ACADEMIC SOCIALIZATION AND THE TRANSITION TO ELEMENTARY SCHOOL:
PARENTS’ CONCEPTIONS OF SCHOOL READINESS, PRACTICES,
AND CHILDREN’S ACADEMIC ACHIEVEMENT TRAJECTORIES
By
Jaime Lynn Puccioni

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Educational Policy
Curriculum, Instruction, and Teacher Education
2012

ABSTRACT
ACADEMIC SOCIALIZATION AND THE TRANSITION TO ELEMENTARY SCHOOL:
PARENTS’ CONCEPTIONS OF SCHOOL READINESS, PRACTICES,
AND CHILDREN’S ACADEMIC ACHIEVEMENT TRAJECTORIES
By
Jaime Lynn Puccioni
By the time children enter kindergarten, significant socioeconomic and racial gaps in
academic achievement exist (Coley, 2002; Rouse, Brooks-Gunn, & Mclanahan, 2005).
Kindergarten is considered to be a pivotal point of educational transition, as academic
achievement upon kindergarten entry is associated with subsequent academic success (Claessens,
Duncan, & Engel, 2009; Hair, Halle, Terry-Humen, Lavelle, & Calkins, 2006). Although
parents’ socioeconomic status and level of education are related to their children’s early
achievement, home-based parental involvement during early childhood also contributes to
children’s school readiness (Rouse et al., 2005). Taylor, Clayton, and Rowley (2004) offer a
model of academic socialization which suggests that parents’ beliefs about schooling influence
parenting practices, both of which contribute to children’s transition to elementary school. The
current study tests the theory that parents’ school readiness beliefs influence parenting practices,
which in turn influence children’s academic achievement during the transition to elementary
school. The impact of academic socialization on children’s achievement trajectories in reading,
mathematics, and general knowledge was examined using latent growth curve modeling.
Parents’ school readiness beliefs predicted children’s beginning academic achievement and
growth over time. Parents’ transition practices partially mediated the relationship between their
own readiness beliefs and children’s beginning academic achievement and growth over time.
The findings broaden our understanding of home-based parental involvement during early
childhood by illuminating the inherent connection between parents’ school readiness beliefs,

parenting practices, and children’s early academic achievement during the transition to
elementary school.

ACKNOWLEDGMENTS

It is my great pleasure to thank those who made the completion of this dissertation
possible. I am indebted and truly grateful to my dissertation chair and advisor, Dr. Kimberly
Maier whose guidance and support have been invaluable throughout the dissertation process. I
would also like to sincerely thank the members of my dissertation committee, Dr. Nell Duke, Dr.
Barbara Schneider, and Dr. Suzanne Wilson, each of whom provided thoughtful critiques and
suggestions. I would also like to express my gratitude to Dr. Alex von Eye, Dr. Rick DeShon,
and Dr. Tenko Raykov for providing answers to my many queries about structural equation
modeling. The folks at the National Center for Education Statistics were always so helpful,
thank you Kristin Flanagan and Jill McCarroll for answering all of my questions about the
ECLS-K data set. Finally, I would like to offer my heartfelt gratitude to my husband for his
unfaltering support and to my son for understanding that I had to work on my dissertation and
had less time to play. I promise to play, laugh, and generally have more fun, now that my
dissertation is complete.

iv

TABLE OF CONTENTS

LIST OF TABLES………………………………………………………………………..……vi
LIST OF FIGURES………………………………………………………………………...…vii
CHAPTER 1
INTRODUCTION………………………………………………………...……………………1
CHAPTER 2
LITERATURE REVIEW…………………………………………………...………………….5
CHAPTER 3
METHODS……………………………………………………………….…...……….………33
CHAPTER 4
RESULTS……………………………………………………………….……...…….………..50
CHAPTER 5
DISCUSSION…………………………………………………………………...…………….72
APPENDICES……………………………………………………………………..………….94
REFERENCES……………………………………………………………………..………..119

v

LIST OF TABLES

TABLE 1

Summary Statistics and Missing Rates for Model Variables……...………….102

TABLE 2

Correlations Among Parents’ Readiness Beliefs, Transition Practices,
and Reading Achievement Outcomes…………...…………………...…..…...105

TABLE 3

Correlations Among Parents’ Readiness Beliefs, Transition Practices,
and Mathematics Achievement Outcomes…………………….……...………106

TABLE 4

Correlations Among Parents’ Readiness Beliefs, Transition Practices,
and General Knowledge Achievement Outcomes………….…………...……107

TABLE 5

Summary of Model Fit Indices for all Measurement and Structural
Models Under Study……………………………………………………...…..108

TABLE 6

Factor Loadings of Indicators for Structural Model of Academic
Socialization and Reading Achievement …………..……….………….….....109

TABLE 7

Structural Model Estimates of Academic Socialization as a Predictor
of Reading Achievement………………….……………………………….....110

TABLE 8

Factor Loadings of Indicators for Structural Model of Academic
Socialization and Mathematics Achievement …………...…..………….…....112

TABLE 9

Structural Model Estimates of Academic Socialization as a Predictor
of Mathematics Achievement….………………………….………..………...113

TABLE 10

Factor Loadings of Indicators for Structural Model of Academic
Socialization and General Knowledge Achievement …………..…....…..…...115

TABLE 11

Structural Model Estimates of Academic Socialization as a Predictor
of General Knowledge Achievement…………………………….…..….…....116

TABLE 12

Tests of Mediation for Structural Equation Models from Parents’
Beliefs to Achievement Outcomes…………………………….……..………118

vi

LIST OF FIGURES

FIGURE 1

Conceptual model representing the influence of academic
socialization on children’s achievement ………………………….…….……..95

FIGURE 2

Measurement model of academic socialization and reading
achievement……………………………. ………………….…………....…….96

FIGURE 3

Measurement model of academic socialization and mathematics
achievement……………………………. ………………....…………….…….97

FIGURE 4

Measurement model of academic socialization and general
knowledge achievement…………………………………....………………….98

FIGURE 5

Structural model of academic socialization as a predictor of reading
achievement……………………………. ……………….…...………….…….99

FIGURE 6

Structural model of academic socialization as a predictor of mathematics
achievement……………………………. ……………….……...……….……100

FIGURE 7

Structural model of academic socialization as a predictor of general
knowledge achievement……………………………. ……..……...…….……101

FIGURE 8

Reading achievement trajectories by race and ethnicity conditional
on socioeconomic status……………………………………………...………111

FIGURE 9

Mathematics achievement trajectories by race and ethnicity conditional
on socioeconomic status………………………………………………...……114

FIGURE 10

General knowledge achievement trajectories by race and ethnicity
conditional on socioeconomic status……………………………………...….117

vii

CHAPTER 1
INTRODUCTION
Although gaps in educational achievement have narrowed over the past three decades, the
persistent demographic divide in educational outcomes continues to be of great concern in the
United States (Jencks & Phillips, 1998; Reardon & Robinson, 2008). These gaps exist in and out
of the classroom, extending from early childhood across the lifespan (Rouse, Brooks-Gunn, &
Mclanahan, 2005). By the time children enter kindergarten, significant socioeconomic and racial
gaps in school readiness already exist (Coley, 2002; Rouse et al., 2005). On average, children
from affluent families are more likely to know the alphabet, be able to identify sounds, and
understand more complex mathematical concepts before kindergarten (Coley, 2002). Data from
the Early Childhood Longitudinal Study (ECLS-K), a large, nationally-representative survey,
show that upon kindergarten entry, the average cognitive scores of affluent children are
approximately three-fifths of a standard deviation higher than those of children from lower
socioeconomic backgrounds (Coley, 2002; Lee & Burkham, 2002).
The transition to kindergarten is regarded as a pivotal developmental period, as patterns
of achievement and behavior established in the initial school years can have profound impacts on
children’s developmental trajectories for school success or failure (Alexander, Entwisle, &
Bedinger, 1994; Entwisle & Alexander, 1999; Entwisle, Alexander, & Olsen, 2005). Recent
findings indicate that cognitive and noncognitive skills upon school entry relate to later
achievement (Bodovski & Farkas, 2007; Claessens, Duncan, & Engel, 2009; Hair, Halle, TerryHumen, Lavelle, & Calkins, 2006; Li-Grining, Votruba-Drzal, Maldonado-Carreno, & Haas,
2010) and virtually all researchers associate socioeconomic status and parental level of education

1

with school readiness and subsequent academic success (Booth & Crouter, 2008; Brooks-Gunn
& Markman, 2005; Duncan & Magnuson, 2005).
Although status variables such as socioeconomic status and parental level of education
are clearly related to children’s early learning and development, parenting characteristics and
parent-child interactions have also been shown to contribute to children’s transition to
elementary school ( Booth & Crouter, 2008; Brooks-Gunn & Markman, 2005). Parents play a
key role in children’s early learning and attitudes toward school and achievement through a
process of socialization (Bempechat, 1992; Ginsberg, Bempechat, & Chung, 1992; Maccoby,
1992). Taylor, Clayton, and Rowley’s (2004) model of academic socialization suggests that
parents’ attitudes, values, and beliefs about school influence parenting behaviors and transition
practices, which in turn influence children’s transition to elementary school. Transition practices
are conceptualized as behaviors that are designed to prepare children for starting school and have
been operationalized as reading to the child, saying the alphabet, practicing counting, and so
forth (Barnett & Taylor, 2009; Taylor, Clayton, & Rowley, 2004).
Despite the increasing emphasis on families' contributions to children’s school readiness
and transitions to elementary school (Booth & Crouter, 2008; Pianta, Cox, & Snow, 2007),
research examining the role of academic socialization on young children’s early achievement has
received less attention (Barbarin et al., 2008; Barnett & Taylor, 2009; Taylor et al., 2004). It
seems plausible that differences in parents’ readiness beliefs and transition practices contribute to
the early academic achievement gap. Establishing a link between parents’ school readiness
beliefs and transition practices may enhance conceptual models of home-based parent
involvement, ultimately highlighting important targets for interventions aimed at maximizing the
positive parental contributions to children’s early learning and transitions to elementary school.

2

The purpose of this study is to examine the relationship between parents’ school
readiness beliefs, transition practices, and children’s academic achievement over time using
nationally representative data drawn from the Early Childhood Longitudinal Study-Kindergarten
cohort (ECLS-K). The primary contribution of this study is to understand the degree to which
parents’ school readiness beliefs influence their transition practices, which in turn influence
children’s academic achievement during the transition to elementary school. In other words, this
study will examine the degree to which parents’ readiness beliefs directly and indirectly
influence children’s academic achievement as they begin elementary school.
The study begins by examining variability within parents’ readiness beliefs and transition
practices as well as children’s achievement trajectories. The first set of research questions
addressed, are as follows.
1. Do parents’ school readiness beliefs vary by socioeconomic status and
race/ethnicity?
2. Do parents’ transition practices differ by socioeconomic status and race/ethnicity?
3. Do children’s academic achievement trajectories vary by socioeconomic status
and race/ethnicity?
The study then examines the degree to which parents’ school readiness beliefs and
transition practices influence children’s academic achievement trajectories. This was
accomplished by estimating the direct and indirect effects of parents’ readiness beliefs and
transition practices on children’s academic performance trajectories. The second set of research
questions addressed, are as follows.
4. To what extent do parents’ school readiness beliefs influence children’s academic
achievement upon kindergarten entry and over time?

3

5. To what extent do parents’ transition practices influence children’s academic
achievement upon kindergarten entry and over time?
6. To what extent do parents’ transition practices mediate the relationship between
their school readiness beliefs and children’s academic achievement upon
kindergarten entry and over time?

