MIND OVER MATTER: INTERGENERATIONAL LINKS TO YOUTH ...
Transcript of MIND OVER MATTER: INTERGENERATIONAL LINKS TO YOUTH ...
The Pennsylvania State University
The Graduate School
College of Health and Human Development
MIND OVER MATTER: INTERGENERATIONAL LINKS TO YOUTH EDUCATIONAL
ASPIRATIONS AND ATTAINMENT
A Dissertation in
Human Development and Family Studies
by
Emily Pressler
© 2014 Emily Pressler
Submitted in Partial Fulfillment of the Requirements
for the Degree of
Doctor of Philosophy
May 2014
ii
The dissertation of Emily Pressler was reviewed and approved* by the following:
Scott Gest Associate Professor of Human Development Dissertation Co-Advisor Co-Chair of Committee Daphne C. Hernandez Assistant Professor of Health and Human Performance Dissertation Co-Advisor Co-Chair of Committee
Eric Loken Research Associate Professor of Human Development
Marianne Hillemeir Professor of Health Policy and Administration and Demography Eva S. Lefkowitz Associate Professor of Human Development Graduate Professor-in-Charge, Human Development and Family Studies
*Signatures are on file in the Graduate School
iii
ABSTRACT
Literature exploring the development of adolescent educational aspirations, and the
impacts these educational aspirations have on youth educational attainment has largely ignored
the developmental foundations of parents’ educational orientations. Specifically most research
exploring parental impacts on youth educational aspirations and attainment have measured
parental influence primarily in adulthood, post-fertility. To address this gap, the current two
studies examined a potential intergenerational transmission of adolescent educational aspirations,
as well as the impacts this transfer has on youth educational attainment in a multi-generational
sample from the National Longitudinal Survey of Youth 1979 and linked Child and Young Adult
Files.
The first study examined whether the educational aspirations women reported during
adolescence and adulthood predict the educational aspirations of their offspring above and
beyond a host of child and maternal characteristics in a sample of 1,851 mothers and their 3,198
children. Further the potential intergenerational transmission of educational aspirations was
explored separately by level of maternal educational attainment at the child’s birth as
demographic trends suggest greater increases in youth aspirations have occurred for children born
to less educated women. Findings indicate that higher maternal aspirations for children predict the
high educational aspirations of their offspring above and beyond youth and parental
characteristics. Yet, in one case this relationship varied by maternal educational attainment,
suggesting that an intergenerational transmission of adolescent aspirations between women and
their children may exist. Specifically, among a sample of youth with the highest educated
mothers, maternal aspirations for children and women’s aspirations during adolescence predicted
her offspring’s aspiration to attain a Bachelor’s degree or more.
iv
The second study explored whether multiple measures of educational aspirations relate to
youth educational attainment and whether these relationships vary by youth race/ethnicity in a
sample of 1,191 mothers and their 1,660 children. In addition, the study explored whether youth
engagement (i.e. hours spent on homework a week) and disengagement (i.e. truancy) mediated
the associated between educational aspirations and the likelihood youth earned their high school
credentials on time. Findings indicate that youth and maternal educational aspirations for her
child were associated with increases in the odds that youth would earn their high school
credentials on time, enroll in college and earn a Bachelor’s degree or more. While maternal and
youth aspirations were significant predictors of the educational attainment of white youth, youth
aspirations were primarily predictive of the educational attainment of black and Hispanic youth.
Further, analyses revealed that neither the time youth spend on homework nor their reports of
truancy in high school mediate the associations between measures of educational aspirations and
the timeliness of their high school credential attainment for all race/ethnicities.
Taken together, findings from this dissertation suggest that the educational aspirations
women have for their children are important determinants for the aspirations their offspring will
have as adolescents. Further, both maternal and youth educational aspirations are important
predictors of later youth educational attainment, and when explored by maternal educational
attainment and race/ethnicity more similarities than differences occur. In sum, this dissertation
suggests that theories of racial and ethnic deficits, or group supported counter-cultural norms,
should be put to rest as positive generational shifts have occurred increasing the educational
aspirations of all youth. Future interventions and education policies should focus on increasing
familial and youth information about higher education through resources and assistance to
increase the educational attainment of all youth.
v
TABLE OF CONTENTS
LIST OF FIGURES .................................................................................................................. vii
LIST OF TABLES ................................................................................................................... viii
ACKNOWLEDGEMENTS ..................................................................................................... ix
Chapter 1 Background ............................................................................................................. 1
Introduction ..................................................................................................................... 1 Paradoxes in educational attainment and aspirations ............................................... 2
Rising tide of youth aspirations ........................................................................ 2 Attitude-achievement paradox ......................................................................... 2
Determinants of youth educational aspirations ........................................................ 4 Focus of dissertation ................................................................................................ 5
Chapter 2 Do as I Say Not as I Did: Youth Educational Aspirations Across Generations .... 6
Introduction ..................................................................................................................... 6 Parent educational aspirations and experiences in childhood ................................. 8 Youth aspirations and maternal educational attainment .......................................... 9
Current study ................................................................................................................... 10 Method ............................................................................................................................ 12
Data .......................................................................................................................... 12 Analytic sample ....................................................................................................... 13 Measures .................................................................................................................. 14
Youth educational aspirations ......................................................................... 14 Maternal educational aspirations for child ...................................................... 15 Maternal educational at birth ........................................................................... 15 Covariates ........................................................................................................ 15
Analytic plan ........................................................................................................... 16 Results ............................................................................................................................. 17
Descriptive statistics ................................................................................................ 17 Multivariate logistic regressions .............................................................................. 18 Sensitivity models ................................................................................................... 20
Discussion ....................................................................................................................... 21 Demographic shifts in educational aspirations ......................................................... 21 Differential impacts of aspirations .......................................................................... 22 Limitations and future directions ............................................................................. 24 Conclusions and implications for the educational pipeline ..................................... 24
Chapter 3 The Power of Positive Thinking: Exploring Educational Attainment, Aspiraitons, Engagement and Race ................................................................................. 27
Introduction ..................................................................................................................... 27 Attitude-achievement paradox .................................................................................. 28
vi
Familial and intergenerational threads of youth aspirations .................................... 29 Engagement-achievement paradox .......................................................................... 31
Current study .................................................................................................................... 33 Method ............................................................................................................................. 36
Data .......................................................................................................................... 36 Analytic sample ....................................................................................................... 37 Measures .................................................................................................................. 38
Youth educational attainment .......................................................................... 38 Youth educational aspiraitons ......................................................................... 38 Maternal educational aspiraitons during adolescence ..................................... 39 Maternal educational aspiraitons for child ...................................................... 39 Youth school engagement ............................................................................... 40 Youth school disengagement ........................................................................... 40 Race/ethnicity .................................................................................................. 40 Covariates ........................................................................................................ 40
Analytic plan ........................................................................................................... 41 Results ............................................................................................................................. 43
Descriptive statistics ................................................................................................ 43 Multivariate logistic regressions .............................................................................. 44
Aspirations and attainment .............................................................................. 44 Aspirations and attainment by race/ethnicity ................................................... 44 School engagement/disengagement ................................................................. 45 Sensitivity analyses ........................................................................................... 47
Discussion ....................................................................................................................... 48 Differential reports of risk by race/ethnicity ........................................................... 49 Uniformity of youth educational aspirations ........................................................... 51 Educational aspirations and youth attainment ......................................................... 52 Educational aspirations, engagement/disengagement and attainment ..................... 55 Limitations and future directions ............................................................................. 56 Conclusions ............................................................................................................. 57
Chapter 4 Conclusion ............................................................................................................. 58
Summary of findings paper 1 .......................................................................................... 58 Summary of findings paper 2 .......................................................................................... 61 Conclusions and implications for youth attainment ........................................................ 64
References ............................................................................................................................... 67
Appendix: Tables and Figures ................................................................................................ 75
vii
LIST OF FIGURES
Figure 1: Educational aspirations by generation . .................................................................... 75
viii
LIST OF TABLES
Table 2-1: Analytic sample descriptive statistics . ................................................................... 76
Table 2-2: Adjusted odds ratios (standard error) [95% CI] of intergenerational trasmission of educational aspiraitons in full sample . ..................................................... 78
Table 2-3: Adjusted odds ratios (standard error) [95% CI] of intergenerational trasmission of educational aspiraitons by maternal educational attainment .................... 81
Table 3-1: Descriptive statistics of dependent, independent and mediating variables for full sample and by race/ethnicity ..................................................................................... 83
Table 3-2: Descriptive statistics of all covariates in full analytic sample and by race/ethnicity .................................................................................................................... 85
Table 3-3: Adjusted odds ratios (standard error) [95% CI] predicting youth educational attainment with educational aspirations ................................................................................. 88
Table 3-4: Testing mediation for high school engagement and disengagement in white sample ....... 89
Table 3-5: Testing mediation for high school engagement and disengagement in black sample ....... 91
Table 3-6: Testing mediation for high school engagement and disengagement in Hispanic sample .. 96
ix
ACKNOWLEDGEMENTS
I would like to acknowledge, my husband John Batterton, my parents, my friends, my
family, my cohort and Daphne for their continued support.
1
Chapter 1 Background
Introduction
In 2013 the National Center for Education Statistics estimated that almost 90% of youth
age 18- 24 would earn their high school diploma or GED. Even though the United States
currently enjoys higher rates of high school graduation compared to decades prior, leaks do exist
in the educational pipeline, so that at each increasing level of educational attainment fewer youth
either enroll in, or graduate from such educational programs (Bureau of Labor Statistics; BLS,
2013a). Unfortunately, these leaks in the educational pipeline do not occur equally across groups
of students, but rather success in the educational pipeline often is correlated with their socio-
economic status or race/ethnicity (NCES, 2013). Although the race/ethnic gap appears to be
narrowing over time (Tienda, 2013), it still has a stubborn hold over youth educational
attainment. This persistent gap is worrisome as generations of students are being blocked from
their full academic potential.
This gap places great risk at the United States economy financially, as well as the well-
being of the United States labor market, as gaps in educational attainment are linked to disparities
in adult participation in the labor market and earnings (BLS, 2103b), as well as disparities across
many domains of adult and child well-being (Davis- Kean, 2005; Ellickson, Saner, & McGuigan,
1997; McCarty et al., 2008; Jemal et al., 2008; Smith, Brooks-Gunn, & Klebanov, 1997). Further,
as the diversity of the racial and ethnic composition of the United States continues to increase
over time (see Perez & Hirschman, 2009; U.S. Census, 2012), the United States’ ability to
succeed in a global labor market will undoubtedly hinge upon the training and success of all its
citizens. Therefore the goal of this dissertation is to explore one potential avenue that can be
exploited to increase the educational attainment of all youth: the intergenerational link between
2
youth educational aspirations and attainment. To do so, I will examine whether intergenerational
links drive two well-documented phenomena: increases in youth aspirations over time and the
attitude-achievement paradox.
Paradoxes in educational attainment and aspirations
Rising tide of youth aspirations
At the same time gaps continue in the educational attainment of United States’ youth and
young adults, the overall tide of youth educational aspirations appears to be increasing. For
example, in her study of three nationally representative cohorts of high school sophomores,
Goyette (2008) found that over time youth educational aspirations have increased. Specifically, in
1980 43% of high school sophomores aspired to attain a Bachelor’s degree or more, while this
number increased to almost 85% of high school sophomores in 2010. Further, the greatest
increases in youth aspirations occurred in adolescents with the least educated parents. This
phenomenon is thought to be driven by a national discourse supporting ‘college for all,’ rather
than parallel increases in youth occupational aspirations, demands set by the labor market, or
increasing levels of parental educational attainment (see, Schneider and Stevenson 1999;
Rosenbaum 2001; for discussions).
Attitude-achievement paradox
The second related phenomena explored in this dissertation is the attitude-achievement
paradox (Mickelson, 1990), or when under-performing (and often under-resourced) youth report
higher, or equally high, educational aspirations compared to their higher achieving (and higher-
3
resourced) counterparts. Research exposing the attitude-achievement paradox largely stemmed
from empirical explorations that sought to validate, or invalidate, the Oppositional Culture
Framework (Fordham & Ogbu, 1986; Ogbu, 1981). Developed to highlight the potential logical
behavior of inner-city urban youth who engaged in illegal activities as employment, rather than
invest in the education system, the Oppositional Cultural Framework suggests that the counter-
educational behavior of young urban black youth was driven through group pressures that
discouraged education as a viable opportunity. Rather, minority youth with histories of
oppression were thought to reject the valued institutions of the majority, as these biased
institutions would provide few benefits or opportunities for minority member advancement.
Yet, when the educational aspirations of black, Hispanic and Asian students were
gathered, rather than support theories such as the Oppositional Culture Framework, the attitude-
achievement paradox was discovered. Minority and under-performing youth consistently reported
high educational aspirations, despite the considerable barriers they face, and in some cases even
higher than their more skilled or privileged counterparts (Ainsworth-Darnell & Downey, 1998;
Blau, Moller, & Jones, 2004; Downey, Ainsworth & Qian, 2009). Further through these studies,
the importance of familial history and educational narratives was suggested to both: (a) alter the
process through which youth chose their comparison groups and also (b) maintain youth
optimism or persistence even in the face of considerable academic obstacles (e.g., Blau et al.,
2004; Moller, Stearns, Blau, & Land, 2006). Therefore this dissertation will also explore whether
the educational aspirations collected from (a) parents as adolescents, (b) mothers in regards to
their hopes for their children, and (c) youth during adolescence are differentially related to the
educational attainment of groups of students by race/ethnicity.
4
Determinants of youth educational aspirations
Historically, research exploring the processes through which youth create their
educational aspirations fall primarily into two camps: (1) research employing status attainment
models and (2) research employing models similar to Bayesian Learning Theories. Status
attainment models (e.g. Blau & Duncan, 1967; Haller & Portes, 1973) find that children largely
adopt the educational aspirations of their parents or ‘significant others’ (see, Andrew & Hauser,
2011; Kao & Tienda, 1998, for discussions). Educational aspirations are then hypothesized to
stabilize almost entirely by adolescence and regardless of current or future youth experiences,
achievements or challenges in school (Haller & Portes, 1973). Research modeled after Bayesian
Learning Theories (Morgan, 2005) posits that student educational aspirations may be subject to
change, as individuals may engage in complex processes to incorporate information regarding
their likelihood of achieving their desired education. Thus, when faced with accumulating
educational challenges, or experiences evoking strong negative stimuli related to school, students
may begin to downgrade their educational aspirations and attitudes.
To date little research has explored the underlying familial processes that serve as a
potential conduit for the transfer of parental educational aspirations to their children or that
support the educational attainment of youth later on. Further, beyond parental educational or
occupational statuses, parents are often cast as minor agents regarding the development of their
children’s educational values. When parents’ perceptions of education are acknowledged, often
they are included only insofar as how parents’ aspirations and attitudes for their child’s education
relate to various outcomes. The life course framework (Elder, 1998) suggests that not only will
children develop across the life course, so too may their parents. In addition, interactionist
perspectives on development (e.g., Bronfrenbrenner & Morris, 1998; Conger & Donnellan, 2007)
find individuals respond to and cause reciprocal reactions in their environment, thus
5
fundamentally changing the worlds they develop in, as well as the course of their own
development. Thus, by not involving parents’ earlier educational aspirations we may be ignoring
an important foundation for the development of their offspring’s educational values and
orientations.
Focus of dissertation
Previous research has not focused on the intergenerational determinants and processes
that may transfer educational aspirations from parents to adolescents. Therefore, this dissertation
addresses this gap by highlighting the potential links that may exist across generations in the
following two papers. The first study examines whether an intergenerational transmission of
youth educational aspirations exists between a sample of women and their adolescent offspring.
Specifically, the study includes a measure of mothers’ earlier educational aspirations when she
was an adolescent and women’s maternal aspirations for their children as forces driving the
formation of their child’s educational aspirations during adolescence. Further, these relationships
are explored separately by maternal educational attainment at the birth of the child as recent
research has found increases in youth aspirations have occurred at a higher rate in students with
lower-educated mothers (Goyette, 2008). In the second study of the dissertation, measures of
mothers’ earlier educational aspirations, mothers’ aspirations for their children, and their child’s
aspirations during adolescence are examined in relation to later youth educational attainment. In
addition, these relationships are explored separately by student race/ethnicity as potentially
intergenerational narratives of education may occur unequally across the sample and relate to
differential relationships between the measures of educational aspirations and youth attainment.
Ultimately, this dissertation contributes to current literature by pulling back the empirical lens
through which the determinants of youth educational aspirations and attainment are developed.
6
Chapter 2
Do as I Say, Not as I Did: Youth Educational Aspirations Across Generations
Introduction
Today there are multiple leaks in the United States’ educational pipeline, whereby most
youth earn their high school credentials, yet fewer enroll in, or graduate from, college. For
example, almost 90% of youth (age 18- 24) earned their high school diploma or GED in 2011
(National Center for Education Statistics, 2011). Yet, only 66% of recent high school graduates
went on to enroll in post-secondary education (Bureau of Labor Statistics, 2013). Sadly, for an
increasing number of students just ‘making it’ to college does not correlate with eventual
completion, as ‘stopping out’ and dropping out are increasingly common among youth enrolled in
two- and four- year institutions (National Center for Education Statistics, 2011; Symonds,
Schwartz, & Ferguson, 2011). These gaps in enrollment and attainment have fueled an
increasingly contentious debate regarding whether it would be more beneficial to mend such
leaks, or instead to restrict or redirect the stream of students enrolling in post-secondary education
to begin with (Domina, Conley, & Farkas, 2011; Rosenbaum, 2011).
Ironically, at the same time leaks persist in the U.S. educational pipeline, youth
reported educational aspirations have largely increased. These contradictory phenomena suggest
that the country’s educational attainment problem may not be a result of a youth educational
aspiration problem. In other words, the country may not have a low educational aspirations
problem. Research has also found these trends are not fully explained by the proportions of
parents with high educational attainment, or entirely reflective of youth occupational aspirations
(Goyette, 2008). Rather, these population trends mimic processes uncovered through more
individual-level research such as the ‘attitude-achievement’ paradox (Mickelson, 1990), where
7
the greatest increases in youth aspirations over time have occurred in populations with the least
educated parents. Thus, adolescents who have fewer resources are found to have educational
aspirations just as high as their more privileged counterparts.
