s smoking and attitude_Q2
Smoking should be banned in public places
N (%)
Parent’s Smoking strongly
disagree
disagree agree strongly
agree
Total
never
11 (1.6) 24 (3.4) 157 (22.5) 209 (30.0) 401 (57.5)
ex
4 (0.6) 43 (6.2) 63 (9.0) 75 (10.7) 185 (26.5)
current
39 (5.6) 36 (5.2) 27 (3.9) 10 (1.4) 112 (16.1)
Total
54 (7.7) 103 (14.8) 247 (35.4) 294 (42.1) 698 (100.0)
88
Chapter Six:
Discussion
A. Summary
While the intergenerational transmission of family culture, values, and behaviors
has long been acknowledged as an important research topic, the existing literature
suffers from some notable limitations. Few studies have examined the persistence
and varying patterns of intergenerational transmission over the human life course.
In addition, there is insufficient research using longitudinal data to empirically
explore the factors which facilitate the transmission process and influence the
dynamics among the various intergenerational mechanisms of transmission. Each of
the three papers in this dissertation contribute both to the existing knowledge about
the stability of intergenerational transmission of values and behaviors and to the
understanding of influential determinants in this process over the life course.
The first paper explored the persistence and the lagged parental influence on
children’s values. Even beyond that, it examined the effects of family solidarity in
order to understand how different types of early parent-child relationships manifest
differently over the life course. The second paper utilized life course perspectives to
understand the changing trajectories of parent-child value transmission across time
and whether heterogeneous patterns among individuals could be explained by the
quality of family relationships and the different and type of relationship with
89
parents. Lastly, the third paper used parent modeling and parent socialization
frameworks to investigate the intergenerational transmission of smoking. By doing
so, the relative importance of parent modeling and parent socialization can now be
understood, with further evidence of the power of intergenerational relationships to
determine behavior.
Overall, this dissertation showed the longitudinal patterns of intergenerational
transmission and demonstrated that parental influence in early childhood tends to
persist through midlife for the majority of individuals. At the same time, the study
also provided evidence of existing patterns of heterogeneous trajectories among
individuals. Emphasizing the important role of family, family-related factors,
especially emotional intimacy in the parent-child relationship, were shown to be
powerful elements in determining values and behaviors in adulthood. Family
factors also magnified the effect of modeling and socialization in the process of
intergenerational transmission. In recent years, there has been increasing concern
about the weakening role of family as a primary socialization agent. This
dissertation has made a meaningful contribution by providing empirical evidence
for the stability of parent-child similarity and the continued effect of family
solidarity. The conceptual frameworks which were adapted to the current
dissertation – bioecological theory, family socialization theory, attachment theory,
and life course perspectives – also provided a useful framework to explain the
findings in this dissertation.
90
B. Limitations and Strength
This dissertation has several limitations in terms of sample, missing data, and
analytical strategy. The Longitudinal Study of Generations (LSOG) is not nationally
representative; the baseline sample was extracted from a regional population in
southern California area in 1971. Hence, the sample consists mostly of middle class
Caucasians and does not adequately represent the ethnic or racial diversity
currently present in the United States population. Future research on this subject
needs to better reflect this diversity.
Furthermore, sub-sample sizes, such as the number of step-families, were
relatively small and some available variables were limited to one time point or
retrospective collection (e.g., with smoking behavior), so it is hard to disentangle
intertwined issues such age, cohort, and period effects and changes across time. On
the other hand, this limitation can be understood in light of the unique advantages it
provides. For instance, the longitudinal and multigenerational design allows dyadic
analysis and a more in-depth examination of family relationships using a variety of
variables.
This dissertation also investigated only one cohort of dyads. It is possible
that results may vary depending on cohort membership. For example, generation 3
(G3) of LSOG consists of baby-boomers, which suggests that caution should be taken
when applying the findings of this study to other cohorts. Future research may
address this limitation by employing a cohort-sequential design.
91
Lastly, it has been known that intergenerational transmission is stronger
between same sex relationships (e.g. father-son, mother-daughter) than between
opposition sex relationships (e.g. father-daughter, mother-son). However, this
dissertation was not able to explore this effect due to low power from the sample
size.
C. Implications and Future Research
This dissertation not only contributes to the current body of knowledge, but
also suggests directions for future research by identifying the complexity in the
relationship between parents’ values and behaviors and their offspring’s, as well as
the diversity of familial, contextual, and temporal factors which influence
transmission. Researchers in the field of social science and family studies who will
explore longitudinal patterns of intergenerational transmission need to consider
these factors. Specifically, this dissertation makes the following contributions:
First, this dissertation demonstrates that parents exert great influence over
their offspring’s values and behaviors and this influence remains over the life course.