4

CHAPTER 2
LITERATURE REVIEW
This chapter begins with a review of existing literature relevant to the prediction of
children’s academic achievement during the transition to elementary school. The review begins
with an overview of racial/ethnic and socioeconomic academic achievement gaps as children
transition to elementary school. Academic socialization provides the conceptual framework for
the present study and is explicated next. Current studies examining parents’ conceptions of
school readiness are discussed. Then, recent empirical research that has tested hypothesized
models using more traditional statistical techniques (e.g., standard multiple regression) is
described, followed by studies that test mediating models using sophisticated, multivariate
statistical techniques (e.g., path analysis). A description of the limitations of existing research
follows, and the chapter concludes with an example of how this study will address these
limitations.
Existing Research
Achievement Gaps as Children Progress through Elementary School
Considerable attention has focused on the persistent racial, ethnic, and socioeconomic
disparities in academic achievement in the United States. National studies repeatedly show that
the average African American student scores below the average European American student on
standardized tests of reading and mathematics skills, as does the average Hispanic student
(Jencks & Phillips, 1998; Reardon & Gallindo, 2009; Reardon & Robinson, 2008). The average
student from a low socioeconomic status family scores much lower on such tests than students
from higher socioeconomic status families (Reardon, 2011; Reardon & Robinson, 2008). The
next section presents information on achievement gaps between African American and European
American children as well as differences between Hispanic and European American children as
5

they progress through elementary school, followed by a discussion of achievement gaps between
children from high- and low-socioeconomic-status homes.
The most commonly cited contemporary evidence on the development of achievement
gaps in elementary school comes from the Early Childhood Longitudinal Study—Kindergarten
cohort of 1998-1999 (ECLS-K), which includes kindergarten through eighth grade data on a
nationally representative sample of children who were enrolled in kindergarten in the fall of
1998. Analysis of the ECLS-K data show that on average, gaps between African American and
European American children in reading and mathematics assessments are sizable upon
kindergarten entry, about one-half and three-quarters of a standard deviation, respectively (Fryer
& Levitt, 2004; Reardon & Robinson, 2008). Measured in standard deviation units, these gaps
widen slightly through kindergarten and first grade, and then widen more rapidly between first
and fifth grade, by which time the gap in reading is about three-quarters of a standard deviation,
and the mathematics gap is a full standard deviation (Reardon & Robinson, 2008).
In regards to the achievement gap between Hispanic and European American students,
the most detailed evidence comes from the ECLS-K study, which includes a large sample
(approximately 4,000) of Hispanic children. In addition, because the ECLS-K study
administered the mathematics test orally in either English or Spanish (depending on the child’s
language proficiency), the estimates for the differences in mathematics scores should not be
biased by differences in English proficiency. Analysis of the ECLS-K data indicates that
differences between Hispanic and European American children’s reading and mathematics
achievement is very similar in magnitude to the observed differences between African American
and European American children’s achievement. However, the gap between Hispanic and
European American children in reading and mathematics decreases during the elementary school

6

years, while the gap between African American and European American children widens during
the same period (Fryer & Levitt, 2004; Reardon & Gallindo, 2009; Reardon & Robinson, 2008).
In reading, the gap shrinks from one-half to about one-third of a standard deviation, and the gap
in mathematics shrinks from three-quarters to one-half of a standard deviation (Reardon &
Gallindo, 2009; Reardon & Robinson, 2008).
The development of achievement gaps based upon socioeconomic status can be difficult
to examine because unlike race, ethnicity, or gender, socioeconomic characteristics of a family
may change over time, and often change quite dramatically. In addition, socioeconomic status
can be measured in several ways by using a variety of different indicators, such as parental level
of education, occupational status, and family income. Using a composite indicator of
socioeconomic status in the ECLS-K data set, Lee and Burkham (2002) found that in reading and
mathematics during the fall term of kindergarten, children from homes in the lowest quintile
(low-income) scored a full standard deviation below children from homes in the highest quintile
(high-income) and a half of a standard deviation below children from homes in the third quintile
(middle-income). Reardon and Robinson's (2008) analysis of the ECLS-K data found that upon
kindergarten entry, children already showed gaps in reading and mathematical abilities, where
children with parents who have less formal education scored more than a standard deviation
below children with parents who have a college education. The gaps narrowed by 10 % during
the first two years of schooling, but then widened slowly through the fifth grade.
It is a well known phenomenon that race/ethnicity and socioeconomic status are closely
intertwined in the United States. Using multiple socioeconomic status indicators, Fryer and
Levitt (2005) were able to show that socioeconomic factors explained almost all of the difference
in African American and European American children’s mathematics achievement (85%) at the

7

beginning of kindergarten, and all of the difference in reading achievement during the same time
period. By the third grade, however, socioeconomic factors only accounted for approximately
60% of the difference in reading and mathematics achievement. This finding was significant
because it suggests that socioeconomic factors explain, in large part, differences in achievement
at the start of formal schooling, but do not account for differences in the rate of growth during
school. Fryer and Levitt (2005) also found that, on average, African American children continue
to lose ground in reading and mathematics relative to European American children, even while
taking into account important sociodemographic covariates as well as school and teacher fixed
effects. Seasonal comparison research offers a plausible hypothesis for these findings by
suggesting that children experience equivalent gains during the school year but decline during
the summer and winter vacations, due to family and neighborhood factors (Entwisle &
Alexander, 1992, 1994). However, in a recent study using multilevel growth curve models,
Downey, Von Hippel, and Broh (2004) found that African American children experienced less
rapid growth in reading and mathematics in comparison to European American children during
kindergarten and first grade, but no significant differences in the rate of learning during the
summer.
The extent to which socioeconomic factors account for differences between Hispanic and
European American children’s academic achievement in elementary school appears to be quite
different. Fryer and Levitt (2005) show that socioeconomic factors account for 75 to 80% of the
gap between Hispanic and European American children in kindergarten and 85 to 100% of the
gap in the third grade. The authors suggest that as Hispanic children’s English proficiency
improves, socioeconomic factors explain an increasing portion of the variance in achievement.
Reardon and Gallindo (2006) also show that differences between Hispanic and European

8

American children conditional on socioeconomic status narrow from kindergarten through fifth
grade, while the corresponding gap between African American and European American children
increases during the same time.
In sum, research clearly demonstrates that racial, ethnic, and socioeconomic achievement
gaps exist during the transition to elementary school. The gap between African American and
European American children is apparent at the onset of formal schooling. This gap continues to
widen during elementary school in ways that are not completely explained by socioeconomic
family background characteristics or school quality. The same patterns are not found for
Hispanic children as they transition to elementary school; socioeconomic differences account for
a large portion of the gap between Hispanic and European American children’s reading and
mathematics achievement and the gap appears to narrow as children progress through elementary
school. Although it is clear that family background and schooling each play some role in the
development of achievement gaps, the extant research is less clear about the processes and
mechanisms that occur prior to formal schooling and produce racial, ethnic, and socioeconomic
achievement disparities. The theory of academic socialization is presented as a plausible
hypothesis in the following section.
Conceptual Framework
Parents are considered to be the primary agents of socialization during early childhood.
The process by which parents shape children’s cognitive and non-cognitive skills is broadly
encompassed by the term socialization (Bempechat, 1992; Ginsberg et al., 1992; Maccoby, 1992;
Maccoby & Martin, 1983). The conceptual model of academic socialization put forth by Taylor,
Clayton, and Rowley (2004) draws upon Brofenbrenner’s (1986) ecological theory of
development as well as Pianta and Walsh’s (1996) contextual systems model. The ecological

9

theory of development suggests that intrafamilal and extrafamilial factors shape children’s
developmental trajectories, while the contextual systems model emphasizes intrafamilial and
school system factors. The conceptual model of academic socialization reflects the tenets of
ecological theory as well as the contextual systems model but extends the conventional
understanding of these relationships across systems by taking into account parents’ beliefs about
schooling. Academic socialization suggests that parents’ cognitions about schooling influence
parenting behaviors, which in turn, influence children’s early outcomes during the transition to
elementary school (Taylor et al., 2004). Taylor, Clayton, and Rowley’s (2004) conceptual model
of academic socialization also reflects tenets of ecological theory and contextual systems model
by suggesting that children’s academic socialization is influenced by one’s socioeconomic and
cultural contexts. Academic socialization provides the conceptual framework for the present
study, as this model integrates several important features of parenting that can influence
children’s early learning and development: who the parents are, what they believe about
schooling, and what they do.
Parental involvement is a multidimensional construct that includes not only direct
involvement in children’s schools, such as volunteering in classrooms and attending parentteacher conferences, but also home-based involvement (Epstein & Connor, 1992; Hill & Taylor,
2004; Suizzo & Stapleton, 2007). Parental involvement for young children entering kindergarten
has been operationalized as attending school meetings, volunteering in classrooms, and serving
on committees (Barnett & Taylor, 2009; Taylor et al., 2004), all of which, has been found to be
associated with their children's early school success (Grolnick & Slowiaczek, 1994; Hill, 2001;
Hill & Craft, 2003). However, there is a home-based component of parent involvement, outlined
by Epstein and Conner (1992), which includes parental interaction with children in the home.

10

Scott-Jones (1995) builds upon this framework of parental involvement by suggesting that
parental interaction can be conceptualized in four ways: valuing, monitoring, helping, and doing.
One important aspect of home-based parental involvement involves helping interactions.
Helping interactions are specific parent-child learning activities focused on the acquisition of
basic academic skills in reading, mathematics, and other subject matter (Scott-Jones, 1995) and
can be thought of as part of the academic socialization process for young children, which Taylor
et al. (2004) define as transition practices. Parental transition practices may include reading to
the child, singing songs, or practicing counting, all of which are intended to prepare children for
starting school (Barnett & Taylor, 2009; Taylor et al., 2004).
The current study also draws upon a social constructivist theoretical formulation
advanced by Wertsch (1979), based on the work by Vygotsky (1978) to further develop the
construct of transition practices. Parents help young children acquire skills and knowledge in an
expert-novice relationship by serving as the supportive, knowledgeable “others” as children
acquire new skills and knowledge. The interactions of children with adults provide other
regulation needed for the child to perform cognitive tasks. From interactions with adults
involving other regulation, children gradually develop the capacity for self-regulation.
According to this theory, children’s mental activity exists first on a social or inter-psychological
level. Gradually, after social interactions with parents or other knowledgeable persons,
children’s mental activity occurs at an individual or intra-psychological level. This model of
children learning from a knowledgeable parent in an everyday social context and gradually
advancing to independent performance is termed apprenticeship in the work of Rogoff (1990).
In the current study, transition practices are conceptualized as a form of home-based parental
involvement that focuses on parental helping interactions which provide a type of other

11

regulation or apprenticeship opportunity to help prepare children for elementary school.
Therefore, transition practices are considered to be targeted parent-child interactions which are
intended to support children’s successful transition to elementary school.
Academic socialization suggests that home-based parental involvement mediates the
relationship between who parents are, what they believe about schooling, and children’s early
outcomes. The main support for the mediating role of parent behaviors comes from literature
that examines the simple associations between pairs of status variables, such as socioeconomic
status and parent behaviors, parent behaviors and children’s academic achievement. The
literature posits that parents, as primary caregivers, are the main agents for the promotion of
positive academic experiences during children’s early years. The evidence comes from
intervention research (e.g., Boyce, Innocenti, Roggman, Norman, & Ortiz, 2010; Olds &
Kitzman, 1993; Olds et al., 2007; Schweinhart et al., 2005; Senechal & Young, 2008; Stormshak,
Kaminski, & Goodman, 2002), qualitative studies (e.g., Heath, 1983; Milne & Plourde, 2006),
and quantitative studies (e.g., Cooper, Crosnoe, Suizzo, & Pituch, 2010; Foster, Lambert,
Abbott-Shim, McCarty, & Franze, 2005; Lamb Parker, Boak, Griffin, Ripple, & Peavy, 1999;
Orr, 2003; Smith, Brooks-Gunn, & Klebanov, 1997; Votruba-Drzal, 2006).
Several aspects of parenting behaviors have been shown to play an essential role in
children’s early achievement. One important aspect, and the most frequently researched area,
focuses on the home literacy environment. Aulls and Sollars (2003) reported that differences in
preschool children’s print awareness, book knowledge, and code knowledge are related to the
quality of the literacy activities provided at home. Similarly de Jong and Leseman (2001)
demonstrated that a high level of home literacy enhances vocabulary and reading achievement in
children. In the seminal longitudinal study conducted by Hart and Risley (1995), the authors

12

observed 42 families over a period of three years, demonstrating that the quantity and quality of
parent-child interactive language during the first three years of life is related to the differences
between vocabulary development in children of higher versus lower socioeconomic status
families. Activities such as telling stories, singing songs, and making books have also been
shown to encourage the acquisition of literacy skills (Boyce et al., 2010; Bryant, Maclean,
Bradley, & Crossland, 1990; Glazer, 1989; Maclean, Bryant, & Bradley, 1987; Sonnenschein,
Brody, & Musterman, 1996). In a meta-analytic review, Senechal and Young (2008) found that
interventions in which parents tutored their children using specific literacy activities produced
better results compared to interventions where parents only read to their child or listened to their
child read. It is clear from the research that parental behaviors, more specifically the quantity
and quality of parent-child interactions, plays an important role in children’s early learning and
development. However, these parenting behaviors vary by important sociodemographic
characteristics.
Research suggests that home-based parental involvement, as evidenced by the frequency
of parents engaging young children in language and literacy-related activities as well as
educational activities, varies by parental socioeconomic status (Hart & Risley, 1995; Heath,
1983; Lareau, 2003), level of education (Farkas & Hibel, 2008), and race/ethnicity (BrooksGunn & Markman, 2005). In a recent analysis of the Panel Study of Income Dynamics, which
used time-diary data to measure how young children spend their time, Phillips (2011) found
significant socioeconomic and racial differences in parenting and the amount of time children
spend in different contexts and activities. The author found that, on average, infants, toddlers,
and preschoolers from high-income families spend more time in novel environments such as
indoor and outdoor recreation places, church, businesses or other institutions, relative to children