Current literature exploring the development of youth educational aspirations is largely
driven by the Status Attainment/Wisconsin Models, which argue that youth educational
aspirations are static across development (Blau & Duncan, 1967; Haller & Portes, 1973; Sewell,
Haller, & Portes, 1969). These models find that youth educational aspirations are largely adopted
from the influence of ‘significant others,’ (parents or peers) via a plethora of intervening
variables. For example, this literature has repeatedly documented that mothers’ educational
aspirations for their children, as well as parents’ eventual occupational and educational statuses,
are robust predictors of child educational aspirations and attainment (De Civita, Pagani, Vitaro, &
Tremblay, 2004; Fan & Chen, 2001; Taylor, Clayton, & Rowley, 2004). While these models have
built the foundation for empirical explorations documenting the processes through which youth
develop educational and occupational aspirations, I argue these models may not go far enough.
Frequently these models incorporate parental influence from constructs measured in adulthood,
post-fertility. Therefore these models implicitly ignore the events that occurred prior to parents
reaching adulthood. The question remains whether or not the aspirations parents held as youth
relate to their children’s educational aspirations.
In addition to extending the framework of the Status Attainment/Wisconsin Models, this
study employs two developmentally based frameworks to examine a potential intergenerational
transmission of educational aspirations in a sample of women and their adolescent offspring from
the National Longitudinal Survey of Youth 1979. First, this study employs the life course
framework (Elder, 1998), which suggests individuals continue to develop across their entire life
span. Thus, models employing current or retrospective reports of parental aspirations may
produce biased estimates of the true relationship between parent and child aspirations by only
8
capturing more current parental aspirations and ignoring the developmental foundations of
parents’ educational orientations. This study is also guided by an interactionist perspective on
development (e.g., Bronfrenbrenner & Morris, 1998; Conger & Donnellan, 2007), which argues
that individuals are active agents in their development, who will both react to, and alter, internal
as well as external stimuli. I argue an individuals’ educational aspirations are not developed in a
vacuum. Children’s educational aspirations should at least partially derive from the orientations
parents themselves had as adolescents, as well as the current educational aspirations parents have
for their children. As both the life course framework, and the interactionist perspective of
development would argue, parents’ current beliefs and values are, for better or for worse, the
result of events and processes set in motion years earlier. While these beliefs and characteristics
are likely to be strongly related across stages of development, as one stage directly and indirectly
impacts the next, these orientations are not likely to be equivalent. Thus, I will test the strength of
both sets of parental aspirations (past and current), and the importance of these orientations in the
development of their offspring’s own educational values.
Parental educational aspirations and experiences in childhood
Literature exploring parents’ own educational orientations, experiences, and attitudes, has
traditionally relied on either retrospective information gained from parents as adults, or on
measures of current parent educational aspirations for their children. While the relevance of
parents’ past experiences on their current behavior is often discussed in ethnographic and
qualitative research (e.g., Lareau, 2003), little quantitative research has explored the relationships
between parents’ attitudes toward school as students themselves, and the attitudes they have
toward their offspring’s education (see, Taylor et al. 2004, for discussions). One similar study did
find parents might relive their childhood social experiences when rearing children, and guide their
9
offspring through similar transitions based on these ‘re-activated’ emotions (Putallaz, Costanzo,
& Smith, 1991). Thus it is possible that parents will engage in similar processes when shaping the
educational orientations of their offspring, based on their own past experiences and thoughts as
students. Yet, the relevance of parents’ earlier experiences compared to their current adult
perceptions remains questionable as this literature is largely based upon retrospective information
that may be biased by parents’ current emotions, recent experiences, and memory recall.
In the only study the I could find where school experiences were collected prospectively
from parents (Kaplan, Liu, & Kaplan, 2000), the negative school experiences of mothers when
they were 7th graders directly predicted the negative experiences of their children at the same age.
Further, maternal negative school experiences resulted in decreased maternal involvement in their
child’s education, as well as decreased maternal encouragement for child education, even after
controlling for later maternal educational attainment. Currently little research employs
prospective reports of parents’ school aspirations while parents (as adolescents) were enrolled in
school. Further, the only study I could find which employed prospective measures of parental
school orientations (Kaplan et al., 2000) focused primarily on emotionally salient negative school
experiences (i.e., conflict with teachers, getting expelled from school), rather than a more global
measure of educational aspirations. Thus, it remains to be seen whether parents’ educational
aspirations during adolescence may serve as a foundation for their children’s later aspirations.
Youth aspirations and maternal educational attainment
Across development, youth frequently report high educational aspirations (Kao &
Thompson, 2003). Yet, recent literature has uncovered that educational aspirations reported by
cohorts of high school sophomores today are significantly higher compared to previous cohorts of
high school sophomores (Goyette, 2008). Discussions regarding the sources of these aspirational
10
trends have focused upon three likely phenomena: (a) the proportion of students with highly
educated parents, (b) the increasing educational demands of the labor market, and (c) a national
discourse supporting a ‘college for all’ viewpoint (see, Schneider and Stevenson 1999;
Rosenbaum 2001; for discussions). Undoubtedly, maternal educational attainment is a driving
force of child educational aspirations, as high maternal education is directly and positively related
to family SES (Mau & Bikos, 2000; Sewell & Shah, 1968), as well as indirectly via the type and
quality of investments parents make in their child’s educational or cognitive development (Davis-
Kean, 2005; Deslandes & Bertrand, 2005; Hoover- Dempsey et al., 2005; Kohl, Lengua, &
McMahon, 2000).
However, Goyette (2008) found that the increases in national rates of parental
educational attainment, and the demands of the labor force, do not fully account for the high
aspirations of recent cohorts of high school sophomores. Instead, she found the greatest increases
in youth educational aspirations occurred in youth with the least educated parents, similar to the
well documented ‘attitude-achievement paradox,’ whereby lesser skilled youth report as high, or
higher aspirations as their more advanced counterparts (Crosnoe, 2001). Thus, high youth
educational aspirations persist even when students may lack the institutional (e.g., from
communities or schools) as well as personal (e.g., from family or individually) resources
necessary to be successful. And this paradox may be further compounded in the ‘college for all’
context thus supporting the higher goals of all students, rather than perhaps more realistic goals,
or varying needs of students.
Current study
The current study examines the intergenerational transmission of educational aspirations
using the National Longitudinal Survey of Youth 1979 and linked supporting child files. By
11
applying the life-course and interactionist perspectives this study will determine whether
prospective maternal educational aspirations (generation 1; G1) at two different time points: as
adolescents and adults shape the educational aspirations of their adolescent offspring (generation
2, G2) in a large multi-generational sample. I hypothesize that the educational aspirations women
hold as adolescents for themselves and as adults for their children will be positively predictive of
the aspirations their adolescent offspring hold at a similar age, above and beyond more proximal
characteristics of mothers and their children. This study will also examine whether the
intergenerational transmission of educational aspirations is specific to a particular level of
maternal educational attainment as previous research has found the greatest increases in
educational aspirations occur among youth with the least educated parents (Goyette 2008).
Therefore, I hypothesize that the adolescent educational aspirations of the offspring of the least
educated women are driven by the aspirations their mother reported during adolescence.
This study contributes to the literature in several ways. First, this is one of a few studies
to incorporate prospective reports of maternal educational aspirations while mothers were
adolescents and adults to examine whether these values are transmitted from mothers to their
children. By not relying on retrospective reports of maternal aspirations, findings from this study
are less likely to be impacted by biases related to mothers’ current emotional state, more recent
adult experiences, and memory recall. Second, the longitudinal nature of the NLSY data provide
this study the unique opportunity to match respondent’s age when reporting educational
aspirations across two generations. This allows for a more nuanced comparison of the educational
aspirations of mothers and their children when they are likely experiencing similar developmental
and educational challenges. Related, the NLSY data provides the opportunity to include a host of
multi-generational (e.g., maternal and youth characteristics) and longitudinally constructed
characteristics (e.g., proportion of childhood exposed to poverty) to obtain more precise estimates
of the possible intergenerational transmission.
12
This study also dives deeper into the development of youth aspirations by incorporating
multiple comparisons of youth educational aspirations, rather than employing a single binary
measure of aspiring to ‘get past college’ or ‘earn a Bachelor’s degree’ (Bozick, Alexander,
Entwisle, Dauber, & Kerr, 2010; Goyette, 2008). Rosenbaum (2011) argues students who are
aware of, and aspire for specific types of degrees (e.g., Associates versus Bachelor’s versus
Masters’), are more likely to have received high quality information regarding college, and the
differences between such programs of study. This information is likely correlated with student
past and future achievement, as well as parent educational attainment, and household or school
resources. Therefore contrasts included in the current study may provide greater insight as to
whether youth interpret differing levels of educational aspirations uniformly across all levels of
maternal educational attainment (proxy for youth and family privilege, resources, and
information).
Method
Data
Data for this paper come from the National Longitudinal Survey of Youth 1979 Cohort
(NLSY79), and the linked Children and Young Adult files (CNSLY). NLSY data are sponsored
by the U.S. Department of Labor and have been compiled through the Ohio State University
Center for Human Resource Research. Beginning in 1979 the NLSY followed a nationally
representative sample of 6,403 men and 6,283 women who were 14 to 21 years old in 1978
(referred to as G1). Biennially starting in 1986, the NLSY79 was expanded to include mother
supplemental surveys, in which G1 women reported additional demographic, and developmental
information for their offspring (referred to as G2 from here on after). Further, beginning in 1988,
13
and biennially thereafter, G2 children who were 10 years or older directly reported information
related to their experiences and development. Across time the NLSY has achieved high rates of
participant retention, and the CNLSY files contain information on over 11,000 children born to
NLSY79 women as of the 2010 survey wave.
Analytic sample
The original G1 cohort is composed of 6,403 men and 6,283 women who were 14- 21 in
1978. However, all men were excluded from the analytic sample, as the NLSY did not collect
information regarding men’s offspring. As the primary goal of the study is to explore the
intergenerational transmission of youth educational aspirations, childless women were excluded
from the study (n = 1,353 cases excluded). Because this study explores the transmission of
adolescent educational aspirations, mothers who were young adults (i.e. 20 years old or older) at
the time they reported their educational aspirations were excluded from the analytic sample (n =
1,468 cases excluded).
In addition, 2,832 of the 8,129 G2 children with eligible G1 mothers were excluded
because youth were missing G2 educational aspirations (769 children where either too old or too
young to have valid educational aspirations in adolescence). Further, 258 G2 children were
excluded as their mother’s educational attainment was missing at the time of their birth, while an
additional 1,814 respondents were excluded for containing high levels of missing data across all
variables included in analyses (over 20% of the analytic data). Thus, the final analytic sample
consists of 1,851 G1 mothers and their 3,198 G2 children.
14
Measures
Youth educational aspirations
G1 educational aspirations were taken from the first NLSY data survey in 1979 where
women still enrolled in high school reported how far in school they expected to reach (i.e. what
year or grade of school) (mean age in years= 16.68). From this continuous measure, an ordinal
measure of G1’s aspirations during adolescence was created where 0= ‘less than high school
diploma/GED,’ 1= ‘high school diploma/GED,’ 2= ‘some college/other training,’ 3= ‘attain a
Bachelor’s degree or more.’ The developmental timing of this measure was chosen in efforts to
maximize the number of G1 women who were still adolescents, as well as reduce the likelihood
that academic milestones (e.g., graduated from high school) and adult experiences would impact
women’s educational aspirations.
Similar to their mothers, G2 youth reported how far in school they expected to reach
during adolescence (mean age in years= 14.90; average calendar year of assessment = 2001).
From these continuous measures, five binary variables were created to indicate which level of
education the youth aspired to complete (leave before high school graduation, graduate from high
school, complete some college or other training, get a Bachelor’s degree, and get more than a
Bachelor’s degree). The categories were then used to create three dichotomous dependent
variables in which the primary focus was to predict the higher level of educational aspirations: (a)
complete some college or more versus high school diploma or less; (b) complete a Bachelor’s
degree or more versus some college; and (c) attain more than a Bachelor’s degree versus a
Bachelor’s degree. The sample sizes in each regression model reflect the number of G2 youth
who reported aspiring each specific category of educational attainment; thus not all G2 youth may
be included in each model.
15
Maternal educational aspirations for child
From 1988- 2010, G1 women reported how far they thought their child would go in
school. To mirror the developmental timing of other key variables, mother’s responses were taken
from the calendar years when children were 12- 15 years old. To parallel the other measures of
educational aspirations, mother’s aspirations for their child’s education was recoded into an
ordinal variable where 0= ‘less than a high school diploma/GED,’ 1= ‘attain a high school
diploma/GED,’ 2= ‘complete some college/other training,’ and 3= ‘attain a Bachelor’s degree or
more.’
Maternal education at birth
Every survey year from 1979- 2010, G1 women reported the highest year or grade of
school they had completed. From this continuous measure, three binary indicator variables were
created to capture G1’s educational attainment at the time of G2’s birth: less than high school
diploma, high school diploma/GED, and completed some college or more.
Covariates
A host of G1 mother and G2 youth characteristics were included in models that are
correlated to youth educational aspirations. G2 youth characteristics include gender [female, male
(reference)], race/ethnicity [White (reference), Hispanic, Black], child’s total standard score on
the math subset of the Peabody Individual Achievement Test scale (PIAT; Dunn and Markwardt,
16
1970) when they were 8 to 9 years old, whether the child participated in a special education
program at school (1= yes, 0= no), whether the child experienced grade retention by 8th grade
(1= held back by 8th grade; 0 = not held back by 8th grade), number of siblings, the proportion of
years from birth to age 13 children were exposed to: poverty and maternal unemployment, and
the average percentile score of their household’s cognitive stimulation and emotional support
when children were 0 – 5 from the Home Observation Measure of the Environment- Short Form
(HOME-SF; Caldwell and Bradley 1984). G1 mother characteristics at child’s birth include age,
educational attainment (less than high school, High School Diploma/GED, some college or
more), marital status (1= not married, 0= married). A measure of maternal cognitive skills
administered in 1981 from the Armed Forces Qualification Test (AFQT) was also included.
Analytic plan
Descriptive and multivariate regression analyses were performed using Stata SE 12.1
(Stata Corp LP, College Station, TX). Weighted descriptive statistics were obtained from non-
imputed data using the NLSY79 2010 youth population weight (Y2615900). The analytic sample
was stratified by G1 educational attainment at G2’s birth and mean differences were conducted
using analysis of variance for continuous variables and chi-square tests for dichotomous
variables. Prior to conducting multivariate logistic regression models, multiple imputation
techniques were used to impute missing data using the ice command for participants missing less
than 20% of overall data. Compared to listwise deletion, multiple imputation processes reduce
biases in parameter estimates and standard errors by maintaining a closer representation of this
sample to the population (Graham, 2009; Graham & Schaefer, 1999). Last, standard errors in all
regression models were adjusted to account for the lack of independence of observations as
children are clustered within families.
17
A series of multivariate logistic regressions were conducted to determine whether G1
adolescent educational aspirations are transmitted to G2 adolescent educational aspirations. In the
first set of regression models, G2 adolescent educational aspirations were regressed on G1
adolescent educational aspirations. In the second model of the series, G2 maternal educational
aspirations for her children, as well as all child and maternal covariates were included in
regression models. These additional models allow us to examine whether G1’s prior aspirations
as an adolescent are as important for the development of her child’s aspirations, compared to the
aspirations she has for her child as an adult. In addition, all models were conducted first on the
full analytic sample and then stratified by maternal educational attainment at the birth of the child
to explore whether this transmission varies by maternal educational attainment.
Results
Descriptive statistics
Figure 1 presents the differing levels of adolescent educational aspirations by generation
status (G1 mothers versus G2 youth, weighted statistics). Generally, a larger portion of G2 youth
reported aspiring to reach higher levels of education than G1 mothers. The majority of G2 youth
aspired to attain a Bachelor’s degree or more (a combined 68%), compared to 33% of G1
mothers. Further, a large portion of G1 mothers aspired to obtain a high school diploma (41%)
compared to only 20% of G2 youth. The largest percent change in youth aspirations occurred in
the percent of youth aspiring to obtain more than a Bachelor’s degree. Comparing G1 and G2
aspirations over time, there was a 225% change among youth aspiring to complete more than a
Bachelor’s degree.
Table 1 presents the weighted descriptive statistics for the full analytic sample (nG1=
18
1,851, nG2= 3,198). The analytic sample was predominantly White (77%), followed by Black
(16%) and then Hispanic (7%), with 13% of G2 youth reported being held back a grade by the
time they were in 8th grade. G2 youth had on average just fewer than two siblings (M= 1.86, SD=
1.32) and lived in households under the federal poverty line (FPL) about 22% of the years from
birth to age 13 (approximately 3 years). At the time of G2’s birth, G1 mothers were
approximately 25 years old (M= 25.00, SD= 3.65), the majority had their high school diploma
(50%) or some college or more education (34%), and were married (75%). On average G1
mothers were unemployed for just under 5 years from the time G2 was born to the time G2 turned
13 (M= 0.38, SD= 0.31).
Compared to G2 youth born to a G1 mother with some college education or more, G2
youth born to a G1 mother with less than a high school diploma were less likely to aspire to attain
a Bachelor’s degree or more. G2 youth born to a G1 mother with less than a high school
education were more likely to be Hispanic, be held back, reside with more siblings, exposed to
greater amounts of poverty, and born to a single mother compared to youth born to G1 mothers
with some college education or more.
Multivariate logistic regressions
Table 2 presents the multivariate logistic regression results in which the intergenerational
transmission of adolescent educational aspirations was explored in the full analytic sample.
Higher G1 adolescent educational aspirations related to a 55% increase in the odds G2 youth
aspired to complete some college or more versus a high school education or less (Model 1).
Further, higher G1 adolescent educational aspirations related to a 32% and 17% increase in the
odds G2 youth aspired to complete a Bachelor’s degree or more versus some college, and more
than a Bachelor’s versus a Bachelor’s degree respectively (Model 1). Once G1 and G2 covariates
19
were included in the analyses these relationships were reduced to non-significance (Model 2).
However, higher G1 maternal aspirations for her child were related to a 67% increase in the odds
G2 youth aspired to complete some college or more versus a high school education or less (Model
2). Higher G1 maternal aspirations for her child were related to a 50% and 30% increase in the
odds G2 youth aspired to attain a Bachelor’s degree or more versus some college, and more than
a Bachelor’s versus a Bachelor’s degree, respectively, even with controls included in models
(Model 2).