As shown by the first paper, children’s religious beliefs and gender role ideology
closely resemble their parents’ beliefs and ideology in early adulthood. The
assertion that religious beliefs are significantly transmitted in early life and remain
over the life course is further supported by the findings in the second paper that
show that the majority of children remain fairly consistent in the extent to which
92
they express similar beliefs to those of their parents. Understanding that the
intergenerational similarity is consistent over the life course has theoretical
implications.
Specifically, these findings expand the applicability of family socialization
theory beyond early childhood. Although many studies have found that parents are
important socialization agents, heavily influencing their offspring’s values and
behavior during childhood, very little empirical research has explored the influence
of parents as conduits for values and behaviors beyond childhood. This is partially
due to limited availability of family level longitudinal data. This dissertation
provided empirical evidence of the longitudinal impact of parents on their
offspring’s values and behaviors. Further, the finding of lagged effects of parents’
values on their children serves as empirical evidence of the usefulness of life course
perspective on the research of intergenerational transmission. Little has been
known about whether parents’ influence is conditional based upon children’s life
stage. These findings show the possible sleeper effect of parents’ influence. This
dormant influence from when children were young activated when children reached
midlife. There is a dearth of research in intergenerational transmission of values
and behaviors exploring children’s later stages of life; consequently, little has been
known about the persistence of intergenerational transmission over time. This
finding supports to family socialization theory and attachment theory even in the
longitudinal framework, further suggesting that parental influence is important to
children not only in early life stages, but also throughout the life course.
93
Next, this dissertation also provides evidence that although parents greatly
influence children, the extent of the intergenerational transmission varies
depending upon the type of values and behaviors being transmitted. Indeed, the
longitudinal pattern of intergenerational similarity varies among individuals, as well.
The first paper demonstrated that both the strength of early transmission and the
extent of persistence were significantly weaker for gender role ideology than they
were for religious beliefs. In addition, the second paper explored heterogeneous
patterns of intergenerational similarity over the life course and found three
different trajectories. These results demonstrate that testing whether values or
behaviors are significantly transmitted in general is important, but in order to fully
understand the complexity of intergenerational transmission, so is exploring 1)
variations across value or behavior domains and 2) variations between individuals.
From a theoretical standpoint, these results support the life course perspective and
its emphasis on change over the life course and human agency. Thus, future
research will benefit from considering the possible within and between group
variations, as well as from exploring the factors involved in producing
heterogeneous trajectories of intergenerational transmission.
Lastly, this dissertation provides evidence that solidarity between parent and
child plays a key role in facilitating intergenerational transmission and in
moderating the effect between parental modeling and parental socialization.
Intergenerational affectual solidarity (i.e., emotional closeness) makes a unique
contribution to the transmission process. As demonstrated by the first paper,
94
transmission of gender role ideology was not found until the offspring reached
midlife, and it was only for those who had very close relationships with their parent
in the early stages of life. This was further supported by the findings from the third
paper which found that affectual solidarity was a significant activator of an
interaction effect. Studies in the 1980s and 1990s examined the moderating effect of
implicit and explicit socialization mechanisms on intergenerational smoking
transmission, but did not find a significant effect. Interestingly, and consistent with
this literature, no moderating effect between parents’ smoking and parents’ anti-
smoking socialization on children’s smoking initially emerged in the third paper’s
analysis overall, either. However, once this sample was divided into high solidarity
groups and low solidarity groups, an effect on children’s smoking emerged for those
who had very close relationships with their parent.
Intergenerational solidarity is widely recognized as crucial in the
transmission process. This dissertation took this concept a step further by
demonstrating that intergenerational solidarity in early life can have a dormant
effect, emerging only in later life stages. In addition, although intergenerational
solidarity itself does not have a direct influence on values and behaviors, it can have
a powerful effect in determining whether parents successfully transmit their values
and behaviors to their children. Knowledge of this enduring importance of high-
quality parent-child relationships in early childhood may inform policy makers’
priorities when formulating new social policies. Future research should attempt to
include various family-related factors such as type of relationship with parents (step
95
or bio) and temporal context, and should consider the effects of time-varying
characteristics of intergenerational solidarity on the transmission process.
Intergenerational transmission should be addressed as a dynamic process
which changes over the life course, with its power dependent upon the type of
values and behaviors being transmitted. It is also important to note both that inter-
individual variations do exist in the pattern of intergenerational transmission over
the life course and that intergenerational solidarity plays a key role in facilitating
the process and in activating the interaction among mechanisms. Future research in
this area would benefit from using globally and nationally representative samples in
order to explore the similarities and differences across different countries and
ethnic groups.
