Life Span Theory

03-Chibucos.
qxd
11/16/2004
5:33 PM
Page 39
3
LIFE-SPAN
DEVELOPMENTAL THEORY
ife-span developmental theory concerns

the study of individual development, or
ontogenesis, from conception to death.
A key assumption of this theory is that development does not cease when adulthood is reached
(Baltes, Lindenberger, & Staudinger, 1998,
p. 1029)1. Life-span researchers and theorists
assume that each major period of life has its own
developmental challenges and accomplishments,
and that adaptive processes are at work within
all periods of the life span. This theoretical
approach is clearly focused on individual development rather than on family development. It is
concerned with comparing an individuals development with that of others and with the individuals own status at various points in time. We
include it in this volume on family theories for
several reasons.
First, since the purview of the theory is
womb to tomb, it of necessity touches upon
all of the family-related issues that characterize
family life course theory, such as the birth of
a child, the development of Alzheimers disease,
the macro-level political and economic slings
and arrows that impact both individuals and
familiesthe latter impacted either directly or
indirectly through one or more of its members.
Of course, as noted, regardless of its study of

development across the human life span, the
focus is still on individuals rather than groups,
and this makes the theory decidedly distinct from
the family life course perspective that focuses on
the family as the unit of analysis.
Second, we believe that this approach has
much to teach others within subdisciplines of
the family and human development sciences. For
example, this theoretical approach is exceptionally rigorous in the way it links theory with
methodology across the life span (youll get a
taste of this from the two readings we have
selected for this chapter) as well as in the breadth
of substantive issues that are studied ontogenetically. In addition, it has progressed from
a largely descriptive focus on developmental
stages (predominant 4050 years ago) to a more
recent focus (beginning in the 1970s) on trying
to determine the mechanisms that determine
developmental change or consistency. (Family
life course theory is undergoing a similar transition as it has evolved from an emphasis on the
family life cycle to a more comprehensive commitment to examining the causes of family
development and change [cf. Chapter 3; Elder,
1996; White & Klein, 2002].) Overall, for the
39
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 40
40 READINGS IN FAMILY THEORY
reasons mentioned as well as because it passes

with flying colors most tests (White & Klein,
2002, p. 232, Table 9.1) of how to evaluate a
theorys usefulness, life-span developmental
theory can serve as an important model for new
students as well as for scholars.
What does it mean to say that the focus of
life-span developmental theory and research is
the individual? Given the identification of a substantive focus (e.g., personality development or
cognitive abilities), this approach aims to study
one or more of the following things:
1. Normative developmental change (For
example, what is the typical course of personality development, or of cognitive capabilities?
What are the mechanisms for this typical course
of development or developmental change?)
2. Interindividual differences in developmental change (For example, are there differences
between people over time, within age periods or
across age periods, in their outgoingness or in
their visual acuity? What are the mechanisms that
produce such differences between people?)
3. Intraindividual change and consistency in
development (For example, what is the course of
development for self-esteem or for the tendency
to take personal risks? Are these things consistent
in individuals over time or are they characterized
by plasticity? [Developmental plasticity is a
general term used in the study of ontogeny to
indicate how malleable or changeable something
is.] What are the mechanisms producing consistency or change?)
It is important to note that these three
emphases apply to individual periods of the life
span (e.g., infancy, adolescence, or old age) as
well as to the life span as a whole. In asking the
above questions, one also begins to get a flavor
of the kinds of methodological implications
of life-span developmental theory. How does
one investigate developmental plasticity or consistency? Does one compare, say, a group of
9-year-olds with a group of 23-year-olds to see if
interindividual changes in cognitive functioning
occur? Or, does one follow longitudinally the

development of cognitive functioning over a protracted period of time? Further, if one substitutes
the word family or families for people or individuals in the above questions, it becomes clear
that life-span developmental theory and family
life course theory are both fundamentally
time-related theoretical approaches. (Do note,
however, that time is really just a shorthand
descriptor for all that happens during it. Like
age, it doesnt explain anything.) This developmental perspectivewhat happens and why,
over timeis an integral part of both conceptualizations, and it is one thing that distinguishes
them from the other theoretical perspectives in
this book. Although some of the other theories
(e.g., social learning theory or social exchange
theory) may imply or make assumptions about
change over time, the concept of developmental
change is not an integral component of these
other perspectives.
THE READINGS
Schoon et al. (2002) use the life-span developmental perspective in combination with concepts
from ecological theory (Bronfenbrenner & Ceci,
1994; Chapter 10) and developmental contextualism (Ford & Lerner, 1992) to examine the relationships between early and continuing social
risk and academic achievement in childhood/
adolescence as well as between earlier risk and
adult attainments. Developmental contextualism
views human development as the dynamic interaction between a changing individual and a
changing context. This reading uses fairly complex data analysis procedures (path analysis
and multiple regression) to explicate a complex
development trajectory over a significant portion
of the life span. (You can safely ignore the statistical details since the articles text is very clear in
stating what the analyses mean and how they
can be interpreted.) Another key feature of the
Schoon et al. reading is what we mentioned
earlier in this chapter: the link between theoretical perspective and methodological choices. In
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 41
Life-Span Developmental Theory 41
this case, data were collected on the same

individuals (longitudinal study) across a significant period of the life span to answer the
specific research questions that the researchers
had formulated. Without longitudinal data, the
researchers questions would not be as adequately
addressed.
The other reading in this chapter concerns
the longitudinal impact of change or continuity
in marital status on the psychological well-being
of adults (Marks & Lambert, 1998). The researchers perspective on this issue is guided by both
the life-span developmental and the family life
course theories, which (as we have said previously) maintain that it is important to examine
the sequelae of discontinuities and continuities
throughout life, not just in early childhood. In
their study, Marks and Lambert reflect on important life-span developmental concepts including
the timing and sequence of life events, the context
in which certain developmental outcomes occur
(e.g., it is much less atypical in 2004 than it was
in 1964 to be a divorced adult), and the developmental readiness of individuals to deal with
positive and potentially negative change.
Note that, besides the overall theoretical
approach, Marks and Lamberts research shares
two features with Schoon et al. First, the use of
longitudinal data to answer developmental questions was appropriately employed. Second, the
analysis of these longitudinal data was, again,
complex. We would urge you to examine the
various data presentations, but not to be overwhelmed by them since the text is very clear in
explaining things. Also, if you suspect a trend
herethat longitudinal data require more complex statistical techniquesyou are correct. But,
the payoff is much greater when such data and
analyses are employed in developmental research.
They allow more direct answers to the research
questions, which in turn allow more direct tests
of theory (i.e., explanation and understanding of
results). The nature of the data (same people over
time) and the analyses take into account many
possibly influential variables and connections
among variables in a way that pays specific

attention to sequencing and timing. This is simply
not possible with nonlongitudinal (i.e., crosssectional) information. This is another example
of the usefulness of life-span developmental theory
to guide research.
ISSUES FOR YOUR CONSIDERATION

1. Do you think developmental contextualism might be a useful concept for family life
course theory? (Hint: The answer is yes!)
Why? How could it be integrated into the family
life course perspective?
2. Discuss an individual developmental issue
for a nonadult (any age under 18) that would
likely impact the childs family and its functioning in positive or negative ways. Then, from the
other direction, consider some discontinuity in
family life that would be likely to impact one or
more of the familys members. What theories
might help you explain these impacts?
3. How do you think of different periods in
the life span? Do you see them as preparatory for
something later, as self-contained, or as being the
result of earlier periods? If you had to make a
guess, what are the relative degrees of continuity
and discontinuity that might be normative across
the life span? What do the readings in this chapter have to say about the last question?
FURTHER READING
Bronfenbrenner and Ceci (1994), Ford and Lerner
(1992).
Note
1. This chapter relies extensively on Baltes,
Lindenberger, and Staudinger (1998), and Goulet and
Baltes (1970).
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 42
THE INFLUENCE OF CONTEXT,

TIMING, AND DURATION OF RISK
EXPERIENCES FOR THE PASSAGE FROM
CHILDHOOD TO MIDADULTHOOD
INGRID SCHOON, JOHN BYNNER, HEATHER JOSHI,
SAMANTHA PARSONS, RICHARD D. WIGGINS, AND AMANDA SACKER
Abstract
Introduction
This study investigated the long-term effects

of social disadvantage on academic achievement and on subsequent attainments in adulthood. The study drew on data collected for
more than 30,000 individuals born 12 years
apart, following their development from birth
to adulthood. The pathways that link social disadvantage to individual development across the
life course were analyzed in a developmental
contextual systems model. The results showed
that the influence of risk factors associated
with socioeconomic disadvantage depended on
the developmental stage of the individual, the
experience of long-term or continuous disadvantage, and the overall sociohistorical context. Early risk had a moderate influence on
the formation of individual competences. The
greatest risk was associated with persisting
and accumulating experiences of socioeconomic disadvantage throughout childhood and
adolescence. Material conditions improved
for the later-born cohort, yet pervasive social
inequalities existed that affected outcomes during childhood and were consequently reflected
in adult attainment.
Children raised in socioeconomically disadvantaged families are at risk for a variety of

adjustment problems, including increased risk for
poor academic achievement (Bolger, Patterson,
& Thompson, 1995; Campbell & Ramey, 1994;
Duncan, Brooks-Gunn, & Klebanov, 1994; Felner
et al., 1995; Pungello, Kupersmidt, Burchinal, &
Patterson, 1996; Ramey & Ramey, 1990; Walker,
Greenwood, Hart, & Carta, 1994) and adjustment
problems in later life, as reflected in occupational
attainment or social position (Blau & Duncan,
1967; Bynner, Joshi, & Tsatsas, 2000; Caspi,
Wright, Moffitt, & Silva, 1998; Rutter & Madge,
1976; Schoon & Parsons, 2002b; Sewell, Hailer,
& Ohlendorf, 1970). Socioeconomic background
is one of the main predictors of cognitive development, which provides the underpinnings of
academic achievement on which much success in
later life depends. The experience of socioeconomic disadvantage may severely strain adaptational abilities of children, and is thus a potential
risk factor for development. The consequences of
growing up in a disadvantaged family environment can continue into adulthood or even into
the next generation (Birch & Gussow, 1970;
42
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 43
Garmezy, 1991). Yet, most research on the

influence of socioeconomic disadvantage on
developmental outcomes has been cross-sectional
in nature, and has assessed the impact of episodic
rather than persistent economic difficulties. The
aim of this study was to examine the long-term
effects of socioeconomic disadvantage on academic achievement and consequent adult attainments. Taking a longitudinal perspective stretching
from birth to adulthood, this study examined the
impact of early and persistent social disadvantage
(which was termed social risk) on academic
achievement during childhood and adolescence
and on adult outcomes in a developmentalcontextual framework.
Fundamental to the idea of risk is the predictability of life changes from earlier circumstances. As expressed through the concept of
a risk trajectory, one risk factor reinforces
another, leading to increasingly restricted outcomes in later life (Rutter, 1990). In this study
the question of the relation between early life
experiences and consequent adjustment patterns
was recast in terms of a testable model of continuities in social disadvantage and individual
adjustment and their interactions over time. By
analyzing data from two cohorts of children
born in 1958 and 1970, the investigation also
took into account the changing sociohistorical
context.
Development and Context
The processes by which the socioeconomic
background influences individual development
are not yet fully understood. It has been argued
that socioeconomic status (SES) at the time of
the childs birth is an indicator of the social context, but may also reflect parental genetic characteristics, which are assumed to have some role
in determining the level of academic functioning
of their children (Plomin & Bergeman, 1991;
Plomin & McClearn, 1993; Scarr, 1992). Proponents of behavioral genetics see the course of
human development as a function of genetically
controlled maturational sequences (Scarr, 1992).
However, findings show that in no case is the
genetic determination so strong that there is

no room for environmental effects (Plomin &
Daniels, 1987; Scarr, 1992). There is now
increasing skepticism about the usefulness of
approaches formulated within behavioral genetics, on scientific as well as social and ethical
grounds (Baumrind, 1993; Hoffman, 1994;
Jackson, 1993; Lerner & von Eye, 1992). Human
development takes place in a social context, and
is therefore influenced by a persons interactions
within that context. Genetic factors are only one
of a much larger series of possible explanations
for human behavior. More recent developmental
approaches have conceptualized genes and other
biological variables as contributors to reciprocal,
dynamic processes that can only be fully understood in relation to sociocultural environmental
contexts (Bronfenbrenner & Ceci, 1994; Gottlieb,
Wahlsten, & Lickliter, 1998; Horowitz, 2000).
Human beings differ in their capacity for realizing individual talents, and it is important to
understand under what circumstances individual
potentials find expression.
This study was not designed to examine the
heritability of certain traits, but rather to investigate the long-term effects of social risk on academic attainment and consequent adult outcomes
in a changing sociohistorical context. Especially
useful for this type of analysis are approaches
developed by proponents of an ecological perspective of the life course that conceptualizes
human development as the dynamic interaction
between a changing individual and a changing
context (Baltes, 1987; Bronfenbrenner, 1979;
Elder, 1985; Featherman & Lerner, 1985; Lerner,
1984, 1996; Sameroff, 1983). For example, in
their bioecological theory of nature-nurture
effects, Bronfenbrenner and Ceci (1994) argue
that the long-term interactions between children
and their environments are a necessary condition
for the expression of any trait. They differentiate
between the proximal environment, which is
directly experienced by the individual (e.g., the
family environment), and more distal cultural
and social value systems that have an indirect
effect on the individual, and are often mediated
by the more proximal context. Proximal
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 44
processes constitute the basic mechanisms that

produce effective developmental functioning.
They reflect the immediate day-to-day experiences that most directly shape adaptation in the
face of adversity. The form, power, content, and
direction of the proximal processes that affect
development very systematically as a joint function of the person, the environment (both immediate and more remote) in which the processes
are taking place, and the nature of the developmental outcome under consideration.
Ecological explanations of the association
between socioeconomic origin, educational
achievement, and occupational attainment have
emphasized the role of different opportunities and
socialization processes that exist across SES
levels. Individuals from more privileged homes
have more educational opportunities, greater
access to financial resources when they are needed
(e.g., to pay for higher education), and more role
models, occupational knowledge, and informal/
kinship networks (Schulenberg, Vondracek, &
Crouter, 1984). The same cumulative effect, in the
opposite direction, can occur for those who are not
born so lucky and who consequently acquire an
enhanced likelihood of risk. However, early experiences, whether good or bad, do not determine an
invariant life path. For example, in Werner and
Smiths (1992) longitudinal study of high-risk
children, one third had made satisfactory adjustments in adult life, despite being born into highly
disadvantaged circumstances. Some individuals
succeed despite the odds and break the vicious
cycle (Clarke & Clarke, 2000; Elder, Pavalko, &
Hastings, 1991; Pilling, 1990), whereas others
from privileged backgrounds fail to succeed.
Individual differences in response to adversity can
lead to some form of intensification or increased
vulnerability, or to the amelioration or protection
against risk factors that could lead to maladaptive
outcomes (Garmezy, 1985; Rutter, 1990; Werner
& Smith, 1992). Thus, individuals are not
passively exposed to experiential factors, but can
become producers of their own development
(Bronfenbrenner, 1979).
There is consistent evidence of three broad
sets of variables implicated in the development
of positive adjustment in the face of adversity,

a phenomenon also referred to as resilience
(Luthar, Cicchetti, & Becker, 2000; Masten,
Best, & Garmezy, 1990; Rutter, 1990; Werner &
Smith, 1992). These protective factors include
(1) attributes of the individual (e.g., ability, temperament, and motivation), (2) characteristics of
the family (e.g., parental interest and support),
and (3) aspects of the wider social context (e.g.,
neighborhoods and social support system), thus
lending support to the proposition that proximal
processes are fueled by the joint function of
the person and the context (both immediate and
more remote), as suggested by Bronfenbrenner
and Ceci (1994). Resilience is not a personality
attribute, but a dynamic process of positive adaptation despite the experience of significant adversity or risk (Luthar et al., 2000; Masten et al.,
1990; Rutter, 1990; Werner & Smith, 1992).
Recent discussions have raised questions
regarding the multidimensional nature of positive
adaptation or resilience, suggesting that it is necessary to differentiate between constellations of specific risks and specific outcomes (Bronfenbrenner
& Ceci, 1994; Luthar et al., 2000). At-risk children
can show considerable heterogeneity in functioning across different domains (i.e., academic,
emotional, and behavioral adjustment), and successful adaptation in one domain does not imply
positive adaptation in another (Cicchetti &
Garmezy, 1993). The focus of this study was on
academic adjustment, because of its importance as
a context for adaptation in Western culture.
Success or failure in school can have serious
individual and social consequences, and lays the
foundation for future careers.
Timing and Duration of Risk Experiences
There is evidence of great diversity in the
temporal dimension of disadvantage: much
socioeconomic disadvantage is short term,
although a great amount lasts for most of the
childhood years (Duncan & Rodgers, 1988). The
experience of socioeconomic disadvantage does
not always have an immediate impact, and
vulnerabilities may emerge only later in life
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 45
(Clarke & Clarke, 1981). Early adversity may be

overcome by improved circumstances, but may,
nevertheless, leave the individual potentially
more vulnerable to any disadvantage experienced at a later stage (Cicchetti & Tucker, 1994).
Recent studies have shown that persistent socioeconomic disadvantage has stronger effects than
intermittent adversity on individual outcomes
(Ackerman, Schoff, Levinson, Youngstrom, &
Izard, 1999; Bolger et al., 1995; Duncan et al.,
1994; Pungello et al., 1996) and it has been
argued that chronically stressful environments
hinder the development of successful adaptation
(Hammen, 1992). On the other hand, there is evidence that adversity during early childhood, as
opposed to during later developmental periods,
has a crucial impact on later adjustment, especially for academic attainment (Axinn, Duncan,
& Thornton, 1997; Duncan, Yeung, BrooksGunn, & Smith, 1998; Haveman & Wolfe, 1994).
There is, however, also evidence that current
contextual adversity determines current adjustment (Campbell, Pierce, Moore, Marakovitz, &
Newby, 1996; Feiring & Lewis, 1996; Tizard,
1976). Differences in findings can be explained
by methodological variations in the studies that
involve different developmental periods, different indicators of adversity and adjustment, and
different analytical strategies. The relative
contribution of early, concurrent, or persistent
effects can only be elucidated by drawing on
longitudinal studies, which provide detailed
information of individuals followed over time.
Assessing Socioeconomic Risk
Socioeconomic disadvantage is associated
with a variety of cofactors, such as poor living
conditions, overcrowding, or lack of material
resources that pose risks for adaptive development (Ackerman et al., 1999; Conger et al., 1993;
Duncan & Brooks-Gunn, 1997; Fergusson,
Horwood, & Lawton, 1990; Fitzgerald, Lester,
& Zuckerman, 1995; Huston, McLoyd, & Coll,
1994).
Individual risk factors do not exert their effect
in isolation, but rather in interaction with other
influences. It has been suggested that it is the
number of these factors and their combined