Table 3 presents the multivariate logistic regression results that examines whether the
transmission of adolescent educational aspirations is specific to a particular level of maternal
educational attainment. Among a sample of G2 youth whose G1 mothers had less than a high
school education at the time of their birth (Panel A), G1 adolescent educational aspirations were
not significantly related to G2’s youth educational aspirations either before (Models 1) or after
covariates were included in the models (Models 2). However, higher G1 maternal aspirations for
her child were related to a 43% increase in the odds G2 youth aspired to complete some college
or more versus a high school education or less (Model 2).
Next, the relationship between G1 and G2 educational aspirations was explored among a
sample of youth whose mothers had a high school diploma/GED at the time of their birth (Panel
B). Higher G1 adolescent educational aspirations were related to a 17% increase in the odds G2
youth aspired to complete some college or more versus a high school education or less (Model 1).
Again this relationship was reduced to non-significance when G1 maternal aspirations for her
child and other covariates were included in the model (Model 2). However, higher G1 maternal
aspirations for her child were significantly related to the odds G2 youth aspired to complete some
college or more versus a high school education or less (74% increase), a Bachelor’s degree or
more versus some college (61% increase), as well as more than a Bachelor’s degree versus a
Bachelor’s degree (33% increase) (Models 2).
20
Finally, the relationship between G1 and G2 educational aspirations was explored among
a sample of G2 youth whose G1 mothers’ had some college education or more at the time of their
birth (Panel C). Higher G1 adolescent educational aspirations were significantly related to a 58%
increase in the odds G2 youth aspired to compare a Bachelor’s degree or more versus some
college (Model 1). Once child and maternal covariates were entered in the model, G1 adolescent
educational aspirations were significantly related to a 47% increase in the odds G2 youth aspired
to complete a Bachelor’s degree or more versus some college (Model 2). Further, higher G1
maternal aspirations’ for her child related to a two fold increase in the odds G2 youth aspired
some college or more versus high school education or less and a Bachelor’s degree or more
versus some college, and a 41% increase in the odds G2 youth aspired more than a Bachelor’s
degree versus a Bachelor’s degree (Model 2).
Sensitivity models
Two additional sets of sensitivity analyses were conducted to explore the robustness of
the main findings and potential threshold effects (available upon request). First, two binary
indicator variables were included in models in which a value of 1 signified (1) G1 as an
adolescent aspired to complete some college or more, and (2) G1 aspired for her child to
complete some college or more. Multivariate logistic regression models with these versions of G1
aspirations support the main findings presented in Table 2 as well as Table 3. Next, binary
indicators were included in models in which a value of 1 indicated (1) G1 as an adolescent
aspired to attain a Bachelor’s degree or more, and (2) G1 aspired for her child to attain a
Bachelor’s degree or more. Multivariate logistic regression results paralleled the main findings
with one exception, G1’s aspirations as an adolescent to attain a Bachelor’s degree or more was
related to a 77% and 40% increase in the odds G2 aspired to complete a Bachelor’s degree or
21
more, as well as more than a Bachelor’s degree, above and beyond other covariates.
Discussion
This study explored the intergenerational transmission of youth educational aspirations
by employing measures of prospective educational aspirations during adolescence for a sample of
mothers and their offspring, as well as a measure of mothers’ maternal educational aspiration for
her child. When the full set of maternal and child covariates were included in regression models,
G1 maternal aspirations for G2 youth were found to be a significant predictor of G2 youth
aspirations across all three levels of child aspired educational attainment. Women’s own
aspirations reported during adolescence were not significantly related to their offspring’s
aspirations. Thus, women’s maternal aspirations for her child are a strong indicator of a child’s
own aspirations during adolescence. Parallel to the Status Attainment/Wisconsin models, these
results suggest that children largely aspire to “do as their mothers say, not as they did.” However,
the relationships between women’s adolescent aspirations, women’s maternal aspirations, and
their offspring’s aspirations, vary by maternal educational attainment. Below the differential
relationships are described in terms of recent demographic shifts in youth educational aspirations.
Demographic shifts in educational aspirations
Parallel with literature examining the generational shifts among cohorts of high school
students’ educational aspirations (e.g., Goyette 2008), descriptive analyses reveal the educational
aspirations of G2 youth (on average born in 1987) are much higher than the educational
aspirations of G1 mothers (on average born in 1962). Specifically, G2 youth were less likely to
aspire to attend ‘some college/other training’ compared to their G1 mothers. Instead, G2 youth
22
are more likely to aspire to achieve a Bachelor’s degree or more. G2’s high aspirations may not
be surprising as the majority of their G1 mothers also aspired for their children to attain a
Bachelor’s degree or more. Empirical research suggests these increases in youth educational
aspirations are driven primarily by a national discourse supporting all youth to attend college (see
Goyette 2008; Rosenbaum 2001). As the nation supports the high educational aspirations of
youth, students are increasingly less often presented with viable career or educational programs
outside of a traditional 4-year institution at home, as well as school. Therefore, it is not surprising
that G2 youth were more likely to aspire to reach higher educational attainment levels than their
G1 mothers. The concern is that the discourse may present more risks to student educational
attainment if students’ are not also provided high quality information and student support.
Differential impacts of aspirations
Parallel to existing literature (see, Fan and Chen 2001; Taylor et al. 2004, for reviews),
this study finds in the full analytic sample, the maternal aspirations women report for their
children in adulthood, and not the aspirations they held for themselves as adolescents, are
strongly related to the educational aspirations children report for themselves. The positive
relationship between women’s maternal aspirations and their child’s own aspirations remained
statistically significant even when other important maternal and child covariates were included in
models.
However, when the sample was stratified by women’s educational attainment, a slightly
different story emerged. Mirroring discussions of recent demographic trends (i.e. Goyette, 2008),
the high aspirations of children born to the least educated women, do not appear to be driven by
their mother’s aspirations as an adolescent, nor the aspirations their mother holds for them as an
adult. For example, among samples of adolescents with the least educated mothers, the
23
aspirations women held during adolescence were not predictive of their children’s own adolescent
aspirations, and the maternal aspirations women held for their children were only related to
increases in the odds their children aspired to complete some college or more versus high school
or less. Yet, descriptively, 45% of the children born to women without a high school degree
aspired to attain a Bachelor’s degree or more, and 55% of these children aspired to complete
some college/other training or more. Therefore it appears other sources, not those stemming from
their mothers, are driving the increase in these children’s educational aspirations. In order to
dismantle the attitude-achievement paradox, adolescents with fewer resources must be given the
proper career guidance, along with emotional and financial support. For example, more guidance
and information from school staff related to the educational requirements and qualifications of
specific occupational and career goals adolescents report in late high school. Thus, allowing
youth to design their educational aspirations to more closely mirror their occupational goals,
rather than maintain loose or vague educational aspirations that may or may not help youth
achieve their occupational goals.
On the other hand, among adolescents with the highest educated mothers (i.e. mothers
who completed some college or more), the high educational aspirations women held as
adolescents are significantly related to the odds their offspring will aspire to attain a Bachelor’s
degree or more versus some college, above and beyond her maternal aspirations and other
maternal and child covariates. Taken together, these findings suggest that children of higher
educated women may adopt the high educational aspirations their mothers have for them. These
processes may be set into motion even decades prior when their mothers were adolescents
themselves. Thus, there is an intergenerational transmission of educational aspirations that
appears to be driven by parental discourse and resources among families where mothers aspired
and attained higher education.
24
Limitations and future directions
While the current study contributes greatly to understanding the intergenerational
transmission of educational aspirations, several limitations of should be acknowledged. First, this
study explored the aspirations of youth when both generations were close to the same age (when
possible); yet recent research finds youth aspirations may begin to crystallize as early as 4th grade
and have important implications for youth educational attainment (Bozick et al., 2010). Further,
the stability or volatility of youth educational aspirations may have important implications for
students’ abilities to reach their desired level of educational attainment. Thus, an exciting
extension of this study, which is beyond the scope of the current paper, would be to examine
whether an intergenerational transmission of educational aspirations relates to the stability or
volatility of youth and maternal educational aspirations.
Finally, due to the NLSY study design, the impacts fathers have on youth educational
aspirations was completely ignored. Unfortunately the fathers of NLSY children, as well as the
children of the original cohort of NLSY79 men were not interviewed. Thus, the role of both
parents’ educational aspirations on the formation of their offspring’s aspirations cannot be
accounted for, and the possible differing perspective and impacts G1 fathers have on their
children is not included. Future empirical explorations should include the perspectives and
impacts both parents have on the development of youth educational aspirations, as one half of this
relationship was left uncovered in the current study.
Conclusions and implications for the educational pipeline
Taken together, results from this study suggest that while the United States may have an
educational attainment problem, the country may not have a low aspirations problem. Yet it is
25
also important to remember the political and fiscal context in which such high youth aspirations
occur today. For example, many school systems face drastic cuts in budgets, and often these
budget cuts are first felt in reduced student services, increased class sizes, and less individual
attention. Thus, at a time when many students, aspire to attend college, the resources these
students may depend upon to navigate these complicated pathways, or connect students to support
(i.e. financial or academic) are being undermined. Therefore the gap between students’
educational attainment and aspirations may be related to the limited resources available to schools
and students, and especially those students who depend upon resources at school because they
lack these supports elsewhere.
Currently exciting avenues attempting to increase the college enrollment and attendance
of high school graduates facing risk focus upon reducing inconsistent information and increasing
student support. In one example, researchers randomized varying levels of information regarding
the cost of local universities and supported completing the Free Application for Federal Student
Aid (FAFSA) form when participants came to an H&R block to fill out their 2008 tax returns
(see, Bettinger, Long, Oreopoulos, & Sanbonmatsu, 2012, for details). Results from this study
demonstrate that increases in information and support to families can increase rates of college
enrollment, attendance, and persistence, as well as receipt of financial aid, and submissions of the
FAFSA. However, conditions of increased information without additional support did not result
in improved outcomes, suggesting that in order to patch leaks in the attainment pipeline,
increased information paired with support may be necessary.
By taking advantage of the multi-generational nature of the NLSY79, this study was able
to empirically determine that the educational aspirations mothers hold for their children are
perhaps more important for the educational aspirations of their children above and beyond child
and maternal characteristics. Further, for the children of higher educated women, a transmission
of youth aspirations may exist, suggesting that the positive orientations women held as youth
26
continue through to their future children even decades later. This study points to the importance
of providing more support to students who may be first-generation high school graduates or the
first in their families to pursue a post-secondary education. Thus, as more generations of youth
look to further their educational attainment beyond high school, and attain more complex
programs of study, communities, parents and schools should be aware of the differing needs each
student may possess in order to increase the likelihood students will reach their educational
aspirations.
27
Chapter 3
The Power of Positive Thinking: Exploring Educational Attainment, Aspirations, Engagement and Race
Introduction
Currently on average about 90% of youth earn their high school credentials; yet
compared to about 95% of white students, only 89% of black and 75% of Hispanic students attain
their high school diploma or GED (National Center for Education Statistics; NCES, 2013a).
These educational inequalities persist, and even expand across the educational pipeline so that
40% of white students earned their Bachelors in 2012 compared to only 23% of black and 15% of
Hispanic students (NCES, 2013a). In essence, compared to their white peers, black and Hispanic
students face a heightened risk of prematurely ending their educational programs prior to
completion, and this risk only increases with each level of educational attainment. One promising
avenue of research finds that on the whole all students report high educational aspirations (see,
Kao & Thompson, 2003, for review). Unlike many other educational outcomes, as the
educational disadvantage of a student increases, the aspirations reported by youth also increases,
or are reported to be just as high as their more advantaged peers (Ainsworth-Darnell & Downey,
1998). Therefore unpacking the race and ethnic differences related to the association between
educational aspirations and attainment may help to further our understanding of the gaps in the
educational pipeline.
28
Attitude-achievement paradox
The ‘attitude-achievement’ paradox (Mickelson, 1990), describes a phenomenon in
which black and Hispanic students’ school achievement, success and setbacks, are inversely
related to their educational aspirations (Blau, Moller, & Jones, 2004; Mau & Bikos, 2000; Moller,
Stearns, Blau, & Land, 2006; Stearns, Moller, Blau, & Potochnick, 2012). Often this paradox is
incorporated into discussions of the racial/ethnic achievement gaps, or that black and Hispanic
students may be underperforming but are not necessarily less inspired, than their white peers. For
example, Crosnoe (2001) found a decline in student educational aspirations across high school
was attributed to higher achieving white, rather than lower achieving black and Hispanic youth.
In addition, Blau and colleagues find black students are less likely to drop out after grade
retention (Moller et al., 2006), and more likely to attend college despite low-test scores (Blau et
al., 2004), compared to their their white peers. These researchers argue that perhaps due to
cultural differences, and differential exposures to bias, white students may interpret low-test
scores or grade retention as true indicators of their academic potential, while black students may
attribute these events to systematic discrimination and bias (see Moller et al., 2006, for
discussions). These differences in interpretation may explain why some students may downgrade
their aspirations in the face of academic setbacks, while others may not. Taken together these
findings contradict the Oppositional Culture Explanation (Fordham & Ogbu, 1986; Gibson &
Ogbus, 1991; Ogbu, 1981), as the racial and ethnic related gaps in student educational attainment
do not appear to be driven by low student aspirations or negative school orientations.
29
Familial and intergenerational threads of youth aspirations
Students’ with racial/ethnic identities in which historical oppression and bias are present
often hold very high aspirations and view education as an important pathway for upward mobility
(Blau et al., 2004; Spera, Wentzel, & Matto, 2009). In their comprehensive review Fredricks and
colleagues’ (2004) found students who reported experiencing discrimination, yet felt they had the
support to deal with these experiences were more likely to be engaged in school. Perhaps then it
is through these intergenerational narratives, and parental support that students who face
considerable obstacles in education still value education and remain engaged at school. Research
has found that the decision to drop out of at least high school often begins with the increased
likelihood students partake in other disengaged behaviors such as truancy and that familial and
household characteristics can play an important part in pushing students to stay in, rather than
drop out of, school (see, Tyler & Lofstrom, 2009, for discusions). Therefore this paper explores
the educational aspirations mothers reported for themselves as adolescents, as well as the
aspirations they report for their children, as both measures may (1) support her adolescent
offspring’s aspirations and (2) encourage her offspring to move further in education against all
odds.
Regardless of family race/ethnicity, or SES, research has found most parents report high
educational aspirations for their children (Spera et al., 2009). However, research has
demonstrated parental encouragement and involvement in youth education does vary in
combination with an individual’s race/ethnicity as well as the overall ability of students. For
example, Crosnoe (2001) found black parents were more involved in their child’s education
across all levels of student ability (measured by academic ‘tracks,’) than white parents, and this
difference was especially pronounced for students in lower tracts. He argues that this involvement
is likely driven by the importance parents ascribe to their child’s education, as well as the
30
educational aspirations parents may hold for their children, regardless of their child’s current
abilities. These findings coupled with the previously discussed academic perseverance and
resilience of minority students over their low-test scores or grade retention (as described in Blau
et al., 2004; Moller et al., 2006) calls to attention the lasting impacts that parental aspirations for
children may have when helping some children overcome risk. Further, as black and Hispanic
parents are likely to have their own experiences with racism or bias as an adolescent, these
parents’ aspirations during adolescence could be an indicator of the support and the importance
their parents, and generations before them, placed on education, and the importance of
overcoming obstacles against all odds.
Explorations of similar relationships within Hispanic families are more limited than
empirical explorations contrasting the aspirations of black and white parents, and the differential
impacts these parental aspirations may have on youth attainment. Further, the studies that do
explore the relationships between parental investment and support of youth academics in
Hispanic families are often confounded with issues of acculturation or immigration. In one study
which employed qualitative interviews of Hispanic parents and their children who recently
immigrated to the United States, Behnke and colleagues (2004) found a remarkable degree of
continuity in the educational aspirations reported across generations. However, some youth did
not share their parents’ optimism and reported lower educational aspirations than the aspirations
their parents reported for them. These interviews also underscored the importance of parental
support for education and parent led strategies for overcoming discrimination as adolescents
frequently reported racism and bias as barriers that would keep them from reaching their
academic goals. Other literature exploring important mechanisms driving the academic
attainment of Hispanic youth focus on the importance of the family when shaping youth
outcomes. For example, another study exploring the relationships among Mexican women across
three generations found that each generation named their mother as a significant source of support
31
for their educational attainment, and across generations aspirations mothers held for their
daughters increased (Hernandez, Vargas-Lew, & Martinez, 1994). Yet in literatures discussed
here, the majority of each study’s sample recently immigrated to the United States, while the
current study includes a sample of Hispanic participants in which both parents and their offspring
were born in the United States. Therefore this study also fills gaps in the current literature by
exploring the importance multiple generational measures of educational aspirations among a
sample of Hispanic-American families have on youth educational attainment. Further, by
extension this study can begin to uncover whether or not parental focus and support of youth
education has as important an impact on youth attainment in American born Hispanic families as
they do in immigrant families (see, Behnke et al, 2004; Kao, 2004, for discussions).
Engagement-achievement paradox
In an attempt to explain the ‘attitude-achievement’ paradox, researchers began
investigating the school engagement of minority youth, as doing well in school requires more
from a student than positive thinking alone. Researchers began to question whether black and
Hispanic youth engaged in counter-productive and risky school behaviors, and if it was due to
these behaviors that racial and ethnically based attainment gaps persist. Students with higher
academic achievement are more likely to be engaged in coursework by completing class
assignments, putting more effort into homework, participating in class discussions, or arriving to
class on time (Finn, 1993; 2006). On the other hand, students who disengage from school by
engaging in truancy, missing class, or not completing homework, are more likely to experience
grade retention and lower levels of attainment (Allensworth & Easton, 2005; Fredricks,
Blumenfeld, & Paris, 2004; Hickman et al., 2008). Further these disengaged behaviors often
32
disrupt classroom procedures, and risk the relationships students have with their classroom
teachers (Klem & Connell, 2004; Skinner & Belmont, 1993; Vallerand, Forier, & Guay, 1997).