96
References
Acock, A. , & Bengtson, V. (1978). On the relative influence of mothers and fathers: A
covariance analysis of political and religious socialization. Journal of Marriage
and the Family, 40, 519.
Acock, A. (1984). Parents and their children: The study of inter-generation
influence. Sociology and Social Research, 68, 151.
Ainsworth, M. (1973). Development of infant–mother attachment. In Caldwell, B. &
Ricciuti, H. (Eds.), Review of Child Development Research, 3. Chicago: University
of Chicago Press.
Alwin, D. (1984). Trends in parental socialization values: Detroit, 1958-1983.
American Journal of Sociology, 90, 359.
Alwin, D. (1990). Historical changes in parental orientations to children. In Mandell,
N. & Cahill, S. (Eds.), Sociological Studies of Child Development. 3. Greenwich,
CT: JAI Press.
Axinn, W. & Thornton, A. (1993). Mothers, children and cohabitation: The
intergenerational effects of attitudes and behaviors. American Sociological
Review, 58, 233.
Avenevoli , S. & Merikangas, K. (2003). Familial influences on adolescent smoking.
Addiction, 98, 1.
97
Bailey, L., Ennett, T., & Ringwalt, L. (1993). Potential mediators, moderators, or
independent effects in the relationship between parents’ former and current
cigarette use and their children’s cigarette use. Addictive Behaviors, 18, 601.
Baltes, P. (1979). Life span developmental psychology: some converging
observations on history and theory. In Baltes, P. & Brim, O. (Eds). Life Span
Development and Behavior, 2. New York: Academic Press.
Bengtson, V. & Schrader, S. (1982). Parent-child relations. In by Mangen, D. &
Peterson, W. (Eds.) Research Instrument in Social Gerontology: Social roles and
social participation, 2. Minneapolis: University of Minnesota Press.
Bengtson, V. & Kuypers, J. (1971). Generational difference and the developmental
stake. Aging and Human Development, 2, 249.
Bengtson, V. & Roberts, L. (1991). Intergenerational solidarity in aging families: An
example of formal theory construction. Journal of Marriage and Family, 53, 856.
Bengtson, V. (2001) Beyond the nuclear family: The increasing importance of
multigenerational bonds. Journal of Marriage and the Family, 63, 1.
Bengtson, V., Biblarz, J., & Roberts, L. (2002). How families still matter: A
Longitudinal Study of Youth in Two Generations. New York: Cambridge,
University Press.
Bengtson, V., Copen, C., Putney, N., & Silverstein, M. (2009). A longitudinal study of
the intergenerational transmission of religion. International Sociology, 24, 325.
Bengtson, V., Putney, N., & Harris, S. (2013). Families and Faith: How religion is
passed down across generations. New York: Oxford University Press.
98
Bock, G. & Goode, J. (2006). Understanding Nicotine and Tobacco Addiction.
Hoboken, NJ: John Wiley & Sons.
Booth, A., Johnson, D., Branaman, A., & Sica, A. (1995). Beliefs and behavior: Does
religion matter in today’s marriage? Journal of Marriage and Family, 57, 661.
Bowlby, J. (1969). Attachment and loss. New York: Basic Books.
Bronfenbrenner, U. (1994). Ecological models of human development. In
International Encyclopedia Of education, 3. Oxford: Elsevier.
Boardman, J., Blalock, C., & Pampel, F. (2012). Trends in the genetic influences on
smoking. Public, Environmental & Occupational Health, 51, 108.
Bumpass, L. & Lu, H. (2000). Trends in cohabitation and implications for children’s
family contexts in the United States. Population Studies: A Journal of
Demography, 51, 29.
Bucx, F., Raaijmakers, Q., & Van Wel, F. (2010). Life course stage in young adulthood
and intergenerational congruence in family attitudes. Journal of Marriage and
Family, 72, 117.
Chassin, L., Presson, C., Sherman, J., Corty, E., & Olshavsky, W. (1984). Predicting the
onset of cigarette smoking in adolescents: A longitudinal study. Journal of
Applied Social Psychology, 14, 224.
Chassin, L., Presson, C., Rose, J., & Sherman, S. (1998). Maternal socialization of
adolescent smoking: Intergenerational transmission of smoking-related beliefs.
Psychology of Addictive Behaviors, 12, 206.
99
Chassin, L., Presson,C., Rose, J., & Sherman, S. (2002). Parental smoking cessation
and adolescent smoking. Journal of pediatric psychology, 27, 485.