13

from lower-income families. In addition, among school-age children, low-income children
seemed to converse less frequently with parents compared to children from high-income
families. Holding measures of socioeconomic status as well as child and parent health constant,
Phillips (2011) demonstrates that, on average, African American infants, toddlers, and
preschoolers spend four hours less time per week conversing with adults in comparison to young
European American children. The author also demonstrates that, on average, high-income
children, from infancy to age six, spend more time engaged in literacy activities than their lowincome counterparts. Furthermore, African American infants are read to less often than their
European American counterparts from similar backgrounds, but this disparity decreases as
children age. Phillips (2011) notes that although these differences are seemingly small, the
disparities in weekly reading time implies a gap of over 100 hours of reading exposure by the
time children enter school. These recent findings are consistent with previous research which
suggests that children from different sociodemographic backgrounds have different learning
experiences during early childhood.
Academic socialization as proposed by Taylor et al., (2004) has received limited attention
in the literature, but it is widely assumed that parents’ school readiness beliefs will influence
their behaviors, and ultimately affect children’s early learning and development (Barbarin et al.,
2008; Miller, 1988; West, Germino-Hausken, & Collins, 1995). Studies have empirically
examined the relationship between hypothesized predictors within the academic socialization
framework, that is, who the parents are, what they believe, what they do, and their children’s
early outcomes. In the next section, this research is reviewed. The focus of the review is on the
findings of complex, multivariate models that have explored hypothesized mechanisms
underlying children’s early achievement. Sophisticated statistical modeling techniques that

14

allow for the investigation of interrelationships among multiple predictors simultaneously offer
potential advantages over methods that explore fewer predictors. The following section will
begin by reviewing studies which employ traditional multiple linear regression models to test the
relationships between parental status variables, behaviors, and early child outcomes. Then the
review focuses on empirical studies that examine the relationship between parents’ school
readiness beliefs, behaviors, and children’s early academic outcomes using standard linear
regression techniques as well as path analysis to examine the mediation of parental beliefs and
early child outcomes by parenting behaviors.
Empirical Studies
Who parents are, what parents do, and early child outcomes. Studies have
empirically analyzed the relationship between family background, home-based parent
involvement, and early developmental outcomes for children using large-scale nationally
representative data sets. In the seminal study on this topic, Smith et al. (1997) analyzed data
drawn from the Children of the National Longitudinal Survey of Youth (CNLSY) and the Infant
Health and Development Program (IHDP) studies to examine the consequences of living in
poverty on young children’s early cognitive, verbal, and early academic achievement. Results
from multiple regression analyses found that parental income was associated with child
outcomes and measures of home-based parental involvement (measured by the HOME inventory
scale, which is a global measure of parenting practices), partially mediated the correlation
between parental income and child outcomes as well as maternal education and child outcomes.
In another study using the same data, Phillips, Brooks-Gunn, Duncan, Klebanov, and Crane
(1998) examined the relationship between family background, parenting practices, and young
children’s verbal and cognitive performance. Standard multiple regression analyses were also

15

conducted; findings indicated that parental income, maternal education, and measures of homebased parent involvement were all significant predictors of children’s vocabulary development.
Studies analyzing the more recent ECLS-K data by Lee and Burkham (2002), as well as
Farkas and Hibel (2008), have also found that family status variables, such as socioeconomic
status, the maternal level of education, and measures of home-based parent involvement, were
significant predictors of children’s overall kindergarten achievement in reading and mathematics.
Cooper et al., (2010) employed multilevel modeling techniques to analyze ECLS-K data, and
found that home-based parental involvement in education partially mediates the association
between family poverty and children’s mathematics and reading achievement, but also
discovered differences paralleling racial/ethnic heritage.
Several quantitative studies have used mediation models (path analysis) to investigate the
mediating role of distal (socioeconomic) and proximal (early home-based parental involvement)
predictors and children’s school achievement. Votruba-Drzal (2006) used New York
Longitudinal Study (NYLSY) data, in which an abbreviated version of the HOME inventory
measured home-based parental involvement as evidenced by the amount of cognitive stimulation
and emotional support the parent provided the child. The author investigated the mediating role
of early childhood parent involvement in the association between family income and reading,
mathematics, and behavior problems in middle childhood. The findings indicate that the quality
of early childhood home-based parent involvement mediates the influence of family income
during early childhood on reading, mathematics, and behavior problems, and shows that the
effects of family interaction are even more influential on reading skills and behavior problems
than on mathematics achievement. Other studies have also found that measures of home-based
parental involvement mediates the relationship between risk factors and pre-kindergarten

16

achievement, self-regulation, and social behavior (Mistry, Benner, Biesanz, Clark, & Howes,
2010), between income and emergent literacy (Foster et al., 2005), and between income and
children’s ability to pay attention (Dilworth-Bart, Khurshid, & Vandell, 2007).
These studies contribute to the literature in several important ways. These studies clearly
indicate that distal and proximal predictors contribute to children’s early learning and
development. Path analysis models also provide evidence that home-based parental involvement
mediates the relation between parents’ sociodemographic factors and early child outcomes. The
question that remains, what influences the nature and degree of home-based parental
involvement? The conceptual model of academic socialization suggests that parental beliefs or
expectations about schooling influence parenting behaviors, which then influence child
outcomes. The subsequent section reviews studies which examine the relationship between
parental school readiness-related beliefs and expectations, behaviors, and children’s early
academic outcomes using standard correlation or regression methods, followed by a review of a
study that employs path analysis to examine the mediation of parental beliefs and child outcomes
by parental involvement.
What parents believe about readiness, what parents do, and early child outcomes.
Studies examining parental beliefs can generally be divided into two major categories: beliefs
about children’s developmental processes, and beliefs about children’s specific abilities
(Goodnow & Collins, 1990; Miller, 1988). These studies have focused on parental beliefs about
the ages that children are expected to perform certain tasks (Hess, Kahiwagi, Azuma, Price, &
Dickson, 1980; Rosenthal & Gold, 1989), traits or values parents want to develop in their
children (Kohn, 1969; Levine, 1988; Okagaki & Frensch, 1998; Okagaki & Sternberg, 1993),
theories about child development (Goodnow, 1998; Keels, 2009; Kinlaw, Kurtz-Costes, &

17

Goldman-Fraser, 2001), and more recently, conceptions of school readiness (Barbarin et al.,
2008; Diamond, Reagan, & Bandyk, 2000). To understand the relationship between what
parents believe about school readiness and early academic outcomes, it is useful to begin with a
discussion about the role of parents’ educational expectations on children’s academic
achievement because it is assumed that parents’ expectations influence parent behaviors, which
in turn, influence children’s achievement.
Although the term parental expectation has been defined in various ways in the literature,
most researchers characterize parental expectations as realistic beliefs or judgments that parents
have about their children’s future achievement as reflected in course grades (Alexander et al.,
1994; Entwisle & Alexander, 1996; Hill, 2001), highest level of schooling attained (Goldenberg,
Gallimore, Reese, & Garnier, 2001), college attendance (Glick & White, 2004), as well as
expectations for mastery of early developmental tasks (Hess, Holloway, Dickson, & Price, 1984;
Hess et al., 1980). Much of the research evidence for the connection between parents’
expectations and academic outcomes has focused on adolescents from various ethnic groups
(e.g., Chen & Lan, 1998; Goyette & Xie, 1999; Reynolds & Gill, 1994; Smith-Maddox, 1999).
A few studies have focused specifically on how parental educational expectations influence
outcomes for elementary school-age children (Hess et al., 1984; Hill, 2001; Wentzel, 1998),
including children from diverse backgrounds (Davis-Kean, 2005; Entwisle & Alexander, 1996;
Halle, Kurtz-Costes, & Mahoney, 1997), and how these early expectations affect outcomes
through children’s school careers (Goldenberg et al., 2001; Hess et al., 1984). In general,
parental expectations have been found to play an important role in students’ academic success.
Students whose parents hold high expectations receive higher grades, achieve higher scores on

18

standardized tests, and persist longer in school than do those whose parents hold relatively low
expectations (Davis-Kean, 2005; Halle et al., 1997; Hess et al., 1984; Pearce, 2006).
Despite the role parental educational expectations play in children’s academic success, it
seems logical that parents’ conceptions of school readiness are more likely to shape early
academic outcomes through home-based parental involvement before children transition to
elementary school. The concept of school readiness has evolved over the past few decades, from
its formal introduction by the National Education Goals Panel (NEGP 1997), to more recent
efforts by Ackerman and Barnett (2005) and Snow (2006). The general conceptualization is that
school readiness encompasses the skills, knowledge, and dispositions associated with later
success in school (Ackerman & Barnett, 2005; Crinic & Lamberty, 1994; Snow, 2006). Despite
the growing emphasis on school readiness, research examining parents’ school readiness beliefs
is somewhat limited (Barbarin et al., 2008). The extant literature is descriptive, often comparing
parents’ readiness beliefs to teachers’, and rarely examines the link between school readiness
beliefs and children’s academic achievement.
Parents conceive of school readiness largely in terms of academic skills, such as naming
letters and numbers (Barbarin et al., 2008; Harradine & Clifford, 1996; West et al., 1995).
Several studies have shown that parents and teachers have different views of school readiness,
with parents placing more emphasis on academic skills, such as knowing letters and numbers, in
comparison to kindergarten teachers who place more emphasis on social and emotional
development (Haines, Fowler, Schwartz, Kottwitz, & Rosenhoetter, 1989; Harradine & Clifford,
1996). These differing parent-teacher beliefs are most pronounced when parents are from lowincome homes (Piotrkowski, 2004; Piotrkowski, Botsko, & Mathews, 2000) and have had less
formal education (West et al., 1995).

19

Parents’ readiness beliefs may be influenced by their social or cultural contexts (Graue,
1993; West et al., 1995). Graue’s (1993) ethnographic study found that working class parents
living in a small rural community were more likely to base school readiness beliefs and
expectations on their own kindergarten experiences, where identifying the children's elementary
school entrance date was parents' most important school-related activity. In contrast, the more
affluent parents in the community were active data collectors who held and utilized highly
developed ideas about educational practices and school readiness issues, based on information
gathered from conversations within their social network of teachers, school administrators, and
parents.
Using data from the National Household Education Survey, West et al. (1995) examined
how parents’ readiness beliefs were associated with sociodemographic characteristics. The
National Household Education Survey included seven school readiness-related questions focused
on behavioral and pre-academic tasks. These questions asked parents to rate on a 5-point Likert
scale how important it is for the child to be able to take turns and share, communicate needs, be
curious about approaching new activities, sit still and pay attention, know the letters of the
alphabet, count to 20, and be able to use pencils and paint brushes before entering kindergarten.
Chi-square analyses indicated that parents with lower levels of formal education placed greater
emphasis on school readiness attributes such as sitting still and paying attention, counting,
knowing the alphabet, and using pencils in comparison to parents with higher levels of formal
education. Diamond, Reagan and Bandyk (2000) also analyzed the NHES data to examine the
relationship between parents’ conceptions of readiness to race, ethnicity, and development.
Using principal component analysis, the authors developed a unitary construct representing
parents’ conceptions of school readiness. Comparison of mean factor scores revealed that all

20

parents, regardless of race or ethnicity, placed relatively high levels of importance on children
displaying certain skills and attributes prior to kindergarten.
These studies contribute to the literature by offering some insight into parents’
conceptions of school readiness. Overall, the studies suggest that all parents have concerns about
their child’s transition to elementary school. There is some evidence that parents’ school
readiness beliefs are linked to one’s social and cultural contexts, yet the evidence is inconclusive.
Although the literature around parents’ current school readiness views is limited (Barbarin et al.,
2008), there are a handful of studies which examine the relationship between parents’ school
readiness beliefs or expectations for mastery of developmental tasks, and children’s early
academic outcomes. This research is reviewed next.
Parents’ expectations for mastery of early developmental skills have also been shown to
be related to later achievement (Hess et al., 1984; Hess et al., 1980). In a longitudinal study
examining the predictive relationship between maternal characteristics and academic
performance within a small sample of European American families, Hess et al. (1984) found that
developmental expectations and maternal behaviors were related to measures of school readiness
and sixth grade achievement. The measure of maternal developmental expectations was a
composite score based upon items from a developmental expectation questionnaire that asked
mothers to rate the age range during which their child would master types of developmental tasks
(e.g., take care of own clothes, help with household tasks, eat at table), and acquire schoolrelevant skills (e.g., tell time up to a quarter of an hour), predict their child’s educational and
occupational status as an adult, and specify concerns about hypothetical challenges their child
may have in school. Measures of maternal behaviors assessed at the onset of the study focused
on teaching and communication style, as well as affective interaction between mother and child.