Some literature exploring the classroom behavior of minority youth has pointed to an
‘engagement-achievement’ paradox (Shernoff & Schmidt, 2008) whereby the lower achievement
of black students compared to white students is not explained through corresponding gaps in
student engagement. For example, when compared to white and Hispanic students, research has
found black students were more likely to report being engaged in school, hold positive school
attitudes, and were less likely to endorse ‘cheating’ or ‘breaking the rules’ as acceptable
(Ainsworth-Darnell & Downey, 1998; Johnson, Crosnoe, & Elder, 2001). Therefore it is possible
that the high aspirations of black students may result in higher levels of classroom engagement,
which one would expect to correspond to higher levels of attainment. Yet in one study when the
relationships between engagement and youth achievement were explored, the engagement of
black students was not associated with equally high returns to academic achievement as their
White and Asian peers (Shernoff & Schmidt, 2008). Rather, school engagement was negatively
associated with the GPA of black students (Shernoff & Schmidt, 2008). Potentially one link
explaining the pro-school behavior yet lower achievement of black youth are the perceptions and
biases of students’ teachers, as teachers (especially non-black teachers) were found to report that
their black students engage in more problem (and distracting) classroom behaviors than their
white peers (Downey & Pribesh, 2004).
While the educational engagement of black student and the impacts their school behavior
may have on their academic achievement is complex, the literature exploring the school behavior
of Hispanic youth is more straightforward. For example one study found Hispanic students were
more attached to school, compared to their black and white peers, while black and white students
reported equal levels of school attachment (Johnson et al., 2001). However, the organization of
the classroom may make a difference in student engagement as well as the type of activity or
33
instructional method applied in each lesson: another study found that Hispanic students were
more engaged during group work and class discussions than other peers, while black students are
highly engaged across all activities (Uekawa, Borman, & Lee, 2007). However, the question
remains what impacts the aspirations parents have for their children as well as the aspirations
parents had during adolescence themselves may have on youth engagement. Further, it is possible
that consistent and positive aspirations collected across multiple stages of parental development
(e.g., adolescence and adulthood) as well as between multiple members of families (e.g. mothers
and youth) may both propel students into positive behavior, as well as higher levels of
educational attainment. Therefore in an attempt to contribute to current literature exploring the
engagement-achievement paradox, this paper explores if the race/ethnic differences in
relationships between intergenerational and youth educational aspirations and attainment are
mediated by the engagement and disengagement behaviors of youth in high school.
Current study
The overarching goal of this paper is to unpack the racial and ethnic related differences in
multiple measures of educational aspirations and student behavior that may drive the
race/ethnicity gaps in youth educational attainment. Specifically, the first aim is to descriptively
explore profiles of youth and their mothers and how these profiles of educational aspirations,
behaviors, and demographic characteristics, may vary by student race/ethnicity. The second aim
of this study is to determine whether youth educational aspirations (G2 youth aspirations),
mother’s educational aspirations reported during adolescence (G1 youth aspirations), and the
maternal aspirations women report for their children (G1 maternal aspirations) predict youth
educational attainment. I hypothesize that in the full sample all three measures of educational
aspirations will be positively associated with all three measures of youth educational attainment.
34
The third aim is to determine whether the relationships between the three measures of
educational aspirations and youth educational attainment vary by student race/ethnicity. I
previously found in the first paper of my dissertation (see Panel C of Table 3, pg. 36) that G1
aspirations during adolescence were significantly related to the educational aspirations of G2
youth born to higher educated women. As white students are more likely to have higher educated
parents than their black and Hispanic peers (NCES, 2010), and as the status attainment models
find youth and maternal aspirations for children are predictive of youth educational attainment
(Spera, Wentzel, & Motta, 2009), I hypothesize all three measures of aspirations will predict G2
youth attainment among white students.
Research links intergenerational narratives of social mobility, as well as parental support
for education as important potential mechanisms for the academic achievement of black youth
(see, Blau et al., 2004). Therefore, I hypothesize that similar to their white peers the educational
attainment of black students will be predicted by all three measures of educational aspirations.
However, these relationships may occur through intergenerational narratives of the value of
education and overcoming structural and racially based obstacles to social mobility, rather than
the accumulation of advantage across generations of white families.
The development of my hypotheses for Hispanic students is slightly more complex. On
the one hand, research finds that some Hispanic youth may have lower educational aspirations
compared to the aspirations their parents report for them (Behnke et al., 2004). These lower
aspirations may be driven by concerns students have for the financial burdens college may place
on their families, or reflective of less clear occupational goals or little information Hispanic
students may have about college in general (Kao & Tienda, 1998). However, other literature finds
that the familial links and parental support for higher educational attainment are especially salient
for Hispanic youth (see, Kao, 2004; Hernandez et al., 1994, for discussions). Similarly parental
led strategies for overcoming racial barriers in school are just as important for the educational
35
success of Hispanic students as they are for black students (Behnke et al., 2004; Fredricks et al.,
2004). Taken together, I hypothesize the educational aspirations Hispanic G2 youth hold for
themselves will be more predictive of their educational attainment than either G1 adolescent or
G1 maternal aspirations. As in this case, youth may hold more realistic educational goals than
their mothers, or potentially only aspire to higher levels of education if their family has the plans
in place, or financial means, to allow for such dreams.
Finally, this paper will explore whether student engagement and disengagement in high
school mediates the relationship between educational aspirations and the likelihood the student
will earn their high school diploma/GED ‘on time’ (aim 4). Explorations of student engagement
and disengagement by race/ethnicity present a complex story between the relationships of student
race/ethnicity, school behavior, and the returns these behaviors may bring to their educational
attainment (see Fredricks et al., 2004, for discussions). Therefore mediation models will be
conducted for each race/ethnicity separately. In efforts to streamline these complex relationships,
only the educational aspirations that significantly predict the likelihood G2 youth earned their
high school credentials on time in aim 3 will be included in models.
For this aim I hypothesize that the high school engagement and disengagement of white
and Hispanic students will at least partially mediate the association between aspirations and
educational attainment. However, I hypothesize that the relationship between educational
aspirations and the odds black students earn their high school credentials on time will not be
mediated by the time they spend on homework or whether or not they engage in truancy in high
school. I do not doubt that the time youth spend on homework or whether they skip school (or
not) will impact their odds for earning their high school degree on time. Rather, I hypothesize the
consequences of high school disengagement, or the payoffs of high school engagement, are not
going to explain, nor alter the aspirations of black youth. Instead I hypothesize, beyond potential
consequences or payoffs to students’ engagement/disengagement, black students’ aspirations and
36
attainment will be strongly linked. This link is likely to stem from an intergenerational narrative
or parental supports for education that may not be closely aligned with youth ability or obstacles
in school (see, Crosnoe, 2001).
The current study extends literature exploring educational inequalities, youth behavior
and youth aspirations in multiple ways. First, literature has suggested for some students their
cultural and historical backgrounds may provide an important support network for their values in
education and persistence to achieve higher levels of attainment (Ainsworth-Darnell & Downey,
1998; Blau et al., 2004). This study is one of the first to incorporate measures of educational
aspirations across two generations to explore whether the positive orientations women had during
adolescence, as well as the aspirations women hold for their children, and youth hold for
themselves may work together when predicting G2 youth’s final educational attainment. Further,
although the racial and ethnic diversity of the United States continues to increase (Perez &
Hirschman, 2009; U.S. Census, 2012), the empirical explorations of these relationships for
Hispanic students has occurred relatively less often than for white youth, or between white and
black youth alone. Therefore the expansion of this study to explore these relationships for white,
black and Hispanic students can present a broader picture of the sources and mechanisms driving
youth attainment in three of the largest racial/ethnic groups in the United States.
Method
Data
Data for this paper come from the National Longitudinal Survey of Youth 1979 Cohort
(NLSY79), and the linked Children and Young Adult files (CNSLY). NLSY data are sponsored
by the U.S. Department of Labor and have been complied through the Ohio State University
37
Center for Human Resource Research. Beginning in 1979 the NLSY followed a nationally
representative sample of 6,403 men and 6,283 women who were 14 to 21 years old in 1978
(referred to as G1 from here on after). Biennially starting in 1986 the NLSY79 was expanded to
include mother supplemental surveys, in which G1 women reported additional demographic, and
developmental information for their offspring (referred to as G2 from here on after). Beginning in
1988, and biennially thereafter, G2 children who were 10 or older directly reported information
related to their experiences and development. Across time the NLSY has achieved high rates of
participant retention, and the CNLSY files contain information on over 11,000 children born to
NLSY79 women as of the 2010 survey wave.
Analytic sample
Between 1970 and 2010 11,503 children were born to 4,924 NLSY79 mothers. The
sample was first restricted to youth who were at least 22 years old by 2010 as younger
participants would be less likely to have the opportunity to achieve the high levels of youth
educational attainment explored in the paper (3,868 cases excluded). In addition the sample was
restricted to cases with valid data across all three measures of youth educational attainment
(2,295 cases excluded). Next, cases were excluded if their mothers were older than 19 during the
time they reported their own educational aspirations (1,721 cases excluded), and if cases were
missing values on any measure of the three educational aspirations (1,112 cases excluded). An
additional 837 cases were excluded for missing valid information on their engagement and
disengagement behaviors during high school, and 10 cases were excluded for missing high levels
of information across all variables (missing 20% or more data on analytic variables). Thus the
final analytic sample consists of 1,660 youth and their 1,191 mothers.
38
Measures
Youth educational attainment
Beginning in 1994, youth who were turning 15 (or older) reported biennially at each
survey wave whether they had earned their high school diploma/ GED, the date they earned their
high school credentials, whether they were enrolled in college (and whether they were enrolled
part-time or full time), and the highest grade or year of school they had completed. From these
measures of educational attainment and the timeliness of these achievements three dichotomous
dependent variables were created in which the primary focus was to predict: (1) whether the
youth earned their high school credentials ‘on-time’ (prior to their 19th birthday), (2) whether the
youth ever enrolled in college, and (3) whether the youth earned a Bachelor’s degree or more.
These measures were created to mirror similar measures developed by NCES and the Department
of Education that capture national rates of the timeliness of high school graduation, persistence in
college enrollment and overall educational attainment levels (NCES, 2013b).
Youth educational aspirations
Biennially from 1986 to 2010, when G2 youth were at least 10 years old, children
reported how far in school they thought they would reach. From this continuous measure, an
ordinal measure of G1’s aspirations during adolescence was created where 0= ‘less than high
school diploma/GED,’ 1= ‘high school diploma/GED,’ 2= ‘some college/other training,’ 3=
‘attain a Bachelor’s degree,’ and 4= ‘attain more than a Bachelor’s degree.’ To decrease the
likelihood youth would increase or decrease their aspirations as they were closer to each
educational attainment milestone, or consequently alter their school engagement and
39
disengagement G2 measures of youth educational aspirations were taken from the survey in
which youth were 14 to 15 years old.
Mother educational aspirations during adolescence
G1 educational aspirations were taken from the first NLSY data survey in 1979 where
women still enrolled in high school reported how far in school they expected to reach (i.e. what
year or grade of school) (mean age in years= 16.78). Similar to their offspring’s’ educational
aspirations, G1’s aspirations during adolescence was recoded into an ordinal measure where 0=
‘less than high school diploma/GED,’ 1= ‘high school diploma/GED,’ 2= ‘some college/other
training,’ 3= ‘attain a Bachelor’s degree,’ and 4= ‘attain more than a Bachelor’s degree.’ The
developmental timing of this measure was chosen in efforts to maximize the number of G1
women who were still adolescents, as well as reduce the likelihood that academic milestones
(e.g., graduated from high school) and adult experiences would impact women’s educational
aspirations.
Maternal educational aspirations for child
From 1988- 2010, G1 women reported how far they thought their child would go in
school. To mirror the developmental timing of other key variables, mother’s responses were taken
from the calendar years when children were 12- 15 years old. To parallel the other measures of
educational aspirations, mother’s aspirations for their child’s education was recoded into an
ordinal variable where 0= ‘less than a high school diploma/GED,’ 1= ‘attain a high school
diploma/GED,’ 2= ‘complete some college/other training,’ 3= ‘attain a Bachelor’s degree’ and 4=
‘attain more than a Bachelor’s degree.’
40
Youth school engagement
Youth school engagement is captured by youth report on how many hours per week they
worked on their homework (a) inside and (b) outside of school. These two measures were
aggregated and averaged across surveys when youth were 16 to 18 years old, thus indicating the
average hours per week youth reported spending on homework.
Youth school disengagement
Youth school disengagement is represented by youth reports of ‘skipping school or
classes without permission,’ when they were 16- 18 years old. Youth are given a value of ‘1’
indicating they engaged in truancy if they positively reported skipping school during this time.
Race/ethnicity
In 1979 NLSY79 participants reported their race/ethnicity. Children were then assigned
the same race/ethnicity as their mothers: white (reference), black and Hispanic.
Covariates
Models include child, maternal, and household characteristics that are correlated to
educational aspirations, student engagement/disengagement, and youth educational attainment.
G2 youth characteristics include gender [female, male (reference)], race/ethnicity [White
(reference), Hispanic, Black], child’s total standard score on the math subset of the Peabody
Individual Achievement Test scale (PIAT; Dunn and Markwardt, 1970) when they were 8 or 9
years old, whether the child participated in a special education program at school (1= yes, 0= no),
41
whether the child experienced grade retention by 8th grade (1= held back by 8th grade; 0 = not
held back by 8th grade), and number of siblings. Child affirmative responses to five items
indicating negative peer groups (e.g., my friends pressure me to ‘try cigarettes,’ ‘try alcohol,’
‘commit violence or crime,’) when youth were 12 to 13 years old were also included in models.
Items were averaged and higher scores indicate higher levels of negative peer pressures. The
proportion of years from birth to age 13 children were exposed to: poverty and maternal
unemployment, and the average percentile score of their household’s cognitive stimulation and
emotional support when children were 0 – 5 from the Home Observation Measure of the
Environment- Short Form (HOME-SF; Caldwell and Bradley 1984) were also included in
models. G1 mother characteristics at child’s birth include: age, educational attainment [less than
high school, High School Diploma/GED (reference), some college or more], marital status (1=
not married, 0= married) were included in models. Further, a measure of maternal cognitive skills
administered in 1981 from the Armed Forces Qualification Test (AFQT) was also included in
models.
Analytic plan
To address aim 1, descriptive and multivariate regression analyses were performed using
Stata SE 12.1 (Stata Corp LP, College Station, TX). Weighted descriptive statistics were obtained
from non-imputed data using the NLSY79 2010 youth population weight (Y2615900). The
analytic sample was stratified by participant race/ethnicity and mean differences were conducted
using analysis of variance for continuous variables and chi-square tests for dichotomous
variables. Prior to conducting multivariate logistic regression models, multiple imputation
techniques were used to impute missing data using the ice command for participants missing less
than 20% of overall data. Compared to listwise deletion, multiple imputation processes reduce
42
biases in parameter estimates and standard errors by maintaining a closer representation of this
sample to the population (Graham, 2009; Graham & Schaefer, 1999). Last, standard errors in all
regression models were adjusted to account for the lack of independence of observations as
children are clustered within families.
For aim 2, a series of multivariate logistic regressions were conducted to explore the
relationships between multiple measures of educational aspirations across generations and
offspring educational attainment. For aim 3, the sample was then stratified by race/ethnicity to
explore whether the relationships between the three measures of educational aspirations and G2
educational attainment vary by race/ethnicity. For aim 4, methods described in Barron and Kenny
(1986) were followed to explore whether student high school engagement and disengagement
mediated the relationship between measures of educational aspirations and youth educational
attainment separately for each race/ethnicity. First, multivariate logistic regressions were
conducted to explore whether measures of educational aspirations predict the odds G2 youth
would earn their high school diploma/GED ‘on time.’ Next, OLS and logistic regressions were
conducted to explore whether the measures of educational aspirations were significantly
associated with youth engagement (i.e. hours spent on homework a week) and youth
disengagement (i.e. truancy) in high school when G2 youth were 16 – 18 years old. Multivariate
logistic regressions were then conducted to determine whether measures of youth engagement
and disengagement in high school were related to the odds G2 youth earned their high school
credentials ‘on time.’ Finally measures of youth engagement and disengagement in high school
were included in a model where educational aspirations predict youth educational attainment, and
the significance of potential mediating effects were tested following Sobel (1982). In all models
the full set of child and maternal characteristics were included.
43
Results
Descriptive statistics
Table 1 and Table 2 present the weighted descriptive statistics for the full analytic sample
(nG1= 1,191, nG2= 1,660). The majority of G2 youth earned their high school diploma/GED ‘on
time,’ (79%), 58% of the analytic sample enrolled in college, and 19% of the analytic sample
earned a Bachelor’s degree or more. Almost 70% of G2 youth aspired to attain a Bachelor’s
degree, compared to only 26% of their G1 mothers. However, G1 reported high aspirations for
their children as 60% of G1 women aspired for their children to attain a Bachelor’s degree or
more. Thirty-three percent of G2 youth reported reporting skipping school when they were 16- 18
years old, and during the same time spent on average just less than 8 hours (M= 7.81, SD= 7.53) a
week on homework. White G2 youth were more likely to earn their high school credentials ‘on
time,’ enroll in college and earn a Bachelor’s degree or more compared to their black and
Hispanic counter parts. Black G2 youth reported spending the fewest hours a week on homework
and were also less likely to report skipping school than their white and Hispanic peers.
The analytic sample was predominantly White (76%), followed by Black (17%) and then
Hispanic (7%), with 6% of G2 youth participating in special education and 13% reported being
held back a grade by the time they were in 8th grade (Table 2). G2 youth had on average just
fewer than two siblings (M= 1.82, SD= 1.23) and lived in households under the federal poverty
line (FPL) about 26% of the years from birth to age 13 (a little more than 3 years). At the time of
G2’s birth, G1 mothers were approximately 23 years old (M= 22.68, SD= 2.95), the majority had
their high school diploma (57%) or some college or more education (23%), and were married
(70%). On average G1 mothers were unemployed for just under 5 years from the time G2 was
born to the time G2 turned 13 (M= 0.41, SD= 0.30). Black G2 youth spent more time under the
44
poverty line from birth to 13, and in households with lower ratings of emotional support from
birth to 5 than their white and Hispanic peers. G2 white youth were born more often to higher
educated women who had higher cognitive scores than black and Hispanic youth in the sample.
Multivariate logistic regressions
Aspirations and attainment
Table 3 presents the multivariate logistic regression results in which the associations
between the three measures of educational aspirations and G2 youth educational attainment were
explored. Higher G2 youth educational aspirations were associated with a 34% increase in the
odds G2 youth would attain their high school diploma/GED ‘on time’ (Panel A). Further, higher
G2 youth educational aspirations were associated with a 39% increase in the odds G2 youth
would enroll in college as well as attain a Bachelor’s degree or more (Panel A). Higher G1 youth
educational aspirations were not significantly related to the odds G2 youth would attain their high
school diploma/GED ‘on time,’ enroll in college, or attain a Bachelor’s degree or more. However,
higher G1 aspirations for her child were related to a 19% increase in the odds G2 youth would
attain a high school diploma/GED ‘on time,’ a 32% increase in the odds G2 youth would enroll in
college, and a 56% increase in the odds G2 youth would attain a Bachelor’s degree or more.