effect that shape development (Sameroff, Seifer,
Baldwin, & Baldwin, 1993; Rutter, 1979). The
relation between any single risk factor and subsequent outcomes tends to be weak, and usually
many variables are involved in determining an
outcome (Ackerman et al., 1999; Rutter, 1990;
Sameroff et al., 1993; Szatmari, Shannon, &
Offord, 1994). Serious risk emanates from the
accumulation of risk effects (Robins & Rutter,
1990). In comparison with single-risk models,
multiple-risk models have been shown to be good
predictors of individual outcomes (Ackerman
et al., 1999; Caprara & Rutter, 1995; Fergusson,
Horwood, & Lynskey, 1994; Sameroff et al.,
1993). Most multiple-risk studies use a single
index, summing the number of risk factors
present. Summing the number of risk factors in a
single index, however, gives equal weight to all
risk factors and does not take into consideration
the relative contribution or overlap in risk factors
(Greenberg, Lengua, Coie, & Pinderhughes,
1999; Szatmari, Shannon, & Offord, 1994).
There is now an increasing awareness that the
processes that link socioeconomic disadvantage
to individual development operate at varying
levels of specificity, and that there is a need to
distinguish between economic disadvantage per
se and other associated aspects of environmental
adversity (Ackerman et al., 1999; Duncan et al.,
1994; McLoyd, 1990; Szatmari et al., 1994).
Furthermore, Ackerman et al. (1999) argue that
a single multiple-risk index aggregates a set of
variables that may relate differently to child
functioning, or that may function differently for
advantaged and disadvantaged families, and that
it does not distinguish between persistent and
transitory experiences. Thus, they recommend
the use of discrete groupings of indicators,
narrowing the focus of the variables involved, or
isolating specific factors that pose risks for individual adjustment. The usefulness of such an
approach has been demonstrated in a number of
studies (Ackerman et al., 1999; Deater-Deckard,
Dodge, Bates, & Pettit, 1998; Szatmari et al.,
1994) and was adopted in the present study.
Instead of aggregating various cofactors of
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 46
socioeconomic disadvantage, only indicators

of the socioeconomic family background were
included to assess the unique effect of socioeconomic disadvantage on individual development.
The study drew on indicators of socioeconomic
family status as well as indicators of living conditions and material resources available to the
family. This approach thus went beyond studies
that used social status or income as sole indicators of socioeconomic risk and more accurately
reflected the everyday experiences within the
proximal family context.
Effects of the Wider Sociohistorical Context
Another issue to be addressed concerns the
impact of the wider sociohistorical context in
shaping individual development. For example,
Elder (1999) demonstrated the crucial impact of
the Great American Depression and the outbreak
of World War II on the developmental pathways
of individuals born between 1920 and 1921 in
Oakland, CA. The present study compared the
development of two British birth cohorts born 12
years apart, thereby taking into account contextual
effects that might help to explain differences in
response to disadvantage. The study drew on data
collected for two birth cohorts born in 1958 and
1970, respectively. Changes in social, economic,
and education policies between 1960 and 1980
resulted in the cohort members growing up in
very different environments. Between 1979 and
1986, and again between 1989 and 1993, the
sharpest rise in unemployment since World War II
took place in the United Kingdom. The mid1980s saw the virtual disappearance of the youth
labor market in Britain (Banks et al., 1992;
Bynner, Elias, McKnight, & Pan, 1999). Many
have argued that children born in the 1970s experienced a major shift in life expectations across
the generations. This generation X grew up at
a time when the prospects of achieving employment directly after leaving school and maintaining a continuing career were increasingly in
question, especially for those young people without qualifications (Bynner, Ferri, & Shepherd,
1997; Schoon, McCulloch, Joshi, Wiggins, &
Bynner, 2001; Schoon & Parsons, 2002a). In

response to the changing nature of labor markets
and employment opportunities, young people
are under increasing pressure to continue fulltime education beyond the age of 16 years, and
to acquire formal qualifications. Poor academic
achievement, which presented no significant
barrier to employment in the past, now predicts
real difficulties in finding employment, and ultimately exclusion from the labor market (Bynner,
Joshi, & Tsatsas, 2000).
A Developmental-Contextual
Model of Cumulative Risk Effects
The aim of this study was to examine the
extent of continuity of socioeconomic disadvantage from birth to midadulthood and the maintenance of academic adjustment in the face of that
risk. To better understand the long-term effects
of social risk, the extent to which risk effects persist and how they interact with individual adjustment was considered. Using Structural Equation
Modeling (SEM; Bollen, 1989), the long-term
influence of socioeconomic disadvantage on
individual development was assessed, taking into
consideration the influence of context, timing,
and duration of risk experiences. The pathways
through which the experience of socioeconomic
risk influences the development and maintenance of individual adjustment (i.e., academic
attainment) during childhood and adolescence
as well as the pathways that link childhood
conditions to adult outcomes were investigated.
Figure 3.1 gives a diagrammatic representation
of the developmental-contextual model for
assessing the long-term impact of socioeconomic disadvantage. The model is an explicit
developmental model that assesses the timing
and the duration of the interactions between
individual and context. The variables shown are
all latent or unobserved variables.
The model specifies that conditions at birth
(parental social class) influence the consequent
experience of socioeconomic risk (i.e., low social
class of parents and material disadvantage in the
family home), as well as the level of individual
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 47

b
a2
ACA
ACA
ACA
f
Parental
Social
Class
c
d
SOC
c
d
e
a1
RISK
RISK
Birth
Figure 3.1
Early Childhood
RISK
a1
a1
Late Childhood
Adolescence
Adulthood
Developmental-contextual model of psychosocial adjustment. ACA = academic adjustment;

Risk-socioeconomic risk; SOC = social position in adulthood. See text for further details.
adjustment (i.e., academic achievement). The path

from parental social class at birth to consequent
socioeconomic risk is labeled a1, and the path
from parental social class at birth to academic
attainment is labeled a2. It was hypothesized that
socioeconomic disadvantage at one time point
would predict the experience of socioeconomic
disadvantage at a later time point. The arrows
labeled a1 that link the latent risk variables indicate this continuity of socioeconomic adversity
over time. It was furthermore hypothesized that
social class at birth would be an important predictor of academic adjustment (path a2), and that
academic attainment at one time point would influence academic attainment at a later time point
(paths b). It was also assumed that the experience of concurrent socioeconomic disadvantage
would be associated with the development and
maintenance of academic adjustment. These concurrent associations during early and late childhood, as well as during adolescence, are indicated
as c. The model thus tested the persistence of
socioeconomic risk and the maintenance of academic adjustment in the face of that risk. The model
furthermore assessed the cumulating effects of
social risk on academic adjustment by estimating
the additional incremental risk effects at subsequent time points (early childhood, late childhood,
and adolescence). The paths labeled d indicate
these time-lagged effects of earlier risk on later
academic attainment. The model furthermore
examined the independent influence of social risk
experienced during adolescence on adult social
status (path e), as well as the influence of academic adjustment during adolescence on adult
social status (path f). To establish whether
parental social class at birth had an influence on an
individuals adult social status, independent of
intervening experiences, path g was included,
which was an indicator of cohort effects not
accounted for by the risk and adjustment levels
carried forward in time. The model also considered
the effects of a changing sociohistorical context,
which is not shown in the diagram but was taken
into account in the analyses by applying the model
to two cohorts of young people growing up 12
years apart. The importance of the life-course
developmental perspective for this study lay in its
scope to integrate process and structure and to link
individual time with historical time. The longitudinal approach had a number of methodological
advantages with regard to the research questions:
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 48
reports of early life events were not influenced by

knowledge of the subsequent personal history;
the cohort included appropriate controls; and the
conjoint impact of different factors experienced at
different time points could be analyzed in a multivariate, multicausal model.
Method
This study used data collected for the National
Child Development Study (NCDS) and the
British Cohort Study (BCS70), two of Britains
richest research resources for the study of
human development. The participants of the
NCDS included all persons born in Great Britain
between March 3 and March 9, 1958. In five
follow-up studies, data were collected on the
physical, psychosocial, and educational development of the cohort at ages 7, 11, 16, 23, and 33
years (Shepherd, 1995), and, most recently, at age
42 (Bynner, Ferri, Shepherd, & Smith, 2000). At
each sweep between 1958 and 1974, wide-ranging information was collected from parents,
teachers, school medical officers, and, at later
stages, from the cohort members themselves via
personal interviews. Satisfactory response rates
have been reported for each sweep, and comparison of data has shown that the achieved samples
did not markedly differ from the target samples,
or from other survey samples of the British
population (Shepherd, 1993, 1995).
The BCS70 has followed children born in
the week April 511, 1970, from birth through to
adulthood. Data sweeps took place when the
cohort members were ages 5, 10, 16, and 26 years
(Ekinsmyth, Bynner, Montgomery, & Shepherd,
1992), and most recently at age 30 (Bynner, Ferri,
Shepherd, & Smith, 2000). In the birth survey,
information was collected by means of a questionnaire that was completed by the midwife present
at birth, and supplementary information was
obtained from clinical records. In 1975 and 1980,
parents of the cohort members were interviewed
by Health Visitors, the cohort members themselves
undertook ability tests, and the school health
service gathered medical information on each
child. This was supplemented at ages 10 and 16 by
information gathered from head and class teachers

who completed questionnaires. A low response
rate at age 16 occurred because of a teacher strike,
which coincided with the fieldwork. All school
children were afflicted in the same way, and the
demographic characteristics of the sample at age
16 remained representative of the target population
(Shepherd, 1997). The follow-up study at age 26
was conducted via a mail survey. At age 30, data
were collected by interview with the cohort
members (Bynner, Ferri, et al., 2000). An analysis
of response bias showed that the achieved samples
did not differ from the target samples across a
number of key variables (social class, parental education, and gender), despite a slight underrepresentation of the most disadvantaged groups
(Butler, Despotidou, & Shepherd, 1997; Davie,
Butler, & Goldstein, 1972; Shepherd, 1997).
Sample
The sample consisted of all individuals for
whom complete data were collected at birth. The
following analyses were based on a total sample
of 16,994 cohort members for the NCDS and
14,229 cohort members for the BCS70. In both
cohorts, there were 52% men and 48% women.
Data collected between birth and age 16 were
linked with data collected at age 30 (BCS70)
and age 33 (NCDS) when the cohort members
reached midadulthood. Potential bias due to
missing variable information in both cohorts was
addressed in the section on estimating the model.
Measurement of Socioeconomic Risk
Risk factors associated with socioeconomic
disadvantage were indicated by parental social
class and material conditions in the family
household. The indicator variables were measured at ages 7, 11, and 16 for the NCDS, and
ages 5, 10, and 16 for the BCS70.
Parental Social Class

In both the NCDS and the BCS70, social class
was measured by the Registrar Generals measure of social class (RGSC). The RGSC is
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 49
defined according to job status and the

associated education, prestige (OPCS, 1980), or
lifestyle (Marsh, 1986), and is assessed by the
current- or last-held job. The RGSC is coded on
a 6-point scale: I, professional; II, managerial
and technical; IIINM, skilled nonmanual; IIIM,
skilled manual; IV, partly skilled; and V,
unskilled (Leete & Fox, 1977). The occupational
categories used in the U.S. census and other
European countries are similarly based on the
skills and status of different occupations
(Krieger & Williams, 1997). Class I represents
the highest level of prestige or skill and Class V
represents the lowest. In cases in which the
father was absent, the social class (RGSC) of
the mother was used in the BCS70. The same
applied to the NCDS at ages 7, 11, and 16; however in cases in which there was no father at
birth, the mothers fathers social class was used.
Material Conditions
Material conditions in the family environment
were assessed on the basis of a summative index,
in which the presence or absence of four indicator variables (listed below) is summed. The scale
gives an overall score of material disadvantage
that ranges between 0 and 4.
Overcrowding. This is a dichotomous variable
based on the ratio of people living in the household to the number of rooms in the household.
One or more persons per room was coded 1,
and less than one person per room was coded 0.
Household amenities. This is a dichotomous
scale based on the cohort members family
having sole use of a bathroom, toilet, and hot
water. The same three questions were asked in
both studies. Sole access to all of these amenities
was coded 0, and shared use or no access to
any of these amenities was coded 1.
Housing tenure. The tenure of the home was
defined as owner-occupier (0) or other (1).
Receipt of state benefits. Receipt of state benefits
is an indicator of financial hardship within the
family environment (Fogelman, 1983). The

assessed benefits include payment of unemployment benefit, income support, and housing benefit, but exclude payment of pension or child
benefit. Parents were coded as either not in
receipt of benefits (0) or in receipt of benefits in
the last 12 months (1).
Individual Adjustment
Individual adjustment was measured by the
childs academic attainment at each measurement point. To reflect the changing competencies
of the growing child, academic attainment was
assessed differently during early childhood (age
5 or 7), middle childhood (age 10 or 11), and
adolescence (age 16).
Academic Attainment at
Age 5 (BCS70) and 7 (NCDS)
The Human Figure Drawing Test used in the
present study was a modified version of the Drawa-Man test originally devised by Goodenough
(1926) and developed further by Harris (1963).
The Harris-Goodenough Test has good reliability
(.94; Osborn, Butler, & Morris, 1984). It has been
evaluated as a measure of intelligence and significant correlations, rs averaging between .4 and .5,
with conventional IQ tests (Binet, Wechsler)
have been reported (Scott, 1981). The scoring
of the drawings produced by the children is
based on 30 developmental items suggested by
Koppitz (1968) and uses the Harris (1963) point
system of scoring. One point is scored for each
item represented in the drawing, giving a maximum possible score of 30. In both cohorts the
children had to draw two figures. In the BCS70,
however, only one figure was coded. Thus, the
maximum score in the NCDS was 60, and in the
BCS70 was 30. The achieved scores ranged
from 0 to 53 in the NCDS, and 0 to 23 in the
BCS70.
The Copy-a-Design test (Davie et al., 1972)
assesses the cohort members perceptual-motor
ability. The ability to copy designs or geometric
shapes is included as one element of assessment
in many standard intelligence tests. The test used
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 50
in the cohort studies assumed that children

had reached a certain level of conceptual development to be able to recognize the principles
governing different geometric forms and to
reproduce them (Osborn et al., 1984). The test
has satisfactory reliability (.70; Osborn et al.,
1984). In the NCDS, test scores ranged from 0 to
12; in the BCS70, the range was 0 to 8.
The Southgate Reading Test (Southgate,
1962), a test of word recognition and comprehension particularly suited to identifying problems with reading in young children, was used
in the NCDS. The test has good reliability (.94;
Southgate, 1962). The range of scores in the
NCDS was 0 to 30.
The English Picture Vocabulary Test, an
adaptation of the American Peabody Picture
Vocabulary Test (Brimer & Dunn, 1962), was
used for the BCS70 cohort. The test has good
reliability (.96; Osborn et al., 1984). It consists
of 56 sets of four different pictures with a particular word associated with each set of four pictures. The child is asked to indicate the one
picture that corresponds to the given word, and
the test proceeds with words of increasing difficulty, until the child makes five mistakes in a run
of eight consecutive items.
The Problem Arithmetic Test was used in
the NCDS cohort only, because cohort members
in the BCS70 were too young to have started
formal training in arithmetic. This test has
a satisfactory reliability of .85 (Pringle et al.,
1966).
Academic Attainment at
Age 10 (BCS70) and 11 (NCDS)
The National Foundation for Educational
Research in England and Wales (NFER) constructed a reading comprehension test specifically for use in the NCDS (Fogelman, 1983).
Good test reliability has been reported (.82;
Goldstein, 1979), and scores range from 0 to 35.
In the BCS70, a shortened version of the Edinburgh Reading Test, a test of word recognition,
was used after consultation with the tests
authors (Godfrey Thompson Unit, 1978). The
shortened test version contained 54 items that

examined vocabulary, syntax, sequencing, comprehension, and retention. The test has good
reliability (.87) and the items discriminate well
between good and poor readers (Butler et al.,
1997).
The NFER developed an arithmeticmathematics test specifically for use in the NCDS
(Fogelman, 1983). The scores range from 0 to 40.
The test has good reliability (.94; Goldstein,
1979). The lack of a fully acceptable mathematics test appropriate for 10-year-olds also led to
the development of a special test for the BCS70
cohort. It consisted of a total of 72 multiple
choice questions and covered in essence the rules
of arithmetic, number skills, fractions, measures
in a variety of forms, algebra, geometry, and
statistics. The test has good reliability (.92) and
the items have adequate discrimination (Butler
et al., 1997).
Academic Attainment
at Age 16 (Both Cohorts)
Two measures of academic achievement in
secondary school were considered: the highest
level of secondary school examinations passed
by the students at age 16, and the exam scores
that students obtained when they were 16 years
old (16 was the minimum legal age at which a
child could leave school).
There are essentially two types of secondary
school examinations that a student can pass at
age 16: the Certificate of Secondary Education
(CSE) examination and the ordinary (0 level)
examinations within the General Certificate of
Education (GCE) examinations. The GCE is the
accepted examination for children of above average intelligence and caters to approximately 20%
of the total age group, whereas the CSE examination is designed to cover a wider range of ability than does the GCEan additional 40%so
that the two exams combined are intended for
some three fifths of the population (Rutter,
Maughan, Mortimore, & Ouston, 1979). Both
the GCE and the CSE are subject based, and
grades are awarded on the basis of performance
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 51
with a range from 1 to 5 (or A to E). Generally,

GCE grades of D and E are classified as failures.
There is an accepted equivalence between the
two examination systems with a grade 1 on the
CSE examination being seen as equal to at least
a grade C pass on the GCE examination. For
both cohorts, the highest level of qualifications
obtained at age 16 was recorded, ranging from
none (0), CSE grade 25 (1), and CSE grade 1 or
0 level (2).
An overall exam score could also be calculated from the examination performance at age
16. The actual examination results of the NCDS
cohort were collected from schools in 1978,
whereas the BCS70 cohort members selfreported their examination results in a follow-up
study in 1986. The examination system was the
same for both cohorts, with the BCS70 being one
of the last cohorts to sit the two-tiered examination structure of 0 levels and CSEs. A simple
scoring technique was applied to the results, in
which a score of 7 was given to a grade 10 level
and a score of 1 was given to a grade 5 CSE.
Scores ranged from 0 to 106 in the NCDS and
from 0 to 97 in the BCS70.
Attainments in Adulthood
Adult attainment was indicated by two
measures of social position: the RGSC and
the Cambridge Scale (CS) assessed at age 30 for
the BCS70 cohort members and at age 33 for the
NCDS cohort members. The 6-point RGSC scale,
developed by the Office of Population and
Census Surveys (OPCS; 1990), was described
above. For ease of interpretation the coding was
reversed, so that a high score indicated a high
social position. The CS was conceptualized as
an indicator of general social advantage and
lifestyle (Prandy, 1990). It is based on the analysis of friendship choices, judged to be the most
accurate indication of perceived and experienced
social distance between members of different
occupations. The scale is measured on a 100point continuum, whereby high scores indicate a
higher level of social advantage.
Statistical Analysis:
Modeling Cumulative Risk Effects
Structural equation modeling (Bollen, 1989)
was used to formulate theoretically derived
hypotheses about variable relations and to test
postulated pathways between the variables and
the assumed mediating processes involving latent
variables with multiple indicators. Latent variables represent hypothetical concepts that cannot
be observed or measured directly. Instead, a set of
observed variables are hypothesized to be imperfect indicators of the latent variable. Because the
study involved a cross-cohort comparison of data
that had been collected for different surveys,
some of the data were similar but not necessarily
identical for the two cohorts. Great care was
taken to define the latent constructs in as similar
a way as possible in the two cohorts. Table 3.1
summarizes the selection of comparable indicator variables in both cohorts.
All analyses were carried out using the
SEM program AMOS 4.01 (Arbuckle, 1999).
The AMOS program uses maximum likelihood
estimation that can be based on incomplete
data, known as the full information maximum
likelihood (FIML) approach. This approach is
preferable to estimation based on complete data
(the listwise deletion [LD] approach) because
FIML estimates tend to show less bias and
be more reliable than LD estimates, even when
the data deviate from missing at random and
are nonignorable (Arbuckle, 1996). In the LD
approach, the complete data covariance matrix is
the data source for the latent variable analysis. In
the FIML approach, estimation is based on the
many covariance matrices between observed
variables for all patterns of missing data in the
other observed variables. Thus, it is not possible
to present a single correlation matrix for the
observed variables. Instead, Appendices A and B
give the FIML estimates of the correlations
between the observed variables, the means and
SDs for the observed variables, and the FIML
estimates of means and SDs.
In line with current practice, several criteria
were used to assess the fit of the model to the
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 52

Table 3.1
Variables and Observed Indicators in Both Cohorts
Variables
Socioeconomic risk (RISK)
Parental social class
Material conditions
Academic adjustment (ACA)