21

School readiness assessments were administered to children ages 5 and 6 and measured
knowledge in the areas of reading and mathematics. When the children were 12 years old, they
took the vocabulary and mathematics concepts subtest of the Iowa Tests of Basic Skills (ITBS).
Results from regression analyses indicated that maternal expectations for achievement and
behaviors (specifically, communication and the use of descriptors during play) assessed during
the preschool years were positively related to measures of school readiness. Measures of
maternal teaching style, communication, affective tone, and maternal expectations continued to
predict children's later performance.
Using data from 452 children participating in a public pre-kindergarten program,
Barbarin et al. (2008) examined the relationship between parents’ conceptions of school
readiness, race and ethnicity, socioeconomic status, and kindergarten outcomes. In order to
ascertain parents’ conceptions of school readiness, the authors asked the following questions in
an open-interview process: (a) In your opinion, what must your child know or be able to do by
the time she or he starts school in kindergarten?; (b) What knowledge or ability must a child
possess to be ready for school? The authors used content analysis to develop the following
response clusters of parents’ school readiness beliefs: (a) nominal knowledge (the ability to name
objects, letters, colors, and numbers); (b) inferential reasoning (higher order thinking skills); (c)
independence (the capacity for self-care and autonomy); (d) motor skills (the ability to engage in
activities that require fine and gross motor skills); (e) language/early literacy (skilled use of
language and emergent literacy, e.g., use vocabulary, link letters to sounds, read; (f) numeracy
(beginning skills in mathematics, e.g., add, subtract, and count money); (g) social competence
(the ability of a child to develop positive relationships with peers and adult caregivers); (h) selfregulation (the capacity of the child to comply with behavioral expectations of the school); and

22

(i) general knowledge (information that the child needs for problem solving, self-identification,
and handling situations when away from home, e.g., knowledge of name, age, and home
address).
Frequency analyses revealed that parents conceived of readiness largely in terms of
nominal knowledge, followed by language/early literacy, with inferential skills rarely mentioned.
Chi-square analyses indicated no significant association between parents’ readiness beliefs and
poverty status. There were significant differences by parental employment status. In general,
employed parents placed more importance on school readiness and were more likely to cite
inferential reasoning as an important readiness trait in comparison to non-employed parents.
There were also statistically significant differences in parents’ beliefs in the areas of social
competence and self-regulation. In general, African American parents placed less emphasis on
social competence and self-regulation in comparison to European Americans and Hispanic, but
these differences were only significant for low-income parents. Correlations between parents’
readiness beliefs and child outcomes indicated that parents who believed numeracy and general
knowledge to be important school readiness attributes had children with greater knowledge of
letters, numbers, and colors, while parents who emphasized inferential skills had children with
greater vocabularies, phonemic awareness, and applied problem solving skills.
How do parents’ beliefs about schooling influence home-based parental involvement? In
a recent study, Barnett and Taylor (2009) examined the relationship between parents’ previous
schooling experiences and current kindergarten transition practices with a sample of 79 mothers
from a diverse community using structured interviews. Mothers’ recollections were of their own
experiences with schools, teachers, and peers as well as their parents’ school involvement. The
mothers' statements were coded for emotional positivity or negativity. Transition practices were

23

operationalized as academic-oriented activities, such as reading to the child or saying the
alphabet. Using multiple linear regression techniques, the authors found that parents who had
positive recollections of their parents’ school involvement engaged in more transition practices,
while accounting for parental income, self-esteem, and self-efficacy constant.
Sy and Schulenberg (2005) tested a mediating model which hypothesized that parental
beliefs and expectations influence measures of parental involvement which, influence children’s
achievement over time in European American and Asian American families using latent growth
curve modeling. Using data drawn from the ECLS-K data set, the authors developed latent
constructs representing parental beliefs, expectations, and involvement. The construct
representing parental beliefs included three measures that assessed the importance parents placed
on children's early academics, such as knowing the alphabet, counting, and using writing
materials. The parents’ educational expectations factor consisted of one measure which asked
parents about the highest level of education they expected their child to attain. Parent
involvement was divided into two categories: in-the-home and outside-of-the-home involvement.
The constructs representing parent involvement in the home included creating a home literacy
environment and setting rules for watching television. The home literacy environment construct
was made up of two measures that assessed the frequency that parents read and told stories to the
child. One indicator variable was used to develop the construct representing family rules for
television viewing. Parent involvement outside of the home was divided into two categories:
school participation and educational activities outside of the home. The construct representing
school participation included three indicator variables that measured the number of times parents
volunteered in the classroom, attended open house night, participated in back-to-school events,
or attended general school events. Educational activities outside of the home were based on

24

three measures asking if parents took the child to the zoo, aquarium or petting farm, library,
museum, art gallery, or historical site. The model tested the theory that parents’ beliefs and
expectations influenced measures of parental involvement, which in turn, influenced children’s
achievement.
Results indicated that Asian American parents had higher mean beliefs and expectation
scores in comparison to European American parents. Analysis of the multigroup model
indicated that European American parents’ beliefs had a positive influence on home literacy
activities, television viewing rules, and school participation. Asian American and European
American parents' educational expectations had a positive influence on home literacy activities,
television rules, and educational activities, while the relationship between expectations and
parents' school participation varied. There was a significant negative relationship between
educational expectations and school involvement for Asian American parents, in contrast to the
significant positive association found for European American parents.
Measures of parental involvement were positively related to European American and
Asian American children’s achievement. Home literacy activities had a significant positive
relationship with children’s beginning reading achievement in kindergarten and growth through
the first grade. There was also a positive relationship between home literacy activities and
mathematics achievement in kindergarten, but a negative relationship with growth. In addition,
there was a significant positive relationship between school participation and reading or
mathematics growth over time. There was no statistically significant relationship between
educational activities outside of the home and children’s reading or mathematic growth in
kindergarten. Overall, parental beliefs were positively related to parental involvement, more

25

specifically home literacy and school participation, which in turn had a positive influence on
European American and Asian American children’s academic achievement.
Although our knowledge about parents’ current views of readiness is somewhat limited
(Barbarin et al., 2008), these studies contribute to the literature in several important ways. The
findings reveal that parents do hold certain beliefs and expectations about children’s readiness
for school. There is also some evidence that parents’ conceptions of school readiness vary by
social and cultural contexts, yet these findings are inconclusive. Although much of the research
is based upon analyses of large-scale secondary data, it has been argued that survey questions
may fail to capture parents’ nuanced ideas around school readiness (Diamond et al., 2000).
Barbarin et al. (2008) used open-ended interview questions, which allowed parents to share their
conceptions of readiness in contrast to simply answering survey questions. Results from the
content analysis confirmed previous findings which suggest that parents placed greater emphasis
on nominal knowledge, such as knowing the letters of the alphabet and counting, in comparison
to other areas of school readiness (Harradine & Clifford, 1996; Piotrkowski et al., 2000; West et
al., 1995). Therefore, one could argue there is some evidence that survey instruments are
somewhat valid measures of parents’ school readiness beliefs.
The studies reviewed thus far have significance for parents of young children
transitioning to elementary school. There was some evidence to suggest that parents’
conceptions of school readiness are positively related to children’s early academic achievement.
In addition, there was some initial evidence to support the idea that parental beliefs and
expectations are positively related to parental involvement (operationalized as the home literacy
environment, school participation, and educational activities outside of the home), which then
influences European American and Asian American children’s early academic achievement.

26

However, additional research is warranted to examine these relationships in other populations of
interest.
The studies reviewed have several limitations, some of which are related to
generalizability or methodological issues. The Hess et al. (1984) study examined the
relationship between parents’ expectations for early developmental skills and academic
achievement drawing upon a small and select sample of European American children; therefore,
the findings could not be generalized to a larger population. Although the Barbarin et al. (2008)
findings were based on a larger and more diverse sample in comparison to the study conducted
by Hess et al. (1984), there were still issues of generalizability. The study sample was drawn
from participants of a publicly funded pre-kindergarten program in several states and the authors
reported that there were a disproportionate number of low-income parents in the sample.
Thererfore, findings from the Barbarin et al. (2008) study can only be generalized to parents
participating in publicly funded pre-kindergarten programs. Methodologically, the study
conducted by Barbarin et al. (2008) provided a more nuanced understanding of parental
conceptions of school readiness; however, the study uses correlation methods to estimate the
relationship between parents’ conceptions of school readiness and child outcomes. The use of
correlation allows the investigator to analyze the association between parent beliefs and child
outcomes, but does not control confounding variables or estimate the relative contribution of
predictors, as is the case in standard linear regression. Although the studies could not generalize
their findings to the larger national population, they do offer initial insight about the relationship
between parents’ school readiness-related beliefs and children’s academic outcomes in
elementary school.

27

The study conducted by Sy and Schulenberg (2005) employed sophisticated modeling
techniques to analyze the relationship between parental beliefs, involvement, and early academic
achievement based on a nationally representative sample of European American and Asian
American children in the United States. Therefore, the findings from the study have greater
generalizability. There were, however, several methodological issues. For example, the ECLSK study used a multistage sampling design where children are clustered within schools, schools
are clustered within strata, and strata are clustered within primary sampling units. The
researchers, however, did not address the hierarchical nature of the data, which may result in
biased estimates. Second, all of the latent exogenous factors (predictors) were comprised of
ordinal variables, yet the authors treated them as continuous variables instead of using a
continuous/categorical variable methodology (CVM) as described by Muthén (1984). Failure to
treat categorical/ordinal data appropriately may also result in biased parameter estimates.
Finally, the study does not include any formal tests of mediation, so there are no estimates of
direct, indirect, or total effects available. Despite the methodological shortcomings, this study
illustrates the ways in which a mediating model can be used to examine the complex relationship
between parental beliefs, parental involvement, and children’s achievement over time.
It is broadly assumed that parents’ school readiness beliefs influence parenting practices,
and ultimately children’s academic achievement. Research, however, has not provided clear and
strong support for this pathway (Barbarin et al., 2008; Graue, 1992; Murphey, 1992; West et al.,
1995). Establishing a link between parents’ school readiness beliefs and transition practices may
help us understand additional factors contributing to the academic achievement gap and may
ultimately help us highlight potential targets for interventions. This study fills a gap in the
literature and builds upon previous research by using a nationally representative data set to

28

examine the relationship between parents’ conceptions of school readiness, transition practices,
and children’s academic achievement during the transition to elementary school. The study tests
the theory that parents’ conceptions of school readiness influence parents’ transition practices,
and therefore influence children’s academic achievement trajectories. The study also explores
the ways in which parents’ socioeconomic and cultural contexts shape academic socialization as
a possible explanation for the early academic achievement gap.
In this study, academic socialization is conceptualized as parents’ school readiness beliefs and
transition practices designed to prepare children for school. Parents’ school readiness beliefs are
operationalized as the importance parents place on a child mastering developmental skills, such
as knowing letters of the alphabet, counting, and communicating as well as the importance
placed on displaying social skills, such as sharing and being calm. Transition practices are
operationalized as targeted interactive parent-child learning experiences in the home which are
designed to support children’s successful transition to elementary school. Academic
achievement is measured in domains of reading, mathematics, and general knowledge during
kindergarten and first grade. The conceptual model tested in this study is illustrated in Figure 1.
The model suggests that parents’ school readiness beliefs influence parents’ transition practices
and children’s academic achievement. The model also suggests that parents’ transition practices
mediate the relationship between parents’ conceptions of school readiness and children’s
academic achievement. Finally, the model proposes that parents’ conceptions of school
readiness and transition practices are influenced by one’s race, ethnicity, and socioeconomic
status.
Purpose of Study

29

The purpose of the current study is to examine the relationship among parents’ school
readiness beliefs, transition practices, and children’s academic performance over time. The first
task is to describe variation within parents’ readiness beliefs, transition practices, and children’s
academic achievement by socioeconomic status and race/ethnicity. The first set of research
questions are as follows.
1. Do parents’ school readiness beliefs vary by socioeconomic status and
race/ethnicity?
2. Do parents’ transition practices vary by socioeconomic status and race/ethnicity?
3. Do children’s academic achievement trajectories vary by socioeconomic status
and race/ethnicity?
The second task is to examine the degree to which parents’ readiness beliefs and
transition practices influence children’s academic achievement trajectories. This was
accomplished by estimating the direct and indirect effects of parents’ readiness beliefs and
transition practices on children’s academic performance trajectories. The second set of research
questions are as follows.
4. To what extent do parents’ school readiness beliefs influence children’s academic
achievement upon kindergarten entry and over time?
5. To what extent do parents’ transition practices influence children’s academic
achievement upon kindergarten entry and over time?
6. To what extent do parents’ transition practices mediate the relation between
parents’ school readiness beliefs and children’s academic achievement upon
kindergarten entry and over time?