Chassin, L., Presson, C., Sherman, S., Seo, D., & Macy, J. (2010). Implicit and explicit
attitudes predict smoking cessation: Moderating effects of experienced failure
to control smoking and plans to quit. Psychology of Addictive Behaviors, 24, 670.
Carlson, D. & Knoester, C. (2011). Family structure and the intergenerational
transmission of gender role ideology. Journal of Family Issues, 32, 709.
Chodorow, N. (1978). The Reproduction of Mothering: Psychoanalysis and the
Sociology of Gender. Berkeley: University of California Press.
Cooper, C. R., Grotecant, H. D. & Gondon, S. M. (1983). Individuality and
connectedness in the family as a context for adolescent identity formation and
role-taking sill. In H. Grotevant & C. Cooper (Eds.), Adolescent Development in
the Family 43. San Francisco: Jossey-Bass.
Copen, C., & Silverstein, M. (2008). The transmission of religious beliefs across
generations: Do grandparents matter? Journal of Comparative Family Studies,
39, 59.
Cotter, D., Hermsen, J., & Vanneman, R. (2010). The end of the gender revolution?
Gender role attitudes from 1977 to 2008. The American Journal of Sociology. 3,
116.
Cunningham, M. (2001). The influence of parental attitudes and behaviors on
children’s attitudes toward gender and household labor in early adulthood,
Journal of Marriage and Family, 63, 111.
100
Darling , N. & Cumsille, P. (2003). Theory, measurement, and methods in the study
of family influences on adolescent smoking, Addition, 98, 21.
De Houwer, J., Thomas, S., & Baeyens, F. (2001). Associative learning of likes and
dislikes: A review of 25 years of research on human evaluative conditioning.
Psychological Bulletin, 127, 853.
Dishion, T. & McMahon, R. (1998). Parental monitoring and the prevention of child
and adolescent problem behavior: A conceptual and empirical formulation.
Clinical Child and Family Psychology Review, 1. 61.
Doherty, W., & Allen, W. (1994). Family functioning and parental smoking as
predictors of adolescent cigarette use: A six-year prospective study, Journal of
Family Psychology, 8, 347.
Dudley, L. (1978). Alienation from religion in adolescents from fundamentalist
religious homes. Journal for the Scientific Study of Religion, 17, 389.
Edgell, P. (2006). Religion and family in a changing society. Princeton: Princeton
University Press.
Elder, G. (1974). Children of the great depression: Social Change in Life Experience.
Chicago: University of Chicago Press.
Elder, H. (1978). Family history and the life course. In by Hareven, T. (Eds.)
Transitions: The family and the life course in historical perspective. New York:
Academic Press.
Enders, C. (2001). A primer on maximum likelihood algorithms available for use
with missing data. Structural Equation Modeling, 8, 128.
101
Ennett, S., Bauman, K., Foshee, V., Pemberton, M., & Hicks, K. (2004). Parent-child
communication about adolescent tobacco and alcohol use: What do parents say
and does it affect youth behavior? Journal of Marriage and Family, 63, 2004.
Erikson, E. (1950). Childhood and Society. New York: W. Norton.
Erickson, A. (1992). Adolescent religious development and commitment: A
structural equation model of the role of the family, peer group and educational
influences. Journal for the Scientific Study of Religion, 31, 131.
Fan, P. & Marini, M. (2000). Influences on gender-role attitudes during the transition
to adulthood. Social Science Research, 29, 258.
Farkas, A., Gilpin, E., White, M., & Pierce, J. (2000). Associations between household
and workplace smoking restrictions and adolescent smoking. Journal of the
American Medical Association, 284, 717.
Flannelly, K., Galek, K., Kytle, .J, & Silton, N. (2010). Religion in America 1972-2006:
Religious affiliation, attendance, and strength of faith. Psychological Report, 106,
875.
Francis, J. & Brown, L. (1991). The influence of home, church, and school prayer
among 16-year old adolescents in England. Review of Religious Research, 33, 112.
Francis, J. & Gibson, M. (1993). Parental influence and adolescent religiosity: A study
of church attendance and attitude toward Christianity among adolescents 11 to
12 and 15 to 16 years old. International Journal for the Psychology of Religion, 3,
241.
102
Ganong, H., & Coleman, M. (2004). Stepfamily Relationships: Development, dynamics,
and interventions. New York, NY: Kluwer Academic Plenum.
Glass, J., Bengtson, V. L., & Dunham, C. C. (1986). Attitude similarity in three
generation families: Socialization, status inheritance, or reciprocal influence?
American Sociological Review, 51, 685.
Graham, J.W. (2003). Adding missing-data-relevant variables to FIML-based
structural equation models. Structural Equation Modeling, 10, 80.