Academic attainment
Social position in adulthood (SOC)

Social position at age 30 (BCS70)
and age 33 (NCDS)
Observed Indicators
Fathers social class
Mothers social class
No housing tenure, overcrowding (>1 person per
room), household amenities (shared use of
bathroom), receipt of state benefits
Ages 5 and 7: Human Figure Drawing,
Copy-a-Design, Reading and vocabulary
(NCDS and BCS70), and Arithmetic (NCDS
only) tests
Ages 10 and 11: Reading and mathematics tests
(NCDS and BCS70)
Age 16: Exam scores (NCDS and BCS70)
Registrar Generals measure of social class (RGSC)
Cambridge Scale (CS)
NOTE: NCDS = National Child Development Study; BCS70 = British Cohort Study.
data. The x2 statistic is overly sensitive to model

misspecification when sample sizes are large
or the observed variables are non-normally
distributed. The root mean square error of
approximation (RMSEA) gives a measure of
the discrepancy in fit per degrees of freedom
(Steiger, 1990). It is bounded below by 0, only
taking this value if the model fits exactly. The
RMSEA is useful because it encompasses the
idea that a model is only expected to provide an
approximation to the data rather than an exact fit.
If the RMSEA is <.05, the model is considered
a close fit to the data. Another advantage of the
RMSEA is that confidence intervals may be
calculated, which give further information on the
reliability of the goodness of fit. The Consistent
Akaike Information Criterion (CAIC) is a measure of parsimonious fit recommended for large
samples (Bollen & Long, 1993). The CAIC considers both the fit of the model and the number
of estimated parameters whereby smaller values
indicate a more parsimonious fit (Bollen &
Long, 1993; Bozdogan, 1987). The final index of
choice was the comparative fit index (CFI)
whose values are restricted to lie on a 0 to 1
continuum, with higher values indicating a better

fit (Bentler, 1990). The CFI is a populationbased index that compares the model to a null
model. The null model is a model in which there
are no relations between any of the observed
variables, but their variances are not constrained
and are free to be estimated by the fitting procedure. The CFI of a model is normally tested
against a minimum criterion value of .95.
Modeling Strategy
Four separate models were run for each
cohort. Model 1 was the Full Developmental
Contextual Model, which included all paths indicated in Figure 3.1. In addition to Model 1 three
other models were fitted to the data, to test
the increase of fit depending on the pathways
included, and to identify the most parsimonious
model for describing the long-term influences
of socioeconomic adversity on academic adjustment. Model 2 was the DevelopmentalContextual Model without path g (which assessed
the direct effect of parental social class on an
individuals adult social status independent of
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 53
the intervening variables). Model 3 was the

DevelopmentalContextual Model without paths
d (which indicated the time-lagged effects
from social risk to academic attainment). Model
4 was the DevelopmentalContextual Model
without the correlations c (which described the
concurrent associations between social risk and
academic attainment).
Results
The distribution of the risk indicator variables
is shown in Table 3.2. Generally, material conditions improved for the later-born BCS70 cohort.
In comparison with cohort members born in 1958,
more families owned their home, there was less
overcrowding, and fewer households had to
share amenities. The percentage of families who
were receiving state benefits had remained stable. Also noted was upward mobility among the
parents of cohort members born in1970, whereas
Table 3.2
the social position of parents of the earlier born

NCDS cohort remained stable over the years. If
one compares the risk prevalence in the 1958
NCDS cohort at age 16 to the one in the 1970
BCS70 of the younger cohort at age 5, which
were assessed at roughly the same time (i.e.,
1974 and 1975, respectively), Table 3.2 shows
that the distribution of social status of the parents
was comparable in the two samples, as were
housing tenure and shared use of amenities.
Differences in the rate of overcrowding might be
explained by different stages of family formation
in the two cohorts.
In the next step, a set of nested alternative
SEM models were run to test the increase in fit
depending on the pathways included in the
analysis, and to identify the most parsimonious
model for describing the long-term influences
of socioeconomic adversity on academic adjustment. The goodness-of-fit indicators for the different models are shown in Table 3.3.
Distribution of the Risk Variables in the 1958 National Child Development Study (NCDS) and
the 1970 British Cohort Study (BCS70)
Cohort (%)
Risk Variable
NCDS
BCS70
22
24
23
55
66
19
19
44
40
7
Age 11/10 (1969/1980)

Social class (% in RGSC IV and V)
No housing tenure
Overcrowding (1+ person per room)
Shared use of amenities
Family receives benefits
23
54
51
7
27
18
39
30
3
22
Age 16 (1974/1986)
No housing tenure
Family receiving benefits
22
50
60
7
23
13
28
17
1
27
Birth (1958/1970)
Social class at birth (% in RGSC IV and V)
Age 7/5 (1964/1975)
No housing tenure
NOTE: RGSC = parental social position.
54
723.73
730.92
848.12
1,022.88
77
78
79
80
.998
.998
.998
.997
.998
.998
.996
.997
CFI
1,347
1,344
1,450
1,614
1,866
1,863
2,510
2,329
CAIC
.024
.024
.026
.029
.027
.027
.033
.032
RMSEA
90% CI
.023.026
.023.026
.024.028
.027.030
.025.028
.025.028
.032.035
.030.033
Test of Close Fit
1 (M2-M1)
2 (M3-M1)
3 (M4-M1)
1 (M2-M1)
2 (M3-M1)
3 (M4-M1)
df
7.19
124.39
299.15
7.63
665.44
495.46
x2
< .01
< .001
< .001
< .01
< .001
< .001
Step-Down Goodness of Fit
NOTE: df = degrees of freedom; CFI = comparative fit index; CAIC = Consistent Akaike Information Criterion; RMSEA = root mean square error of approximation;
CI = confidence interval.
1,221.47
1,229.10
1,886.91
1,716.93
93
94
95
96
df
5:33 PM
Goodness of Fit
Comparative Goodness of Fit of Structural Equation Models Run for the 1958 National Child Development Cohort (NCDS) and the
1970 British Cohort Study (BCS70)
11/16/2004
NCDS
Model 1: The full model
Model 2: Excluding path g
Model 3: Excluding paths d
Model 4: Excluding correlations c
BCS70
Model 1: The full model
Model 2: Excluding path g
Model 3: Excluding paths d
Model 4: Excluding correlations c
Table 3.3
03-Chibucos.qxd
Page 54
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 55
In both cohorts Model 2 was judged to be

the best model for describing the relations in the
data. Because the x2 statistic is overly sensitive
to model misspecification when sample sizes are
large, the x2 statistic in conjunction with the
CAIC were used to identify the best-fitting
model for the data. Model 2 was run without path
g suggesting that there were no cohort effects
independent of intervening variables. Given that
Model 2 was the most parsimonious in both
cohorts, it is described in detail below.
Table 3.4 gives the standardized parameter
estimates of the measurement model for both
cohorts. The standardized regression weights of
the indicator variables on their latent variable
differed slightly for the two cohorts. Social class
had a similar weighting in both cohorts, and was
the most important indicator of socioeconomic
disadvantage. Material deprivation gained more
importance as the NCDS cohort members grew
older, whereas for the BCS70 cohort there was a
peak at age 10. The most important indicators of
a childs adjustment in the NCDS were generally
reading and math test scores, and at age 5 also
the Human Figure Drawing and Copy-a-Design
tests. The BCS70 cohort had slightly different
indicators: at age 5 the Copy-a-Design test and
the reading test were the principal indicators,
whereas at age 10, the reading and math tests
were the principal indicators. For both cohorts,
test performance at age 10/11 was more crucial
than earlier performance as a determinant of a
childs adjustment. At age 16, the exam score
was the key indicator of a childs adjustment in
both cohorts, particularly so in the later-born
cohort. Social status for the 33-year-old NCDS
cohort members was slightly better identified by
the CS than by the RGSC, whereas in the BCS70
at age 30 years, the RGSC was a better indicator
than the CS.
Figures 3.2 and 3.3 show the pathways
between the latent or unobserved variables,
which represent continuities and interactions of
social risk and academic adjustment, and give
the standardized coefficients for the structural
model, fitted separately for both cohorts. Several
covariances between the error terms for the
observed variables were included a priori to

account for the autocorrelations over time.
Parental social class and material deprivation
were hypothesized to covary for the measurement points at consequent ages, as were the academic attainment scores. The variables shown
were all latent or unobserved variables.
The hypothesized pathways were supported
by the data, with the parameter estimates all
being significantly different from 0, p < .005,
and in the predicted direction. The effect sizes of
the parameter estimates are described as small,
r = .10, medium, r = .30, and large, r = .50, following Cohens 1992 power primer. In both
cohorts a stark chaining, or continuity of risk
factors, was found: parental social class at birth
predicted the experience of risk at subsequent
ages, and the experience of risk at one time point
increased the probability that risk would also be
encountered at a later time point. Parental social
class also had a moderate influence on academic
adjustment. The experience of social risk at birth
influenced the level of later academic attainment.
Also observed were continuities in academic
adjustment level over time. Academic attainment
at one time point was a significant predictor
of academic attainment at a later time point.
Continuities occurred because current adjustment encompassed previous adjustment as well
as earlier structural and functional change. The
detrimental effect of experiencing disadvantage
at one measurement point was carried forward
into the future via decreased individual adjustment levels.
The model also shows concurrent associations between social risk and academic attainment. In the NCDS, these current associations
(coefficients c) were of moderate size at age 7
and 16, and were only of small size at age 11. In
the BCS70, the concurrent association between
social risk and academic attainment at age 5 was
of moderate size, whereas at age 10 and 16 the
associations were only small.
In addition, there were small time-lagged risk
effects (paths d) which indicated the added
negative influence of social risk on subsequent
attainment not accounted for by the risk carried
56
.62
.34
.55
.80
.41
.48
.50
.80
.66
Academic adjustment (ACA)

ACACopy-a-Design
ACAHuman Figure Drawing
ACAReading
ACAMath
ACAExam score
ACAHighest Qualification at age 16
Social status in adulthood

Social statusRegistrar Generals
measure of social class
Social statusCambridge Score
.80
.41
BCS70,
age 5
.85
.88
.78
.53
NCDS,
age 11
.83
.83
.76
.50
BCS70,
age 10
.69
.61
.74
.53
NCDS,
age 16
.87
.63
.76
.45
BCS70,
age 16
.95
.88
NCDS,
age 33
.65
.74
BCS70,
at 30
5:33 PM
NCDS,
age 7
The Measurement Model. Estimated Standardized Regression Weights for the National Child Development Study (NCDS) and the
1970 British Cohort Study (BCS70)
11/16/2004
Risk
RiskParental social class
RiskMaterial deprivation
Table 3.4
03-Chibucos.qxd
Page 56
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 57

.73
.78
.32
Parental
Social
Class
ACA
RISK
SOC
.26
RISK
.97
.99
Late Childhood
Early Childhood
RISK
.05
Adulthood
Adolescence
The National Child Development Study: Developmental-contextual model of accumulating risk

effects from birth to adulthood. ACA = academic adjustment; Risk = socioeconomic risk;
SOC = social position in adulthood. See text for further details.
.71
.81
.45
Parental
Social
Class
Birth
ACA
ACA
.45
ACA
RISK
Early Childhood
.64
.09
.15
.07
.91
Figure 3.3
.57
.49
.19
Birth
ACA
.21
.38
.88
Figure 3.2
ACA
SOC
.22
RISK
.93
Late Childhood
.94
RISK
Adolescence
.14
Adulthood
The British Cohort Study: Developmental-contextual model of accumulating risk effects from
birth to adulthood. ACA = academic adjustment; Risk = socioeconomic risk; SOC = social
position in adulthood. See text for further details.
forward in time. In both cohorts these time-lagged

risk effects were greatest at the transition from late
childhood to adolescence.
The best predictor of attained social status in
early adulthood was academic attainment at age
16, which confirmed the crucial role of education
in determining adult outcomes. In both cohorts,
the direct influence of social risk experienced at
age 16 on adult social status was only small, yet

in comparison with the NCDS the effect size had
nearly trebled in the BCS70. It is important to
remember that the experience of social risk does
have additional indirect effects on social status
operating through the childs academic adjustment. For the NCDS cohort, the combined effect
of the variables in the model explained 37% of
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 58
the variation in adult social status; and for the

BCS70 cohort, the model explained 54% of the
variation.
Discussion
In the study of human development, laboratories are rare. The closest one can get to them is
when the opportunity arises to compare and contrast experiences in different societies or in different sociohistorical contexts. The present study
used data collected for two national birth cohort
studies born 12 years apart. On comparison of
the prevalence of risk indicators that were
assessed at roughly the same time (e.g., parental
social class or housing tenure experienced by 16year-olds born in 1958 and 5-year-olds born in
1970, assessed in 1974 and 1975, respectively),
it appears that the samples represented well the
state of affairs over the historical period covered.
The data suggest that there was a real-time
secular shift in material resources. For cohort
members born in 1970, the indicator variables
point to improved material circumstances for the
cohort as a whole. Furthermore, the parents of
cohort members born in 1970 experienced
upward social mobility while bringing up their
children. Those cohort members born in 1970
who suffered deprivation, however, appear to
have been more disadvantaged relative to other
children in the same cohort than similarly
affected children born earlier, in 1958. A strong
continuity of social risk effects was observed,
which was of similar strength for both cohorts.
The strength of the association can be explained
by the fact that in addition to indicators of
material disadvantage, indicators of SES, which
denote relative position in society, were also
used. However, despite improved material conditions, and the experience of upward social
mobility among parents of the BCS70 cohort, the
relative social position remained remarkably
stable. Furthermore, the influence of parental
social class at birth on academic adjustment was
greater for cohort members born in 1970 than for
those born in 1958, and the direct influence of
social risk experienced in adolescence on adult
attainment was only small, but nearly trebled

for the later-born BCS70 cohort. These findings
suggest that for cohort members born in 1970,
contextual factors, to some extent, became
more important than for cohort members born in
1958 in shaping the development of academic
adjustment, and in influencing attainments in
adulthood. Material conditions improved, yet
socioeconomic disadvantage continued to be a
barrier for individual achievements.
It has been argued that SES at the time of a
childs birth is an indicator of the social context,
but also reflects hereditary influences, which
are assumed to have some role in determining
the level of academic functioning (Plomin &
Bergeman, 1991; Plomin & McClearn, 1993;
Scarr, 1992). The present studys data show that
the influence of social class at birth on academic
adjustment was of moderate size, and differed
slightly for the two cohorts. Socioeconomic
disadvantage experienced consequently during
childhood had an additional detrimental influence on the level of academic adjustment, suggesting that contextual factors have a role in
shaping the level and maintenance of academic
adjustment throughout childhood and adolescence. Furthermore, there were smalltime-lagged
effects of earlier risk on later academic achievement, indicating that social risk does not always
have an immediate impact, and that vulnerabilities may emerge only later in life. On the other
hand a considerable stability of individual
adjustment was also seen. These results suggest
that adaptation is a product of both developmental history and current circumstances (Clarke &
Clarke, 1981, 2000; Sroufe, Egeland, & Kreuzer,
1990), shaped by the interactions between
individual and context (Bronfenbrenner &
Ceci, 1994).
Generally, the findings of this study concur
with a large body of previous research, which
has reported consistent correlations between
measures of social disadvantage and measures
of individual academic achievement and occupational attainment (Blau & Duncan, 1967; Bolger
et al., 1995; Bynner, Joshi, & Tsatsas, 2000;
Campbell & Ramey, 1994; Duncan et al., 1994;
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 59
Felner et al., 1995; Pungello et al., 1996; Ramey

& Ramey, 1990; Rutter & Madge, 1976; Sewell
et al., 1970; Walker et al., 1994). The approach
adopted in the present study begins to provide an
understanding of the ways in which constellations of social risk and individual adjustment
emerge early in the life course, and how they
interrelate over time. The findings suggest that
pervasive social inequalities exist that influence
academic attainment during childhood, and which
are consequently reflected in adult achievements.
Risk experiences are not randomly distributed in
a population, and being born into a relatively disadvantaged family increases the probability of
accumulating risks associated with that disadvantage, setting a child onto a risk trajectory
(Rutter, 1990). On the other hand, great stability
in individual adjustment is also seen, and
academic attainment at one time point is a strong
predictor of academic attainment at a later
time point.
The best predictor of adult social status was
academic attainment, which confirmed the crucial role of educational achievement in determining adult outcomes. This is not to deny the
influence of social risk on individual development. Experience of early social risk influences
the level of academic adjustment, which in turn
influences adult attainment. Despite the importance of early childhood, the data did not conclusively support the assumption that parental
social class at birth has a strong influence on an
individuals adult social status independent of
intervening experiences. The early family environment is important in shaping subsequent
development, yet intervening processes also have
to be considered to obtain a better understanding
of adult outcomes. The whole life course is
important, not just the early years. The findings
of this study lend support to the proposition
made by Bronfenbrenner and Ceci (1994) that
individual adaptation across the life span is a
joint function of the characteristics of the developing person and the contextboth proximal
and more remotein which development
takes place. The experience of early disadvantage weakens individual adjustment, and this
detrimental effect is then carried forward into

the future. Subsequent experiences of adversity
add to the deterioration of already reduced
adjustment. A general premise of life-course
studies postulates that adaptations to change are
influenced by what people bring to the new situation. If individual adjustment is already weakened
at a very early age, it becomes increasingly difficult to fully develop ones potential. This negative
chain effect undermines the academic adjustment
of the young person, and ultimately the individual
attainments in adulthood. Generally the results
imply that cumulative adversity has effects
beyond those associated with current or early
adversity (Ackerman et al., 1999; Bolger et al.,
1995; Duncan et al., 1994; Pungello et al., 1996).
As suggested by Sroufe et al. (1990), adjustment
problems possibly do not lie with the individual
per se but in the persistent adjustment of the individual to adverse conditions over time. Children
growing up in households in which they have no
room of their own, or possibly even a desk or
table at which to complete their homework, are
less likely than their more privileged peers to do
well in school.
The data further suggest differences in the
timing of risk effects. Among cohort members
born in 1958 (NCDS), the influence of concurrent social risk was greatest during early childhood, at age 7, and during adolescence, at age
16, when important decisions about future
careers are made. Among cohort members born
in 1970 (BCS70), the greatest effect of concurrent risk on individual adjustment was found at
age 5, whereas at later ages the cohort members
appeared to be comparatively unaffected by
current socioeconomic family circumstances
over and above those already accounted for by
previous experiences of socioeconomic adversity. These differences in timing of risk effects
might be explained by changes in the sociohistorical context. Cohort members turned 16 in
1974 and 1986, respectively. In consequence to
the virtual disappearance of the youth labor
market that occurred between 1979 and 1986, the
later-born cohort encountered more complex and
varied education, training, and employment
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 60
choices. Although in the 1970s the predominant

pattern was to leave school at the minimum age
and move directly into a job, by the 1990s most
young people continued in full-time education
after the age of 16 (Bynner et al., 1999; Bynner,
Joshi, & Tsatsas, 2000). Cohort members born in
1970 were under increasing pressure to acquire
formal qualifications, whereas most young
people born in 1958 could expect to obtain
employment regardless of their educational
attainment (Bynner et al., 2000). Thus, parents
of cohort members born in 1958, especially less
privileged parents, might not have pushed their
children to obtain good grades, and rather might
have encouraged their children to leave school
early to earn a wage. Parents in the later-born
BCS70 cohort, in contrast, might have generally
put more emphasis on achieving good examination results, whatever their socioeconomic circumstances. It could also be that in the later-born
BCS70 cohort, other factorssuch as support
from the school environment or contact with
peerswhich were not captured in the model,
might have played a more important role in influencing academic achievement than in the earlierborn NCDS cohort.
In interpreting the findings, some limitations
of the study should be noted. This work involved
a cross-cohort comparison of the constellations
between adversities and adjustment of individuals growing up 12 years apart. A latent-variable
modeling approach was used to convey our
theoretical framework for assessing the impact
of contextual risk on individual development
from birth to early adulthood. The emphasis was
on investigating relations between latent variables in the two cohorts, rather than on the relations between observed variables. Great care was
taken to measure the latent variables in as similar a way as possible in the two cohorts. As with
all research using cohort studies, this work was
constrained by data collected up to 40 years ago
in light of research practice prevalent at that
time. There are always limitations placed on
comparative analysis across cohort studies that
were not designed to measure the same variables,
and stronger associations might have been
obtained by using different indicator variables. It