30

Several studies have examined differences in parents’ readiness beliefs by
sociodemographic characteristics, and findings are inconclusive (Barbarin et al., 2008; Diamond
et al., 2000; Graue, 1993b; West et al., 1995). I expected parents’ readiness beliefs to vary by
sociodemographic characteristics. It is hypothesized that socioeconomic status will be positively
related to parents’ school readiness beliefs. That is, more affluent parents will place greater
emphasis on children’s school readiness skills and attributes. It is also hypothesized that
European American parents will place greater emphasis on school readiness skills and attributes
in comparison to African American and Hispanic parents.
Consistent with previous research, it is hypothesized that parents’ transition practices will
vary by sociodemographic characteristics (Brooks-Gunn & Markman, 2005; Phillips, 2011). It is
hypothesized that socioeconomic status will be positively related to parents’ transition practices.
In other words, affluent parents will report engaging in more transition practices. It is also
hypothesized that European American parents will report engaging in greater amounts of
transition practices in comparison to African American and Hispanic parents.
Based upon previous findings, children’s beginning knowledge (Coley, 2002; Lee &
Burkham, 2002; Zill & West, 2001) and growth over time is expected to vary by
sociodemographic characteristics (Bodovski & Farkas, 2007; Fryer & Levitt, 2004; Reardon,
2011; Reardon & Gallindo, 2009; Reardon & Robinson, 2008). It is hypothesized that
socioeconomic status will exert a strong positive influence on children’s beginning achievement
and growth over time. Children from more affluent homes will begin school with higher
achievement scores and experience more rapid academic growth. It is also hypothesized that
European American children will have higher beginning achievement scores and greater rates of
growth in comparison to African American and Hispanic children.

31

Consistent with previous research examining the relationship between parents’ beliefs
and expectations on children’s academic performance (Barbarin et al., 2008; Hess et al., 1984),
parents’ readiness beliefs are expected to be positively related to children’s academic
achievement upon kindergarten entry and over time. It is hypothesized that readiness beliefs will
exert a positive influence on children’s beginning knowledge and rate of academic growth. It is
also hypothesized that parents’ school readiness beliefs will exert a stronger influence on
children’s beginning knowledge in comparison to the influence exerted on academic growth over
time. Based on the home environment and parent involvement literature it is hypothesized that
parents’ transition practices will be related to children’s achievement trajectories. It is
hypothesized that parents’ transition practices will be positively related to children’s
kindergarten achievement and rates of academic growth.
Given the shortage of empirical studies which use mediating models to investigate
parents’ conceptions of school readiness and their relative impact on transition practices and
children’s achievement, analyses investigating these relationships in this study are somewhat
exploratory and hypotheses are based on logical assumptions. It seems reasonable to argue that
the degree to which parents emphasize school readiness skills and attributes will influence their
use of transition practices, which in turn influence children’s early academic achievement. It is
hypothesized that parents’ transition practices will mediate the relationship between parents’
beliefs and children’s academic achievement in kindergarten and over time. The indirect effect
of parents’ readiness beliefs on children’s achievement will exert greater influence on children’s
beginning knowledge in comparison to the influence exerted on academic growth over time.

32

CHAPTER 3
METHODS
Data Source and Sample
The sample for this study consisted of a subset of children and parents who participated
in the Early Childhood Longitudinal Study-Kindergarten cohort (ECLS-K), which was
developed under the sponsorship of the U.S. Department of Education, National Center for
Education Statistics (National Center for Education Statistics, 2000). The design of the ECLS-K
was guided by an ecological framework of children’s development and schooling that
emphasizes the interrelationships between child, family, school, and community (NCES, 2000).
This multisource, multi-method study includes interviews with parents, teachers, school
administrations, and direct child assessments in the cognitive, psychosocial, and physical
domains. The ECLS-K data set allows researchers to analyze the complex relationships between
children’s early learning and school experiences and compare them to later elementary school
performance.
The ECLS-K study used a multistage probability sample design to select a nationally
representative sample of children attending kindergarten in 1998-1999. In the base year children
were selected for the ECLS-K study using a multistage probability design. During the first stage
of sampling, counties or groups of counties were selected with probability proportional to size.
The primary sampling unit (PSU) measure of size was based on the number of 5 year old
children in a county, with a slight modification to facilitate the oversampling of children from
Asian/Pacific Islander heritage. This procedure resulted in a total of 100 primary sampling units.
The second stage of sampling involved selecting private and public schools that offered
kindergarten programs. For each PSU, a frame of public and private schools offering
kindergarten programs was constructed using existing school universe files. The selection of
33

schools was systematic, with probability proportional to a weighted measure of size based on the
number of kindergartners enrolled in the school. As with the PSU sample, the measure of size
was constructed by taking into account the desired oversampling of Asian and Pacific Islander
children. Public and private schools constituted distinct sampling strata. In total, 1,280 schools
were sampled for the original frame. Of these, 953 were public schools and 460 were private
schools.
The third stage of sampling involved the selection of children within sampled schools.
From each school, an independent sampling stratum was formed within each school based on a
complete roster of enrolled children. One stratum contained Asian and Pacific Islander children,
and the second, all other students. Within each stratum, students were selected using equal
probability systematic sampling, using a higher rate for the stratum with Asian and Pacific
Islander children. In general, the target number of children sampled at any one school was 24.
Once the sampled child was identified, parent contact information was obtained from the school
in order to gain consent for the child assessment and parent interview.
The ECLS-K spring first grade data collection targeted all base year respondents. While
all students still enrolled in their base year schools were re-contacted, only 50% of children who
transferred from their kindergarten school were followed for data collection. The spring first
grade sample was freshened to include current first graders who had not been enrolled in
kindergarten during 1998-1999, and therefore, had no chance of being included in the base year
kindergarten sample. This freshening process used a half-open interval sampling procedure
which was implemented in the same 50% subsample of ECLS-K base year schools where
transfer students were followed for data collection. This process added 133 schools to the
second stage of sampling and 165 first grade students in the ECLS-K sample. This resulted in a

34

sample of 18,249 children. As a result of the complex sampling design, claims made about the
national kindergarten population of 1998 and first grade population of 1999 as a whole from this
data set require the use of a sample weight.
Criteria for Selecting Data File Sample
A subsample of children was selected for the present study. Since the focus of the study
is on the achievement gap between historically underachieving children and their more affluent
European American counterparts in kindergarten, only children of African American, Hispanic,
and European American racial and ethnic heritage were included. The present study made use of
cognitive assessments in reading, mathematics, and general knowledge in the fall of kindergarten
in 1998, the spring of kindergarten in 1999, and the spring of first grade in 2000, in addition to
parent interview data collected during the fall of kindergarten.
The analytic sample was selected based on several steps. First, children who were of
European American, African American, or Hispanic racial/ethnic heritage were selected. This
resulted in a sample size of 15,913. Second, children who were assigned a weight of zero in the
ECLS-K data set due to missing cognitive, psychosocial, and physical assessment data during
one of the data collection periods were excluded from the sample. This step resulted in a sample
of 13,130 children.
Instruments
Data for the ECLS-K were collected from a variety of sources. These included direct
assessments of students, parent/guardian interviews, and teacher and school administrator
surveys. This study uses data drawn from child assessments collected in the fall and spring of
kindergarten and spring of first grade, as well as parent interview data collected in the base year.
Although direct child assessments were collected at four points, from kindergarten through the

35

first grade, the fall of first grade assessments were limited to a 30% subsample and inclusion of
this data would limit the sample size to 30% of the original sample. Therefore, data from third
wave of data collection was excluded from this analysis. The instruments, data collection
processes, and measures are described below.
Direct child assessments. Direct one-on-one child assessments were conducted in the
cognitive, psychomotor, and physical domains. The ECLS-K cognitive battery consisted of
questions in three subject areas: language and literacy, mathematical thinking, and general
knowledge. Prior to administering the cognitive assessment battery in the fall of the base year, a
language screening assessment was administered to students identified by their school records or
teacher as coming from a home in which English was not the primary language. The Oral
Language Development Scale (OLDS) was given to those children who had a non-English
language background. The language screener determined if the children understood English well
enough to receive the direct child assessments in English. Children who passed the OLDS
screening received the full direct battery of assessments, while children who spoke Spanish were
administered the Spanish translation form of the mathematics and psychomotor assessments, and
an alternate form of the OLDS.
The direct cognitive assessments consisted of a set of two-stage assessments. The first
stage consisted of a routing section for each subject area, and was followed by several alternative
second-stage forms depending on the child’s performance on the first stage assessment. The
routing sections consist of 12 to 20 items with a broad range of difficulty. The second-stage
forms varied by level of difficulty so that a child would be administered questions appropriate to
his or her current level of ability for each cognitive domain. The reading and mathematics

36

assessments had low, medium, and high difficulty second-stage options, while the general
knowledge assessment had two second-stage alternatives.
The cognitive assessment scores were equated with Item Response Theory (IRT). IRT
made it possible to calculate scores that could be compared regardless of which second-stage
form a child took. This scaling method places children’s scores on a continuous ability scale
using items from the routing test, plus a core set of items shared among the different secondstage forms. This process estimates the score a child would have achieved if all of the items on
all forms of the test had been administered (NCES, 2000)
Parent/guardian interview. The majority of parents participating in the base year data
collection were interviewed in the fall of 1998 and again in the spring of 1999. Parents or
guardians were asked to provide important information about the sampled child, the home
environment, parent expectations and behaviors, and family characteristics by using a computerassisted telephone interview or a computer-assisted personal interview for families without a
telephone. Typically, the respondent for the parent interview was the mother of the child;
however, the respondent could have been a father, adoptive parent, foster parent, grandparent,
another relative, or a non-relative guardian living in the household with the child. Preference for
respondents was given to the child’s mother, then another parent or guardian, followed by
another household member.
The parent interview was conducted primarily in English, but provisions were made to
interview parents who spoke other languages. The questionnaire was translated into Spanish,
which was then printed on hardcopy. Bilingual interviewers were trained to conduct the parent
interview in either English or Spanish. If the interview was conducted in Spanish, the

37

interviewer used the hardcopy questionnaire and then entered respondents’ answers in the
computer-assisted interview program.
Measures
Child outcome indicators.
Reading. The language and literacy (reading) assessment measured children’s basic
literacy skills which include print familiarity, letter recognition, beginning and ending sounds of
a word, rhyming, word recognition, receptive vocabulary, and comprehension, which includes
listening comprehension and understanding words in context. Comprehension items were
targeted to measure skills in initial understanding, developing interpretation, personal reflection,
and demonstrating critical stance. Reliability estimates for the reading assessments were .93 in
the fall of kindergarten, .95 in the spring of kindergarten, and .97 for the spring of first grade
(NCES, 2000).
Mathematics. The mathematics assessment items were designed to measure conceptual
and procedural knowledge as well as problem solving skills. Approximately one half of the test
consisted of questions focusing on number sense, number properties, and operations, while the
remainder addressed understanding patterns and number relationships as well as formulating
conjectures and identifying solutions. Reliability estimates for the mathematics assessments
were .92 in the fall of kindergarten, .94 in the spring of kindergarten, and .94 for the spring of
first grade (NCES, 2000).
General knowledge. The general knowledge assessment measured understanding of
social studies and science content. The items captured information on children’s conceptions of
the social, physical, and natural world, as well as their ability to draw inferences and comprehend
implications. The social studies subdomain included questions which measured children’s

38

knowledge in a wide range of disciplines including history, government, culture, geography,
economics, and law. The science subdomain included questions from the fields of life, earth,
space, and physical science. Reliability estimates for the general knowledge assessments were
.88 in the fall of kindergarten, .89 in the spring of kindergarten, and .89 for the spring of first
grade (NCES, 2000).
Parental beliefs and practices.
Parents’ school readiness beliefs. Development of the parents’ school readiness beliefs
construct was guided by empirical research. According to previous research, parents of young
children hold certain beliefs about school readiness (Barbarin et al., 2008; Diamond et al., 2000;
Graue, 1993), as well as expectations for mastery of developmental milestones (Hess et al., 1984;
Hess et al., 1980). The latent construct representing parents’ school readiness beliefs was
comprised of seven measures that asked parents to rate the importance of a child exhibiting
specific academic skills and behavioral attributes before entering kindergarten. For example,
questions measuring the importance parents placed on academic skills focused on counting,
knowing the alphabet, and drawing before entering kindergarten (e.g., How important do you
think it is that a child knows the letters of the alphabet before going to kindergarten?). Questions
measuring the importance parents placed on behavioral attributes focused on being calm,
sharing, and communicating. Parents were asked to respond to each prompt using a 5-point
Likert-type scale, which was coded into five categories with scores in parentheses: (1) not
important, (2) not very important, (3) somewhat important, (4) very important, and (5) essential.
Cronbach’s alpha for the six school readiness beliefs items was .77.
Parents’ transition practices. The latent construct representing parents’ transition
practices was based on theory as well as empirical research. According to Vygotsky (1978) and