Aspirations and attainment by race/ethnicity
Panel B of Table 3 presents the multivariate logistic regression results where the
associations between three measures of educational aspirations and G2 youth educational
attainment were explored in the white sub-sample. For white youth, higher G2 youth aspirations
45
were associated with a 36%, a 47%, and a 44% increase in the odds G2 youth would attain a high
school diploma/GED ‘on time,’ enroll in college, and attain a Bachelor’s degree or more
(respectively). Higher G1 aspirations for her child, in the white sample were associated with a
36%, a 50% and a 55% increase in the odds G2 youth would attain their high school
diploma/GED ‘on time,’ enroll in college, and attain a Bachelor’s degree or more (respectively).
Higher G1 youth educational aspirations were not significantly associated with any measure of
G2 educational attainment in the white sample.
Panel C and Panel D of Table 3 present the multivariate logistic regression results for the
black (Panel C) and Hispanic (Panel D) samples where the associations between three measures
of educational aspirations and G2 youth educational attainment were explored. In the black
sample, higher G2 youth aspirations were related to a 28% and 36% increase in the odds G2
youth would attain a high school diploma/GED ‘on time,’ and enroll in college. Higher G1
educational aspirations for her child is associated with a two fold increase in the odds black G2
youth would attain a Bachelor’s degree or more; however G1 youth educational aspirations were
not significantly related to the educational attainment of G2 black youth. Finally, in the Hispanic
subsample, higher G2 youth educational aspirations were associated with a 39%, 32%, and 75%
increase in the odds Hispanic G2 youth would attain a high school diploma/GED ‘on time,’ enroll
in college, and attain a Bachelor’s degree or more (respectively). G1 maternal aspirations and G1
youth aspirations were not significantly associated with any measure of Hispanic G2 youth
educational attainment.
School engagement/disengagement
Table 4 presents the multivariate regression results in which the mediating mechanisms
of youth high school engagement (Panel A) and disengagement (Panel B) on the relationship
46
between G2 youth aspirations, G1 maternal aspirations and the likelihood G2 would earn their
high school diploma/GED ‘on time’ were explored for white youth. In Panel A, G2 youth
aspirations and G1 maternal aspirations were associated with a respectively 35% and a 34%
increase in the odds G2 white youth would attain their high school credentials on time controlling
for child and maternal covariates as well as youth school disengagement (Model 1). Further, G2
youth aspirations and G1 maternal aspirations were positively and significantly associated with
the time G2 white youth reported spending on homework (Model 2). However, the school
engagement of white youth (measured by hours/week on homework) was not significantly
associated with the odds of G2 earning their high school credentials on time; therefore mediation
was not established (Model 3). In panel B, higher G2 youth aspirations and higher G1 maternal
aspirations was related to a 37% and 35% increase in the odds that G2 earned a high school
diploma/GED ‘on time’ (Model 1). However, G2 disengagement in high school (measured by
truancy) was determined not to mediate the relationship between G1 maternal and G2 youth
educational aspirations and attainment of a high school diploma/GED on time, as neither G2
youth educational nor G1 maternal educational aspirations significantly predict G2
disengagement (Model 2).
Table 5 present the multivariate regression results where the mediation of youth
engagement and disengagement from high school was explored in the black sample. Although G2
youth educational aspirations were associated with a 30% increase in the odds black youth would
earn their high school credentials on time (Model 4, Panel A and Panel B), G2 youth aspirations
were not significantly associated with G2 engagement (Model 2, Panel A) nor disengagement
(Model 2, Panel B). Therefore, the engagement and disengagement of black students in high
school did not mediate the relationships between G2 youth aspirations and the likelihood G2
would earn their high school credentials on time.
47
Table 6 presents the multivariate regression results where the mediation of youth
engagement and disengagement from high school on the relationship between G2 youth
aspirations and G2 earning their high school credentials on time was explored in the Hispanic
sample. The regression results reveal that the aspirations of G2 Hispanic youth were positively
and significantly associated with the hours/week students spent on homework (Model 2, Panel
A). Further the school engagement of Hispanic students was associated with a 6% increase in the
odds students would earn their high school credentials on time (Model 3, Panel A). Last, the
relationship between G2 youth educational aspirations and the likelihood G2 would earn their
high school credentials on time decreased, albeit slightly, with the inclusion of youth school
engagement in the model (Model 1, OR= 1.36, p<.05, to Model 4, OR = 1.32 p<.05). However,
the Sobel test determined the mediating effect did not reflect a statistically significant change.
Finally, the disengagement of Hispanic youth in high school does not mediate the relationship
between G2 youth aspirations and G2 attainment of a high school diploma/GED on time. G2
youth aspirations were not significantly associated with high school disengagement (Model 2,
Panel B).
Sensitivity analyses
Two additional sets of sensitivity analyses were conducted to test the robustness of the
main findings. Multivariate logistic regressions were conducted to determine whether the
relationships between the three measures of educational aspirations and G2 youth educational
attainment varied by (1) maternal educational attainment and (2) G2 exposure to poverty.
Findings were consistent with the main findings presented (results available upon request). Next,
mediating models were tested similar to those presented in Tables 4- 6 yet where only the
relationship between truancy or hours spent on homework were included in models. Again, these
48
results again supported the main findings presented (available upon request). Finally, additional
mediation models were explored with a transformed version of the hours youth spent on
homework (by taking the log of the square root to adjust for the non-normal distribution of the
variable). Results from models including this transformed variable support the main findings
discussed (available upon request).
Discussion
This study explored racial and ethnic differences in the relationships between youth
educational attainment and measures of youth and maternal educational aspirations across
multiple generations. Multivariate logistic regressions conducted with the full sample revealed
that G2 youth educational aspirations and G1 maternal educational aspirations for her child were
predictive of the likelihood G2 youth would (a) earn their high school credentials ‘on time,’ (b)
enroll in college and, (c) earn a Bachelor’s degree or more. However, when the sample was
stratified by student race/ethnicity, the above findings appear to be driven largely by white youth
and their families. On the other hand, the educational aspirations of black and Hispanic G2 youth
appeared to be more consistent predictors of youth educational attainment than the aspirations
their mother’s hold for them, or the aspirations their mothers had as adolescents. Further, this
study explored whether the relationships between the intergenerational and familial measures of
educational aspirations and the timeliness of youth high school educational attainment were
explained by the engaged and disengaged behaviors of youth in high school separately for white,
black and Hispanic families. Overall findings indicate that the hours youth reported spending on
homework, and their reported truancy in high school, do not mediate the positive relationships
between measures of educational aspirations and youth high school attainment. This did not vary
by student race or ethnicity. Below the racial and ethnic variations in relationships between G2
49
youth, G1 youth and G1 maternal educational aspirations and G2 youth attainment, as well as
possible explanations for why measures of youth engagement and disengagement in high school
did not mediate these associations across all race/ethnicities are discussed.
Differential reports of risk by race/ethnicity
Consistent with the well documented racial and ethnic gaps in youth educational
attainment (e.g. NCES, 2013; Tienda, 2013), descriptive statistics revealed higher percentages of
white G2 youth in the analytic sample earned their high school credentials on time, enrolled in
college, and earned a Bachelor’s degree or more, compared to their black or Hispanic peers.
Therefore while trends in youth educational attainment suggest more racial and ethnic minorities
are enrolling in and graduating from college (Tienda, 2013), descriptive findings from this study
indicate that black and Hispanic students still lag behind their white peers in reaching higher
levels of educational attainment. Further, when groups of students are collapsed across cohorts
these racial and ethnic gaps in educational attainment are striking.
Yet, descriptive statistics also suggest a slightly different story in regards to patterns of
student engagement (hours per week spent on homework on average) and disengagement
(skipped a day of school without permission) by race/ethnicity. While black students reported
spending the least time on their homework (compared to their white and Hispanic peers) they
were also the least likely to report engaging in truancy during high school. Parallel with other
literature, these findings suggest that it may be necessary to explore the relationships between
student race/ethnicity and student engagement and disengagement via multiple constructs, as
when only explored by student engagement or student disengagement, an incomplete portrait of
student behavior may emerge (Ainsworth-Darnell & Downey, 1998). However, more work is
needed to explore both student reports of engagement and disengagement behaviors with more
50
measures spanning the three domains of engagement (see Fredricks et al, 2004, for discussion) as
well as engaging reports of student behavior from administrators, teachers, or peers, as other
literature finds contradictory trends emerge when student versus teacher or peer reports of student
classroom behavior are explored (Downey & Pribesh, 2004).
Additionally, half of the Hispanic students in this sample reported skipping school during
high school. At the same time while these students were engaging in high rates of truancy, they
were also spending significantly less time on average on their homework than their white peers.
Therefore descriptive statistics from this study highlight the need for more literature to explore
the engaged and disengaged behaviors of Hispanic students, as the current literature surrounding
the engagement and disengagement of diverse samples of youth are often limited, or when
applied to Hispanic youth focus more upon special issues related to immigrant families and their
youth (e.g., Behnke et al., 2004; Hernandez et al., 1994; Kao, 2004; Kao & Tienda, 1998). Thus,
understanding more about the processes that lead Hispanic students to disengage in school and
instead engage in risky school behaviors (such as truancy) compared to their peers may have
exciting implications for educators and policy makers looking to increase the academic success of
this growing population of youth.
Mirroring the racial/ethnic differences in the rates of educational attainment, and most
likely contributing to them were striking differences in the profiles of youth educational and
environmental adversity. More than twice the numbers of black and Hispanic students, compared
to white students in this sample were held back by 8th grade and reported participating in their
schools’ special education program. Likely one source contributing to the racial/ethnic
differences in grade retention and special education placement was the nearly 10 point difference
in students’ scores on the math section of the PIAT (Dunn and Markwardt, 1970) when black and
Hispanic (compared to white) students were 9 or 10 years old. This finding is particularly
worrisome as recent literature finds student math skills are an important indicator of later student
51
achievement in math as well as reading (Duncan et al., 2007). Further, the combined risk of low
math skills, grade retention and special education placement likely present serious implications
for students’ later adolescent and adult outcomes, as other research find persistent academic and
behavior problems are highly and negatively associated with lower rates of graduation and
college enrollment (Duncan and Magnuson, 2011).
The varied experiences of students based on race and ethnicity are not just limited to the
school context, since descriptive statistics found black and Hispanic students were exposed both
to higher levels of household poverty and maternal unemployment than their white peers. Further,
the average emotional and cognitive quality of their household environments (as captured by the
HOME scales, Caldwell and Bradley 1984) was also lower in black and Hispanic households
from birth to age 5 compared to their white peers. Black and Hispanic children in the sample were
also less likely to have highly educated, or married mothers, perhaps contributing both to
economic and emotional strain mirrored throughout these measures.
Uniformity of youth educational aspirations
When exploring the educational aspirations of youth in the sample, especially in light of
the racial/ethnic related differences in youth educational attainment and exposure to adversity, the
overall consistency and high level of youth educational aspirations is all the more astounding.
The majority G2 youth in the sample aspired to at least attain a Bachelor’s degree when they were
14- 15 years old, and these high aspirations were consistent across groups of white, black and
Hispanic youth. This finding mirrors recent literature uncovering demographic trends of high
school sophomores in which increasingly the educational aspirations of youth are rising, and the
greatest increases in youth educational aspirations occur in youth with the least privileged
backgrounds (Goyette, 2008). The high aspirations of youth in this sample, despite their
52
economic or educational adversity parallel the battery of literature uncovering the attitude-
achievement paradox (e.g., Kao & Thompson, 2003). Further, similar to other literature (e.g.,
Spera et al., 2009), the majority of G1 women in the sample aspired for their children to at least
attain a Bachelor’s degree. This was true across all race/ethnicities, and the aspirations were at
much higher rates than the aspirations they held for themselves decades earlier.
Educational aspirations and youth attainment
The second aim of this study was to explore the role G2 youth educational aspirations,
G1 youth educational aspirations and G1 maternal aspirations for her child have when predicting
the likelihood G2 would earn their high school credentials on time, enroll in college and attain a
Bachelor’s degree or more. In the full sample, my hypothesis that all three measures of
educational aspirations would predict G2 youth educational attainment was partially supported.
Parallel to other literature (e.g., De Civita, Pagani, Vitaro, & Tremblay, 2004; Downey,
Ainsworth, & Qian, 2009; Spera, 2005) multivariate logistic regressions revealed that G2 youth
and G1 maternal educational aspirations were predictive of the likelihood G2 youth would attain
all three levels of educational attainment explored in the study. However, the educational
aspirations G1 women reported during adolescence were not predictive of any level of their
offspring’s educational attainment above and beyond the other two measures of educational
aspirations and a host of child and maternal characteristics. These findings provide at least
preliminary evidence of the arguments developed in Spera and colleagues’ (2009) study. They
found once other characteristics of households and parents were controlled for (i.e. parental
educational attainment, exposure to poverty, employment), the impacts mother’s earlier
aspirations have on youth attainment were reduced to non-significance. Thus, the results highlight
the importance that other family and household resources can have on later youth educational
53
attainment. At the same time, as G1 maternal and G2 educational aspirations were significantly
associated in the full sample with the odds G2 youth would earn all three levels of educational
attainment even after the inclusion of other characteristics, the power in which more recent
positive aspirations of parents and youth have for youth attainment is impressive.
Next, the sample was stratified by student race/ethnicity. This was done because the
literature suggests that the educational aspirations women hold for their children and the
aspirations youth hold for themselves may matter more for minority youth aspirations and
attainment than the aspirations and attainment of white students (e.g., Blau et al., 2004). Contrary
to my hypotheses, multivariate logistic regressions revealed that the educational aspirations of G2
youth and the maternal aspirations G1 women reported for their children were positively and
significantly associated with the odds white G2 youth would reach all three measures of youth
educational attainment. Whereas, overall the educational aspirations black and Hispanic G2 youth
reported were more consistently related to youth educational attainment. While G1 maternal
aspirations reported by black women were significantly associated with the odds their offspring
would attain a Bachelor’s degree or more, the G1 maternal aspirations of Hispanic women and
G1 youth aspirations of all women were not significantly associated with any measure of youth
attainment. Taken together these findings suggest that perhaps in order for gains to be seen in
youth educational attainment beyond those found in youth educational aspirations, high maternal
aspirations may need to be paired with equally high household or parental resources such as
education, income or earlier student educational skill.
As briefly discussed above, demographic trends suggest the educational aspirations of
youth over time appear to be increasing at unprecedented levels (Goyette, 2008). Therefore, these
youth-level forces may be washing out the influences of family or parental-level narratives
pushing youth educational attainment, as the aspirations of entire cohorts of youth may be rising.
Thus, the historical narratives of social mobility and overcoming racially charged obstacles or
54
prejudices may still be occurring; yet at the same time all youth regardless of their race/ethnicity
may simply aspire to go to college along with the rest of their peers. On the other hand, it is also
possible that the past aspirations women report for themselves simply do not matter as much as
the current aspirations they have for their children, or above and beyond the impacts other
parental and household resources have on their child’s education. However, it is also possible that
high G1 educational aspirations may relate to more stability in supports or investments parents
later make in the educational attainment of their children, and those were not able to be explored
in the current study. Thus, more work exploring the relationships between earlier measures of
parental educational aspirations and stability or change over time should be explored.
Taken together these findings suggest that the high educational aspirations youth report
as adolescents should not be discounted when predicting youth educational attainment, as
consistently these aspirations were highly predictive of youth attainment across all races and
ethnicities. Further, these findings suggest that the racial and ethnic-based leaks in the educational
pipeline do not appear to be driven by parallel racial/ethnic gaps in the educational orientations of
youth. Instead, all youth in this study reported high educational aspirations, and these aspirations
appear to be important for the later educational attainment of all students. Unfortunately, this
study did not find an intergenerational link between G1 youth aspirations and G2 youth
educational attainment, and this lack of a statistically significant relationship did not vary by
student race/ethnicity. Yet, maternal aspirations were robust predictors of the educational
attainment of white youth, and in one instance associated with increased educational attainment
for black youth. Consequently white students in this study were found to experience much less
economic and educational adversity leading up to high school. Therefore addressing racial/ethnic
variations in exposure to adversity and economic inequality are likely to serve as a better lever to
close the racial/ethnic gaps in youth educational attainment.
55
Educational aspirations, engagement/disengagement and attainment
The last aim of this study was to explore whether the hours youth spent on average per
week on homework and truancy in high school mediated the associations between measures of
educational aspirations and the likelihood G2 youth would graduate high school on time, by
student race/ethnicity. Across all groups of students reports of the time they spent on homework a
week, as well as their truant behavior, did not mediate the relationships between educational
aspirations and the likelihood G2 youth would earn their high school credentials on time. Further,
the break down of potential mediating relationships occurred at different points in the test of
mediation for white, black and Hispanic students. For example, high G2 youth aspirations were
not significantly predictive of the time black students spent on homework or their reports of
truancy thus suggesting that the relationship between G2 aspirations and G2 attainment simply
does not flow through these two measures of youth engagement/disengagement. While for white
and Hispanic students, G2 youth aspirations (and G1 maternal aspirations in the case of the white
sample) were predictive of the time youth spent on homework, the time they spent on homework
was either not associated with their attainment (for white students) or did not reduce the direct
effect from aspirations to attainment (for Hispanic students). As discussed in Fredrick and
colleagues’ (2004) review of youth engagement and disengagement, it is possible that the time
youth spend on homework is not correlated with deeper and more nuanced constructs of student
engagement that are also important for youth academic success, such as student effort, quality of
work, or interest. Further, these deeper constructs of emotional and cognitive engagement are
both harder to measure, yet potentially more relevant for youth academic success, than a simple
count measure of time spent on homework. In regards to youth truancy it is possible that this
measure was too far along the spectrum of behavioral disengagement. Future work could instead
explore measures of more ‘minor offenses’ of student disengagement that may ultimately
56
aggregate and lead to truancy or low educational attainment. Examples of minor offenses include
rates of attendance, participation in class discussion, or tardiness, all which are not currently
captured in the NLSY79.