may also be that missing data at the individual
level and at the variable level affected the validity
of the results. Response bias at the individual
level would tend to underestimate the magnitude
of the effects of social disadvantage on individual
adjustment, because sample attrition is greatest
among individuals in more deprived circumstances. The results may thus provide a conservative estimate of social inequalities experienced in
childhood. Missing data at the variable level may
also be nonrandom. The FIML approach has been
adopted as a best effort technique for dealing
with these problems, but bias in the model estimates may still be present. Nonetheless, the data
offers a unique opportunity to investigate the
processes that link the experience of early
socioeconomic disadvantage, academic achievement, and adult occupational attainment in two
cohorts growing up in a changing sociohistorical
context. Comparisons with other studies on the
long-term effects of socioeconomic risk are compromised by the lack of consistency in the measures of social risk and individual adjustment
across the life course. This study was designed to
measure the extent to which ill effects of social
disadvantage persist over time, and the way in
which the constellations of social risk and academic adjustment emerge in a changing sociohistorical context. Future research should explore
in more detail the factors and processes that
modify the constellations between risk and
adjustment (e.g., individual temperament or personality characteristics, parenting styles, or characteristics of the school environment) that can
act as protective factors by impeding or halting
the negative chain process, and enabling the
child to move into positive directions (Masten
et al., 1990; Rutter, 1990; Schoon & Parsons,
2002b; Werner & Smith, 1992). The present
study has shown that the constellations of social
risk and academic adjustment vary by age, context, and duration of the experiences, suggesting
that different factors and processes might be
important at different developmental stages,
indifferent contexts, and for different developmental outcomes.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 61
It can be concluded that the contextual

perspective within life-course research offers a
useful framework that contributes to a fuller
understanding of the processes that link social disadvantage to individual development. The impact
of risk factors depends on characteristics of the
individual, but also on the context, including the
proximal family environment as well as the wider
sociohistorical context that dictates opportunities
and possibilities. The effects of socioeconomic
disadvantage are cumulative. Both the timing and
duration of risk experiences play a role in shaping
individual adjustment, and for a better understanding of successful adaptation it is necessary
to consider the dynamic interaction between a
changing individual and a changing context. The
whole life path is important in shaping individual
development, not just the early years. This study
has confirmed the vital role of contextual experiences on individual adjustment, and thus underlines the importance of an ecological approach
within developmental psychology.
Acknowledgments
This work was funded by the U.K. Economic and Social Research Council (Grant No.
R000238051). Previous versions of this article
were presented at the XXVII International
Congress of Psychology in Stockholm (July
2000), the 42nd Congress of the German
Psychological Society in Jena (September 2000),
and the London Conference of the British
Psychological Society (December 2000). Thanks
are due to the anonymous reviewers of an earlier
draft of this paper for their helpful and constructive comments.
Addresses and Affiliations

Corresponding author: Ingrid Schoon, Department of Psychology, City University, London,
Northampton Square, London EC1V OHB,
U.K.; e-mail: i.schoon@city.ac.uk. John Bynner,
Heather Joshi, and Samantha Parsons are at
The Institute of Education, London; Richard
D. Wiggins is also at the City University,

London; and Amanda Sacker is at Royal Free
and University College Medical School, London.
References
Ackerman, B. P., Schoff, K., Levinson, K.,
Youngstrom, E., & Izard, C. E. (1999). The relations between cluster indexes of risk and promotion and the problem behaviours of 6- and
7-year-old children from economically disadvantaged families. Developmental Psychology, 6,
13551366.
Arbuckle, J. C. (1996). Full information estimation in the presence of incomplete data. In
G. A. Marcoulides & R. E. Schumacker (Eds.),
Advanced structural equation modelling techniques. Mahwah, NJ: Erlbaum.
Arbuckle, J. C. (1999). AMOS for Windows. Analysis
of moment structures. Version 4.01 [Computer
software]. Chicago: SmallWaters.
Axinn, W., Duncan, G. J., & Thornton, A. (1997). The
effects of parents income, wealth, and attitudes on
childrens completed schooling and self-esteem.
In G. J. Duncan & J. Brooks-Gunn (Eds.),
Consequences of growing up poor (pp. 518540).
New York: Russell Sage Foundation.
Baltes, P. B. (1987). Theoretical propositions of lifespan developmental psychology: On the dynamics between growth and decline. Developmental
Psychology, 23, 611626.
Banks, M., Bates, I., Breakwell, G., Bynner, J., Emler,
N., Jamieson, L., & Roberts, K. (1992). Careers
and identities. Milton Keynes, UK: Open
University Press.
Baumrind, D. (1993). The average expectable
environment is not good enough. A response to
Scarr. Child Development, 64, 12991317.
Bentler, P. M. (1990). Comparative fit indices in
structural models. Psychological Bulletin, 107,
238246.
Birch, H. G., & Gussow, J. D. (1970). Disadvantaged
children: Health, nutrition and school failure.
New York: Grune & Stratton.
Blau, P. M., & Duncan, O. D. (1967). The American
occupational structure. New York: Wiley.
Bolger, K. E., Patterson, C. J., & Thompson, W. W.
(1995). Psychosocial adjustment among children
experiencing persistent and intermittent family
economic hardship. Child Development, 66,
11071129.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 62

Bollen, K. A. (1989). Structural equations with latent
variables. New York: Wiley.
Bollen, K. A., & Long, J. S. (Eds.). (1993). Testing structural equation models. Newbury Park, CA: Sage.
Bozdogan, H. (1987). Model section and Akaikes
Information Criteria (AIC): The general theory
and its analytical extensions. Psychometrika, 52,
345370.
Brimer, M. A., & Dunn, L. M. (1962). English Picture
Vocabulary Test. Bristol, UK: Education
Evaluation Enterprises.
Bronfenbrenner, U. (1979). The ecology of human
development: Experiments by nature and design.
Cambridge, MA: Harvard University Press.
Bronfenbrenner, U., & Ceci, S. J. (1994). Naturenurture re-conceptualized in developmental
perspective: A bioecological model. Psychological
Review, 101, 568586.
Butler, N., Despotidou, S., & Shepherd, P. (1997).
1970 British Cohort Study (BCS70) ten-year
follow-up: A guide to the BCS70 10-year data
available at the Economic and Social Research
Unit data archive. London: City University,
Social Statistics Research Unit.
Bynner, J., Elias, P., McKnight, A., & Pan, H. (1999).
The changing nature of the youth labour market
in Great Britain (Report to the Rowntree
Foundation). York, UK: Joseph Rowntree
Foundation.
Bynner, J., Ferri, E., & Shepherd, P. (Eds.). (1997).
Getting on, getting by, getting nowhere. Twentysomething in the 1990s. Aldershot, UK: Ashgate.
Bynner, J., Ferri, E., Shepherd, P., & Smith, K. (2000).
The 19992000 surveys of the National Child
Development Study and the 1970 British Cohort
Study (Working paper No. 1). London: Centre for
Longitudinal Studies.
Bynner, J., Joshi, H., & Tsatsas, M. (2000). Obstacles
and opportunities on the route to adulthood.
London: Smith Institute.
Caprara, G. V., & Rutter, M. (1995). Individual development and social change. In M. Rutter &
D. J. Smith (Eds.), Psychosocial disorders in
young people (pp. 3566). Chichester, UK: Wiley.
Campbell, F. A., & Ramey, C. T. (1994). Effects of early
intervention on intellectual and academic achievement. A follow-up study of children from lowincome families. Child Development, 65, 684698.
Campbell, S. B., Pierce, E. W., Moore, G.,
Marakovitz, S., & Newby, K. (1996). Boys
externalizing problems at elementary school age:
Pathways from early behavioural problems,

maternal control, and family stress. Development
and Psychopathology, 8, 701719.
Caspi, A., Wright, B. R. E., Moffitt, T. E., & Silva, P. A.
(1998). Early failure in the labor market:
Childhood and adolescent predictors of unemployment in the transition to adulthood.
American Sociological Review, 63, 424451.
Cicchetti, D., & Garmezy, N. (Eds.). (1993).
Milestones in the development of resilience.
Development and Psychopathology, 5, 497774.
Cicchetti, D., & Tucker, D. (1994). Development
and self-regulatory structures of the mind.
Clarke, A. D. B., & Clarke, A. M. (1981). Sleeper
effects in development: Fact or artefact.
Developmental Review, 1, 344360.
Clarke, A. D. B., & Clarke, A. M. (2000). Early experience and the life path. London: Jessica Kingsley.
Cohen, J. (1992). A power primer. Psychological
Bulletin, 112, 155159.
Conger, R. D., Conger, K. J., Elder, G. H., Lorenz, F. O.,
Simons, R. L., & Whitbeck, L. B. (1993). Family
economic stress and adjustment of early adolescent girls. Developmental Psychology, 29,
206219.
Davie, R., Butler, H., & Goldstein, H. (1972). From
birth to seven: The second report of the National
Child Development Study (1958 Cohort).
London: Longman.
Deater-Deckard, K., Dodge, K. A., Bates, J. E., &
Pettit, G. S. (1998). Multiple risk factors in the
development of externalising behavior problems:
Group and individual differences. Development
and Psychopathology, 10, 469494.
Duncan, G. J., & Brooks-Gunn, J. (Eds.). (1997).
Consequences of growing up poor. New York:
Russell Sage Foundation.
Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K.
(1994). Economic deprivation and early childhood development. Child Development, 65,
296318.
Duncan, G. J., & Rodgers, W. L. (1988). Longitudinal
aspects of poverty. Journal of Marriage and the
Family, 50, 10071021.
Duncan, G. J., Yeung, W. J., Brooks-Gunn, J., &
Smith, J. R. (1998). How much does childhood
poverty affect the life chances of children?
Ekinsmyth, C., Bynner, J., Montgomery, S., &
Shepherd, P. (1992). An integrated approach to
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 63

the design and analysis of the 1970 British Cohort
Study (BCS70) and the National Child Development Study (NCDS) (Working paper No. 1).
London: City University, Social Statistics Research
Unit Cohort Studies.
Elder, G. H., Jr. (Ed.). (1985). Life course dynamics.
Ithaca, NY: Cornell University Press.
Elder, G. H., Jr. (1999). Children of the Great
Depression. Social change in life experience (25th
anniversary ed.). Boulder, CO: Westview Press.
Elder, G. H., Pavalko, E. K., & Hastings, T. H. (1991).
Talent, history and the fulfillment of promise.
Psychiatry, 54, 251267.
Featherman, D. L., & Lerner, R. M. (1985).
Ontogenesis and sociogenesis. Problematics for
theory about development across the life-span.
Feiring, C., & Lewis, M. (1996). Finality in the
eye of the beholder: Multiple sources, multiple
time points, multiple paths. Development and
Psychopathology, 8, 721733.
Felner, R. D., Brand, S., DuBois, D. L., Adan, A.,
Mulhall, P. F., & Evans, E. G. (1995). Socioeconomic disadvantage, proximal environmental
experiences, and socioemotional and academic
adjustment in early adolescence: Investigation
of a mediated effects model. Child Development,
66, 774792.
Fergusson, D. M., Horwood, L. J., & Lawton, J. M.
(1990). Vulnerability to childhood problems and
family social background. Journal of Child
Psychology and Psychiatry, 31, 11451160.
Fergusson, D. M., Horwood, L. J., & Lynskey, M. T.
(1994). The childhoods of multiple problem adolescents: A 15-year longitudinal study. Journal of
Child Psychology and Psychiatry, 35, 11231140.
Fitzgerald, H. E., Lester, B. M., & Zuckerman, B. S.
(1995). Children of poverty: Research, health,
and policy issues. New York: Garland.
Fogelman, K. (Ed.). (1983). Growing up in Great
Britain. Papers from the National Child
Development Study. London: Macmillan (for the
National Childrens Bureau).
Garmezy, N. (1985). Stress-resistant children: The
search for protective factors. In A. Davids (Ed.),
Recent research in developmental psychopathology (pp. 213233). Elmsford, NY: Pergamon.
Garmezy, N. (1991). Resilience and vulnerability to
adverse developmental outcomes associated
with poverty. American Behavioural Scientist,
34, 416430.
Godfrey Thompson Unit, University of Edinburgh.

(1978). Edinburgh Reading Test. Sevenoaks, UK:
Hodder and Stoughton.
Goldstein, H. (1979). Some models for analyzing longitudinal data on educational attainment. Journal
of the Royal Statistical Society (Series A), 142,
407442.
Gottlieb, G., Wahlsten, D., & Lickliter, R. (1998). The
significance of biology for human development:
A developmental psychobiological systems view.
In R. M. Lerner (Ed.), W. Damon (Series Ed.),
Handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed.,
pp. 233274). New York: Wiley.
Goodenough, F. (1926). Measurement of intelligence by
drawings. New York: Harcourt, Brace & World.
Greenberg, M. T., Lengua, L. J., Coie, J. D., &
Pinderhughes, E. E. (1999). Predicting developmental outcomes at school entry using a multiple-risk model: Four American communities.
Developmental Psychology, 35, 403417.
Hammen, C. (1992). Cognitive, life stress, and interpersonal approaches to a developmental psychopathological model of depression. Development and
Psychopathology, 4, 189206.
Harris, D. B. (1963). Childrens drawings as measures
of intellectual maturity. New York: Harcourt,
Brace & World.
Haveman, R., & Wolfe, B. (1994). Succeeding generations: On the effects of investments in children.
New York: Russell Sage Foundation.
Hoffman, L. W. (1994). A proof and a disproof questioned. Social Development, 3, 6063.
Horowitz, F. D. (2000). Child development and the PITS:
Simple questions, complex answers, and developmental theory. Child Development, 71, 110.
Huston, A. C., McLoyd, V. C., & Coll, C. G. (1994).
Children and poverty: Issues in contemporary
research. Child Development, 65, 275282.
Jackson, J. F. (1993). Human behavioral genetics.
Scarrs theory and her views on interventions: A
critical review and comments on their implications for American children. Child Development,
63, 13181332.
Koppitz, E. M. (1968). Psychological evaluation of
childrens human figure drawings. New York:
Grune & Stratton.
Krieger, N., & Williams, D. R. (1997). Measuring
social class in US public health research:
Concepts, methodologies and guidelines. Annual
Review of Public Health, 18, 341378.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 64

Leete, R., & Fox, J. (1977). Registrar Generals social
classes: Origins and users. Population Trends, 8,
17.
Lerner, R. M. (1984). On the nature of human plasticity. New York: Cambridge University Press.
Lerner, R. M. (1996). Relative plasticity, integration,
temporality, and diversity in human development:
A developmental contextual perspective about
theory, process, and method. Developmental
Lerner, R. M., & von Eye, A. (1992). Sociobiology
and human development: Arguments and evidence. Human Development, 35, 1233.
Luthar, S. S., Cicchetti, D., & Becker, B. (2000). The
construct of resilience. A critical evaluation and
guidelines for future work. Child Development,
71, 543562.
Marsh, C. (1986). Social class and occupation. In
R. Burgess (Ed.), Key variables in social investigation (pp. 123152). London: Routledge.
Masten, A. S., Best, K. M., & Garmezy, N. (1990).
Resilience and development: Contributions from
the study of children who overcome adversity.
McLoyd, V. C. (1990). The impact of economic hardships on black families and children: Psychological
distress, parenting, and socioemotional development. Child Development, 61, 311346.
OPCS (Office of Population Censuses and Surveys)
and Employment Department Group. (1980).
Standard Classification of Occupations (SOC).
London: Her Majestys Stationery Office.
OPCS (Office of Population Censuses and Surveys)
and Employment Department Group. (1990).
Standard Classification of Occupations (SOC).
London: Her Majestys Stationery Office.
Osborn, A. F., Butler, N. R., & Morris, A. C. (1984).
The social life of Britains five-year-olds. London:
Routledge.
Pilling, D. (1990). Escape from disadvantage. London:
Palmer Press.
Plomin, R., & Bergeman, C. S. (1991). The nature of
nurture: Genetic influence on environmental
measures. Behavioral and Brain Sciences, 14,
373427.
Plomin, R., & Daniels, D. (1987). Why are children in
the same family so different from one another?
Behavioral and Brain Sciences, 10, 116.
Plomin, R., & McClearn, G. E. (Eds.). (1993). Nature,
nurture and psychology Washington, DC: American
Psychological Association.
Prandy, K. (1990). The revised Cambridge Scale of

occupations. Sociology, 24, 629655.
Pringle, M. L. K., Butler, N. R., & Davie, R. (1966).
11,000 seven year olds. London: Longman.
Pungello, E. P., Kupersmidt, J. B., Burchinal, M. R., &
Patterson, C. J. (1996). Environmental risk factors
and childrens achievement from middle childhood to early adolescence. Developmental
Ramey, C. T., & Ramey, S. L. (1990). Intensive education intervention for children of poverty.
Intelligence, 14, 19.
Robins, L. N., & Rutter, M. (Eds.). (1990). Straight
and devious pathways from childhood to adulthood. Cambridge, UK: Cambridge University
Press.
Rutter, M. (1979). Protective factors in childrens
responses to stress and disadvantage. In
M. W. Kent & J. E. Rolf (Eds.), Primary prevention in psychopathology (Vol. 3, pp. 4974).
Hanover, NH: University Press of New England.
Rutter, M. (1990). Psychosocial resilience and protective mechanisms. In J. Rolf, A. S. Masten,
D. Chichetti, K. H. Nuechterlin, & S. Weintraub
(Eds.), Risk and protective factors in the development of psychopathology (pp. 181214). New
York: Cambridge University Press.
Rutter, M., & Madge, N. (1976). Cycles of disadvantage: A review of research. London: Heinemann
Educational Books.
Rutter, M., Maughan, B., Mortimore, P., & Ouston, J.
(1979). Fifteen thousand hours. Secondary
schools and their effects on children. London:
Open Books.
Sameroff, A. J. (1983). Developmental systems:
Contexts and evolution. In W. Kessen (Ed.),
P. H. Mussen (Series Ed.). Handbook of child
psychology: Vol. 1. History, theory and methods
(pp. 237294). New York: Wiley.
Sameroff, A. J., Seifer, R., Baldwin, A., & Baldwin, C.
(1993). Stability of intelligence from preschool
to adolescence: The influence of social and
family risk factors. Child Development, 64,
8097.
Scarr, S. (1992). Developmental theories for the
1990s: Development and individual differences.
Child Development, 63, 119.
Schoon, I., McCulloch, A., Joshi, H., Wiggins, R. D.,
& Bynner, J. (2001). Transitions from school
to work in a changing social context. Young, 9,
423.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 65

Schoon, I., & Parsons, S. (2002a). Teenage aspirations
for future careers and occupational outcomes.
Journal of Vocational Behavior, 60, 262288.
Schoon, I., & Parsons, S. (2002b). Competence in the
face of adversity: The impact of early family
environment and long-term consequences.
Children & Society, 16, 260272.
Schulenberg, J., Vondracek, F. W., & Crouter, A. C.
(1984). The influence of the family on vocational
development. Journal of Marriage and the
Family, 10, 129143.
Scott, L. H. (1981). Measuring intelligence with the
Good-enough-Harris drawing test. Psychological
Bulletin, 89, 483505.
Sewell, W. H., Hailer, A. O., & Ohlendorf, G. W.
(1970). The educational and early occupational
status attainment process: Replication and
revision. American Sociological Review, 35,
10141027.
Shepherd, P. (1993). Analysis of response bias. In
E. Ferri (Ed.), Life at 33. The fifth follow-up
of the National Child Development Study
(pp. 184188). London: National Childrens
Bureau and City University.
Shepherd, P. (1995). The National Child Development
Study. An introduction, its origins and the
methods of data collection (Working paper
No. 1). London: City University, Social Statistics
Research Unit.
Shepherd, P. (1997). Survey and response. In J. Bynner,