39

Wertsch (1979), children develop through interaction and dialogue with adults such as parents,
teachers, or older siblings. Through engagement in challenging activities and play with a more
knowledgeable other, children will perform activities successfully (Vygotsky, 1978; Wertsch,
1979). Empirical studies also indicate that engaging children with interactive learning
opportunities improves children’s cognitive outcomes (Senechal & Young, 2008). Interactive
learning opportunities or helping interactions (Scott-Jones, 1995) during early childhood are
conceptualized as transition practices. This construct was comprised of eight variables
measuring the frequency of targeted interactive parent-child learning activities in the home.
Parents responded to prompts that asked about the frequency in which they engaged their child in
a variety of interactive activities that included: reading, telling stories, singing songs, playing
games, playing sports, teaching about nature, making art, and building things together. The
coding scheme for these measures was divided into four categories with scores in parentheses:
(1) not at all, (2) once or twice a week, (3) 3 to 6 times a week, and (4) everyday. Cronbach’s
alpha for the eight transition practice items was .72.
Covariates.
Socioeconomic status. Family socioeconomic status (SES) was a composite variable
available in the ECLS-K data set. The components used to create the SES composite were: (a)
father/male guardian’s level of education; (b) mother/female guardian’s level of education; (c)
father/male guardian’s occupation; (d) mother/female guardian’s occupation; and (e) household
income. Occupation was re-coded to reflect the average of the 1989 General Social Science
Survey (GSS) prestige score. This was computed as the average of the corresponding prestige
score for the 1980 Census occupational categories covered by the ECLS-K occupation variable

40

(NCES, 2000). The resulting composite was the average of the five measures, each of which
was standardized to have a mean of 0 and a standard deviation of 1 (NCES, 2000).
Child’s race and ethnicity. The parent/guardian reported the child’s race during the
baseline parent interview. The categories for this variable included: European American, nonHispanic; African American, non-Hispanic; Hispanic, race specified; Hispanic, no race specified;
Asian; Native Hawaiian or other Pacific Islander; American Indian or Alaska Native and more
than one race specified, non-Hispanic. Categories for Hispanic, race specified and Hispanic, no
race specified were combined into one category, and American Indian or Alaska Native with
more than one race specified, non-Hispanic were combined into another category, resulting in
five categories: European American, African American, Hispanic, Asian American (Asian), and
other. A final race/ethnicity category was created with the following categories: European
American, African American, and Hispanic racial or ethnic heritage. Dummy variables were
created with European American as the reference category.
Data Analytic Strategy
The statistical analysis technique used to test hypotheses was latent growth curve
modeling (LGCM), which falls within the broad family of structural equation modeling (SEM).
Latent growth curve models allow the researcher to model change over time by estimating
individual growth trajectories as latent intercept and slope factors across the sample while
including predictors of change (Duncan, Duncan, & Stryker, 2010; Singer & Willet, 2003). This
technique is therefore aligned with the study’s purpose, which is to describe and predict change
in children’s academic achievement during the transition to elementary school. The LGCM for
this study used raw data analyzed in Mplus version 5 with a mean and covariance structure
(Muthén & Muthén, 1998-2007). Descriptive analyses were conducted using Stata 11

41

(StataCorp., 2009). Means and standard deviations were computed for continuous variables and
frequencies were computed for ordinal variables.
There are several aspects of SEM which made it a particularly advantageous statistical
technique for this analysis. First, SEM takes a confirmatory or hypothesis-testing approach to
the analysis of a structural theory bearing on some phenomenon (Byrne, 2011). Since the pattern
of relations was specified a priori, SEM lends itself well to the analysis of data for inferential
purposes. Second, SEM provides estimates of error variance parameters, which allows for the
assessment and correction of measurement error in contrast to traditional multivariate procedures
(Kline, 2011; Raykov & Marcoulides, 2006). Finally, SEM is also capable of testing theoretical
models known as structural regression models which test relationships between unobservable
constructs, termed latent variables (Kline, 2011; Raykov & Marcoulides, 2006). These
characteristics of SEM make this statistical technique aligned with the study’s purpose, which
aims to examine the relationship between several latent constructs (i.e., parents’ readiness beliefs
and transition practices, and children’s achievement trajectories).
The full structural equation model is comprised of two parts, the measurement model (or
factor analytic model) and the structural model (Kline, 2011; Raykov & Marcoulides, 2006).
The measurement model focuses solely on the extent that observed variables are linked to their
underlying latent factors by analyzing the covariation among the set of observed variables for a
latent construct (Kline, 2011; Raykov & Marcoulides, 2006). The measurement model uses a
confirmatory rather than an exploratory factor analysis approach, implying that latent constructs
are based on knowledge of a theory, empirical research, or both. The relations between observed
measures and underlying constructs that are represented by factor loadings (i.e., regression

42

paths) must be postulated a priori. The squared factor loading represents the percent of variance
in the observed variable explained by the latent variable.
The structural model allows for the specification of a regression structure among the
latent variables (Kline, 2011; Raykov & Marcoulides, 2006). In other words, one can
hypothesize the impact of one latent variable on another in the modeling of causal direction. The
measurement model depicts the link between the latent variables and their observed measures,
while the structural model depicts the predictive relationship among latent variables themselves.
The full structural model is capable of simultaneously assessing the magnitude and significance
of relations among exogenous (i.e., predictors) and endogenous (i.e., mediators and outcomes)
variables in the model which are represented by a series of structural (i.e., regression) equations.
Structural equation models are schematically portrayed using particular configurations of
four geometric symbols: a circle, square, single-headed arrows, and double-headed arrows
(Kline, 2011; Raykov & Marcoulides, 2006). By convention, circles or ellipses represent
unobserved latent variables or factors, and squares or rectangles represent observed variables
(Kline, 2011; Raykov & Marcoulides, 2006). Single-headed arrows represent the impact of one
variable on another, while a double-headed arrow represents covariance or correlation between
pairs of variables (Kline, 2011; Raykov & Marcoulides, 2006). The schematic representation of
a model under study is represented by a path diagram, which is equivalent to the mathematical
representation whereby a set of equations relates endogenous (dependent) variables to their
exogenous (explanatory) variables.
Estimation
In accordance with Kline (2011) and Raykov and Marcoullides (2006) the latent growth
curve model was tested in several steps in order to identify sources of potential problems. The

43

first step involved estimating the measurement model, which includes the unconditional growth
model, whereas the second stage involved estimating the full structural model. Estimation and
determination of the fit for both the measurement model and the full structural model involved
the same considerations; therefore, these are outlined first.
The ECLS-K data was collected using a stratified cluster sample design in order to obtain
a nationally representative sample of the population of kindergarten students in 1998-1999.
Without accounting for this sampling design, design effects may occur that decrease the accuracy
of estimates. To account for this design, unbiased estimates and standard errors can be
calculated in Mplus version 5 (Muthén & Muthén, 1998-2007) with sample design information,
which was provided in the ECLS-K data set via cluster, strata, and weight variables. Therefore,
all analyses, including descriptive and latent growth curve models, took into account the sample
design by employing the stratum, cluster, and child panel weights (necessary for longitudinal
study). Use of the sample weight allowed a generalization of the findings to African American,
Hispanic, and European American children who entered kindergarten in 1998-1999 and first
grade children in 1999-2000.
Another important consideration is the scale of observed variables in the model. The
current study includes ordinal and continuous variables; therefore, a continuous/categorical
variable methodology (CVM) as described by Muthén (1984) must be utilized. This approach
allows for the estimation of models with any combination of dichotomous, ordinal, or continuous
variables (Kline, 2011). In CVM, bivariate associations among observed variables are estimated
with polychloric correlations, which assume that a normal, continuous process underlies each
observed variable (Flora & Curran, 2004). The CVM approach described by Muthén and
Asparouhov (2002) is implemented in Mplus version 5 by specifying the mean and variance

44

adjusted weighted least square (WLSMV) estimator. The WLSMV is a robust estimator which
accounts for issues of non-normality and non-independence in the data (Kline, 2011).
Missing data is almost always an issue in longitudinal studies (Singer & Willet, 2003).
Failure to appropriately address missing data in longitudinal research can substantially reduce the
sample size and compromise the accuracy of model fit and parameter estimates (Newman, 2003).
Table 1 presents the proportion of missing data for all variables in the models. All models are
estimated with the WLSMV estimator which uses a combination of full information likelihood
method and pairwise present method to handle missing data.
Based on recommendations from Kline (2005) and Hancock and Mueller (2010), model
fit was evaluated using several indicators to assess the goodness-of-fit of hypothesized models:
Chi-square goodness-of-fit index, Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), and
Root Mean Squared Error of Approximation (RMSEA). A non-significant Chi-square value
indicates adequate model fit, however, this value can be very sensitive to large sample sizes
(Kline, 2011) and is viewed by most as overly strict given its power to detect even trivial
deviations of data from the proposed model (Mueller & Hancock, 2010). RMSEA values less
than .05 indicate good fit, values greater than .05 and less than .08 indicate reasonable fit, values
between .08 and .10 indicate mediocre fit, and values greater than .10 indicate poor fit (Browne
& Cudeck, 1993). CFI and TLI values greater than .95 signify a good fit (Hu & Bentler, 1999).
Measurement Model
The measurement models in Figures 2-4 represent the following latent constructs:
parents’ school readiness beliefs, parents’ transition practices, latent intercept, and latent slope.
In the figures, squares represent observed variables and ellipses represent latent constructs. The
single-headed arrows or paths represent the effect of one variable on another while the double-

45

headed arrows represent a correlation or covariation between variables. Although these are all
latent constructs there were important differences in terms of identification due to differences in
the measurement of observed variables as well as scaling procedures. The discussion of
identification begins with parents’ readiness beliefs and transition practices as they required
similar treatment.
Identification of latent variables requires that the researcher identify observed variables
and link these variables to the observed construct. Latent variables are unobserved and therefore
have no definite metric scale. Therefore, the researcher must fix one unstandardized factor
loading for the direct effect on an observed variable equal to 1.0; this is referred to as the
reference variable scaling method (Kline, 2011). The indicator with the unit loading
identification (ULI) is referred to as the reference variable. Specifying a reference variable
assigns a scale to the factor which is related to the explained variance of the reference variable
(Kline, 2011).
Figures 2-4 display the path diagrams, representing the measurement models under study.
Parents’ readiness beliefs, was comprised of six observed ordinal variables, while transition
practices was comprised of eight observed ordinal variables. The parents’ readiness beliefs
factor has a reference variable represented by the numeral 1 that appears next to the path
coefficient for the direct effect of the readiness belief factor on the importance of knowing the
alphabet indicator. The frequency of a parent reading to child is the reference variable for the
transition practices factor.
Modeling change or growth requires that the continuous dependent variable be measured
on at least three occasions; scores have the same units across time and can be said to measure the
same construct (achievement) at each assessment; and data must be time structured, that is, cases

46

are all tested at the same intervals. The ECLS-K direct child assessments in reading,
mathematics, and general knowledge met all of those requirements. Each observation or
measurement is represented as an indicator of two latent growth factors, the intercept (initial
status) and slope factor (linear change) (Kline, 2011; Singer & Willet, 2003). The intercept
factor represents the baseline level of achievement (kindergarten achievement) corrected for
measurement error. Because the intercept is analogous to the intercept in a regression equation,
the unstandardized loadings on this factor are all fixed to 1 (Kline, 2011; Singer & Willet, 2003).
The loadings on the slope factor are fixed to constants which correspond to times of
measurement or the approximate intervals at which data were collected across the two years
under study, excluding wave 3 which is a 30% subsample. This specification depicted in Figures
2-4 begin with 0 for the first measurement and end with 3, which is consistent with other
research analyzing latent growth curve models using academic achievement data from ECLS-K
(e.g., Sy & Shulenberg, 2005). The specification depicts a linear trend and can be interpreted as
change per year.
The means of the intercept and slope factors in this model are free parameters. The mean
of the intercept factor represents the average achievement adjusted for measurement error for the
whole sample of students in the fall of kindergarten 1998. The variance of the intercept factor
represents the amount of variation there was in achievement of individual children at wave 1.
The mean of the slope factor represents the average rate of change in achievement each year for
the entire sample of students. The variance of the slope factor indicates the amount of variation
for individual students’ growth each year. The change model allows the intercept and slope
factors to covary which is illustrated by the double arrow between the two factors. Finally, the

47

measurement model as a whole contains all four latent factors under study, and as such they are
correlated with each other and this is illustrated by double arrows between each factor.
Full Structural Model
The full structural model, also known as the structural regression model, included the
measurement model with specific tests of hypotheses about effect (Kline, 2011). Figures 5-7
represent mathematical equations depicting the full structural regression models under study.
The single-headed arrows from parents’ readiness beliefs to the intercept and slope represent the
direct effect of parental beliefs on children’s beginning kindergarten achievement (intercept) and
growth over time (slope). The single-headed arrows from parents’ transition practices to the
intercept and slope represent the direct effect of transition practices on beginning kindergarten
achievement and growth over time. The single-headed arrow from readiness beliefs to transition
practices represents the direct effect of parents’ readiness beliefs on transition practices. The
path from parents’ beliefs through transition practices to the intercept and slope allows the tests
for mediation. There were also tests for the direct effect of socioeconomic status and
race/ethnicity on parents’ readiness beliefs, transition practices, intercept, and slope (not shown
in Figures 5-7). These full structural models simultaneously assess the magnitude and
significance of relations among exogenous (i.e., parents’ readiness beliefs, race/ethnicity, and
socioeconomic status) and endogenous (i.e., transition practices, intercept, and slope) variables in
the model.
Summary of Analytic Approach
To describe and examine differences in parents’ school readiness beliefs and transition
practices while simultaneously assessing the degree to which academic socialization predicts
children’s academic achievement trajectories during the transition to elementary school, a latent

48

growth model was tested in two steps. In order to estimate structural model estimates, the
measurement model must first show reasonable model fit. Therefore, reading, mathematics, and
general knowledge measurement models will be estimated first to determine the adequacy of
model fit. Goodness-of-fit indices for each model will be examined as measures of model fit.
Each measurement model will be revised as necessary to produce reasonable model fit as
measured by these indices. This analysis will be followed by estimation of the full structural
model to determine the relationship between academic socialization and children’s academic
achievement trajectories. The full structural model will include race/ethnicity and
socioeconomic status variables as covariates, which enables the variation of parents’ readiness
beliefs, transition practices, and children’s achievement trajectories to be examined. All
statistical hypothesis testing will use α = 0.05 to determine statistical significance.