Limitations and future directions
While the current study extends literature exploring the determinants of youth
educational attainment, as well as the variations in attainment, engagement/disengagement, and
multiple measures of educational aspirations by race/ethnicity several limitations of the study
should be acknowledged. First, reviews of studies exploring youth engagement and
disengagement in school find that youth behavior may be differentially interpreted and reported
by teachers, students and their peers. However, this study was only able to incorporate measures
of youth behavior as reported by youth directly, as teacher, parent or peer perceptions of youth
behavior in the sample was not collected by the NLSY79. Literature also suggest that youth
engagement and disengagement in education may be better captured through measures across
multiple domains, such as youth emotional engagement, youth cognitive engagement and youth
behavioral engagement in the classroom (see Fredricks et al., 2004; Kelly, 2008). Therefore, it is
possible that the measures included in this study may not be as related to the timeliness of youth
educational attainment compared to other measures of youth engagement. Thus, more work
exploring other constructs of youth engagement in education broadly, as well as more finitely in
the classroom should be explored. Last, the NLSY also does not include measures of youth
engagement/disengagement at higher levels of youth educational attainment, such as effort or
interest in college courses, continuous enrollment, and credits earned each semester. Therefore
examining youth and parent educational aspirations on the engagement/disengagement and
57
persistence of youth educational attainment at higher levels of educational attainment will be an
exciting avenue of research; however, this was out of the scope of this particular study.
Conclusions
This study finds that profiles of youth academic engagement/disengagement and
adversity vary by student race/ethnicity. While the educational aspirations of youth and the
educational aspirations that mothers have for their children are high for all students, white
students continue to complete higher levels of education. Both youth and maternal educational
aspirations for her child appear to be robustly and positively related to the likelihood that white
youth will earn their high school credentials on time, enroll in college and earn a Bachelor’s
degree or more. For black and Hispanic youth, the educational aspirations that they hold for
themselves, and not the aspirations that their mothers have for them, are related to their
educational attainment. Overall, youth’s positive educational aspirations may not be enough to
close the educational attainment gap. To improve the educational attainment of black and
Hispanic students, we must take advantage of the positive attitudes of black and Hispanic
individuals and strategically target these inequalities through focused education policies and
educational support systems.
58
Chapter 4
Conclusion
The overarching goal of this dissertation was to explore whether an intergenerational
transmission of educational aspirations (collected from two generations when both were
adolescent students) relates to (a) the formation of youth educational aspirations and (b) youth
later educational attainment, as well as the behavior of students in high school. In addition, both
papers explored the similarities and differences in these relationships by the educational
attainment mothers’ had at the birth of their child, and the race/ethnicity of families. Both
characteristics (i.e. race/ethnicity and maternal educational attainment) are linked to discrepancies
in our understanding of the formation of youth aspirations in recent cohorts of students (see
Goyette, 2008; Roderick, Coca, & Nagoka, 2011) as well as the link between aspirations and
youth attainment and achievement (see Mickelson, 1990). Further, these two characteristics are
also highly correlated with individual socio-economic status, and incidences of family and
student level adversity (e.g., Anderson, 2012; Fantuzzo, LeBoeuf, Rouse, & Chen, 2012; NCES,
2011; U.S. Census Bureau, 2013). Therefore examining the potential similarities or differences
that may emerge between these groups related to intergenerational forces shaping youth
aspirations, behavior, or attainment may potentially inform educational policies and institutions
that look to increase the educational success of students facing risk.
Summary of findings paper 1
Descriptive findings from the first study highlight the positive shift that has occurred over
time between the adolescent educational aspirations of a sample of women and the educational
59
aspirations of their adolescent offspring. Further, in the full sample, youth aspirations appear to
be positively associated with the maternal aspirations their mothers reported for them during
adulthood, not the aspirations their mothers reported for their own education as adolescents.
However, differences did emerge when the analytic sample was stratified by the level of maternal
educational attainment at the birth of the child. While, children of least educated women were
found to have higher educational aspirations during adolescence than their mothers at similar
ages, these aspirations were not associated with (a) the aspirations their mothers reported during
adolescence and (b) many times the aspirations their mothers held for their offspring. The
aspirations women reported for their children were only significantly associated with the
likelihood G2 youth would aspire to complete at least some college education compared to less
than high school or their high school diploma/GED. Contrasted to these findings, the maternal
aspirations women held for their children appear to be an important predictor of their offspring’s
adolescent educational aspirations in the sample of women who had a high school diploma/GED
at the time of their child’s birth and women who had at least some college education at the time
of their child’s birth. In addition, for children of the highest educated women, the educational
aspirations women held for their own education as adolescents was positively associated with
their offspring’s aspiration to attain a Bachelor’s degree or more, compared to some college.
Paper 1 was presented with a unique opportunity to explore how the aspirations women
held as adolescent students may be transferred to their offspring decades later when their
offspring were of similar ages. Further, the data employed by this study also included other
important characteristics of women and their children that may serve as mechanisms allowing for
or blocking this intergenerational transfer. For example, the high adolescent educational
aspirations of offspring born to women who did not have their high school credentials at the time
of their child’s birth appear not to be driven by the aspirations their mothers held as adolescents
or the aspirations their mothers have for their children. This may not be particularly surprising as
60
these mothers reported rather low aspirations as adolescents (34% aspired to leave prior to
completing high school, 52% aspired to attain their high school credentials). Further, these
mothers were often unable to attain their aspired level of education (i.e. completing high school).
Therefore, potentially the children of these women turned to other sources of influence such as
their peers or schools when developing their own aspirations and goals regarding their future
education or academic potential. Perhaps due to other characteristics of these women (that are
correlated both with their lower aspirations and lower educational attainment) their children view
them to be less useful or less important sources of information related to education specifically.
As a mother’s level of educational attainment increases it appears her influence on her
child’s aspirations increases as well. The maternal educational aspirations women reported for
their children who had a high school diploma/GED at the time of their child’s birth emerged as a
significant predictor of their child’s own aspirations as adolescents. Maternal educational
aspirations remained significant for children of the highest educated women (those with at least
some college experience or more at the time of their child’s birth), and in one case the educational
aspirations these women reported as adolescents was also predictive of her child’s own
aspirations decades later. Women who had their high school credentials at the time of their child’s
birth were for the most part able to attain the level of education they on average aspired to attain
as adolescents (their high school diploma/GED). Perhaps these women’s thoughts are more
influential on the formation of their child’s educational aspirations as they themselves navigated
the completion of at least one level of education. Potentially this milestone allowed these mothers
to better understand the obstacles their child would need to overcome in order to reach higher
levels of educational attainment.
Similarly, women who had at least some college experience by the birth of their child
likely have first-hand knowledge and experiences that can benefit their children when navigating
the complex pathways from high school to college, or high school to college completion. Further,
61
these women likely have wider networks containing valuable resources that families can draw
upon that may increase the likelihood their children will achieve their aspired educational
attainment, and are likely similar resources these women drew upon when they were also students
decades prior. Therefore it is possible an intergenerational transmission may be contributing to
the high aspirations of some youth today; however results from this study suggest that this
transmission only exists for families in which parents were able to at least complete the level of
education they aspired to complete as adolescents. This paper was unable to pinpoint what is
driving the aspirations of youth with lesser-educated women and will have to be explored in
future literature that may contain other information on important sources of influence beyond
mothers, such as children’s peers, siblings, fathers or schools.
Summary of findings paper 2
Findings from paper 2 highlight the similarities (rather than the differences) in the
relationships between youth educational aspirations and later attainment. Specifically in the full
analytic sample, the aspirations women reported during adolescence were not predictive of the
overall educational attainment of their offspring. This remained true when the sample was
stratified by student race/ethnicity. Rather, in the full analytic sample youth adolescent and
maternal educational aspirations were positively and significantly associated with youth
educational attainment. For white families, maternal aspirations women held for their children as
well as the youth’s own educational aspirations were predictive of youth educational attainment.
For black families youth educational aspirations were consistently predictive of their educational
attainment, and maternal educational aspirations were predictive of the odds their child would
attain a Bachelor’s degree or more. For Hispanic families, youth educational aspirations (not
maternal aspirations for her child) were predictive of youth educational attainment.
62
The exploration of student aspirations and student educational attainment in a sample of
black, Hispanic and white, native-born U.S. adolescents contributed to two areas in educational
literature. First, these findings contradict studies which question the validity of the aspirations
black youth report compared to their white peers, as black students are found to report higher or
equally high aspirations than their white peers yet have lower levels of academic achievement
(see Downey, Ainsworth, & Qian, 2009, for discussions). Further, this study also contributed to
literature by exploring the relationships between youth aspirations and attainment in a diverse
sample of adolescents who all were born in the United States. Currently, limited research exists
which explores these relationships in samples of American-born Hispanic students and their
families. Contrary to literature focusing on immigrant students and their families (e.g., Kao, 2007;
Kao & Tienda, 1998), this study found Hispanic students reported high educational aspirations,
and aspirations similar to their black and white peers. Thus suggesting researchers should pay
mind to the immigration status of their samples, particularly as youth aspirations and the returns
these aspirations bring to the educational attainment of Hispanic youth may vary depending upon
when their families immigrated to the United States.
In this study, whether explored in the full sample, or by student race/ethnicity youth
educational aspirations were positively linked to the overall educational attainment of students,
ranging from the odds they will earn their high school credentials ‘on time,’ to enroll in college,
or attain their Bachelor’s degree or more. Perhaps as youth educational aspirations are increasing
for recent cohorts of youth so too may the potential positive implications these aspirations may
have on youth educational attainment. Or perhaps, compared to students of the same
race/ethnicity, students with higher educational aspirations do indeed reach higher levels of
attainment compared to their peers with lower educational aspirations. Whatever the case, parallel
to Downey and colleagues’ (2009) study, the current study found that high educational aspirations
are associated with positive returns to educational attainment for black, white and Hispanic
63
students. Thus, this paper reinforces the importance a student’s own educational aspirations may
have for their later educational attainment, and this study casts doubt on the allegations that
returns to positive aspirations may not exist for all students.
In keeping with this study’s theme of racial/ethnic similarities rather than differences, this
study did not find evidence for an intergenerational narrative of educational aspirations driving
the educational attainment of black, Hispanic or white students. While the aspirations white
mother’s reported holding for their children were positively associated with the educational
attainment of their offspring, the maternal aspirations of black mothers reported were only
significantly related with the highest level of educational attainment for black youth. Further, the
educational aspirations women of all races reported, as adolescents did not emerge as a driving
factor for the educational attainment of their offspring. It is possible that an intergenerational
narrative does exist, and this narrative relates to increasing levels of youth educational attainment
across generations. Yet, through the measures employed in this paper, there does not appear to be
any significant association between the aspirations women held in adolescence prior to
motherhood and their child’s later academic success. While there is no evidence of a positive
association between women’s adolescent aspirations and their offspring’s later educational
attainment, we can at least feel relief that there is not a negative association potentially driving
racial or ethnic gaps in education for future generations.
The next set of analyses in the second study explored whether youth behavior, such as
time youth spent on homework or reports of truancy, explained (a) the associations between
measures of educational aspirations and youth educational attainment and (b) if these
relationships varied by student/race ethnicity, as racial/ethnic gaps in youth educational
attainment persist over time (Tienda, 2013). When I explored whether measures of youth
behavior explained the relationships between youth and maternal educational aspirations and
attainment for white students, and the youth aspirations of black and Hispanic students,
64
similarities rather than differences emerged. Specifically, in all samples neither measure of youth
behavior mediated the association between measures of educational aspirations and the odds the
youth earned their high school credentials ‘on time.’ The good news is that disparities in youth
educational attainment do not appear to be related to differential levels of the time youth spend on
homework or their reports of truancy. However, as we know these gaps in educational attainment
exist, more work will have to explore what is driving racial and ethnic gaps in education, as well
as what student behaviors serve as a mechanism between youth aspirations and attainment.
Conclusions and implications for youth attainment
Potentially the increasing levels of youth educational aspirations may have positive
implications for the educational attainment of students. Combining the findings from both papers,
I see (1) recent cohorts of students have high aspirations, even higher than the aspirations their
mother’s reported as adolescents and (2) higher youth aspirations are positively associated with
youth educational attainment, and this is consistent for white, black and Hispanic students. This is
great news as overall median earnings increase and unemployment levels decrease as an
individual’s level of educational attainment increases (BLS, 2013). Further, as individual’s
employment and income increases, their odds of living in poverty (and the risks associated with
poverty) decrease (U.S. Census Bureau, 2013), thus providing benefits to other domains of
individual and familial well-being. However, one substantial concern remains. Rates of youth
educational attainment and graduation are not the same for white, black and Hispanic students, or
students coming from lower and higher-SES households (maternal educational attainment is used
as a proxy for household SES in this paper).
While youth aspirations may be a potential lever for increasing the educational
enrollment and attainment of all groups of students, it is not likely the most efficient mechanism.
65
Intervention and other empirical literature suggest that the high aspirations students’ hold may not
be enough for them to achieve their goals. One intervention found the rates of misinformation
related to the costs of college as well as potential funding opportunities are remarkably high
among populations of low-income families and students (Bettinger, Long, Oreopoulos, &
Sanbonmatsu, 2012). Therfore many students who are qualified to apply for and enroll in college
may avoid doing so as they are unaware of current sources of financial aid that are available to
them. A study by Roderick and colleagues (2011) found that the school context also plays an
important role in the post-secondary educational attainment of graduating students. The authors
found qualities and characteirstics of the school environment (such as student services, guidance
and support) play a significant role in helping the students who had high aspirations and high
academic skills to apply for and attend college. Therefore, although their high aspirations may not
hurt their chances for reaching higher levels of educational attainment, in order for students in the
two studies of this dissertation to reach their goals, they would need more from their families and
schools than just positive orientations or encouragement.
Additionally other literature finds that many recent high school graduates are not
prepared for the coursework of their post-secondary education programs, and must spend the first
few semesters in remedial classes reinforcing concepts they should have mastered in high school
(see, Attewell, Lavin, Domina, & Levey, 2006). Likely these remedial courses ultimately increase
student frustration and the costs of college, as students spend more time away from coursework
they might find conceptually exciting, or necessary to complete their training. Thus it is
potentially no wonder that the dropout or ‘stop-out,’ rates of students enrolled in such background
coursework are higher than students who begin college with the necessary skills (Adelman 2004;
Deli-Amen & Rosenbaum 2002; Symonds, Schwartz, & Ferguson, 2011). These studies point to
the importance that the quality, services, and resources of earlier learning contexts (such as in
grades K- 12) have for the later educational success of students.
66
In conclusion, the current dissertation suggests that despite the leaks that are currently
present in the U.S. educational pipeline, students today have high educational aspirations. These
high student aspirations are promising as additionally this dissertation found that high student
aspirations were met with positive returns to the educational attainment of white, black and
Hispanic students. Yet, in order to begin to close the race/ethnic or even socio-economic
educational attainment gaps, it is time that we begin to pair student aspirations with the
appropriate levels of educational and financial support. It may be that only by properly
reinforcing the high goals of low-income students, first-generation college-bound students, and
black and Hispanic students, with student services and higher quality educational experiences that
the educational attainment gap may become ‘mind over matter.’
67
References
Adelman, C. 2004. Principal indicators of student academic histories in postsecondary
education, 1972-2000. Washington, DC: U.S. Department of Education, Institute of
Education Sciences. Retrieved online from:
http://www2.ed.gov/rschstat/research/pubs/prinindicat/prinindicat.pdf
Ainsworth-Darnell, J. W., & Downey, D. B. (1998). Assessing the Oppositional Culture
Explanation for racial/ethnic differences in school performances. American Sociological
Review, 63(4), 536–553.
Anderson, E. (2012). Reflections on the ‘black-white achievement gap.’ Journal of School
Psychology, 50, 593- 597.
Andrew, M., & Hauser, R. M. (2011). Adoption? Adaption? Evaluating the formation of
educational expectations. Social Forces, 90, 497- 520.
Attewell, P., Lavin, D., Domina, T., & Levey, T. (2006). New evidence on college remediation.
The Journal of Higher Education, 77, 886- 924.
Behnke, A. O., Piercy, K. W., & Diversi, M. (2004). Educational and Occupational Aspirations of
Latino Youth and Their Parents. Hispanic Journal of Behavioral Sciences, 26(1), 16–35.
doi:10.1177/0739986303262329
Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2012). The role of application
assistance and information in college decisions: Results from the H&R Block FAFSA
Experiment. The Quarterly Journal of Economics, 127(1205-1242), 1205–1242.
doi:10.1093/qje/qjs017.Advance
Blau, P. M., & Duncan, O. D. (1967). The American occupational structure. New York: Wiley.
Blau, J. R., Moller, S., & Jones, L. V. (2004). Why test? Talent loss and enrollment loss. Social
68
Science Research, 33(3), 409–434. doi:10.1016/j.ssresearch.2003.09.002
Bozick, R., Alexander, K., Entwisle, D., Dauber, S., & Kerr, K. (2010). Framing the future:
Revisiting the place of educational expectations in status attainment. Social Forces,
88(5), 2027– 2052.
Bronfenbrenner, U., & Morris, P. A. (1998). The ecology of developmental process. In W.
Damon (Series Ed.), & R. M. Lerner (Vol. Ed.), Handbook of child psychology: Vol. 1.
Theoretical models of human development (5th ed., pp. 993–1028). New York: Wiley.
Bureau of Labor Statistics (2013a). College enrollment and work activity of 2012 high school
graduates. Retrieved online: http://www.bls.gov/news.release/pdf/hsgec.pdf.
Bureau of Labor Statistics (2013b). Earnings and unemployment rate by educational attainment.
Retrieved online: http://www.bls.gov/emp/ep_chart_001.htm.
Conger, R. D., & Donnellan, M. B. (2007). An interactionist perspective on the socioeconomic
context of human development. Annual Review of Psychology, 58, 175–99.
doi:10.1146/annurev.psych.58.110405.085551
Crosnoe, R. (2001). Academic orientation and parental involvement in education during High
School. Sociology of Education, 74(3), 210–230. doi:10.2307/2673275
Davis-Kean, P. E. (2005). The influence of parent education and family income on child
achievement: The indirect role of parental expectations and the home environment.
Journal of Family Psychology, 19(2), 294–304. doi:10.1037/0893-3200.19.2.294
De Civita, M., Pagani, L., Vitaro, F., & Tremblay, R. E. (2004). The role of maternal educational
aspirations in mediating the risk of income source on academic failure in children from
persistently poor families. Children and Youth Services Review, 26(8), 749–769.
doi:10.1016/j.childyouth.2004.02.019
Deli-Amen, R. & Rosenbaum, J. E. (2004). The unintended consequences of stigma-free
remediation. Sociology of Education, 75, 249- 268.