Grusec, J., Goodnow, J., & Kuczynski, L. (2003). New directions in analyses of
parenting contributions to children’s acquisition of values. Child Development,
71, 205.
Hagestad, G. (2003). Interdependent lives and relationships in changing times: A
life-course view of families and aging. In Settersten, R. (Eds.), Invitation to the
Life Course: Toward new understandings of later life. Amityville, NY: Baywood .
Hareven, T. (1994). Aging and Generational Relations: A historical and life course
perspective. Annual Review of Sociology, 20, 437.
Henriksen, L., & Jackson, C. (1998). Anti-smoking socialization: Relationship to
parent and child smoking status. Health Communication, 10, 87.
Hess, B. & Warning, J. (1978). Parent and child in later life: rethinking and
relationship. In by Lerner, R. & Spanier, Graham. (Eds). Child Influence on
Marital and Family Interaction. New York: Academic Press.
Hill, K., Hawkins, D., Catalano, R., & Abbott, R., & Guo, J. (2005). Family influences on
the risk of daily smoking initiation. Journal of adolescent health, 37, 202.
103
Hitlin, S. (2006). Parental influences on children's values and aspirations: Bridging
two theories of social class and socialization. Sociological Perspectives, 49, 25.
Hoge, R., Petrillo, H., & Smith, I. (1982). Transmission of religious and social values
from parents to teenage children. Journal of Marriage and Family, 44, 569.
Hoffman, J. A. (1984). Psychological separation of late adolescents from their
parents. Journal of Counseling Psychology, 31, 170.
Jackson, C. & Dickinson, D. (2003). Can parents who smoke socialize their children
against smoking? Results from the smoke-free kids intervention trial. Tobacco
control, 12, 52.
Jackson, C. & Henriksen, L. (1997). Do as I say: Parent smoking, antismoking
socialization, and smoking onset among children. Addictive Behaviors, 22, 107.
Jung, T. & Wickrama, K. (2008). An introduction to latent class growth analysis and
growth mixture modeling. Social and Personality Psychology Compass, 2, 302.
Kandel, D., & Wu, P. (1995). The contributions of mothers and fathers to the
intergenerational transmission of cigarette smoking in adolescence. Journal of
Research on Adolescence, 5, 225.
Kodl, M. & Mermelstein, R. (2004). Beyond modeling: Parenting practices, parental
smoking history, and adolescent cigarette smoking. Addictive Behaviors, 29, 17.
Kroger, J., & Haslett, S. J. (1988). Separation–individuation and ego identity status in
late adolescence: A two-year longitudinal study. Journal of Youth and
Adolescence, 17, 59.
104
Lehrer, E. (1999). Religion as a determinant of educational attainment: An economic
perspective. Social Science Research, 28, 358.
Levin, J. (1996). How religion influences morbidity and health: Reflections on
natural history, salutogenesis and host resistance. Social Science & Medicine, 43,
849.
Li, D. (2003). The genetics of smoking related behavior: A brief review. The
American Journal of the Medical Sciences.
Maruyama,G. M. (1998), Basics of Structural Equation Modeling, Thousand Oaks, CA:
Sage.
Merz, E., Schuengel, C., & Schulze, H. (2007). Intergenerational solidarity: An
attachment perspective. Journal of Aging Studies, 21, 175.
Miller, B., & Glass, J. (1989). Parent–child attitude similarity across the life course.
Journal of Marriage and Family, 51, 991.
Milevsky, I., Szuchman, L., & Milevsky, A. (2008). Transmission of religious beliefs in
college students. Religion and Culture, 11, 423.
Miller, R. B., & Glass, J. (1989). Parent-child attitude similarity across the life course.
Journal of Marriage and Family, 51, 991.
Miller, W. & Thoresen, C. (2003). Spirituality, religion, and health: An emerging
research field. American Psychologist, 58, 24.
Min, J., Silverstein, M., & Lendon, J. (2012). Intergenerational transmission of values
over the family life course. Advances in Life Course Research, 17, 112.
105
Moen, P., Erickson, M., & Dempster-McClain, D. (1997). Their mother's daughters?
the intergenerational transmission of gender attitudes in a world of changing
roles. Journal of Marriage and the Family, 59, 281.
Myers, S. M. (1996). An interactive model of religious beliefs inheritance: The
importance of family context. American Sociological Review, 61, 858.
Myers, S.M. (2004) ‘Religion and intergenerational assistance: Distinct differences
by adult children’s gender and parent’s marital status’, Sociological Quarterly,
45, 67.