49

CHAPTER 4
RESULTS
This chapter begins with a description of the sample characteristics. Latent growth
curves were modeled in the content areas of reading, mathematics, and general knowledge. As
described in the previous section, analyses for this study were conducted in two stages.
Therefore, findings for each model will begin with a description of the measurement model and
be followed by a discussion of the structural model.
Descriptive Statistics
Demographic characteristics of the sample are displayed in Table 1. The analytic sample
consisted of 13,133 children. This sample consisted of 49% females, and had a racial/ethnic
composition of 62% European American, 17% African American, and 21% Hispanic.
Achievement scores for reading, mathematics, and general knowledge are also presented. The
correlation matrices for each model are presented in Tables 2-4.
Modeling Growth: Academic Socialization and Reading Achievement
Given that reasonable model fit is needed to obtain trustworthy parameter estimates, the
goodness-of-fit for the reading measurement model was assessed first (see Figure 2). An initial
Chi-square value of 742.593 (p < .001) was obtained and review of Lagrange Multipliers
(modification indices) indicated that modeling covariance between error terms for importance
alphabet and importance count variables would result in a significant decrease in the Chi-square
value. The next specification allowed error terms for the importance alphabet and importance
count variables to covary. This specification resulted in a Chi-Square value of 552.58 (p < .001)
which indicates that the hypothesized measurement model differs significantly from a perfectly
fitting model. However, the Chi-square value is affected by a large sample size (Kline, 2011).

50

More specifically, if the sample size is large, the value of the Chi-square may lead to rejection of
the model even though differences between observed and predicted covariances are slight (Kline,
2011). Therefore, additional fit statistics must be considered and are summarized in Table 5. An
RMSEA of .036 was obtained, which suggests good model fit, according to the
recommendations of Brown and Cudeck (1993). A CFI of .96 and TLI of .97 were obtained
which also suggest good model fit (Hu & Bentler, 1999). Given the large sample size of this
model (N = 13,108), and because the RMSEA, CFI, and TLI values suggest good model fit, this
measurement model was retained.
The intercept value for the reading measurement model was 26.35 (SE = .293), p < .001.
This indicates that the fall kindergarten mean reading score for the entire population was 26.35.
The slope value for the measurement model was 13.27 (SE = .104), p < .001. This value
represents the mean rate of growth for the entire population, which would be interpreted as an
increase of 13.27 points at each data collection period. The variance for the intercept was 83.96
(SE = 1.47), p < .001. The statistically significant variance estimate represents the amount of
variation in the intercept factor. The variance for the slope was 18.30 (SE = .770), p < .001.
This value also represents the amount of unexplained variance for the slope factor. Therefore,
the next logical step would be to include exogenous (predictor) variables into the model to
further explain the amount of variance in the intercept and slope terms.
The next step in the analysis involved estimating the full structural model to determine
the relationship between academic socialization and children’s academic performance
trajectories in reading (see Figure 5). The full structural model was estimated and Table 5
presents a summary of fit statistics. The obtained Chi-square value was 578.73 (p < .001).
Examination of alternative fit statistics indicates the model is a good fit for the data with an

51

RMSEA of .034, CFI of .96, and a TLI of .96. Table 6 presents the unstandardized and
standardized factor loadings for the latent factors in the model. Factor loadings were all
statistically significant at the p < .001 level. The standardized factor loadings for parents’
readiness beliefs indicators were between .60 and .70, while factor loadings for transition
practices fell between .46 and .60. The R² values represent the amount of explained variance for
that indicator variable by the factor. The full structural model was retained and parameter
estimates were evaluated. Results will be discussed as they pertain to the sets of research
questions addressed by this study.
Research questions 1-3: variation within beliefs, behaviors, and reading achievement
The first task was to describe the ways in which parents’ readiness beliefs, transition
practices, and reading achievement varied by socioeconomic status and race/ethnicity. Table 7
presents a summary of unstandardized and standardized parameter estimates. In this table,
columns represent exogenous factors or covariates (explanatory variables) and rows represent
endogenous factors (dependent variables).
It was hypothesized that parents’ school readiness beliefs would vary by socioeconomic
status and race/ethnicity. Examination of parameter estimates in Table 7, row 3, column 5
indicates that there was no statistically significant relationship between socioeconomic status and
parents’ school readiness beliefs. Holding socioeconomic status constant, there was a significant
relationship between race/ethnicity and parents’ school readiness beliefs, but only for African
American parents. On average, African American parents placed greater importance on
children’s school readiness skills and attributes in comparison to European American parents. In
general, Hispanic parents’ school readiness beliefs were not statistically different from European
American parents. The R² value for the latent factor representing parents’ readiness beliefs was
52

.003, indicating that only .3% of the variation in parents’ readiness beliefs was explained by
socioeconomic status and race/ethnicity.
It was hypothesized that parents' transition practices would vary by socioeconomic status
and race/ethnicity. Examination of parameter estimates in Table 7, row 4, column 5 indicate that
there was a positive, statistically significant relationship between socioeconomic status and
parents’ transition practices, holding all else constant. In general, affluent parents reported
engaging young children in more transition practices. Holding socioeconomic status constant,
there was a statistically significant relationship between race/ethnicity and transition practices,
but only for Hispanic parents. On average, Hispanic parents reported engaging in fewer
transition practices with their child in comparison to European American parents. There was no
statistically significant difference between the amount of transition practices reported by African
American and European American parents. The R² value for the latent factor representing
parents’ transition practices was .08, showing that 8% of the variation in transition practices was
explained by socioeconomic status and race/ethnicity.
The next step was to describe children’s reading achievement and examine the ways in
which performance varied by socioeconomic status and race/ethnicity. The mean of the intercept
factor was 27.08 (SE = .265), p < .001, indicating that the sample mean score for European
American children from a home with an average socioeconomic status (SES = 0) had a mean
reading score of 27.08 in the fall of kindergarten. The slope factor was 13.78 (SE = .113), p <
.001. Thus, the average increase in reading for a European American child from an average
socioeconomic status household was 13.08 at each data collection point (i.e., spring of
kindergarten, spring of first grade). The residual covariance between the slope and intercept
terms was 8.88, p < .001. This indicates that children who have higher intercepts also have
53

higher slopes and children with lower intercepts have lower slopes. In other words, a child who
begins kindergarten with a high initial level of reading will have a more rapid rate of growth.
Alternatively, a child who begins kindergarten with a low level of reading will experience a less
rapid rate of growth.
It was hypothesized that children from affluent homes begin kindergarten with higher
reading scores. It was also hypothesized that European American children begin kindergarten
with higher reading scores in comparison to African American and Hispanic children. Parameter
estimates of interest are presented in Table 7, row 1, columns 5-7. There was a significant
positive relationship between socioeconomic status and the intercept factor, holding race and
ethnicity constant. In general, children from more affluent homes begin kindergarten with higher
scores on the reading assessment. Interpretation of the unstandardized estimate would indicate
that a one-unit change in socioeconomic status would result in an increase of 4.13 points on the
fall kindergarten reading assessment. There was also a significant negative relationship between
race/ethnicity and the intercept factor, conditional on socioeconomic status. On average, African
American and Hispanic children began kindergarten with lower reading assessment scores in
comparison to European American children. Interpretation of these estimates would indicate that
on average, African American and Hispanic children scored 1.75 and 2.17 points lower,
respectively, on the fall kindergarten reading assessment in comparison to European American
children from similar socioeconomic status backgrounds.
It was hypothesized that children from affluent homes would experience more rapid rates
of growth in reading. It was also hypothesized that African American and Hispanic children
would experience lower rates of growth in reading in comparison to European American
children. Parameter estimates of interest are presented in Table 7, row 2, columns 5-7. Holding

54

race and ethnicity constant, there was a significant positive association between socioeconomic
status and the slope factor. This indicates that children from more affluent homes experienced
more rapid rates of growth in reading. Holding socioeconomic status constant, there was a
significant, negative association between race/ethnicity and the slope factor. In general, African
American and Hispanic children had less rapid rates of growth in comparison to European
American children. Interpretation of the path coefficients would indicate that on average,
African American and Hispanic children from average socioeconomic status homes had a growth
rate of 12.38 (13.78-1.47) and 13.18 (13.78-.60) points respectively, in comparison to European
American children from an average socioeconomic status home who had a 13.78 point growth
rate.
Based on fitted trajectories Figure 8 provides a visual representation of the estimated
differences in children’s reading achievement trajectories by race and ethnicity, conditional on
socioeconomic status. Examination of the figure depicts the rate of reading growth for European
American, African American, and Hispanic children as they transition to elementary school. On
average, European American children begin kindergarten with higher reading scores and
experience faster rates of growth in comparison to African American and Hispanic children. The
figure also illustrates that Hispanic children begin kindergarten with slightly lower average
reading scores in comparison to African American children, yet Hispanic children’s average rate
of growth is large enough to close this gap by the end of first grade.
Research questions 4-6: academic socialization and reading achievement
The next task was to examine the direct effects of parents’ readiness beliefs on children’s
academic performance trajectories. It was hypothesized that parents’ readiness beliefs would
exert a positive influence on children’s academic performance in kindergarten and rate of

55

growth. However, this effect would decrease over time; in other words, parents’ readiness
beliefs would exert a greater influence on children’s initial kindergarten reading performance in
comparison to the effect exerted on growth over time.
Parents’ readiness beliefs had a significant positive direct effect on children’s beginning
reading achievement, conditional on socioeconomic status and race/ethnicity. Holding
socioeconomic status and race/ethnicity constant, there was a significant positive relationship
between parents’ readiness beliefs and the slope factor. Interpretation of the unstandardized
estimates presented in Table 7, column 3 suggests that a one unit change in parents’ readiness
beliefs would result in a 1.87 point increase in beginning reading achievement and a .48 increase
in growth.
Table 7 also presents the standardized coefficients for the structural model of academic
socialization and children’s reading achievement, which enables one to compare the intensity of
effects of the different exogenous variables on the endogenous variables within the same model.
The standardized regression coefficient describes how many standard deviations of change in the
dependent variable are associated with a change of 1 SD in the independent variable (Hedges,
2008). Interpretation of the path coefficient would indicate that a 1 SD increase in the
importance parents place on school readiness would result in a .13 SD increase in the fall of
kindergarten reading assessment score. There was also a positive significant association between
parents’ readiness beliefs and the slope factor, β = .07, comparison of the standardized
coefficients indicates that the influence of parents’ readiness beliefs on children’s reading
performance declines over time.
As was hypothesized for parental beliefs, transition practices would have a positive effect
on children’s academic performance in kindergarten and over time. Holding socioeconomic

56

status and race/ethnicity constant, parents’ transition practices exerted a significant positive
direct effect on children’s beginning reading achievement and no effect on growth over time.
Interpretation of the unstandardized estimate indicates that a one unit change in transition
practices would result in a .49 point increase in beginning reading achievement. Examination of
the standardized estimates in Table 7 indicates that parents’ readiness beliefs are a stronger
predictor of children’s beginning reading achievement in comparison to transition practices. The
overall model explained 17% of the variation in children’s beginning reading achievement and
11% of variation in growth.
The final task was to examine whether transition practices mediated the relation between
parents’ readiness beliefs and children’s academic performance in kindergarten and over time. It
was hypothesized that parents who placed more importance on school readiness would report
engaging in more transition practices, which in turn influence children’s reading achievement.
Parameter estimates in Table 7 indicate there was a significant positive relationship between
parents’ readiness beliefs and transition practices. Formal tests of mediation were conducted and
confirmed this to be the case. Estimates are presented in the top row of Table 12, which include
unstandardized path coefficients, standard errors, and estimates of the total, direct, and indirect
effects. Indirect effects of parent beliefs on children’s beginning reading achievement, although
small, were nonetheless significant, but the significant direct effect from parents’ beliefs to the
intercept suggests only partial mediation. The unstandardized total effect estimate indicates that
a one unit increase in the importance parents place on school readiness would result in an
approximate two point gain in beginning reading achievement and a half of a point increase in
growth over time. The standardized estimate for the total effect of parents’ beliefs and transition
practices on children’s beginning reading achievement was β = .14 and β = .07 for growth over