69
Deslandes, R., & Bertrand, R. (2005). Motivation of parent involvement in secondary-level
schooling. The Journal of Educational Research, 98(3), 164–175.
Domina, T., Conley, A., & Farkas, G. (2011). The Link between educational expectations and
effort in the college-for-all era. Sociology of Education, 84(2), 93–112.
doi:10.1177/1941406411401808
Downey, D. B., Ainsworth, J. W., & Qian, J. (2009). Rethinking the attitude-achievement
paradox among blacks. Sociology of Education, 82, 1-19.
Elder, G. H. (1998). The life course as developmental theory. Child Development, 69(1), 1–12.
Ellickson, P., Saner, H., McGuigan, K. (2008). Profiles of violent youth: Substance use and
concurrent problems. American Journal of Public Health, 87, 985- 991.
Fan, X., & Chen, M. (2001). Parental Involvement and Students ’ Academic Achievement : A
Meta-Analysis. Educational Psychology Review, 13(1), 1–22.
Fantuzzo, J., LeBoeuf, W., Rouse, H., & Chen, C. C. (2012). Academic achievement of African
American boys: A city-wide, community-based investigation of risk and resilience.
Journal of School Psychology, 50, 559- 579.
Finn, J. D. (2006). The adult lives of at-risk students: The roles of attainment and engagement in
high school (NCES 2006- 328). Washington, DC. doi:10.1037/e501552006-001
Fordham, S., & Ogbu, J. U. (1986). Black students’ school success: Coping with the burden of
‘acting White.’ The Urban Review, 18, 176- 206.
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement : Potential of the
concept, state of the evidence. Review of Educational Research, 74(1), 59–109.
Goyette, K. A. (2008). College for some to college for all: Social background, occupational
expectations, and educational expectations over time. Social Science Research, 37(2),
461–484. doi:10.1016/j.ssresearch.2008.02.002
Graham, J. W. (2009). Missing data analysis: making it work in the real world. Annual Review of
70
Psychology, 60, 549–576. doi:10.1146/annurev.psych.58.110405.085530
Haller, A. O., & Portes, A. (1973). Status Attaimnent Processes. Sociology of Education, 46(1),
51–91.
Hernandez, A., Vargas-Lew, L., & Martinez, C. L. (1994). Intergenerational academic aspirations
of Mexican-American females: An examination of mother, daughter, and grandmother
triads. Hispanic Journal of Behavioral Sciences, 16(2), 195–204.
Hoover- Dempsey, K. V, Walker, J. M. T., Sandler, H. M., Whetsel, D., Green, C. L., Wilkins, A.
S., & Closson, K. (2005). Why do parents become involved? Research findings and
implications. The Elementary School Journal, 106(2), 105–130.
Jemal, A., Thun, M., Ward, E., Henley, J., Cokkinides, V., & Murray, T. (2008). Mortality from
leading causes by education and race in the United States. American Journal of
Preventitive Medicine, 34, 1-8.
Johnson, M. K., Crosnoe, R., & Elder, G. H. (2001). Students’ attachment and academic
engagement: The role of race and ethnicity. Sociology of Education, 74(4), 318–340.
Kao, G. (2004). Parental influences on the educational outcomes of immigrant youth.
International Migration Review, 38(2), 427–499.
Kao, G., & Thompson, J. S. (2003). Racial and Ethnic Stratification in Educational Achievement
and Attainment. Annual Review of Sociology, 29(1), 417–442.
doi:10.1146/annurev.soc.29.010202.100019.
Kao, G., & Tienda, M. (1998). Educational aspirations of minority youth. American Journal of
Education, 106, 349- 384.
Kaplan, D. S., Liu, X., & Kaplan, H. B. (2000). Family structure and parental involvement in the
intergenerational paralleism of school adversity. The Journal of Educational Research,
93(4), 235–244.
Klem, A. M., & Connell, J. P. (2004). Relationships matter: Linking teacher support to student
71
engagement and achievement. Journal of School Health, 74(7), 262–274.
Kohl, G. O., Lengua, L. J., & McMahon, R. J. (2000). Parent involvement in school
conceptualizing multiple dimensions and their relations with family and demographic risk
factors. Journal of School Psychology, 38(6), 501–523.
Mau, W.-C., & Bikos, L. H. (2000). Educational and vocational aspirations of minority and
female students: A longitudinal study. Journal of Couseling Nad Development, 78, 186–
194.
McCarty, C., Mason, W., Kosterman, R., Hawkins, J., Lengua, L., & McCauley, E. (2008).
Adolescent school failure predicts later depression among girls. Journal of Adolescent
Health, 43, 180-187.
Mickelson, R. A. (1990). The Attitude- Achievement Paradox Among Black Adolescents.
Sociology of Education, 63(1), 44–61.
Moller, S., Stearns, E., Blau, J. R., & Land, K. C. (2006). Smooth and rough roads to academic
achievement: Retention and race/class disparities in high school. Social Science
Research, 35(1), 157–180. doi:10.1016/j.ssresearch.2004.08.001.
Morgan, S. L. (2005). On the edge of commitment: Educational attainment and race in the Untied
States. Stanford University Press.
Ogbu, J. (1981). Origins of human competence: A cultural-ecological perspective. Child
Development, 52, 413- 429.
Perez, A. D., & Hirschman, C. (2009). The changing racial and ethnic composition of the US
population: Emerging American identities. Population and Development Review, 35, 1-
51.
Putallaz, M., Costanzo, P. R., & Smith, R. B. (1991). Maternal recollections of childhood peer
relationships: Implications for their children’s social competence. Journal of Social and
Personal Relationships, 8(3), 403–422. doi:10.1177/0265407591083006.
72
Roderick, M., Coca, V., & Nagoka, J. (2011). Potholes on the road to college: High school effects
in shaping urban students’ participation in college application, four-year college
enrollment and college match. Sociology of Education, 84, 178-211.
Rosenbaum, J. (2001). Beyond College For All. New York: Russell Sage Foundation.
Rosenbaum, J. E. (2011). The Complexities of college for all : Beyond fairy-tale dreams.
Sociology of Education, 84(2), 113–117.
Schneider, B., & Stevenson, D. (1999). The ambitious generation: America’s teenagers motivated
but not directionless. New Haven: Yale University Press.
Sewell, W. H., Haller, A. O., & Portes, A. (1969). The Educational and Early Occupational
Attainment Process. American Sociological Review, 34(1), 82–92.
Sewell, W. H., & Shah, V. P. (1968). Parents’ education and children's educational aspirations
and achievements. American Sociological Review, 33(2), 191–209.
Shernoff, D. J., & Schmidt, J. A. (2008). Further Evidence of an Engagement–Achievement
Paradox Among U.S. High School Students. Journal of Youth and Adolescence, 37(5),
564–580. doi:10.1007/s10964-007-9241-z
Smith, J., Brooks-Gunn, J., & Klebanov, P. (1997). The consequences of living in poverty on
young children's cognitive development. In G. J. Duncan & J. Brooks- Gunn (Eds.),
Consequences of growing up poor (pp. 132-189). New York: Russell Sage Foundation.
Spera, C., Wentzel, K. R., & Matto, H. C. (2009). Parental aspirations for their children’s
educational attainment: relations to ethnicity, parental education, children's academic
performance, and parental perceptions of school climate. Journal of Youth and
Adolescence, 38(8), 1140–52. doi:10.1007/s10964-008-9314-7
Stearns, E., Moller, S., Blau, J., & Potochnick, S. (2012). Staying back and dropping Out : The
relationship between grade retention and school dropout. Sociology of Education, 80(3),
210–240.
73
Taylor, L. C., Clayton, J. D., & Rowley, S. J. (2004). Academic socialization: Understanding
parental influences on children’s school-related development in the early years. Review of
General Psychology, 8(3), 163–178. doi:10.1037/1089-2680.8.3.163
Tienda, M. (2013). Diversity ≠ inclusion: Promoting integration in higher education. Educational
Researcher, 42, 467-475.
Tyler, J. H., & Lofstrom, M. (2009). Finishing high school: Alternative pathways and dropout
recovery. The Future of Children, 19(1), 77–103.
Symonds, W. C., Schwartz, R. B., & Ferguson. R. (2011). Pathways to prosperity: Meeting the
challenge of preparing young Americans for the 21st century. Retrieved online: http://
www.gse.harvard.edu/news_events/features/2011/Pathways_to_Prosperity_Feb2011.pdf
Uekawa, K., Borman, K., & Lee, R. (2007). Student engagement in U.S. urban high school
mathematics and science classrooms: Findings on social organization, race, and ethnicity.
The Urban Review, 39(1), 1–43. doi:10.1007/s11256-006-0039-1
U.S. Census Bureau (December, 2012). 2012 National Population Projections. Retrieved online
from: http://www.census.gov/population/projections/data/national/2012.html.
U.S. Census Bureau (2013). Income, poverty, and health insurance coverage in the United States:
2012 (P60-245). Retrieved online from: http://www.census.gov/prod/2013pubs/p60-
245.pdf
U.S. Department of Education, National Center for Education Statistics (2010). Status and Trends
in the Education of Racial and Ethnic Groups (NCES 2010- 015). Retrieved online from:
http://nces.ed.gov/pubs2010/2010015.pdf
U.S. Department of Education, National Center for Education Statistics (2011). Youth indicators
2011 America’s youth: Transitions to adulthood (NCES 2012-026). Retrieved online
from: http://nces.ed.gov/pubs2012/2012026/chapter3_31.asp.
U.S. Department of Education, National Center for Education Statistics (2013a). The Condition
74
of Education 2013 (NCES 2013- 037). Retrieved online from:
https://nces.ed.gov/fastfacts/display.asp?id=27.
U.S. Department of Educaiton, National Center for Educaiton Statistics (2013b). Digest of
Educaiton Statistics: 2012 ( NCES 2014- 015). Retrieved online from:
https://nces.ed.gov/programs/digest/d12/.
Vallerand, R. J., Fortier, M. S., & Guay, F. (1997). Self-determination and persistence in a real-
life setting: Toward a motivational model of high school dropout. Journal of Personality,
72, 1161- 1176.
75
Appendix A
Tables and Figures
Figure 1. Educational aspirations by generation
6
41
20
25
8
3
20
9
42
26
0
5
10
15
20
25
30
35
40
45
Less than High School
High School Diploma/
GED
Some College
Bachelor's Degree
More than a Bachelor's
Degree
Per
cent
Youth educational aspirations
Generation 1 mothers
Generation 2 youth
76
Table 2-1. Analytic sample descriptive statistics
Full analytic
sample
Maternal educational attainment at child’s birth
Less
than high school
High school
diploma/GED
Some
college or more
G2 adolescent aspirations
Less than high school 3% 5% 3% b 1% b
High school diploma/ GED 20% 40% 21% b 9% bc
Some college / other training 9% 10% 11% 7% bc
Attain Bachelor’s degree 42% 29% 41% b 48% bc
More than a Bachelor’s 26% 16% 24% b 35% bc
G1 adolescent aspirations
Less than high school 6% 34% 2% b 1% bc
High school diploma/ GED 41% 52% 59% 9% bc
Some college / other training 20% 7% 21% b 24% bc
Bachelor’s degree or more 33% 7% 18% b 66% bc
G1 aspirations for child
Less than high school 1% 4% 1% b 1% bc
High school diploma/GED 14% 36% 15% b 4% bc
Some college / other training 21% 22% 26% 13% bc
Bachelor’s degree or more 64% 38% 58% b 83% bc
G2 covariates
Gender (female) 49% 46% 50% 50%
Race/ethnicity
White a 77% 62% 78% b 83% bc
Black 16% 22% 15% 13% c
Hispanic 7% 16% 7% b 4% bc
PIAT math skills 103.04(13.21) 95.87 (12.81) 102.66 (12.40) 108.07 (12.63) c
Special education 7% 10% 7% 6%
Was held back by 8th grade 13% 27% 13% b 4% bc
Number of siblings 1.86 (1.32) 2.47 (1.81) 1.75 (1.18) b 1.73 (1.17) b
77
Proportion of years from child birth to 13
Poverty (< 100% FPL) 0.22 (0.32) 0.53 (0.36) 0.23 (0.32) 0.07 (0.18) bc
G1 unemployed 0.38 (0.31) 0.59 (0.28) 0.37 (0.29) 0.31 (0.32) b
HOME-SF percentile score
Cognitive stimulation 51.08 (23.33) 36.73 (25.63) 50.67 (22.84) 57.90 (19.84) bc
Emotional support 49.68 (24.32) 41.01 (25.12) 48.70 (24.25) 54.95 (22.77) b
G1 covariates at child’s
birth
Age 25.00 (3.65) 22.85 (3.65) 24.37 (3.47) 26.87 (3.03) bc
Educational attainment
Less than high school 15% 100% -- --
High school diploma/ GED a 50% -- 100% --
Some college or more 34% -- -- 100%
Marital status
Married a 75% 52% 73% b 87% bc
Not married 25% 48% 27% b 33% bc
G1 covariates
AFQT cognitive skills 43.17 (27.07) 18.22 (16.32) 37.79 (22.29) b 61.40 (25.12) bc
Unweighted nG2 3198 639 1591 968
Note. Weighted descriptives are presented using the population sampling weight Y2615900. G2= Generation 2, G1=
Generation 1, GED = General Educational Development, PIAT = Peabody Individual Achievement Test total standard score,
FPL = Federal Poverty Line, HOME-SF= Home Observation Measure of the Environment- Short Form, AFQT = Armed
Forces Qualification Test.
a Reference category in the multivariate regression models.
b Significantly different from less than high school group (p <.05).
c Significantly different from high school group (p< .05).
78 Table 2-2. Adjusted odds ratios (standard error) [95% CI] of intergenerational transmission of educational aspirations in full sample
G2 aspires to complete some
college or more vs. high school/ <
high school
G2 aspires to attain a bachelor’s
degree or more vs. some college
G2 aspires to attain more than a
bachelor’s degree vs. a
bachelor’s degree
Model 1a Model 2 b Model 1a Model 2 Model 1a Model 2 b
G1 adolescent aspirations 1.55 (0.08) ***
[1.41, 1.71]
1.04 (0.06)
[0.93, 1.17]
1.32 (0.08) ***
[1.86, 1.18]
1.01 (0.08)
[0.86, 1.18]
1.17 (0.06)**
[1.06, 1.29]
1.09 (0.07)
[0.96, 1.23]
G1 aspirations for child --- 1.67 (0.10) ***
[1.48, 1.89]
--- 1.50 (0.13) ***
[1.27, 1.77]
--- 1.30 (0.10) **
[1.11, 1.52]
G2 covariates
Gender (female) --- 1.28 (0.12) **
[1.07, 1.54]
--- 1.16 (0.15)
[0.90, 1.50]
--- 1.45 (0.14) ***
[1.20, 1.75]
Race/ethnicity
White --- --- --- --- --- ---
Black --- 1.24 (0.17)
[0.94, 1.62]
--- 1.37 (0.27)
[0.93, 2.01]
--- 1.16 (0.17)
[0.86, 1.55]
Hispanic --- 1.31 (0.18) *
[1.00, 1.72]
--- 0.94 (0.16)
[0.67, 1.31]
--- 1.03 (0.15)
[0.77, 1.37]
PIAT math skills --- 1.03 (0.00) ***
[1.02, 1.03]
--- 1.01 (0.01)
[1.00 1.03]
--- 1.01 (0.00) *
[1.00, 1.02]
79
Special education --- 0.86 (0.14)
[0.63, 1.18]
--- 0.95 (0.24)
[0.57, 1.56]
--- 1.13 (0.24)
[0.74, 1.72]
Held back by 8th grade --- 0.65 (0.08) **
[0.51, 0.83]
--- 0.99 (0.20)
[0.67, 1.47]
--- 0.87 (0.15)
[0.62, 1.22]
Number of siblings --- 1.01 (0.04)
[0.94, 1.08]
--- 1.03 (0.05)
[0.93, 1.14]
--- 0.96 (0.04)
[0.88, 1.04]
Proportion of years birth to 13:
Poverty (< 100% FPL) --- 0.62 (0.13) *
[0.41, 0.93]
--- 0.70 (0.20)
[0.40, 1.23]
--- 1.46 (0.34)
[0.92, 2.31]
G1 unemployed --- 1.35 (0.27)
[0.91, 2.01]
--- 1.10 (0.27)
[0.68, 1.79]
--- 0.99 (0.19)
[0.67, 1.45]
HOME- SF score:
Cognitive stimulation --- 1.00 (0.00)
[0.99, 1.00]
--- 1.01 (0.00)
[1.00, 1.01]
--- 1.00 (0.00)
[0.99, 1.00]
Emotional support --- 1.00 (0.00)
[1.00, 1.00]
--- 1.00 (0.00)
[0.99, 1.00]
--- 1.00 (0.00) *
[1.00, 1.01]
G1 covariates at child’s birth
Age 1.03 (0.01) *
[1.01, 1.06]
1.02 (0.02)
[0.98, 1.06]
1.02 (0.01)
[0.99, 1.05]
Education:
80
Less than high school --- 0.69 (0.09) **
[0.54, 0.89]
--- 0.97 (0.18)
[0.68, 1.39]
--- 1.03 (0.1)
[0.74, 1.44]
High school diploma/ GED --- --- --- --- --- ---
Some college or more --- 1.18 (0.16)
[0.91, 1.54]
--- 1.41 (0.26)
[0.98, 2.02]
--- 1.05 (0.13)
[0.83, 1.34]
Marital status:
Married --- --- --- --- --- ---
Not married --- 1.02 (0.12)
[0.81, 1.29]
--- 0.78 (0.13)
[0.56, 1.09]
--- 1.06 (0.13)
[0.83, 1.36]
G1 covariate
AFQT cognitive skills --- 1.00 (0.00)
[1.00, 1.01]
1.00 (0.00)
[0.99, 1.00]
1.00 (0.00)
[1.00, 1.01]
Unweighted nG2 3,037 2,369 2,091
Note. * p< .05, ** p<. 01, *** p<.001.
CI = Confidence Interval, G2= Generation 2, G1= Generation 1, GED = General Educational Development, PIAT = Peabody Individual Achievement Test total standard score,
FPL = Federal Poverty Line, HOME-SF= Home Observation Measure of the Environment- Short Form, AFQT = Armed Forces Qualification Test.
a Model 1: Regress G1 adolescent aspirations on G2 youth aspirations.
b Model 2: Model 1 + G1 aspirations for child + covariates on G2 youth aspirations.