E. Ferri, & P. Shepherd (Eds.), Getting on, getting
by, getting nowhere. Twenty-something in the
1990s (pp. 129136). Aldershot, UK: Ashgate.
Southgate, V. (1962). Southgate Reading Tests:
Manual of instructions. London: University of
London Press.
Sroufe, L. A., Egeland, B., & Kreuzer, T. (1990). The
fate of early experience following developmental
change: Longitudinal approaches to individual
adaptation in childhood. Child Development, 61,
13631373.
Steiger, J. H. (1990). Structural model evaluation
and modification: An internal estimation approach.
Multivariate Behavioural Research, 25, 173180.
Szatmari, P., Shannon, H. S., & Offord, D. R. (1994).
Models of multiple risk: Psychiatric disorder and
poor school performance. International Journal
of Methods in Psychiatric Research, 4, 231240.
Tizard, J. (1976). Psychology and social policy.
British Psychological Society Bulletin, 29,
225234.
Walker, D., Greenwood, C., Hart, B., & Carta, J.
(1994). Prediction of school outcomes based on
early language production and socio-economic
factors. Child Development, 65, 606621.
Werner, E. E., & Smith, R. S. (1992). Overcoming the
odds: High risk children from birth to adulthood.
Ithaca, NY: Cornell University Press.
66
.23
.16
.37
.35
.24
.16
.23
.28
.17
.17
.24
.12
.32
.30
.29
.63
.32
.18
.22
.19
.13
.58
.25
.21
.20
.35
.62
.13
.16
.10
.10
.34
.39
.72
.34
.23
.53
.56
.35
.33
.34
.18
.33
.22
.45
.58
.33
.32
.33
.17
.37
.21
.25
.15
.15
.69
.28
10
.14
.17
.10
.10
.37
.48
11
.51
.31
.30
.22
.11
12
.34
.32
.27
.14
13
.36
.16
.08
14
.16
.08
15
.30
16
1.17
3.62
.86 16.74 16.06
3.68
.84
5.13 23.40 23.83
7.02
3.77
.81
.83
1.27
.96 10.29
6.21
1.27
.89
2.48
7.10 7.07
2.07
1.24
.72
9,114 10,320 10,665 13,125 13,129 12,605 12,790 14,065 14,098 13,839 14,056 13,749 13,756
.77
.26
.18
.44
.41
.28
.18
.27
.33
.20
.19
.28
.14
NOTE: RGSC = Registrar General Social Class; CS = Cambridge Score.
17.78
18.14
9,114
.28
.24
.23
.16
.35
.34
.25
.16
.23
.27
.17
.16
.25
.13
17
3.83
3.34 33.86 16.74
1.13 3.62
.87 16.65 16.00
3.69
.85
.85 23.38 23.81
7.02 3.77
.82
1.23
1.33 19.26 18.22
.84 1.28
.97 10.35
6.25
1.28
.89
2.48 7.11
7.07
2.01 1.24
.72
16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994 16,994
.81
.26
.22
.21
.15
.32
.31
.23
.15
.21
.25
.15
.15
.23
.12
FIML estimates
M
SD
N
.21
.23
.26
.22
.49
.35
.30
.29
.56
.37
.18
.21
.13
.13
.59
.30
3.24 34.98
1.31 19.11
10,693 10,838
Parental RGSC birth

Own RGSC at age 33
Own CS at age 33
Exam score at age 16
Highest qualifications at age 16
Parental RGSC at age 16
Material conditions at age 16
Math test at age 11
Reading test at age 11
Arithmetic test at age 7
Reading test at age 7
Drawing test at age 7
Copy test at age 7
5:33 PM
Observed variables
M 3.83
SD 1.23
N 16,994
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
11/16/2004
Bivariate Correlations, Means, SDs, and Sample Sizes for the Variables in the 1958 National Child Development Dataset, Including the Full
Information Maximum Likelihood (FIML) Estimates from the AMOS Modeling
APPENDIX A
03-Chibucos.qxd
Page 66
.14
.07
.15
.18
.10
.23
.11
.19
.23
.12
12. Reading test at age 5
13. Drawing test at age 5
14. Copy test at age 5
.15
.31
.20
.39
.31
.18
.30
.73
.16
.31
.24
.38
.33
.18
.30
.26
.66
.13
.23
.19
.33
10
.42
.32
.08
.14
.11
11
.76 43.98 40.25

3.48
.86
1.06 12.33 12.64
1.26
1.05
9,032 11,251 11,258 12,730 10,051
.32
.26
.06
.11
.10
.44
.27
.15
.15
.13
.26
.36
13
.07
.14
14
.30
15
16
32.78 11.34
4.73
3.55
.65
9.76
3.28
1.97
1.27
.92
9,805 12,177 12,355 11,929 12,056
.11
.21
.21
.30
12
2.98 48.41 18.00

1.32
3.48
.79 43.91 40.17
3.50
.87 33.18 11.29
4.70
3.58
.66
1.25 17.13 16.28
.70
1.26
1.08 12.40 12.71
1.27
1.05
9.83
3.29
1.98
1.28
.92
14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229 14,229
3.25
1.26
6,590
.21
.52
.10
.19
.15
.26
.56
.24
.24
.24
3.77
1.20
14,229
1.40
.69
6,457
.10
.20
.10
.19
.16
.12
.20
.30
.29
.10
.18
2.97
1.20
8,002
21.94
16.65
5,603
.14
.29
.15
.27
.20
.17
.28
.41
.41
.15
.25
.53
3.78
1.20
14,229
49.44
16.12
5,183
.11
.09
.16
NOTE: RGSC = Registrar General Social Class; CS = Cambridge Score.
Observed variables
M
SD
N
FIML estimates
M
SD
N
.30
.06
.13
.33
11. Material conditions at age 10
.08
.22
.55
10. Parental RGSC at age 10
.60
.13
.27
.29
9. Reading test at age 10
.16
.27
.28
8. Math test at age 10
.07
.11
.27
.20
5. Highest qualifications at age 16
.13
.11
.19
.28
.44
.21
.13
4. Exam score at age 16
.30
.20
.22
3. Own CS at age 30
.18
5:33 PM
2. Own RGSC at age 30
1. Parental RGSC birth
11/16/2004
Bivariate Correlations, Means, SDs, and Sample Sizes for the Variables in the 1970 British Birth Cohort Dataset, Including the Full Information
Maximum Likelihood (FIML) Estimates from the AMOS Modeling
APPENDIX B
03-Chibucos.qxd
Page 67
67
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 68
MARITAL STATUS CONTINUITY AND

CHANGE AMONG YOUNG AND MIDLIFE
ADULTS: LONGITUDINAL EFFECTS ON
PSYCHOLOGICAL WELL-BEING
NADINE F. MARKS AND JAMES DAVID LAMBERT
Abstract
Using a life course theoretical framework, this
study examined longitudinal effects of continuity
and transitions in marital status on multiple
dimensions of psychological well-being. Data
came from National Survey of Families and
Households 19871993 respondents ages 19 to 65
(N = 6,948). Differences between men and women
as well as between young and midlife adults were
investigated. Multivariate analyses revealed a
complex pattern of effects depending on the contrast and the outcome examined. Although marriage continued to promote well-being for both
men and women, in some casesfor example,
autonomy, personal growththe single fared
better than the married. The effects of continuity
in single status were not very different for women
in contrast to men. The transition to divorce or
widowhood was associated with somewhat more
negative effects for women. Midlife adults evidenced more psychological resilience than young
adults did in facing the challenges of a marital
transition or remaining single over time.
The social institution of marriage and its
influence on adult well-being remains an enduring interest of family researchers. Historically,
marriage has been quite consistently associated
68
with better psychological well-being than being

single, particularly among men (e.g., Gove,
Hughes, & Style, 1983; Gove & Shin, 1989;
Gove, Style, & Hughes, 1990; Lee, Seccombe, &
Shehan, 1991; Ross, Mirowsky, & Goldsteen,
1990). However, there continues to be an ongoing reassessment of the role marriage plays in
determining well-being as dramatic changes in the
norms, meaning, and dynamics of marriage and
marital stability have swept across America
during the past several decades. Now that about
one in every two new marriages ends in divorce
(Castro-Martin & Bumpass, 1989)sexuality
and even parenthood are increasingly less tied to
marriage (Bumpass, 1990); gendered aspects of
marital, parenting, and employment roles have
come under increased scrutiny and influence in
marital choice and satisfaction (Goldscheider &
Waite, 1991); and the prevalence of single adults
and the proportion of the adult lifetime spent
single has increased to make it a statistically less
deviant adult social status (Schoen & Weinick,
1993; Schoen, Urton, Woodrow, & Baj, 1985)
it might be hypothesized that the importance of
marriage for contemporary adults psychological
well-being is changing (Glenn & Weaver, 1988).
The life course and life span developmental perspectives (Baltes, 1987; Elder, 1992; Featherman,
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 69
1983) suggest that human development is lifelong

and that it is important to examine the sequelae
of continuities and changes in adulthood as well
as childhood. Child developmentalists track the
importance of continuity and change (e.g., loss) in
the primary attachment tie with a parent (usually
the mother) for a childs well-being (Bowlby,
1969, 1973, 1980; Bretherton, 1992). During
adulthood, continuity and change in the primary
attachment with a marital partner might also be
expected to have important consequences for adult
well-being.
Although marriage has been generally associated with better mental health, most of the evidence for the positive effects of marriage on
psychological well-being is based on crosssectional evidence, samples with limited generalizability, or both. A few longitudinal studies of
marriage have been used to confirm that the transition to divorced status has negative effects on wellbeing (e.g., Booth & Amato, 1991; Doherty, Su, &
Needle, 1989; Mastekaasa, 1995; Menaghan &
Lieberman, 1986). Similarly, the psychological
distress accompanying the adjustment to widowhood has been confirmed longitudinally (Stroebe
& Stroebe, 1987; Wortman, Silver, & Kessler,
1993; see Kitson, Babri, Roach, & Placidi, 1989,
for a review). However, no large-scale national
longitudinal analysis has simultaneously examined the effects of continuity in varying marital
statuses and varying types of marital status
change on well-being. Specifically, longitudinal
population analyses comparing the well-being
effects of the transition from being never married
to first married, formerly married to remarried,
and married to widowed with being continuously
married do not exist. A systematic examination of
gender differences in the effects of these different
marital statuses and transitions has never been
undertaken. In addition, multiple well-being
contrasts among persons remaining continuously
in a marital status or making a marital transition in
young adulthood, in contrast to middle adulthood,
have never been carefully examined, even though
a considerable number of marital transitions
occur during both periods (Uhlenberg, Cooney, &
Boyd, 1990). Taking a life course or life span
developmental perspective, we might expect that

differential timing of a transition would make a
difference in its effect on well-being (Hagestad,
1990; Hagestad & Smyer, 1982; Neugarten,
1979).
Limited outcome measures of psychological
well-being plague most studies of gender,
marital status, and psychological well-being.
Depression, life satisfaction, and global happiness
are the most common outcomes examined.
However, the multiple dimensions of psychological well-being are becoming increasingly well
mapped and well measured (Bryant & Veroff,
1982; Ryff, 1989, 1995; Ryff & Essex, 1991; Ryff
& Keyes, 1995). For a more differentiated and
comprehensive understanding of the contemporary effects of the marital role on well-being, it is
desirable to consider several dimensions of psychological well-being, because marriage may be
associated with well-being constraints as well as
well-being enhancement.
The focus of this research was to use recent
longitudinal national survey data, which included
measurement across a wide range of positive
and negative psychological well-being dimensions, to examine the effects of marital status continuity and marital status transitions (change) on
psychological well-being, and to examine gender
differences and age (young adult vs. midlife
adult) differences in these effects.
Theoretical and
Empirical Background
Continuity and Change in Life
Course and Life Span Development
The life course and life span developmental
perspectives suggest that adult development
is characterized by a complex interplay of continuity and change no less than child development
(Baltes, 1987; Elder, 1992; Featherman, 1983).
Family life transitions as well as family life continuity are important components of the process
that helps constitute adult development and
adult well-being (Bengtson & Allen, 1993;
Elder, 1991). Social context and how it changes
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 70
over time (Riley, Foner, & Waring, 1988) also

help determine the social meanings, rewards,
and sanctions for family life continuities and
changes, which are critical for determining the
developmental impact of continuity and transitions in family roles.
In addition, the life course and life span perspectives suggest that social clocks help determine an expectable normative sequencing of
events for the life course (Bengtson & Allen,
1993). Transitions that are socially normative
and on time are expected to be more easily
incorporated into ones identity and more supported by social institutions, therefore yielding
more beneficial effects on well-being than transitions that are non-normative or off time
(Hagestad, 1990; Hagestad & Smyer, 1982;
Neugarten, 1979). The social script for young
adulthood in most societies, including 20th-century America, has included entry into marriage.
Becoming married as a young adult would be
expected to have beneficial effects due to its
fulfillment of normative expectations and its
societal support.
Continuity, too, can have significant developmental effects. For example, in a social context
where remaining never married is considered
deviant, remaining a never-married person beyond
the socially normative and statistically normative
age of first marriage (early to mid-20s in the
contemporary United States) (Schoen & Weinick,
1993) might be expected to cause a decline in
well-being and development. However, as a larger
proportion of young adults are remaining single
longer (Schoen & Weinick, 1993) and the modern
social order provides them with more opportunities and encouragement to live autonomously
(Goldscheider, Thornton, & Young-DeMarco,
1993), remaining never married may no longer
carry such social stigma and negative effects.
Dissolution of marriage by divorce has, at
least until recently, been viewed as an unanticipated and non-normative transition in the adult
life course. The economic and social strains
associated with divorce (McLanahan & Sandefur,
1994; Pearlin & Johnson, 1977), coupled with its
being viewed as non-normative, might well lead
to the expectation that both the transition to

divorce and continuity over time in this status
would lead to poorer well-being and development. Yet again, as the texture of social life
changes and as more adults have become
divorced and have remained divorced longer
(Schoen & Weinick, 1993), the social stigma
once associated with divorce that helped lead to
a decline in well-being among previous divorce
cohorts may have diminished for contemporary
divorce cohorts.
Widowhood can be anticipated in older age,
but given current life expectancy and cumulative
survival rates (Schoen & Weinick, 1993), contemporary adults view widowhood in young or
middle age to be a non-normative transition and
a non-normative status, and therefore we would
expect continuity in this marital status or a
transition to this marital status to be deleterious
to well-being. However, given the relative absence
of stigma and generally better social welfare
(e.g., Social Security) associated with being a
widow in contrast to being divorced, we might
expect that widows would do less poorly than
divorces.
Entry into first marriage for an American
adult might be expected to lead to an increase in
well-being, because it fulfills one social expectation for the adult life course and tends to be
associated with economic and social support
advantages (Ross et al., 1990). Entry into remarriage, although now relatively common, remains
incompletely institutionalized (Cherlin, 1978)
and may therefore be less beneficial for wellbeing than entry into first marriage. The life
course of the person entering remarriage has
included a history of loss or disappointment
to some extent, and remarriage often includes
a complex reordering of relationships with
children and other kin (Pasley & IhingerTallman, 1987).
Another way in which timing of continuity or
a transition might be considered important from
a life course or life span perspective concerns the
developmental readiness and resources a person
has to deal with a transition or time spent in a
status. For example, the transition to fatherhood
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 71
after age 30 in contrast to earlier in adulthood

has been found to be associated with greater
confidence in the paternal role (Nydegger, 1986),
more positive paternal affect (Cooney, Pederson,
Indelicato, & Palkovitz, 1993), more paternal
involvement (Cooney et al., 1993; Daniels &
Weingarten, 1982; Heath, 1994), and a more
effective balancing of work and family demands
(Coltrane, 1990; Frankel & Wise, 1982). These
differences have been hypothesized to be the
result of greater psychological maturity, selfknowledge, and life-management skills possessed by midlife fathers in contrast to young
adult fathers.
Differences in the effects of marital status
continuity and transitions at different developmental periods during adulthood have not been
carefully studied previously. Age is most often
included as a control variable but is not considered a potential moderator of effects. Although
a prototypical version of normative family
development (e.g., Duvall, 1957) might lead us
to expect that transitions into marriage occur in
young adulthood, transitions out of marriage
(through death) occur in later adulthood, and
middle adulthood is characterized by marital
continuity for the vast majority of persons, the
complexity of the modem marital career for
a sizable number of contemporary adults now
includes transitions both into and out of marriage
in middle adulthood (Bumpass, 1990; Bumpass,
Sweet, & Martin, 1989; Uhlenberg et al., 1990).
Marital status continuity or change might
have a different impact in middle adulthood in
contrast to younger adulthood for a variety of
reasons. One reason is the relative age normativeness of the transition or marital status as
noted above. Another reason is the relative
resources each age period provides for adaptation. Young adults might have a well-being
advantage in the face of marital loss due to their
better prospects of remarriage (Bumpass et al.,
1989). On the other hand, differences in psychological maturity and pragmatic life expertise
accrued over time (Baltes, 1987; Baltes &
Staudinger, 1993; Brim, 1992) might make it
easier for midlife adults than for younger adults
to handle the changes associated with, for

example, the loss of a spouse due to divorce or
death.
The Multiple Dimensions
of Psychological Health
Bradburns (1969) analyses of positive and
negative affect provided some of the first empirical evidence that positive and negative wellbeing were related, yet distinct, components of
psychological well-being. Empirical evidence
for the legitimacy of differentiating positive
aspects of psychological well-being and psychological distress was further supported by factor
analytic work done pooling national survey
items by Bryant and Veroff (1982). The results of
their analysis led them to conclude that psychological well-being, as measured in the national
surveys of previous decades, included three distinct components: positive affect, psychological
distress, and self-evaluation.
Ryff (1989, 1995), a life span developmentalist, questioned the adequacy of traditional
positive measures of psychological well-being
(e.g., one-item assessments of happiness and life
satisfaction), which have little developmental or
theoretical basis, to cover the range of positive
mental health and wellness. Drawing from several human development theories, Ryff (1989)
generated and provided evidence of discriminant
validity (in relation to each other as well as in
comparison to prior measures of well-being) of
six new measures of distinct dimensions of psychological wellness: positive evaluation of oneself and ones past life (self-acceptance), a sense
of continued growth and development as a
person (personal growth), the belief that ones
life is purposeful and meaningful (purpose in
life), the possession of quality relations with
others (positive relations with others), the capacity to manage effectively ones life and surrounding world (environmental mastery), and a
sense of self-determination (autonomy). Further
confirmatory factor analyses undertaken by Ryff
and Keyes (1995), using national data, provided
additional evidence that these six components
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 72
of well-being are more appropriately considered