Newman, I. M., & Ward, J. M. (1989). The influence of parental attitude and behavior
on early adolescent cigarette smoking, Journal of School Health, 59, 150.
Nolte, A. E., Smith, B. J., & O’Rourke, T. (1983). The relative importance of parental
attitudes and behavior upon youth smoking behavior. Journal of School Health,
53, 264.
Nylund, K., Asparouhov, T., & Muthen, B. (2007). Deciding on the number of classes
in latent class analysis and growth mixture modeling: a Monte Carlo simulation
study. Structural Equation Modeling, 14, 535.
Parsons, T., & Baltes, R. (1955). Family, socialization, interaction process. Glencoe, Ill:
Free Press.
Pearce, L. & Axinn, W. (1998). The impact of family religious life on the quality of
mother-child relations. American Sociological Review, 63, 810.
Pearlin, I. & Skaff, M. (1996). Stress and the life course: A paradigmatic alliance. The
Gerontologist, 36, 229.
106
Petts, R. (2009). Trajectories of religious participation from adolescence to young
adulthood. Journal for the Scientific Study of Religion, 48, 552.
Proescholdbell, J., Chassin, L., & MacKinnon, P. (2000). Home smoking restrictions
and adolescent smoking. Nicotine & Tobacco Research, 2, 159.
Rimal, R. (2003). Intergenerational transmission of health: The role of intrapersonal,
interpersonal, and communicative factors. Health education & Behavior, 30, 10.
Roberts, R., & Bengtson, V. (1996). Affective ties to parents in early adulthood and
self-esteem across 20 years. Social Psychology Quarterly, 59, 96.
Rodgers, R. (1973). Family Interaction and Transaction. Englewood Cliffs, NJ:
Prentice-Hall.
Roest, A., Dubas, J., & Gerris, J. (2010). Value transmissions between parents and
children: Gender and developmental phase as transmission belts. Journal of
Adolescence, 33, 21.
Rohan, M. J., & Zanna, M. P. (1996). Value transmission in families. In C. Seligman, J.
M. Olson, & M. P. Zanna (Eds), The Psychology of Values: The Ontario
symposium, 8, 253. Hilsdale, NJ: Erlbaum.
Ross, A., Schoon, I., Martin, P. & Sacker, A. (2009). Family and nonfamily role
configurations in two British cohorts. Journal of Marriage and the Family, 71, 1.
Rossi, A. (1980). Aging and parenthood in the middle years. In Baltes, P. & Brim, O.
(Eds). Life Span Development and Behavior. New York: Academic Press.
Rossi, S. & Rossi, H. (1990). Human Bonding: Parent-child relations across the life
course. New York: Aldine de Gruyter.
107
Sargent, J. & Dalton, M. (2001). Does parental disapproval of smoking prevent
adolescents from becoming established smokers? Pediatrics, 108, 1256.
Schoen, R. & Canuda-Romo, V. (2005). Timing effects on first marriage: Twentieth-
century experience in England and Wales and the USA. Population Studies, 59,
135.
Schultenberg, E., Maggs, L., & O’Malley, M. (2003). How and why the understanding
of developmental continuity and discontinuity is important: The sample case of
long-term consequences of adolescent substance use. In Mortimer, T., &
Shanahan, J. (Eds.) Handbook of the Life Course. New York: Plenum Press.
Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical
advances and empirical tests in 20 countries. Advances in Experimental Social
Psychology, 25, 1.
Silverstein, M. & Bengtson, V. (1997). Intergenerational solidarity and the structure
of adult child-parent relationships in American families. American Journal of
Sociology, 103, 429.
Simons-Morton, B. (2003). The protective effect of parental expectations against
early adolescent smoking initiation. Health Education Research, 19, 561.
Smith, D. & Self, G. (1980). The congruence between mother’s and daughter’s sex
role attitudes: A research note. Journal of Marriage and Family, 42, 105.
Starrels, M. E. (1992). Attitude similarity between mothers and children regarding
maternal employment. Journal of Marriage and the Family, 54, 91.
108
Steelman, C. & Powell, B. (1991). Sponsoring the next generational: Parental
willingness to pay for higher education. American Journal of Sociology, 96, 1505.
Tarvis, T. W. (2000). Quality of mother-child interaction and the intergenerational
transmission of sexual values: A panel study. The Journal of Genetic Psychology,
161, 169.
Thomson, E., McLanahan, S., & Curtin, B. (1992). Family structure, gender and
parental socialization. Journal of Marriage and Family, 54, 368.
Thornton, A., Orbuch, T., & Axinn, W. (1995). Parent-child relationships during the
transition to adulthood. Journal of Family Issues, 16, 538.