81 Table 2-3. Adjusted odds ratios (standard error) [95% CI] of intergenerational transmission of educational aspiration by maternal education attainment
G2 aspires to complete some college or
more
vs.
high school/ less than high school
G2 aspires to attain a Bachelor’s
degree or more
vs.
some college
G2 aspires to attain more than a
Bachelor’s degree
vs.
a Bachelor’s degree
Model 1a Model 2 b Model 1a Model 2 b Model 1a Model 2 b
Panel A: Mothers with less than a high school education
G1 youth aspirations 1.19 (0.13)
[0.96, 1.48]
1.17 (0.13)
[0.95, 1.45]
1.02 (0.15)
[0.76, 1.36]
0.92 (0.14)
[0.68, 1.26]
1.09 (0.18)
[0.79, 1.51]
1.05 (0.19)
[0.73, 1.50]
G1 aspirations for child 1.43 (0.15) **
[1.16, 1.77]
1.11 (0.18)
[0.82, 1.52]
1.07 (0.18)
[0.78, 1.48]
Unweighted nG2 639 349 227
Panel B: Mothers with high school diploma/ GED
G1 youth aspirations 1.17 (0.09) **
[1.01, 1.34]
1.03 (0.08)
[0.89, 1.21]
0.98 (0.09)
[0.81, 1.18]
0.90 (0.09)
[0.74, 1.10]
1.08 (0.09)
[0.92, 1.26]
1.09 (0.09)
0.92, 1.29]
G1 aspirations for child 1.74 (0.15) ***
[1.47, 2.07]
1.61 (0.16) ***
[1.32, 1.97]
1.33 (0.14) **
[1.08, 1.64]
Unweighted nG2 1,591 1,198 1,025
Panel C: Mothers who completed some college or more
G1 youth aspirations 1.09 (0.015) 0.88 (0.13) 1.58 (0.24) ** 1.47 (0.24) * 1.19 (0.13) 1.21 (0.14)
82
[0.83, 1.44] [0.65, 1.18] [1.18, 2.03] [1.06, 2.03] [0.96, 1.47] [0.97, 1.53]
G1 aspirations for child 2.04 (0.32) ***
[1.50, 2.78]
2.01 (0.46) **
[1.28, 3.16]
1.41 (0.30) *
[1.02, 2.24]
Unweighted nG2 968 838 769
Note. * p< .05, ** p<. 01, *** p<.001.
CI = Confidence Interval, G2= Generation 2, G1= Generation 1, GED = General Educational Development.
a Model 1: Regress G1 youth aspirations on G2 youth aspirations.
b Model 2: Model 1 + G1 aspirations for child + covariates on G2 youth aspirations.
83
Table 3-1. Descriptive Statistics of dependent, independent and mediating variables for full sample and by race/ethnicity.
Full analytic
sample
Race/ Ethnicity
White Black Hispanic
G2 educational attainment
High school diploma/GED ‘on time’ 79% 83% 67% a 68% a
Enrolled in college 58% 61% 52% a 47% a
Earned bachelor’s degree or more 19% 22% 10% a 9% a
G2 youth aspirations
Less than high school 4% 3% 5% 4%
High school diploma/GED 17% 16% 22% a 18%
Some college/other training 10% 10% 10% 14%
Attain bachelor’s degree 42% 43% 37% 40% a
More than a bachelor’s 27% 28% 26% 25%
G1 youth aspirations
Less than high school 8% 8% 6% a, b 12%
84
High school diploma/GED 46% 46% 47% 48% a
Some college/other training 20% 20% 18% a, b 21%
Attain bachelor’s degree 21% 21% 21% b 16% a
More than a bachelor’s 5% 5% 8% a, b 3%
G1 aspirations for child
Less than high school 1% 1% 2% 2%
High school diploma/GED 17% 15% 23% a 24% a
Some college/other training 22% 23% 20% 24%
Attain bachelor’s degree 46% 47% 41% a 38% a
More than a bachelor’s 14% 14% 14% 12%
G2 high school engagement/disengagement (16 – 18 years)
G2 engaged in homework (hrs/week) 7.81 (7.53) 8.19 (7.65) 6.51 (6.81) a 6.89 (7.55) a
G2 skipped school 33% 33% 29% b 50% a
Unweighted nG2 1660 734 589 337
Note. Weighted descriptives are presented using the population sampling weight Y2615900. G2= Generation 2, G1= Generation 1
a Significantly different from white (p <.05).
b Significantly different from Hispanic (p< .05).
85
Table 3-2. Descriptive statistics of all covariates in full analytic sample and by race/ethnicity.
Full analytic
sample
Race/ Ethnicity
White Black Hispanic
G2 characteristics
G2 gender (female) 49% 49% 52% c 45%
G2 race/ethnicity
White a 76% 100% --- ---
Black 17% --- 100% ---
Hispanic 7% --- --- 100%
PIAT math skills 102.14 (12.93) 104.01 (12.32) 95.69 (12.64) b 97.30 (13.91) b
Special education 6% 5% 11% c 4%
Was held back by 8th grade 13% 10% 24% b 22% a
Number of siblings 1.82 (1.23) 1.71 (1.12) 2.15 (1.46) b 2.29 (1.50) b
Negative peer pressures at 0.14 (0.17) 0.13 (0.16) 0.15 (0.20) a, b 0.14 (0.17)
Proportion of years from child birth to 13 years
Poverty (< 100% FPL) 0.26 (0.33) 0.20 (0.29) 0.50 (0.37) b, c 0.36 (0.37) b
86
G1 unemployed 0.41 (0.30) 0.40 (0.29) 0.46 (0.32) b 0.46 (0.33) b
HOME-SF percentile score birth to 5 years
Cognitive stimulation 50.76 (23.95) 54.97 (22.12) 36.43 (24.51) b 37.34 (24.58) b
Emotional support 50.92 (24.57) 54.32 (23.80) 36.33 (22.38) b, c 46.43 (24.99) b
G1 characteristics
G1 age at child’s birth 22.68 (2.95) 22.88 (2.93) 21.88 (2.86) b, c 22.35 (3.06) b
G1 attainment at child’s birth
Less than high school 20% 18% 25% b 35% b
High school diploma/GED a 57% 59% 52% b 48% b
Some college or more 23% 24% 23% 17% b
G1 marital status at child’s birth
Married a 70% 79% 31% b, c 69% b
Not married 30% 21% 69% b, c 31% b
G1 AFQT cognitive skills 39.78 (25.83) 46.15 (25.01) 18.51 (14.95) b, c 22.36 (18.89) b
Unweighted nG2 1660 734 589 337
87
Note. Weighted descriptives are presented using the population sampling weight Y2615900. G2= Generation 2, G1= Generation 1, GED = General Educational
Development, PIAT = Peabody Individual Achievement Test total standard score, FPL = Federal Poverty Line, HOME-SF= Home Observation Measure of the
Environment- Short Form, AFQT = Armed Forces Qualification Test.
a Reference category in the multivariate regression models.
b Significantly different from white (p <.05).
c Significantly different from Hispanic (p< .05).
88
Table 3-3. Adjusted odds ratios (standard error) [95% CI] predicting youth educational attainment with educational aspirations.
G2 earned high school diploma/
GED ‘on time’
G2 enrolled in college G2 earned Bachelor’s degree or more
Odds Ratio (SE) 95% CI Odds Ratio (SE) 95% CI Odds Ratio (SE) 95% CI
Panel A: Full Sample (n= 1660)
G2 youth aspirations 1.34 (0.08) *** [1.19, 1.50] 1.39 (0.07) *** [1.26, 1.54] 1.39 (0.11) *** [1.19, 1.62]
G1 youth aspirations 0.92 (0.07) [0.79, 1.07] 1.08 (0.07) [0.94, 1.23] 0.92 (0.10) [0.75, 1.13]
G1 aspirations for child 1.19 (0.09) * [1.03, 1.37] 1.32 (0.09) *** [1.16, 1.50] 1.56 (0.17)*** [1.26, 1.94]
Panel B: White (n= 734)
G2 youth aspirations 1.36 (0.14) ** [1.12, 1.66] 1.47 (0.12) *** [1.25, 1.73] 1.44 (0.17) ** [1.15, 1.81]
G1 youth aspirations 0.93 (0.14) [0.70, 1.24] 1.08 (0.13) [0.85, 1.36] 0.87 (0.13) [0.65, 1.17]
G1 aspirations for child 1.36 (0.19) * [1.03, 1.78] 1.60 (0.19) *** [1.24, 1.98] 1.55 (0.24) ** [1.14, 2.08]
Panel C: Black (n= 589)
G2 youth aspirations 1.28 (0.12) ** [1.07, 1.53] 1.36 (0.11) *** [1.15, 1.60] 1.35 (0.22) [0.98, 1.85]
G1 youth aspirations 0.89 (0.11) [0.70, 1.12] 1.01 (0.11) [0.82, 1.24] 0.86 (0.15) [0.61, 1.23]
G1 aspirations for child 1.16 (0.14) [0.92, 1.47] 1.07 (0.11) [0.88, 1.31] 2.00 (0.49) ** [1.24, 3.21]
Panel D: Hispanic (n= 337)
G2 youth aspirations 1.39 (0.18) * [1.07, 1.80] 1.32 (0.16) * [1.04, 1.68] 1.75 (0.47) * [1.03, 2.97]
G1 youth aspirations 1.04 (0.16) [0.77, 1.40] 1.17 (0.18) [0.87, 1.58] 1.11 (0.34) [0.61, 2.03]
G1 aspirations for child 0.95 (0.14) [072, 1.26] 1.31 (0.18) [0.99, 1.72] 1.32 (0.36) [0.77, 2.24]
Note. * p< .05, ** p<. 01, *** p<.001, All models include all child and maternal covariates listed in Table 2.
G2= Generation 2, G1= Generation 1, SE = Standard error. CI = Confidence Interval.
89
Table 3-4. Testing mediation for high school engagement and disengagement in white sample.
Panel A: Testing high school engagement as a mediator between aspirations and G2 earned a high school diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
High school
engagement
G2 earned high school
diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
Model 1 a Model 2 b Model 3 a Model 4 a
G2 youth aspirations 1.35 (0.14) ** [1.11, 1.64] 0.79 (0.24) ** [0.32, 1.25] --- --- 1.36 (0.14) ** [1.11, 1.65]
G1 aspirations for child 1.34 (0.19) * [1.02, 1.76] 1.06 (0.36) ** [0.35, 1.77] --- --- 1.35 (0.19) * [1.01, 1.75]
Engagement
Hours/ Week on H.W. --- --- --- --- 1.01 (0.02) [0.97, 1.04] 0.99 (0.02) [0.96, 1.03]
Disengagement
G2 skipped school 0.66 (0.16) [0.41, 1.05] -0.27 (0.58) [-1.40, 0.86] 0.62 (0.15) * [0.39, 0.98] 0.65 (0.16) [0.41, 1.04]
Panel B: Testing high school disengagement as a mediator between aspirations and G2 earned a high school diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
High school
disengagement
G2 earned high school
diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
Model 1 a Model 2 a Model 3 a Model 4 a
G2 youth aspirations 1.37 (0.14) ** [1.12, 1.67] 0.94 (0.08) [0.79, 1.12] --- --- 1.36 (0.14) ** [1.12, 1.65]
G1 aspirations for child 1.35 (0.19) * [1.02, 1.78] 0.87 (0.10) [0.69, 1.08] --- --- 1.35 (0.19) * [1.02, 1.77]
Engagement
Hours/ Week on H.W. 0.99 (0.02) [0.96, 1.03] 0.99 (0.01) [0.97, 1.02] 1.01 (0.02) [0.97, 1.04] 0.99 (0.02) [0.96, 1.03]
Disengagement
G2 skipped school --- --- --- --- 0.62 (0.15) * [0.39, 0.98] 0.65 (0.16) [0.41, 1.04]
Note. * p< .05, ** p<. 01, *** p<.001, G2= Generation 2, G1= Generation 1, GED = General Educational Development, H.W. = Homework.
90
a Adjusted odds ratios (standard error) [confidence interval]
b Unstandardized beta (standard error) [confidence interval]
Model 1: Educational attainment regressed on educational aspirations + covariates.
Model 2: Mediator regressed on educational aspirations + covariates.
Model 3: Educational attainment regressed on mediator + covariates.
Model 4: Educational attainment regressed on educational aspirations + mediator + covariates.
91 Table 3-5. Testing mediation for high school engagement and disengagement in black sample.
Panel A: Testing high school engagement as a mediator between aspirations and G2 earned a high school diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
High school
engagement
G2 earned high school
diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
Model 1 a Model 2 b Model 3 a Model 4 a
G2 youth aspirations 1.30 (0.12) ** [1.08, 1.55] 0.07 (0.25) [-0.40, 0.55] --- --- 1.29 (0.12) ** [1.08, 1.55]
Engagement
Hours/ Week on H.W. --- --- --- --- 1.02 (0.02) [0.99, 1.06] 1.02 (0.02) [0.99, 1.06]
Disengagement
G2 skipped school 0.63 (0.15) [0.39, 1.00] 0.72 (0.70) [-0.66, 2.10] 0.63 (0.15) [0.40, 0.99] 0.62 (0.15) * [0.39, 0.99]
Panel B: Testing high school disengagement as a mediator between aspirations and G2 earned a high school diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
High school
disengagement
G2 earned high school
diploma/GED ‘on time’
G2 earned high school
diploma/GED ‘on time’
Model 1 a Model 2 a Model 3 a Model 4 a
G2 youth aspirations 1.29 (0.12) ** [1.08, 1.53] 1.01 (0.08) [0.87, 1.18] --- --- 1.29 (0.12) ** [1.08, 1.55]
Engagement
Hours/ Week on H.W. 1.02 (0.02) [0.98, 1.06] 1.02 (0.01) [0.99, 1.04] 1.02 (0.02) [0.99, 1.06] 1.02 (0.02) [0.99, 1.06]
Disengagement
92
G2 skipped school --- --- --- --- 0.63 (0.15)* [0.40, 0.99] 0.62 (0.15) * [0.39, 0.99]
Note. * p< .05, ** p<. 01, *** p<.001, G2= Generation 2, G1= Generation 1, GED = General Educational Development, H.W. = Homework.
a Adjusted odds ratios (standard error) [confidence interval]
b Unstandardized beta (standard error) [confidence interval]
Model 1: Educational attainment regressed on educational aspirations + covariates.
Model 2: Mediator regressed on educational aspirations + covariates.
Model 3: Educational attainment regressed on mediator + covariates.
Model 4: Educational attainment regressed on educational aspirations + mediator + covariates.
93
Table 3-6. Testing mediation for high school engagement and disengagement in Hispanic sample.
Panel A: Testing high school engagement as a mediator between aspirations and G2 earned a high school diploma/GED ‘on time’ G2 earned high school
diploma/GED ‘on time’ High school engagement
G2 earned high school diploma/GED ‘on time’
G2 earned high school diploma/GED ‘on time’
Model 1 a Model 2 b Model 3 a Model 4 a
G2 youth aspirations 1.36 (0.18) * [1.05, 1.76] 0.67 (0.32) * [0.03, 1.30] --- --- 1.32 (0.18) * [1.02, 1.71] Engagement
Hours/ Week on H.W. --- --- --- --- 1.06 (0.03) * [1.01, 1.12] 1.06 (0.03) [1.00, 1.12] Disengagement
G2 skipped school 0.48 (0.15) * [0.26, 0.89] -0.39 (0.80) [-1.98, 1.19] 0.48 (0.15) * [0.26, 0.89] 0.49 (0.16) * [0.26, 0.93] Panel B: Testing high school disengagement as a mediator between aspirations and G2 earned a high school diploma/GED ‘on time’
G2 earned high school diploma/GED ‘on time’
High school disengagement
G2 enrolled in college G2 earned high school diploma/GED ‘on time’
Model 1 a Model 2 a Model 3 a Model 4 a G2 youth aspirations 1.35 (0.18) * [1.04, 1.75] 0.86 (0.09) [0.70, 1.07] --- --- 1.32 (0.18) * [1.02, 1.71] Engagement
Hours/ Week on H.W. 1.06 (0.03) * [1.00, 1.12] 0.99 (0.02) [0.96, 1.02] 1.06 (0.03) * [1.01, 1.12] 1.06 (0.03) [1.00, 1.12] Disengagement
G2 skipped school --- --- --- --- 0.48 (0.15) * [0.26, 0.89] 0.49 (0.16) * [0.26, 0.93] 1.35 (0.18) * [1.04, 1.75] 0.86 (0.09) [0.70, 1.07] --- --- 1.32 (0.18) * [1.02, 1.71]
Note. G2= Generation 2, G1= Generation 1, GED = General Educational Development, H.W. = Homework. * p< .05, ** p<. 01, *** p<.001
a Adjusted odds ratios (standard error) [confidence interval]
b Unstandardized beta (standard error) [confidence interval]
Model 1: Educational attainment regressed on educational aspirations + covariates.
Model 2: Mediator regressed on educational aspirations + covariates.
Model 3: Educational attainment regressed on mediator + covariates.
Model 4: Educational attainment regressed on educational aspirations + mediator + covariates.
VITA
Emily Pressler
Education:
May 2014* Expected PhD, Human Development and Family Studies
The Pennsylvania State University
May 2006 B.S., Psychology
Fordham University
Professional Position:
Aug 2009- Current Graduate Research Assistant
Department of Human Development and Family Studies
The Pennsylvania State University
Publications:
Hernandez, D. C., & Pressler, E. (2014). Accumulation of childhood poverty on young adult weight
status: Race, ethnic, and gender disparities. Journal of Epidemiology and Community Health,
e-pub.
Hernandez, D. C., Pressler, E., Dorius, C., & Stamps Mitchell, K. (2014). Does family instability
make girls fat? Gender differences between instability and weight. Journal of Marriage and
Family, 76, 175-190.
Hernandez, D. C. & Pressler, E. (2013). Maternal union transitions and household food insecurity:
Differences by race and ethnicity. Journal of Family Issues, 34, 373-393.
Raver, C. C., Jones, S. M., Li-Grining, C. P., Zhai, F., Bub, K., & Pressler, E. (2011). CSRP's impact
on low-income preschoolers' pre-academic skills: Self-regulation as a mediating mechanism.
Child Development, 82, 362-378.