different dimensions of wellness than subscales of a single wellness factor; they found
that a model positing six separate factors fit
better than a model positing one global latent
factor.
Examining more complete measurements of
psychological wellness, as well as psychological
distress, offers the potential to reveal a more
comprehensive understanding of the complex
psychological effects stemming from involvement
in significant and often conflictual social roles,
including marriage and parenting (see Umberson
& Gove, 1989, for an illustration of this point in
relation to parenthood status). Therefore, Ryffs
(1989) six dimensions of psychological wellbeing were examined in this analysis along with
the more familiar dimensions of positive affect
(global happiness), self-adequacy (self-esteem
[distinct from self-acceptance, see Ryff, 1989]),
personal mastery (distinct from environmental
mastery, see Ryff, 1989), and psychological distress (depression, hostility).
Marital Status Continuity
and Psychological Well-Being
Most studies examining marital status and
psychological distress have concluded that
married men and women have a mental health
advantage in contrast to their unmarried peers
(Gore & Mangione, 1983; Gove et al., 1983,
1990; Mirowsky & Ross, 1989; Pearlin &
Johnson, 1977). Single, formerly married
personsdivorced and widowedtypically
report poorer well-being and give evidence of
more distress than never-married persons (Gove
& Shin, 1989; Pearlin & Johnson, 1977).
Research examining whether the psychological
benefits of marriage are greater for men or women
has yielded mixed results. Gove and Tudor (1973)
found that marriage protected the mental health of
men more than women. Yet Fox (1980), using data
from three national surveys from 1960, 1970, and
1973, did not find strong support for a gender by
marital status interaction effect. A recent analysis
by Ross (1995) found no gender differences.
However, all of these investigations examined

psychological distress or other psychological
dysfunction as outcomes. There is less evidence
about potential gender differences in positive psychological well-being associated with marriage.
The one positive psychological well-being
outcome that has been extensively studied in
relation to the marital role and across both men
and women is global happiness. Being married
has been consistently associated with more
global happiness (Glenn, 1975; Glenn & Weaver,
1979, 1988; Lee et al., 1991). However, national
trend data from the General Social Survey (GSS)
spanning the 1970s and the 1980s examined by
Glenn and Weaver (1988) revealed a narrowing
of the happiness gap between the married and
the never-married during these years. This trend
was noted particularly for men and for younger
adults (ages 25 to 39). The proportion of nevermarried men indicating they were very happy
increased between 1972 and 1982, whereas the
proportion of younger married women indicating
such high levels of positive well-being decreased
(Glenn & Weaver, 1988). Lee et al. (1991)
extended the analysis of the GSS to 1989 and
found that the gap increased somewhat during
1987 and 1988 but then diminished again in
1989. As before, the changes found in happiness
by marital status were most pronounced among
young adults; specifically, younger nevermarried men and women reported more happiness in the 1980s than in the 1970s, and younger
married women reported less happiness in the
1980s than in the 1970s. Anderson and Stewart
(1994) and Gordon (1994) have also reported
evidence from their recent qualitative studies that
single women report advantages to single status
over marriage in terms of personal autonomy and
growth.
In their meta-analysis of studies of marital
status and well-being, Harding-Hidore, Stock,
Okun, and Witter (1985) found evidence of only
a small positive association between marriage
and subjective well-being. The effects of marriage were smaller for older persons, and they
were also smaller for younger cohorts. The
results of this meta-analysis suggest that it is
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 73
important to examine age differences in the

importance of marriage and marital transitions
and to continue to periodically evaluate associations between marital status and well-being at
different points in time.
Marital Status Transitions and Well-Being
A few longitudinal population-based studies
on the mental health effects of the transition
from marriage to divorce have been conducted.
Menaghan and Lieberman (1986) used a probability sample of more than 1,000 adults from the
Chicago area followed over 4 years (1972 to
1976) to examine the impact of divorce on
change in depressive affect. These researchers
found that, in fact, divorce led to an increase in
depressive affect; greater economic problems,
unavailability of confidants, and a reduction in
living standards accounted for a substantial
amount of the decline in well-being. No difference in change was found for men in contrast to
women. A major strength of this study was its
prospective design. However, it was limited to
an examination of one measure of psychological
well-being (depressive affect), it investigated
only one type of marital transition, and it is now
a story two decades old.
Doherty et al. (1989) conducted a 5-year
(1982 to 1987) longitudinal study of 402 predominantly White, middle-class, middle-aged
couples with teenage children randomly selected
from the enrollment of a Minnesota health maintenance organization in 1982. They found that
women who were separated or divorced during
the study period experienced a decline in psychological mood and an increase in substance
abuse. The transition to dissolution did not result
in declines in well-being for men (although men
who separated or divorced rated lower than continuously married men on psychological mood,
self-esteem, mastery, and substance abuse both
before and after dissolution).
Booth and Amato (1991) analyzed data from
a U.S. national sample of more than 2,000 married
people ages 55 and younger in 1980 who were
followed up longitudinally in 1983 and 1988.
Their analysis of patterns from three time periods

led them to conclude that divorce was associated with a short-term (i.e., less than 2 years postevent), but not long-term (i.e., more than 2 years
postevent), increase in psychological distress and
unhappiness (each outcome measured with a single-item indicator) and that these patterns were
similar for men and women.
Mastekaasa (1995) recently examined national
Norwegian data for 930 persons married in 1980
or 1983 who were also reinterviewed at least
twice subsequent to their initial interview. He
found that persons who separated or divorced
over the longitudinal follow-up period experienced a significant increase in psychological distress (measured with two items), both short-term
(less than 4 years postevent) and long-term (4 to
8 years postevent).
The transition to remarriage has received less
attention. Cross-sectional evidence suggests that
remarried men may be somewhat happier than
once-married men, but remarried women are less
happy than once-married women (White, 1979).
Spanier and Furstenberg (1982) examined the
transition to remarriage longitudinally (1977 to
1979) for their sample of 180 Pennsylvania
respondents. They found that remarriage alone
did not account for well-being differences
between the group that remarried in contrast to
the group that remained divorced during the
period they studied.
The first few months after the death of a
spouse have been consistently associated with
higher levels of depressive symptomology
(Harlow, Goldberg, & Comstock, 1991; Stroebe
& Stroebe, 1987; Wortman, Silver, & Kessler,
1993). Although not totally consistent, current
evidence suggests that widowhood may be more
psychologically problematic for men than for
women (e.g., Gove, 1972; Siegel & Kuykendall,
1990; Stroebe & Stroebe, 1987, 1993) and for
younger widows in contrast to older widows
(e.g., Ball, 1977; Sanders, 1981). Longitudinal
evidence indicates that postbereavement depression effects may be short-lived (1 to 2 years) and
that long-term differences in depression between
widows and married persons may be minimal
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 74
(Harlow et al., 1991; McCrae & Costa, 1993;

Sanders, 1981).
However, most research on widowhood has
been limited by the use of convenience samples
or longitudinal studies begun after the transition
to widowhood, by the use of limited measures of
well-being (e.g., depression only), by a limited
age range (e.g., only the elderly), or by limitation
only to womenmaking gender comparisons
impossible (Kitson et al., 1989). This research
project sought to add to the literature on the
effects of widowhood by (a) using a national
sample, (b) investigating the transition to widowhood prospectively, (c) analyzing the effects
of widowhood on multiple dimensions of wellbeing, (d) exploring gender differences in the
effects of widowhood, and (e) considering a
young adult versus midlife adult contrast for
women (in much of the existing literature,
young widows are actually midlife women).
Research Hypotheses and Questions
In sum, based on life course and life span
developmental theory and previous research, this
study was designed to examine two hypotheses.
Because marriage is a socially normative life
course role for young and midlife adults:
(a) Young and midlife men and women continuously unmarried (separated/divorced, widowed, or
never married) over a 5-year period will evidence
a decline in well-being in contrast to men and
women continuously married; and (b) transitions
out of marriage will lead to a decline in well-being
in contrast to remaining continuously married.
Because research evidence is limited regarding the well-being effects of the transition into
first marriage and remarriage (in contrast to
remaining continuously married), inconsistent
regarding gender by marital status interaction
effects, and scant regarding age by marital status
interaction effects, we also explored three
research questions: (a) Is the transition into
first marriage and remarriage associated with
increased well-being in contrast to remaining
continuously married? (b) Are there gender
differences in the psychological well-being
effects of marital status continuity or change?

and (c) Are there adult age (i.e., young adult vs.
midlife adult) differences in the psychological
well-being effects of marital status continuity or
change?
Methods
Data
The data for these analyses came from the
first and second waves of the National Survey
of Families and Households (NSFH), which
includes information from personal interviews
conducted in 19871988 (Time 1) and in
19921993 (Time 2; 5 years later), with a nationally representative sample of 13,008 noninstitutionalized American adults, 19 years old and
older. This survey included a main sample of
9,643 respondents, with an additional over sample
of 3,374 African Americans, Mexican Americans,
Puerto Ricans, single parents, stepparents,
cohabitors, and recently married persons. The
response rate at Time 1 (19871988) was about
75%. The response rate at Time 2 was about 82%
of first wave respondents. This yielded national
population coverage at a rate of about 62% for
data from both waves. Sampling weights correcting for selection probabilities and nonresponse
allow this sample to match the composition of
the U.S. population on age, sex, and race (see
Sweet, Bumpass, & Call, 1988, for more design
details). The analytic sample for this study consisted of NSFH primary respondents ages 19 to
34 or 40 to 60 in 19871988, who also responded
in 19921993, and who had complete and consistent marital status information for the period
between the two waves of the survey (N = 6,948;
138 cases2% of Time 2 respondents were
excluded due to incomplete information).
Respondents ages 35 to 39 in 19871988 were
excluded from these longitudinal analyses so that
we could make a clear differentiation in the age
group contrast analyses between persons experiencing marital status continuity and change prior
to age 40 and after age 40. (The group ages 35
to 39 between 19871988 and 19921993
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 75
would overlap into their 40s during the 5-year

period investigated; thus, we felt including them
in the analyses would make this distinction less
clear.)
Measures
Outcome measures included a 12-item modified version of the Center for Epidemiological
Studies-Depression (CES-D) index (Radloff,
1977) (alpha = .93), a 3-item measure of hostility/
irritability (alpha = .85), a standard 1-item
measure of global happiness, a 3-item version
of Rosenbergs (1965) self-esteem index
(alpha = .65), a 5-item personal mastery index
consisting of 4 items from the Pearlin Mastery
Scale (Pearlin, Lieberman, Menaghan, & Mullan,
1981) along with a single item of control-mastery
also used in Wave 1 of the NSFH (alpha = .66),
and 3-item versions of Ryffs (1989; Ryff &
Keyes, 1995) six psychological well-being
scales: Autonomy (alpha = .45), Personal Growth
(alpha = .54), Positive Relations With Others
(alpha = .53), Purpose in Life (alpha = .37), SelfAcceptance (alpha = .54), and Environmental
Mastery (alpha = .56). The relatively lower internal consistency of items used for these scales
reflects an a priori decision by Ryff to create short
scales that represent the multi-factorial structure of
the original scales (which consisted of 20 items)
rather than to maximize internal consistency.
These dramatically shortened scales have been
found to correlate from .70 to .89 with the original
highly reliable scales (Ryff & Keyes, 1995).
For three measuresthe CES-D, global happiness, and self-esteemTime 1 assessment of
the measures were available and were controlled
in the respective analyses. For the Personal
Mastery Scale, responses to one item measuring
personal mastery that was included at Time 1 of
the NSFH was included as a Time 1 control (the
correlation of this one item at Time 2 with the
other four items of the scale at Time 2 is .57).
The hostility index and the six Ryff measures
were not included at Time 1, so the CES-D
assessment from Time 1 was entered as a control
for group selection on well-being in all analyses
of these measures to better estimate the likely

longitudinal change in well-being over time due
to marital status continuity or transition. (See
Appendix A at the end of this reading for
descriptives for all variables used in the analysis;
see Appendix B at the end of this reading for a
list of scale items.)
Marital status contrasts were classified into
10 mutually exclusive and exhaustive categories
depending on respondent reports of their marital
history over the 5-year period between Time 1
and Time 2 of the NSFH (see Table 3.5).
Respondents who were continuously married
during this period were classified as married and
used as the contrast category in all analyses;
respondents who were continuously separated,
divorced, or both, were classified as separateddivorced; respondents who were continuously
widowed were classified as widowed; respondents who were continuously never married were
classified as never married; respondents who
were married at Time 1 and separated or
divorced at Time 2 were classified as marriedseparated-divorced; respondents who were
married at Time 1 and widowed at Time 2 were
classified as marriedwidowed; respondents
who were separated, divorced, or widowed at
Time 1 and married at Time 2 were classified as
remarried; respondents who were never married
at Time 1 and married at Time 2 were classified
as first married; respondents who were married
at Time 1 and who experienced both a dissolution of that marriage and a remarriage by Time 2
were classified as marriedunmarriedremarried; and respondents who were never married,
separated, divorced, or widowed at Time 1 and
who experienced both a marriage and a dissolution by Time 2 were classified as unmarried
marriedunmarried.
Respondents were also classified into two
age status categories: young adultsrespondents
who were ages 1934 at Time 1, and midlife
adultsrespondents ages 4060 at Time 1.
Several additional demographic statusesrace
or ethnicity, education, household income,
parental status, and employment statuswere
controlled in all analyses because they are
76
51.0
7.2
1.8
17.0
5.8
1.6
9.2
3.3
1.7
1.3
100.0
3,219
867
184
1,019
430
92
515
386
121
115
6,948
Weighted Percentage
4,166
77
74
242
82
264
240
1,822
647
166
552
Unweighted n
100.0
1.3
1.9
3.3
2.7
7.6
5.9
50.4
9.5
3.1
14.3
Weighted Percentage
Women
2,782
38
47
144
10
251
190
1,397
220
18
467
Unweighted n
100.0
1.3
1.5
3.4
0.4
10.9
5.6
51.6
4.8
0.4
19.9
Weighted Percentage
Men
5:33 PM
NOTE: Percentage columns do not always total 100.0 due to rounding errors.
Continuity
Married
Separated/Divorced
Widowed
Widowed
Change
MarriedSeparated/
Divorced
MarriedWidowed
Never married
First married
Separated/Divorced/
WidowedRemarried
MarriedSeparated/
Divorced/Widowed
Remarried
UnmarriedMarried
Separated/Divorced/
Widowed
Valid cases
Unweighted n
Total Sample
Weighted Percentage Distribution (unweighted n) of 5-Year Marital Status Continuity and Change, National Survey of Families and
Households 19871993, Primary Respondents, Ages 1964 (N = 6,948)
11/16/2004
Marital Status
Table 3.5
03-Chibucos.qxd
Page 76
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 77
associated with both marital status and

psychological well-being and might have
confounded our results (Menaghan & Parcel,
1990; Ross et al., 1990; Voydanoff, 1990). The following variable coding was used: race-ethnicity
(dichotomously coded 1 = African American vs.
0 = all others), education (in years), household
income (continuous measure totaled across all
types of earned and unearned income for all
household members at Time 1), missing on
household income at Time 1 (dichotomous flag
variable to include all respondents missing on
income in the regression analyses), having a
child age 18 or younger in the household at Time
2 (dichotomous, 1 = has child vs. 0 = no child),
and employment status at Time 2 (dichotomous,
1 = employed vs. 0 = not employed). Ordinary
least squares regression models were estimated
throughout using SPSS.
Results
Table 3.6 reports the results of models that estimated the effects of multiple marital status contrasts and Gender x Marital Status interactions on
well-being. Because there was at least one significant Gender x Marital Status interaction effect in
each of the combined gender models estimated (at
least at the trend level), it was deemed appropriate to examine separate models for men and
women to confirm the gender differences in marital status effects. In addition, for these models,
to answer the third research questionDoes age
status (that is young adult vs. midlife adult) make
a difference in the effects of marital status continuity and change on well-being?we included
Age 40+ x Marital Status interaction variables
for each marital status contrast where there were
enough cases to examine contrasts across age
groups. Tables 3.7 and 3.8 provide estimates for
models that examined both genders separately
for each well-being outcome and also added
Age x Marital Status interactions.
We did not create age interactions (a) for continuity in widowhood status, because there were
so few continuously widowed men and women
under age 40; (b) for the transition from never
married to first married for men or women,

because so few cases of this transition were
reported for persons aged 40 and older; (c) for
the transition from married to widowed for
men, because so few cases of this transition were
reported by persons under age 40; and (d) for the
multiple marriage transitionsunmarried
marriedunmarried and marriedunmarried
married, because so few of these cases occurred
at older ages.
Effects of Marital
Status Continuity on Well-Being
The results reported in Table 3.6 suggest that
there are several significant well-being differences between adults who experienced 5-year
continuity as singles in contrast to adults who
experienced continuity as marrieds. The continuously separated or divorced evidenced a decline
in well-being in comparison to continuously
married adults in terms of depression (at a trend
level), global happiness, personal mastery, positive relations with others, purpose in life, selfacceptance, and environmental mastery. These
patterns did not appear to be significantly different for women in contrast to men.
Being continuously widowed over a 5-year
period (almost all respondents in this category
were ages 40 to 59) was associated with an
increase in depression over time in comparison to
being married, but there were no other well-being
differences for this group. No gender differences
in the effect of being continuously widowed were
found for women in contrast to men.
The never married became more depressed
and less happy over time; however, a significant
gender interaction effect (confirmed by subsequent results provided in Table 3.7) indicated
that never-married status led to less unhappiness
for women than for men. Never-married respondents also reported more hostility, less positive
relations with others, and less self-acceptance
than their continuously married counterparts. A
significant gender interaction effect indicated
that never-married women reported even less
self-acceptance than never-married men.
(Text continued on page 82)
78
Change
MarriedSeparated/
Divorced
MarriedWidowed
Never married
First married
Separated/Divorced/
Widowed Remarried
MarriedSeparated/
Divorced/Widowed
Remarried
UnmarriedMarried
Separated/Divorced/
Widowed
Female Married
Separated/Divorced
2.23***
1.58***
.25*
.23
.28
.43*
.59
.13
.27**
.16**
.49*
.01
.11**
.22
.11**
.01
.25
.00
.01
.20
.03
.02
.01
.21
.54
.21*
2.38*
.03
.21
.25***
.16
0.51***
.23
.29***
.02
.26**
.50***
.24
.40***
.19
.08***
.04*
.37
1.73**
.23
.06
2.108
.06
.76**
.38
.33
1.00***
.27
.25
.47
.39***
.46***
.44***
.58***
.45
.43
.30
.51**
.81
.51**
.17
.88
.27
.46
.62
.31
.69***
.18
.69***
.25
.04
.27
.30
.08
.79
.39
.47*** .57***
-.01
.02
.12
.29**
Personal
Autonomy Growth
.02
.24
.96***
.95
.22
.31
.27***
.51***
.31
.80***
2.49***
.80***
.28
.21
1.98*** 1.87***
.16
.52**
.16
.52**
.83*8
1.91***
.918
.56***
.04
.56***
.91***
.56
3.6*
.97***
.19
.32*
.29
.38***
.19*
.72**
1.01*
.07
.64***
.38
.63***
.18
.42
.17
.47*
.37
.08
.27
.04
.07
Purpose SelfEnvironmental
in Life Acceptance Mastery
1.04*** 3.9
.02
.00
-1.12***
.92
.69***
.16
1.04***
.03
Positive
Relations
5:33 PM
1.06
.08
.48
.83
.55*
.29
.15
.62*
.19***
.13
.06
.00
Personal
Self-Esteem Mastery
11/16/2004
.53
.06
.05
.35**
.14***
.08*
Global
Depression Hostility Happiness
Unstandardized Regression Coefficients for the Effects of Marital Status Continuity and Change on Psychological Well-Being
Female
Age 40
Marital Status
Continuity
Married
Separated/Divorced
Widowed
Never Married
Female Separated/
Divorced
Female Widowed
Female Never married
Predictors
Table 3.6
03-Chibucos.qxd
Page 78
.91***
2.04***
.10
.35***
1.47***
.20
4.30***
.09
.23***
.34
2.26***
.18
.37***
.83***
12.74***
.14
1.76**
.32
.19
.44***
.53
.77**
.05
.05
1.33
.84
.24
.25
-.18
.48
14.46*** 12.86***
.03
.10
.27*** .22***
.84
.80
.06
.00
1.10
Personal
Autonomy Growth
.23***
13.37***
.12
.52***
1.00
.57
.10
1.92**
.05
.07
.44
.28
.40
2.49***
13.57***
.09
.50***
1.58**
.50
.02
.34
.16
Purpose SelfEnvironmental
in Life Acceptance Mastery
12.76*** 10.92***
.09
.13
.41***
1.22
.24
.25
.05
.61
Positive
Relations
p .10. *p .05. **p .01. ***p .001 (two-tailed test).
NOTE: All models also included controls for race/ethnicity, employment status, years of education, household income, and presence of a child age 18 in household.
SOURCE: National Survey of Families and Households (19871993), primary respondents (N = 6,948).
Time 1 Well-Being
Depression
Global happiness
Self-esteem
Personal mastery
Constant
R2
.38
.51**
-.55
.72
.24
.20
.03
.23
.13
.42*
Personal
Self-Esteem Mastery
5:33 PM
.30
.11
2.91**
.48
.19
Global
Depression Hostility Happiness
11/16/2004
Female Married
Widowed
Female Never
marriedFirst married
Female Separated/
Divorced/Widowed
Remarried
Female Married
Separated/Divorced/
WidowedRemarried
Female Unmarried
MarriedSeparated/
Divorced/Widowed
Predictors
03-Chibucos.qxd
Page 79
79
80
Change
MarriedSeparated/Divorced
MarriedWidowed
Never marriedFirst married
Separated/Divorced/
WidowedRemarried
MarriedSeparated/Divorced/
WidowedRemarried
Never marriedMarried
Separated/Divorced/Widowed
Age 40
Age 40
MarriedWidowed
.10
.24
.07
.42*
.20
.22
.16
.54**
.24
.63*** .26*
.71*
.21
.14*
.16
.20
.01
.55**
.57***
.62
.28***
.04
.21*
.03
.12
Marital Status
Continuity
Married (omitted)
Separated/Divorced
Widowed
Never married
.08
Men
.14
1.75**
1.52**
.12
2.19***
.29
.35
.03
.05
.72
1.08**
.39
.51
.14
Women
Men
2.68***
2.68***
.08
.04
2.40*
.37
.98
1.05
1.41
1.59**
.76
1.02***
.29
Hostility
1.04*
.12
.12
.27
.54***
.23
.32**
.05
.26
.02
.29*
.18
.14
.06
Women
.55*
.55
.01
.45**
.62
.18
.16
.04
.03
.56*
.25
.44***
.06
Men
Global Happiness
.00
.20*
.20*
.07
.16***
.19
.19***
.03
.01
.07
.04
.05
.07
.01
Women
.32**
.32**
.15
.05
.22
.07
.09
.00
.10
.09
.20
.02
.06
Men
Self-Esteem
2.39*
.02
.02
.02
1.52***
2.73**
.69**
.50
1.16**
.40
.69*
.46
.15
.95***
Women
.81**
.81**
.21
.63
2.24*
.01
.19
.57
.63
1.48**
.35
.25
.14
Men
Personal Mastery
5:33 PM
Age 40
Separated/Divorced
Age 40
Never married
.03
Women
Depression
Unstandardized Regression Coefficients for the Effects of Marital Status Continuity and Change on Psychological
Well-Being by Gender
11/16/2004
Age 40 (Time 1)
Predictors
Table 3.7
03-Chibucos.qxd
Page 80
.11
.20
.10
2.56***
.76***
.36
Men
.10
4.30***
.24***
.37
Women
.21***
.03
Men
.18
2.03***
.39***
.14
Women
.17
2.39***
.34***
.00
Men
Self-Esteem
.15
.83***
12.35***
.32
Women
.13
83***
.28
Men
Personal Mastery
NOTE: All models also include controls for race/ethnicity, employment status, years of education, household income, and presence of a child 18 in household.
p .10. *p .05. **p .01 ***p .001 (two-tailed test).
.10
4.18***
.17
1.40***
1.51*** 1.59***
.77
Women
1.05***
.18
Men
.31***
.39***
.44*
Age 40
Separated/Divorced/
WidowedRemarried
Time 1 Well-Being
Depression
Global happiness
Self-esteem
Personal mastery
Constant
12.69***
R2
Women
Global Happiness
5:33 PM
Predictors
Hostility
11/16/2004
Depression
03-Chibucos.qxd
Page 81
81
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 82
Although these results support, overall, our