Tilson, E., McBride, C., Lipkus, I., & Catalano, R. (2004). Testing the interaction
between parent-child relationship factors and parent smoking to predict youth
smoking. Journal of Adolescent Health, 35, 182.
Tinsley, J. (1992). Multiple influences on the acquisition and socialization of
children’s health attitudes and behavior: An integrative review. Child
Development, 6, 1043.
Uecker, J., Regnerus, M., & Vaaler, M. (2007). Losing my religion: the social sources
of religious decline in early adulthood. Social Forces, 85, 1667.
Vedder, P., Berry, J., Sabatier, C., & Sam, D. (2009). The intergenerational
transmission of values in national and immigrant families: The role of zeitgeist.
Journal of Youth and Adolescence, 38, 642.
109
Vollebergh, W. A. M., Iedema, J., & Raaijmakers, Q. A. W. (2001). Intergenerational
transmission and the formation of cultural orientations in adolescence and
young adulthood. Journal of Marriage and Family, 63, 1185.
Wakefield, M., Chaloupka, F., Kaufman, N., Orleans, C., Barker, D., & Ruel, E. (2000).
Effect of restrictions on smoking at home, at school, and in public places on
teenage smoking: Cross sectional study. British Medical Journal, 321, 333.
Waite, L. & Lehrer, E. (2003). The benefits from marriage and religion in the United
States: A comparative analysis. Population & Development Review, 29, 255.
Weinstein, M., & Thornton, A. (1989). Mother-child relations and adolescent sexual
attitudes and behavior. Demography, 26, 563.
Wen, C., Tsai, S., Cheng, T., Hsu, C., Chen, T., & Lin, H. (2005). Role of parents and
peers in influencing the smoking status of high school students in Taiwan.
Tobacco Control, 14, 10.
West, P., Sweeting, H., & Ecob, R. (2002). Family and friends’ influences on the
uptake of regular smoking from mid-adolescence to early adulthood. Addiction,
94, 1297.
Williams, K. & Umberson, D. (2004). Marital status, marital transitions, and health: A
gendered life course perspective. Journal of Health and Social Behavior, 45, 81.
Willits, K. & Crider, D. (1989). Church attendance and traditional religious beliefs in
adolescence and young adulthood: a panel study. Review of Religious Research,
31, 68.
110
Wills, T., Resko, J., Ainette, M., & Mendoza, D. (2004). Role of parent support and
peer support in adolescent substance use: A test of mediated effects. Psychology
of Addictive Behaviors, 18, 122.
Younnis, J., & Smollar, J. (1985). Adolescents’ Relations with Mothers, Fathers, and
Friends. Chicago: University of Chicago Press.
Abstract (if available)
Abstract
The intergenerational transmission of family culture has been an important topic of research in gerontology and the study of aging families. Scholars from both gerontology and family studies have provided evidence of the impact of parents on their children’s values and behaviors. Whether intergenerational transmission occurs through socialization, didactic training, or modeling, it is clear that parental influences are perpetuated through multiple generations. However, previous studies of intergenerational transmission have primarily examined the issue using cross‐sectional data or retrospective data, making it impossible to examine whether transmitted values and behaviors endure or emerge in children as they grow up. Additionally, only a few research investigations have directly compared mechanisms of intergenerational transmission to explore their relative importance and potential interaction effects. ❧ The overall goal of this dissertation is to describe the timing, trajectory, and dynamics of the intergenerational transmission of beliefs and behaviors over a 34 year period of the family life course. Specifically, three primary questions are investigated. The first looks at the timing of transmission. Do values and behaviors get transmitted early in the family life cycle and persist in children over time, or is transmission lagged such that children show more similarity to their parents’ values and behaviors when they reach middle adulthood than they do early in life? The second question investigates whether there are heterogeneous trajectories of intergenerational transmission. Do various patterns of parent‐child similarity exist? If so, does family solidarity affect life course patterns of intergenerational belief transmission? The third question addresses parents’ influence on specific behaviors of their children. What is the association between the behaviors, values, and attitudes of parents and their children’s behavior? Specifically, what are the potential interactive effects and the relative importance of parents’ behaviors, values, and attitudes on their children’s smoking behavior in adulthood? ❧ The first question of when beliefs are transmitted was investigated using data collected over 34 years from the Longitudinal Study of Generations (LSOG). The LSOG is a four generation longitudinal family study. To explore this question, 775 parent‐child dyads from the 1971 wave and the 2000 wave of the LSOG were analyzed. Time lagged structural equation modeling and multiple group analysis showed values (religious beliefs and gender role ideology) were strongly transmitted from parents to their young adult children, and then remained stable over the life course. This pattern of intergenerational influence was most robustly observed for religious beliefs. Gender role ideologies had a more lagged effect than in other values when children reached middle adulthood and strong emotional closeness within families facilitated value transmission. ❧ The second question about the potential heterogeneous