first hypothesis regarding the well-being benefits
of marriage, an examination of additional
seldom-included other well-being outcomes suggests that the story is more complex. The continuously never-married men and women concurrently
rated themselves as more autonomous and as
experiencing more personal growth than their
married peers.
Overall, we found few gender differences in
the effects of continuity in marital status (in partial answer to Research Question 2). A trend
level gender interaction effect (also confirmed by
subsequent analyses reported in Table 3.7) indicated that never-married women rate themselves
lower on personal growth than never-married
men (but still not significantly lower on personal
growth than the continuously married). These
contrasting results confirm that taking a multidimensional approach to well-being is important
inaccurately understanding the determinants of
psychological well-being, and as life span developmental theory (Baltes, 1987) and structural
analyses of well-being have suggested (Bradbum,
1969; Bryant & Veroff, 1982; Ryff & Keyes,
1995), that multidirectional effects for different
well-being dimensions can occur simultaneously
(Baltes, 1987).
Age Differences in
Marital Status Continuity Effects
A number of interesting age differences in the
effects of marital status continuity emerged from
these analyses (see Tables 3.7 and 3.8). Midlife
men evidenced significantly less of an increase
in depression and less hostility (at a trend level)
over a 5-year period of remaining separated or
divorced than did younger men. Continuously
separated or divorced midlife men also reported
significantly more self-acceptance than younger
men in this marital category. However, these
same separated or divorced midlife men also
reported significantly less personal growth than
did younger separated or divorced men.
Continuously separated or divorced as well
as continuously never-married midlife women
reported significantly more positive relations with

others, and continuously separated or divorced
midlife women reported more autonomy than did
younger women of the same marital status. Nevermarried midlife women also rated their personal
mastery higher than younger never-married
women. Continuously never-married midlife
men, however, reported significantly less selfacceptance than never-married younger men. This
latter case was the only one where age differences
in the effects of remaining single favored young
adults; in general, where age differences occurred,
they suggested that single midlife adults fare
better than single young adults (in response to
Research Question 3).
Effects of Marital
Status Change on Well-Being
Several significant well-being differences
between the continuously married and those
undergoing marital status transitions were also
evident (see Table 3.6). The transition from marriage to separation or divorce was associated with
an increase in depression and a decline in reported
happiness in comparison to remaining married.
Those who separated or divorced also reported
less personal mastery, less positive relations with
others, less purpose in life (at a trend level of significance), and less self-acceptance. Women who
experienced marital dissolution reported significantly more of an increase in depression, more
hostility, more of a decline in self-esteem, less
personal growth (at a trend level), less self-acceptance, and less environmental mastery than men
experiencing marital dissolution (all but the
personal growth trend were further confirmed in
subsequent analyses shown in Table 3.7).
The transition to widowhood in this sample
was associated with lower ratings of hostility,
personal mastery, and purpose in life. Gender
interactions indicated that becoming widowed
was associated with significantly more depression (trend level effect), more hostility, less selfesteem, and more purpose in life for women than
for men. Subsequent analyses reported in Table
3.7 confirm that women, but not men, who were
.20
Men
03
.21
.47***
.48
.70
.14
.42
.49
.53
13.90***14.41***
.09
.04
.54*** .30***
.07
.42
1.12*
.55
.08
.03
.59**
.42
.30
.51** .55**
.18
.26
.28*
Women
.84 .09
.43
.07
.18
.19
.46
.62*
.33
.22
.08
Men
Men
.74
.41
.65***
1.15*
.51
.39
1.18
.17
.25
.65*
.57*
.12
.46
.60
.58
.32
.13
.11
.75
.25
.53*** .69***
.08
.43
.25
.25
.56
.56*** .55***
Women
Personal
Growth
14.47***12.78***13.25***
.03
.12
.10
.24*** .21*** .24***
.18
.53
.58
.52
.36
.42
.73
.33
.33
.84
.33
.05
.17
.16
Men
Autonomy
NOTE: All models also include controls for race or ethnicity, employment status, years of education, household income, and presence of a child 18 in household.
p .10. *p .05. **p .01. ***p .001 (two-tailed test).
.52***
.15
1.13*
12.77*** 13.87*** 10.90*** 11.11*** 13.39*** 13.67*** 13.09***

.11
.07
.14
.13
.14
.11
.09
.41
.66
.16
.63
.52
.46
Constant
R2
.01
.80
.46*** .22*** .23*** .51***
.35
2.23**
.43
2.19***
.20
.98*
.86***
.01
1.07***
.36***
.06
.34
.11 1.97*** .79
.30
.29
.49* .75* 1.77*** 1.45***

1.90 2.58*** .97
.10
.39*
.75***
.73***
.35
.30
.41
.09
Women
1.15*** 1.86*** .15

.61*
.30
.12
.53**
.53**
.14
.03
1.06*
.01
.32
1.07*
.40
.28
Men
Environmental
Mastery
Time 1 Well-Being
Depression
.77
1.80
.56
2.22***
.71
1.62***
.03
.23
.27
.07
1.74*** 1.40***
1.87*
.02
.54**
.34
Women
.57*** .53*** .15
Women
SelfAcceptance
5:33 PM
Change
MarriedWidowed
Separated/Divorced/
WidowedRemarried
WidowedRemarried
UnmarriedMarriedSeparated/
Divorced/Widowed
Age 40 Married
Separated/Divorced
Age 40
MarriedWidowed
Age 40 Separated/
Divorced/WidowedRemarried
.09
Age 40 (Time 1)
Marital Status
Continuity
Married (omitted)
Separated/Divorced
Widowed
Never married
Age 40 Separated/Divorced
Age 40 Never married
Men
Purpose
in Life
11/16/2004

1.86*** 1.73***
.62* .61*
.68*
1.08
.85*** 1.02
.74*** .87***
.01 .31
1.12***
.71
.10 .60
1.06*
.82
.72 .02
Women
Positive
Relations
Unstandardized Regression Coefficients for the Effects of Marital Status Continuity and Change on Psychological Well-Being (Ryff
scales) by Gender
Predictors
Table 3.8
03-Chibucos.qxd
Page 83
83
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 84
widowed within the last 5 years report more

depression. Men, but not women, who were
recently widowed report significantly less hostility than continuously married men (see Table 3.7).
Self-esteem differences between the recently
widowed and the continuously married do not
reach significance for men or women in separate
analyses by gender. Lower reports of purpose in
life are in evidence for both men and women in
comparison to the continuously married, but for
midlife women purpose in life does not appear to
be as compromised as it is for younger women or
men. These results provide partial support for
our second hypothesis regarding the decline in
well-being associated with transitions out of
marriage.
Results reported in Table 3.5 indicate that
becoming married for the first time within the
last 5 years led to more of an increase in wellbeing than remaining continuously married on
all outcomes except hostility and personal
mastery for men (a significant gender interaction
confirmed by subsequent results shown in Table
3.7 indicates that personal mastery was also
increased by first marriage for women). These
consistent beneficial effects among the newly
first married suggest that the answer to the first
part of Research Question 1Is the transition
into first marriage associated with increased
well-being in contrast to remaining continuously
married?is an emphatic yes.
Becoming remarried, however, led to fewer
positive effects (similar to the pattern reported
by Spanier & Furstenberg, 1982), yet remarriage
was also associated with higher reports of autonomy, personal growth, and purpose in life than
those reported by the continuously married (a
more limited affirmative answer to the second
part of Research Question 1). This transition did
not appear to have different effects for men in
contrast to women.
Respondents who experienced both a dissolution or loss of spouse and a remarriage during this
5-year period were not ultimately very different
in well-being from the continuously married,
except they did report lower self-acceptance.
Respondents who went from single to married
and back to single again across this 5-year period

showed clear evidence of poorer well-being in all
dimensions other than autonomy and personal
growth when contrasted with the continuously married. Several significant gender interaction effects
for this group, however, suggest that women actually experienced less negative impact from making
these multiple transitions than did men.
Age Differences in the
Effects of Marital Transitions
Tables 3.7 and 3.8 provide evidence that age
does have a moderating influence on the wellbeing effects of marital status transitions. Midlife
women report less increase in depression and
lower levels of hostility after a marital separation
or divorce than younger women do. Women ages
40 and over who experienced marital separation
or divorce also reported significantly better relations with others and more personal mastery than
younger women. Midlife men experiencing a
separation or divorce reported significantly more
self-acceptance than younger men did.
A trend level effect indicates that midlife
women who experience a transition to widowhood may experience more of a decline in global
happiness than younger women. However,
midlife women who were widowed, as noted
previously, reported significantly more purpose
in life, more personal mastery, and more positive
relations with others (at a trend level of significance) than younger recent widows. In addition,
the transition to remarriage was associated with
significantly less depression for midlife women
than younger women. Overall, in response to
Research Question 3, we found that a transition
out of marriage had less negative impact on the
psychological well-being of midlife adults in
contrast to young adults.
Discussion and Conclusions

These results from longitudinal data provide
considerable support for the continuing importance of marital status for well-being. In support
of our first hypothesis regarding the beneficial
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 85
effects of marriage, we found numerous cases

where a significant negative change in psychological well-being occurred as a result of remaining in
a particular single status over a period of 5 years in
contrast to remaining married over the same
period. We note also that across-sectional analysis
of marital status differences would have clustered
the newly married together with the continuously
married. Our results revealing the considerable
well-being boost that comes with becoming
married for the first time suggest that the newly
married are making a significant contribution to
inflating the mean for psychological well-being
among the married category in most crosssectional studies. By distinguishing between the
newly married and the continuously married, this
study actually yields an even more conservative
examination of marital status differences than is
typical (e.g., using the GSS), because only veterans of marriage (with somewhat lower psychological well-being) were included in the continuously
married comparison group.
The evidence from contrasts between those
who experienced a marital transition out of
marriage and those who remained continuously
married also provides considerable support for
our second hypothesis. We found the transition
to separation, divorce, or widowhood to be associated with negative effects across several components of psychological well-being.
However, if marriage was always a positive
robust influence on all dimensions of well-being,
we would not expect to find any instance where
the well-being of the married and the unmarried
was the same, or where the unmarried evidenced
better well-being than the married. Yet, our
analysis of multiple dimensions of well-being in
many cases does provide evidence of no difference between the continuously married and those
who are unmarried or transitioning out of marriage. Indeed, in a few cases, contrary to our first
and second hypotheses, the unmarried report
better well-being than the married, for example,
in their ratings of autonomy and personal
growth. These inconsistencies in patterns across
outcomes suggest that marriage is not a universal
beneficial determinant of all dimensions of
psychological well-being. It appears wise,

therefore, to continue evaluating the effects of
marriage on well-being with a multidimensional
lens whenever possible, so that we can obtain a
more precise understanding of how and when
marriage is important for mental health.
Likewise, this complex analysis does not
yield a clear and simple answer to our second
research question regarding gender differences
in the effects of marriage on psychological wellbeing. The effects of continuity in single status
are not very different for women in contrast to
men. Yet, in evaluating the effects of recent marital transitions to single status (i.e., married to
separated, divorced, or widowed) we found that
women are somewhat more negatively affected
by such transitions than men (except in the infrequent case of unmarried to married to unmarried
in 5 years). It may be that women experience the
transition to single status with more difficulty
due to a greater decline in income, a greater
share of responsibility for child rearing, and
poorer prospects for remarriage.
Regarding Research Question 3, our analyses
reveal a number of interesting differences in
the effects of marriage on well-being for midlife
adults in contrast to younger adults. In most
cases (the two robust exceptions being personal
growth for separated-divorced men and selfacceptance for never-married men), age group
differences suggest greater adaptability (evidenced by higher psychological well-being)
among mid-lifers facing singleness or transitions
to single statusthat is, separation, divorce, or
widowhood. This is a noteworthy developmental
finding, because marital transitions after age 40
are less statistically normative (Uhlenberg et al.,
1990), and usually non-normativeness is hypothesized to be associated with greater stress
(Neugarten, 1979). In the case of marital status,
however, it may be that a certain degree of expertise in handling life problems and self-management (i.e., wisdom?; see Baltes & Staudinger,
1993; Brim, 1992) has developed by midlife,
allowing for significantly greater maintenance of
psychological well-being while remaining single
or adapting to a transition to single status.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 86
We acknowledge several limitations to this

analysis. We have examined a large array of
psychological well-being outcomes here, yet
we have still omitted other outcomes that might
have been additionally informative in terms of
psychological maladjustment or distress for men,
for example, drinking, drug use, social isolation,
and aggressive behavior (Aneshensel, Rutter, &
Lachenbruch, 1991). Thus, the gender differences that we found due to transitions out of
marriage (e.g., to divorce or widowhood) may
have underestimated the impact of these changes
on men due to our particular selection of psychological well-being outcomes.
Although we have described more marital
status contrasts than any previous study, we still
have been forced to omit additional contrasts
that might have been further enlightening, for
example, cohabitor status and additional duration
in status measures. These differentiations are
surely important (e.g., Mastekaasa, 1994) and
would better describe the continuum of attachment that we agree characterizes contemporary
marriage in the United States (Ross, 1995). To add
these contrasts to our already lengthy list, we
believed would overtax the analysis. A finer-tuned
examination of select differences, looking more
closely at duration and other potential mediating
and moderating factors using the richness of the
NSFH data is certainly now in order. For example,
there is always heterogeneity in marital quality
among the married; we might expect that for some
persons in our analysis, ending a marriage may
have led to less decline in well-being than remaining in a problematic marriage (Wheaton, 1990).
Marital quality distinctions, therefore, might be
usefully explored in future work.
This unique prospective study of the transition
into first marriage, with its unusual and informative contrast to the continuously married,
provided us with an opportunity to address the
question (Research Question 1) of well-being differences between those continuously married and
those transitioning into marriage. Our analyses
yielded interestingly strong evidence that getting
married for the first time leads to a considerable
elevation in well-being. By comparison, remarriage

did not provide nearly so much psychological benefit. What is tempering the rush here? These
intriguing results also deserve further investigation
by subsequent study.
We found it valuable to make distinctions
between the life course transition to first marriage
and transition to remarriage, as well as among
the separated or divorced, widowed, and nevermarried adults, and we encourage the retention of
similar distinctions to future researchers in this
area. We also recommend continued attention to
examining life span developmental (age) differences and both positive and negative dimensions
of psychological well-being, to further a more
complete understanding of the ways in which
marital status continuity and change influences
adult development and mental health.
Note
Support for this research was provided by the John
D. and Catherine T. MacArthur Foundation Research
Network on Successful Midlife Development and a
FIRST Grant Award to the first author from the
National Institute on Aging (AG12731). The National
Survey of Families and Households (NSFH) was
funded by CPR-NICHHD Grant No. HD21009. The
NSFH was designed and carried out at the Center
for Demography and Ecology at the University of
WisconsinMadison under the direction of Larry
Bumpass and James Sweet. We gratefully acknowledge
Larry Bumpass helpful input to this study. Address
correspondence to Nadine F. Marks, Child and Family
Studies, University of WisconsinMadison, 1430
Linden Drive, Madison, WI 537061575; e-mail:
marks@sc.wisc.edu
References
Anderson, C. M., & Stewart, S. (1994). Flying solo:
Single women in midlife. New York: Norton.
Aneshensel, C. S., Rutter, C. M., & Lachenbruch, P. A.
(1991). Social structure, stress, and mental health:
Competing conceptual and analytic models.
Ball, J. F. (1977). Widows grief: The impact of age
and mode of death. Omega, 7, 307333.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 87

Baltes, P. (1987). Theoretical propositions of life-span
developmental psychology: On the dynamics
between growth and decline. Developmental
Baltes, P. B., & Staudinger, U. M. (1993). The search
for a psychology of wisdom. Current Directions
in Psychological Science, 1, 7580.
Bengtson, V. L., & Allen, K. R. (1993). The life course
perspective applied to families overtime. In
P. O. Boss, W. J. Doherty, R. LaRossa, W. R.
Schumm, & S. K. Steinmetz (Eds.), Sourcebook
of family theories and methods: A contextual
approach (pp. 469499). New York: Plenum.
Booth, A., & Amato, P. (1991). Divorce and psychological stress. Journal of Health and Social
Behavior, 32, 396407.
Bowlby, J. (1969). Attachment and loss. Vol. 1:
Attachment (2nd rev. ed., 1982). New York: Basic
Books.
Bowlby, J. (1973). Attachment and loss. Vol. 2:
Separation. New York: Basic Books.
Bowlby, J. (1980). Attachment and loss. Vol. 3: Loss,
sadness, and depression. New York: Basic Books.
Bradburn, N. (1969). The structure of psychological
well-being. Chicago: Aldine.
Bretherton, I. (1992). Attachment and bonding. In
V. B. Van Hasselt & M. Hersen (Eds.), Handbook
of social development: A lifespan approach
(pp. 133155). New York: Plenum.
Brim, O. G. (1992). Ambition. New York: Basic
Books.
Bryant, F. B., & Veroff, J. (1982). The structure of
psychological well-being: Sociohistorical analysis. Journal of Personality and Social Psychology, 43, 653673.
Bumpass, L. L. (1990). Whats happening to the
family? Interactions between demographic and
institutional change. Demography, 27, 483498.
Bumpass, L., Sweet, J., & Martin, T. C. (1989).
Changing patterns of remarriage. Journal of
Marriage and the Family, 52, 747756.
Castro-Martin, T., & Bumpass, L. L. (1989). Recent
trends and differentials in marital disruption.
Demography, 26, 3751.
Cherlin, A. (1978). Remarriage as an incomplete
institution. American Journal of Sociology, 84,
634650.
Coltrane, S. (1990). Birth timing and the division
of labor in dual-earner families. Journal of
Family Issues, 11, 157181.
Cooney, T. M., Pederson, F. A., Indelicato, S., &

Palkovitz, R. (1993). Timing of fatherhood: Is
on-time optimal? Journal of Marriage and the
Family, 55, 205215.
Daniels, P., & Weingarten, K. (1982). Sooner or
later: The timing of parenthood in adult lives.
New York: Norton.
Doherty, W. J., Su, S., & Needle, R. (1989). Marital
disruption and psychological well-being. Journal
of Family Issues, 10, 7285.
Duvall, E. M. (1957). Family development. Philadelphia:
J. B. Lippincott.
Elder, G. H., Jr. (1991). Family transition, cycles, and
social change. In P. A. Cowan & M. Hetherington
(Eds.), Family transitions (pp. 3157). Hillsdale,
NJ: Lawrence Erlbaum.
Elder, G. (1992). The lifecourse. In E. Borgatta &
M. Borgatta (Eds.), Encyclopedia of sociology
(pp. 281311). New York: Macmillan.
Featherman, D. L. (1983). Life-span perspectives in
social science research. In P. B. Baltes & O. G.
Brim, Jr. (Eds.), Life-span behavior and development (Vol. 5, pp. 157). New York: Academic
Press.
Fox, J. W. (1980). Goves specific sex-role theory of
mental illness: A research note. Journal of Health
and Social Behavior, 21, 260267.
Frankel, S. A., & Wise, M. J. (1982). A view of
delayed parenting: Some implications of a new
trend. Psychiatry, 45, 220225.
Glenn, N. D. (1975). The contribution of marriage
to the psychological well-being of males and
females. Journal of Marriage and the Family, 37,
594600.
Glenn, N. D., & Weaver, C. N. (1979). A note on
family situation and global happiness. Social
Forces, 57, 960967.
Glenn, N. D., & Weaver, C. N. (1988). The changing
relationship of marital status to reported happiness.
Journal of Marriage and the Family, 50, 317324.
Goldscheider, F., Thornton, A., & Young-DeMarco, L.
(1993). A portrait of the nest-leaving process in
early adulthood. Demography, 30, 683699.
Goldscheider, F. K., & Waite, L. J. (1991). New
families, no families? Berkeley: University of
California Press.
Gordon, T. (1994). Single women: On the margins?
New York: New York University Press.
Gore, S., & Mangione, T. W. (1983). Social roles,
sex roles and psychological distress: Additive and
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 88

interactive models of sex differences. Journal of
Health and Social Behavior, 24, 300312.
Gove, W. (1972). The relationship between sex roles,
marital status, and mental illness. Social Forces,
51, 3444.
Gove, W. R., Hughes, M., & Style, C. B. (1983). Does
marriage have positive effects on the psychological well-being of the individual? Journal of
Health and Social Behavior, 24, 122131.
Gove, W. R., & Shin, H. (1989). The psychological
well-being of divorced and widowed men and
women: An empirical analysis. Journal of Family
Issues, 10, 122144.
Gove, W. R., Style, C. B., & Hughes, M. (1990). The
effect of marriage on the well-being of adults.
Journal of Family Issues, 11, 435.
Gove, W., & Tudor, J. (1973). Adult sex roles and
mental illness. American Journal of Sociology,
78, 812835.
Hagestad, G. O. (1990). Social perspectives on the
life course. In R. H. Binstock & L. K. George,
Handbook of aging and the social sciences (3rd
ed., pp. 151168). New York: Academic Press.
Hagestad, G. O., & Smyer, M. A. (1982). Dissolving
long-term relationships: Patterns of divorcing in
middle age. In S. Duck (Ed.), Personal relationships, Vol. 4: Dissolving personal relationships
(pp. 155188). New York: Academic Press.
Harding-Hidore, M., Stock, W. A., Okun, M. A., &
Witter, R. A. (1985). Marital status and subjective
well-being: A research synthesis. Journal of
Harlow, S. D., Goldberg, E. L., & Comstock, G. W.
(1991). A longitudinal study of the prevalence of
depressive symptomology in elderly widowed
and married women. Archives of General
Psychiatry, 48, 10651068.
Heath, D. H. (1994). The impact of delayed fatherhood on the father-child relationship. Journal of
Genetic Psychology, 155, 511530.
Kitson, G. C., Babri, K. B., Roach, M. J., & Placidi,
K. S. (1989). Adjustment to widowhood and
divorce. Journal of Family Issues, 10, 532.
Lee, G., Seccombe, K., & Shehan, C. (1991). Marital
status and personal happiness: An analysis of
trend data. Journal of Marriage and the Family,
53, 839844.
Mastekaasa, A. (1994). Psychological well-being and
marital dissolution: Selection effects? Journal of
Mastekaasa, A. (1995). Marital dissolution and

subjective distress: Panel evidence. European
Sociological Review, 11, 173185.
McCrae, R. R., & Costa, P. T., Jr. (1993).
Psychological resilience among widowed men
and women: A 10-year follow-up of a national
sample. In M. S. Stroebe, W. Stroebe, & R. O.
Hansson (Eds.), Handbook of bereavement
(pp. 196207). Cambridge, UK: Cambridge
University Press.
McLanahan, S. S., & Sandefur, G. D. (1994). Growing
up with a single parent: What hurts and what
helps. Cambridge, MA: Harvard University
Press.
Menaghan, E. G., & Lieberman, M. A. (1986).
Changes in depression following divorce: A
panel study. Journal of Marriage and the Family,
48, 319328.
Menaghan, E. G., & Parcel, T. L. (1990). Parental
employment and family life. Journal of Marriage
and the Family, 52, 10791098.
Mirowsky, J., & Ross, C. E. (1989). Social causes of
psychological distress. New York: Aldine.
Neugarten, B. L. (1979). Time, age and the life cycle.
American Journal of Psychiatry, 136, 887894.
Nydegger, C. N. (1986). Timetables and implicit
theory. American Behavioral Scientist, 29,
710729.
Pasley, K., & Ihinger-Tallman, M. (Eds.). (1987).
Remarriage and stepparenting. New York:
Guilford.
Pearlin, L. I., & Johnson, J. S. (1977). Marital status,
life-strains and depression. American Sociological Review, 42, 704715.
Pearlin, L., Lieberman, M., Menaghan, E., & Mullan,
J. (1981). The stress process. Journal of Health
and Social Behavior, 22, 337356.
Radloff, L. (1977). The CES-D Scale: A self-report
depression scale for research in the general population. Applied Psychological Measurement, 1,
385401.
Riley, M., Foner, A., & Waring, J. (1988). Sociology
of age. In N. J. Smelser (Ed.), Handbook of
sociology (pp. 243290). Newbury Park, CA:
Sage.
Rosenberg, M. (1965). Society and the adolescent selfimage. Princeton, NJ: Princeton University Press.
Ross, C. E. (1995). Reconceptualizing marital status
as a continuum of social attachment. Journal of
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 89

Ross, C. E., Mirowsky, J., & Goldsteen, K. (1990).
The impact of the family on health: The decade in
review. Journal of Marriage and the Family, 52,
10591078.
Ryff, C. (1989). Happiness is everything, or is it?
Journal of Personality and Social Psychology, 6,
10691081.
Ryff, C. (1995). Psychological well-being in adult life.
Current Directions in Psychological Science, 4,
99104.
Ryff, C. D., & Essex, M. J. (1991). Psychological
well-being in adulthood and old age: Descriptive
markers and explanatory processes. In K. Warner
Schaie & M. Powell Lawton (Eds.), Annual
review of gerontology and geriatrics (Vol. 11,
pp. 144171). New York: Springer.
Ryff, C. D., & Keyes, C. L. M. (1995). The structure
of psychological well-being revisited. Journal
of Personality and Social Psychology, 69,
719727.
Sanders, C. M. (1981). Comparison of younger and
older spouses in bereavement outcome. Omega,
11, 227232.
Schoen, R., Urton, W., Woodrow, K., & Baj, J.
(1985). Marriage and divorce in twentieth
century American cohorts. Demography, 22,
101114.
Schoen, R., & Weinick, R. M. (1993). The slowing
metabolism of marriage: Figures from 1988 U.S.
marital status life tables. Demography, 30,
737746.
Siegel, J. M., & Kuykendall, D. H. (1990). Loss,
widowhood, and psychological distress among
the elderly. Journal of Consulting and Clinical
Spanier, G. B., & Furstenberg, F. F., Jr. (1982).
Remarriage after divorce: A longitudinal analysis
of well-being. Journal of Marriage and the
Family, 44, 709720.
Stroebe, M. S., & Stroebe, W. (1993). Determinants of

adjustment to bereavement in younger widows
and widowers. In M. S. Stroebe, W. Stroebe, &
R. O. Hansson (Eds.), Handbook of bereavement
University Press.
Stroebe, W., & Stroebe, M. S. (1987). Bereavement
and health: The psychological and physical
consequences of partner loss. Cambridge, UK:
Cambridge University Press.
Sweet, J. A., Bumpass, L. L., & Call, V. (1988). The
design and content of the National Survey of
Families and Households NSFH Working Paper
No. 1. Madison: University of Wisconsin
Madison, Center for Demography and Ecology.
Uhlenberg, P., Cooney, T., & Boyd, R. (1990). Divorce
for women after midlife. Journal of Gerontology:
Social Sciences, 45, S3S11.
Umberson, D., & Gove, W. R. (1989). Parenthood and
psychological well-being: Theory, measurement,
and stage in the family life course. Journal of
Voydanoff, P. (1990). Economic distress and family
relations: A review of the eighties. Journal of
Wheaton, B. (1990). Life transitions, role histories,
and mental health. American Sociological
Review, 55, 209223.
White, L. K. (1979). Sex differentials in the effect
of remarriage on global happiness. Journal of
Wortman, C. B., Silver, R. C., & Kessler, R. C. (1993).
The meaning of loss and adjustment to bereavement. In M. S. Stroebe, W. Stroebe, & R. O.
Hansson (Eds.), Handbook of bereavement
University Press.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 90
APPENDIX A
Descriptive Statistics for Analysis Variables
Total Sample
Mean (SD)
N = 6,948
Women
Mean (SD)
n = 4,166
Demographic Characteristics
Female
Age 40 (Time l)
Black
Employed
Years of education
Household income (in thousands)
Missing on income data
Child 18 in household
.52
.43
.11
.73
13.03 (2.78)
33.37 (41.52)
.32
.42
.44
.12
.64
12.81 (2.64)
32.39 (39.58)
.30
.45
.42
.10
.82
13.27 (2.90)
39.17 (44.52)
.33
.39
Psychological Well-Being
Depression
Depression (Time 2)
Global happiness
Global happiness (Time 2)
Self-esteem
Self-esteem (Time 2)
Personal mastery (1 item)
Personal mastery (Time 2)
Hostility
Autonomy
Environmental mastery
Personal growth
Positive relations
Purpose in life
Self-acceptance
2.18 (1.16)
2.12 (1.13)
5.41 (1.32)
5.36 (1.31)
4.12 (.59)
4.10 (.63)
3.60 (.96)
18.16 (3.4)
3.22 (4.21)
14.48 (2.50)
13.79 (2.74)
15.13 (2.45)
13.70 (3.11)
13.74 (2.87)
13.84 (2.74)
2.32 (1.11)
2.28 (1.11)
5.39 (1.36)
5.33 (1.35)
4.12 (.60)
4.04 (.67)
3.56 (.98)
17.87 (3.48)
3.42 (4.41)
14.36 (2.56)
13.64 (2.81)
15.16 (2.48)
14.02 (3.11)
13.68 (2.89)
13.78 (2.78)
2.04 (1.19)
1.95 (1.12)
5.44 (1.28)
5.39 (1.26)
4.12 (.58)
4.16 (.59)
3.65 (.94)
18.46 (3.28)
3.01 (3.97)
14.61 (2.43)
13.94 (2.65)
15.10 (2.42)
13.35 (3.07)
13.82 (2.84)
13.90 (2.70)
.07
.02
.17
.05
.02
.05
.02
.07
.10
.03
.14
.06
.02
.05
.004
.20
.03
.02
Marital Status
Continuity
Separated/Divorced
Widowed
Never married
Age 40 Separated/Divorced
Age 40 Never married
Female Separated/Divorced
Female Widowed
Female Never married
Men
Mean (SD)
n = 2,782
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 91
Change
MarriedWidowed
Separated/Divorced/
WidowedRemarried
WidowedRemarried
Never marriedMarriedSeparated/
Divorced/Widowed
Age 40 MarriedSeparated/Divorced
Age 40 MarriedWidowed
Age 40 Separated/Divorced/
WidowedRemarried
Female MarriedSeparated/Divorced
Female MarriedWidowed
Female Never marriedFirst married
Female Separated/Divorced/
WidowedRemarried
Female MarriedSeparated/
Divorced/WidowedRemarried
Female UnmarriedMarried
Separated/Divorced/Widowed
Total Sample
Mean (SD)
N = 6,948
Women
Mean (SD)
n = 4,166
Men
Mean (SD)
n = 2,782
.06
.02
.09
.06
.03
.08
.06
.004
.11
.03
.02
.03
.02
.03
.02
.01
.01
.01
.02
.01
.01
.02
.02
.01
.02
.02
.03
.01
.04
.02
.01
.01
SOURCE: National Survey of Families and Households (19871993).

NOTE: Descriptive statistics calculated using weighted data. Dichotomous variable means are proportions.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 92
APPENDIX B
Index Items
I. Ryff Psychological Well-Being Scales (rated on a 6-point scale: strongly disagree to strongly agree)
Autonomy
I tend to be influenced by people with strong opinions.a
I have confidence in my opinions, even if they are different from the way most other people think.
I judge myself by what I think is important, not by the values of what others think is important.
Positive Relations With Others
Maintaining close relationships has been difficult and frustrating for me.a
I have not experienced many warm and trusting relationships with others.a
People would describe me as a giving person, willing to share my time with others.
Purpose in Life
I live life one day at a time and dont really think about the future.a
Some people wander aimlessly through life, but I am not one of them.
I sometimes feel as if Ive done all there is to do in life.a
Self-Acceptance
I like most parts of my personality.
When I look at the story of my life, I am pleased how things have turned out.
In many ways, I feel disappointed about my achievements in life.a
Environmental Mastery
The demands of everyday life often get me down.a
In general, I feel I am in charge of the situation in which I live.
I am quite good at managing the many responsibilities of my daily life.
Personal Growth
I gave up trying to make big improvements or changes in my life a long time ago.a
I think it is important to have new experiences that challenge how you think about yourself and the world.
For me, life has been a continuous process of learning, changing, and growth.
II. Self-Esteem Scale
Please indicate how much you agree or disagree with the following statements:
On the whole I am satisfied with myself.
I am able to do things as well as other people.
I feel that Im a person of worth, at least on an equal plane with others.
III. Personal Mastery Scale
Please indicate how much you agree or disagree with the following statements:
I can do just about anything I really set my mind to.
Sometimes I feel that Im being pushed around in life.a
There is really no way I can solve some of the problems I have.a
I have little control over things that happen to me.a
I have always felt pretty sure my life would work out the way I wanted it to.
a. Item reverse coded.
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 93

IV. Center for Epidemiological Studies Depression Scale (CES-D)
On how many days during the past week did you . . .
Feel you could not shake off the blues even with help from your family and friends?
Feel bothered by things that usually dont bother you?
Feel lonely?
Feel sad?
Feel depressed?
Have trouble keeping your mind on what you were doing?
Not feel like eating, your appetite was poor?
Feel everything you did was an effort?
Feel fearful?
Sleep restlessly?
Talk less than usual?
Feel you could not get going?
V. Hostility Scale
On how many days during the past week did you . . .
Feel irritable, or likely to fight or argue?
Feel like telling someone off?
Feel angry or hostile for several hours at a time?
03-Chibucos.qxd
11/16/2004
5:33 PM
Page 94

Life Span Theory

Uploaded by

Copyright:

Available Formats

Life Span Theory

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Life Span Theory

Uploaded by

Copyright:

Available Formats

03-Chibucos.

ife-span developmental theory concerns

Of course, as noted, regardless of its study of

40 READINGS IN FAMILY THEORY

reasons mentioned as well as because it passes

occur? Or, does one follow longitudinally the

Life-Span Developmental Theory 41

this case, data were collected on the same

among variables in a way that pays specific

ISSUES FOR YOUR CONSIDERATION

THE INFLUENCE OF CONTEXT,

This study investigated the long-term effects

Children raised in socioeconomically disadvantaged families are at risk for a variety of

Life-Span Developmental Theory 43

Garmezy, 1991). Yet, most research on the

genetic determination so strong that there is

44 READINGS IN FAMILY THEORY

processes constitute the basic mechanisms that

of positive adjustment in the face of adversity,

Life-Span Developmental Theory 45

(Clarke & Clarke, 1981). Early adversity may be

number of these factors and their combined

46 READINGS IN FAMILY THEORY

socioeconomic disadvantage, only indicators

Bynner, 2001; Schoon & Parsons, 2002a). In

Life-Span Developmental Theory 47

Developmental-contextual model of psychosocial adjustment. ACA = academic adjustment;

adjustment (i.e., academic achievement). The path

48 READINGS IN FAMILY THEORY

reports of early life events were not influenced by

information gathered from head and class teachers

Parental Social Class

Life-Span Developmental Theory 49

defined according to job status and the

family environment (Fogelman, 1983). The

50 READINGS IN FAMILY THEORY

in the cohort studies assumed that children

shortened test version contained 54 items that

Life-Span Developmental Theory 51

with a range from 1 to 5 (or A to E). Generally,

52 READINGS IN FAMILY THEORY

Variables and Observed Indicators in Both Cohorts

Academic adjustment (ACA)

Social position in adulthood (SOC)

data. The x2 statistic is overly sensitive to model

continuum, with higher values indicating a better

Life-Span Developmental Theory 53

the intervening variables). Model 3 was the

the social position of parents of the earlier born

Age 11/10 (1969/1980)

NOTE: RGSC = parental social position.

Test of Close Fit

Step-Down Goodness of Fit

Life-Span Developmental Theory 55

In both cohorts Model 2 was judged to be

observed variables were included a priori to

Academic adjustment (ACA)

Social status in adulthood

Life-Span Developmental Theory 57