trajectories of intergenerational transmission of religious beliefs was examined by looking at parent‐child similarity over the life course using data from 8 waves of the LSOG (1971, 1985, 1988, 1991, 1994, 1997, 2000, and 2005). A growth mixture model showed three different types of trajectories—“stable similar” (80.3%), “stable dissimilar” (10.5%), and “divergent across time” (9.2%). Most dyads showed a stable similar pattern, providing evidence of the importance of early relationships with parents. However, stable dissimilar and divergent groups were also evident. In addition, family‐related life transitions (such as becoming a parent) and emotional closeness were factors that predicted the pattern of intergenerational similarity over the life course. ❧ The final question was addressed by using parent‐child dyads from the 2000 wave of the LSOG study to explore the mediating and moderating effects among modeling and socialization mechanisms on intergenerational transmission of smoking behavior. Bivariate models showed that both parents’ smoking behavior (modeling) and parents’ anti‐smoking attitude and discouragement (socialization) were significant mechanisms predicting parent‐child smoking behavior transmission. The effect of parents’ smoking on children’s smoking was mediated by parents’ anti‐smoking attitude, and a conditioned moderating effect was found. Parental discouragement of smoking had less of an impact on children’s smoking behavior than these anti‐smoking attitudes did, even when strong parent‐child emotional solidarity was present. ❧ This dissertation aims to extend the empirical development of the literature relating to the intergenerational transmission of values and behaviors. Specifically, it focuses on the importance of exploring the persistence of intergenerational transmission, investigates the dynamics of transmission mechanisms, and uses the longitudinal data to more effectively explain intergenerational transmission over the family life course. These findings have the potential to inform the development of interventions and prevention efforts that address values or behaviors that negatively influence psycho‐socio‐economic outcomes. The final discussion summarizes the above findings and discusses how they can be explained within the theoretical context.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The measurement, life course patterns, and outcomes of intergenerational ambivalence among parent-adult child dyads
PDF
Transmission and transitions: young adults' beliefs, values and life course transitions in familial context
PDF
The effects of wisdom-related personality traits on caregivers’ health: an application of the resilience model
PDF
Childlessness and psychological well-being across life course as manifested in significant life events
PDF
Children 's migration and the financial, social, and psychological well-being of older adults in rural China
PDF
Family relationships and their influence on health outcomes over time among older adults in rural China
PDF
The Cambodian American college experience: intergenerational transmission of trauma, intergenerational conflict, and multiple worlds and identities
PDF
The immediate and long term legacy of relationships with grandparents for the well-being of grandchildren
PDF
Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults
PDF
Incarceration trajectories of mothers in state and federal prisons and their relation to the mother’s mental health problems and child’s risk of incarceration
PDF
Life course implications of adverse childhood experiences: impacts on elder mistreatment, subjective cognitive decline, and caregivers' health
PDF
Intergenerational childhood maltreatment continuity: examining caregiver psychosocial mechanisms, timing and severity characteristics, and dyadic clinical intervention
PDF
Racial/ethnic variation in care preferences and care outcomes among United States hospice enrollees
PDF
Neuroinflammation and the behavioral consequences of air pollution over the life course
PDF
Improvements in women's status, decision-making and child nutrition: evidence from Bangladesh and Indonesia
PDF
Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health
PDF
Family caregiver well-being: generational differences and coping strategies
PDF
Building healthy relationships to end family violence
PDF
Statistical algorithms for examining gene and environmental influences on human aging
PDF
Intergenerational transfers & human capital investments in children in the era of aging
Asset Metadata
Creator
Min, Joohong
(author)
Core Title
Intergenerational transmission of values and behaviors over the family life course
School
Leonard Davis School of Gerontology
Degree
Doctor of Philosophy
Degree Program
Gerontology
Publication Date
05/01/2016
Defense Date
02/03/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
family life course,intergenerational transmission,OAI-PMH Harvest,parent‐child relationships
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Gruenewald, Tara (
committee chair
), Silverstein, Merril (
committee chair
), Chi, Iris (
committee member
)
Creator Email
joohong1234@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-410002
Unique identifier
UC11296208
Identifier
etd-MinJoohong-2485.pdf (filename),usctheses-c3-410002 (legacy record id)
Legacy Identifier
etd-MinJoohong-2485-0.pdf
Dmrecord
410002
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Min, Joohong
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
family life course
intergenerational transmission
parent‐child relationships