57

time which indicates the total effect is stronger on beginning reading achievement and decreases
as children transition to elementary school.
Modeling Academic Socialization and Mathematics Achievement
The goodness-of-fit for mathematics was modeled first (see Figure 3). The initial
mathematics measurement model had a Chi-square value of 720.017 (p < .001). Review of
Lagrange Multipliers indicated that modeling covariance of error terms for importance alphabet
and importance count would result in a significant decrease in the Chi-square value. In the
second mathematics measurement model, error terms for the importance alphabet and
importance count variables were allowed to covary. This specification resulted in a mathematics
measurement model with a Chi-Square value of 517.50 (p < .001). Other fit statistics were
considered and are summarized in Table 5. An RMSEA of .035 was obtained, which suggests
good fit, according to the recommendations of Brown and Cudeck (1993). A CFI of .96 and TLI
of .96 were obtained which also suggest good model fit (Hu & Bentler, 1999). Given that these
fit statistics suggest good model fit, this mathematics measurement model was retained.
The intercept value for the measurement model was 20.68 (SE = .293), p < .001. This
indicates that the mean mathematics score for the entire population was 20.68. The slope value
for the measurement model was 10.97 (SE = .08), p < .001. This value represents the mean rate
of growth for the entire population, which would be interpreted as an increase of 10.97 points at
each data collection period. The variance for the intercept and slope were 73.65 (SE = 1.24) and
11.822 (SE = .391), p < .001, respectively. These statistically significant variance estimates
represents the amount of variation in the intercept and slope factors. Therefore, the next logical
step would be to include exogenous variables into the model to explain the variance in the
intercept and slope terms.

58

The next step in the analysis involved estimating the full structural model to determine
the relationship between academic socialization and children’s academic achievement
trajectories in mathematics (see Figure 6). The full structural model was estimated and Table 5
presents a summary of fit statistics. The obtained Chi-square value was 503.08 (p < .001).
Examination of alternative fit statistics indicates the model was a good fit for the data with an
RMSEA of .031, a CFI of .96, and a TLI of .96. Factor loadings for the full structural
mathematics model were all statistically significant at the p < .001 level and are presented in
Table 8. The standardized factor loadings for parents’ readiness beliefs indicators were between
.60 and .70, while the standardized factor loadings for transition practices fell between .46 and
.60. The full structural model was retained and parameter estimates were evaluated. Results will
be discussed as they pertain to the sets of research questions addressed by this study.
Research questions 1-3: variation within beliefs, behaviors, and mathematics achievement
The first task was to describe the ways in which parents’ readiness beliefs, transition
practices, and children’s academic achievement varied by socioeconomic status and
race/ethnicity. Parameter estimates regarding variation of parents’ beliefs and practices followed
a similar pattern to those discussed in the structural model of reading achievement. Table 9
presents a summary of unstandardized and standardized parameter estimates.
It was hypothesized that parents’ school readiness beliefs would vary by socioeconomic
status and race/ethnicity. There was no statistically significant relationship between
socioeconomic status and parents’ school readiness beliefs. Holding socioeconomic status
constant, there was a significant relationship between race/ethnicity and parents’ school
readiness beliefs, but only for African American parents. In general, African American parents
placed a little more importance on school readiness in comparison to European American

59

parents. In general, Hispanic parents’ school readiness beliefs were not statistically different
from European American parents. The R² value for the latent factor representing parents’
readiness beliefs was .003, indicating that only .3% of the variation was explained by
socioeconomic status and race/ethnicity.
It was also hypothesized that parent’s transition practices would vary by socioeconomic
status and race/ethnicity. Again, there was a positive, statistically significant relationship
between socioeconomic status and parents’ transition practices, holding all else constant. In
general, affluent parents reported engaging young children in more transition practices. Holding
socioeconomic status constant, there was a statistically significant relationship between
race/ethnicity and transition practices, but only for Hispanic parents. On average, Hispanic
parents reported engaging in fewer transition practices with their child in comparison to
European American parents. There was no statistically significant difference between the
amount of transition practices reported by African American and European American parents
from similar socioeconomic status backgrounds. The R² value for the latent factor representing
parents’ transition practices was .08, which indicates that only 8% of the variation was explained
by socioeconomic status and race/ethnicity.
The next step was to describe children’s mathematics achievement and examine the ways
in which performance varied by socioeconomic status and race/ethnicity. The mean of the
mathematics intercept factor was 22.31 (SE = .195), p < .001. Interpretation of the intercept
value would indicate that European American children from an average socioeconomic status
home had a score of 22.31 on the mathematics assessment in the fall of kindergarten. The slope
factor was 11.41 (SE = .079), p < .001. Thus, the average increase in mathematics achievement
for European American children from an average socioeconomic status home was 11.41 points at
60

each data collection point. The residual covariance between the intercept and slope terms was
4.85, p < .001. Interpretation of this estimate indicates that a child who begins kindergarten with
a high initial level of mathematics achievement will experience a more rapid rate of growth.
Conversely, a child who begins kindergarten with a low level of mathematics achievement will
experience a less rapid rate of growth.
It was hypothesized that children from affluent homes would begin kindergarten with
higher mathematics scores. It was also hypothesized that European American children begin
kindergarten with higher mathematics scores in comparison to African American and Hispanic
children. Parameter estimates of interest are presented in Table 9. There was a significant
positive relationship between socioeconomic status and the intercept, holding race and ethnicity
constant. In general, children from more affluent homes begin kindergarten with higher scores
on the mathematics assessment. Interpretation of the unstandardized estimate would indicate
that a one-unit change in socioeconomic status would result in an increase of 3.93 points on the
fall kindergarten mathematics assessment. There was also a significant negative relationship
between race/ethnicity and the intercept factor, holding socioeconomic status constant. On
average, African American and Hispanic children began kindergarten with lower mathematics
scores in comparison to European American children. Interpretation of these estimates would
indicate that on average, African American and Hispanic children from average socioeconomic
status households scored 3.38 and 2.17 points lower, respectively, on the fall kindergarten
mathematics assessment in comparison to European American children from average
socioeconomic status homes.
It was also hypothesized that mathematics growth would be greater for children from
more affluent homes. In addition, African American and Hispanic children would experience

61

less rapid rates of growth in mathematics in comparison to European American children.
Examination of parameter estimates indicates a significant positive association between
socioeconomic status and the slope factor. In general, children from more affluent homes
experience greater rates of mathematics growth. A significant negative association occurred
between race/ethnicity and the slope factor. On average, African American and Hispanic
children have lower rates of growth in mathematics in comparison to European American
children from similar socioeconomic status households. Interpretation of the path coefficient
would indicate that on average, African American children and Hispanic children from average
socioeconomic status homes have a rate of growth of 9.73 and 10.91 points respectively, in
comparison to European American children from similar socioeconomic status homes who have
a growth rate of 11.41 points.
Figure 9 provides a visual representation of the estimated differences in children’s
mathematics achievement trajectories by race and ethnicity, conditional on socioeconomic status.
On average, European American children begin kindergarten with higher mathematics scores and
experience faster rates of growth in comparison to African American and Hispanic children.
Based on fitted trajectories the figure also illustrates that Hispanic children begin kindergarten
with lower average mathematics scores in comparison to African American children; yet,
Hispanic children’s average rate of growth is so rapid that Hispanic children have higher average
mathematics scores in comparison to African American children by the end of the first grade.
Research questions 4-6: academic socialization and mathematics achievement
The next task was to examine the direct effect of parents’ readiness beliefs on children’s
mathematics achievement trajectories. It was hypothesized that parents’ readiness beliefs would
exert a positive influence on children’s mathematics performance in kindergarten and growth.

62

This effect, however, would decrease over time. Holding socioeconomic status and
race/ethnicity constant, there was a significant positive relationship between parents’ readiness
beliefs and the intercept factor. Table 9 presents the unstandardized and standardized
coefficients for the structural model of academic socialization and children’s mathematics
achievement. Parents’ readiness beliefs exert a statistically significant positive effect on
children’s beginning mathematics achievement. Interpretation of the path coefficient would
indicate that a 1 SD increase in the importance parents place on school readiness beliefs would
result in a .10 SD increase in mathematics achievement upon kindergarten entry. There was also
a positive significant association between parents’ readiness beliefs and the slope factor, β =.06,
which indicates that the positive effect exerted by parents’ readiness beliefs on children’s reading
performance declines as children transition to elementary school.
It was also hypothesized that parents’ transition practices would have a positive effect on
children’s academic performance in kindergarten and over time. However, the effect of parents’
transition practices would decrease as children transitioned into elementary school. Holding
socioeconomic status and race/ethnicity constant, there was a small significant positive
relationship between parents’ transition practices and the intercept. There was no statistically
significant relationship between parents’ transition practices and the slope. This indicates that
parents’ transition practices have a small positive effect on children’s mathematics achievement
at the beginning of kindergarten but no influence over time. Examination of the standardized
estimates would indicate that parents’ beliefs exerted a greater influence on children’s beginning
mathematics achievement and growth over time. The overall model explained 24% of the
variation in children’s beginning reading and 10% of variation in growth.

63

The final task was to examine if transition practices mediated the relation between
parents’ readiness beliefs and children’s mathematics achievement in kindergarten and over time.
It was hypothesized that parents who placed more importance on school readiness skills and
attributes would report engaging young children in more transition practices which would
ultimately influence children’s mathematics achievement. Parameter estimates in Table 9
indicate there was a significant positive relationship between parents’ readiness beliefs and
transition practices. Formal tests of mediation were conducted and confirmed that transition
practices mediated the relation between parents’ beliefs and children’s mathematics achievement
upon kindergarten entry. Estimates are presented in Table 12. Indirect effects of parent beliefs
on children’s beginning mathematics performance although small were nonetheless significant,
but the significant direct effect of parents’ beliefs on the intercept suggests only partial
mediation. The unstandardized total effect estimate indicates that a one unit increase in the
importance parents place on school readiness would result in an approximate 1.4 point gain in
beginning mathematics achievement and a quarter of a point increase in growth through the first
grade. The standardized total effects of parents’ beliefs on children’s mathematics achievement
upon kindergarten entry was β=.10 and β=.05 for children’s growth in mathematics indicating
that the effect of parents’ beliefs and transition practices decreases over time.
Modeling Academic Socialization and General Knowledge Achievement
The general knowledge measurement model was estimated (see Figure 4) and followed
an estimation pattern similar to the reading and mathematics measurement models. The initial
general knowledge measurement model had a Chi-square value of 842.43 (p < .001). Review of
Lagrange Multipliers indicated that modeling covariance of error terms for importance alphabet
and importance count would result in a significant decrease in the Chi-square value. In the

64

subsequent general knowledge measurement model, error terms for the importance alphabet and
importance count variables were allowed to covary. This specification resulted in a general
knowledge measurement model with a Chi-Square value of 682.77 (p < .001). Again additional
fit statistics were considered and are summarized in Table 5. This measurement model obtained
an RMSEA of .044, a CFI of .92, and TLI of .93. Based on the RMSEA, CFI, and TLI values
that suggest an adequate fit, the measurement model was retained.
The intercept value for the measurement model was 22.68 (SE = .185), p < .001. This
indicates that the mean general knowledge score for the entire population was 22.68. The slope
value for the measurement model was 3.96 (SE = .035), p < .001. This value represents the mean
rate of growth for the entire population, which would be interpreted as an increase of 3.96 points
at each data collection period. The variance for the intercept was 54.41 (SE = 1.79), p < .001.
The statistically significant variance estimate represents the amount of variation in the intercept
factor. The variance for the slope was 1.89 (SE = .243), p < .001. The small variance estimate
indicates that much of the variance in the slope is already explained by the measurement model.
The next logical step would be to include exogenous variables into the model to explain the
remaining variance of the intercept and slope terms.
The final step in the analysis involved estimating the full structural model to determine
the relationship between academic socialization and children’s academic performance
trajectories in general knowledge (see Figure 7). The full structural model was estimated and a
Chi-square value was 506.04 (p < .001) was obtained. Examination of alternative fit statistics
indicates the model was a good fit for the data with an RMSEA of .034, a CFI of .95, and a TLI
of .95. The unstandardized and standardized factor loadings of model indicators for the full
structural model are presented in Table 10. Factor loadings were all statistically significant at the

65

p < .001 level. The standardized factor loadings for parents’ readiness beliefs indicators were
between .60 and .70, while the factor loadings for transition practices fell between .46 and .60.
The full structural model was retained and parameter estimates were evaluated as they pertain to
the sets of research questions addressed by this study.
Questions 1-3: variation within beliefs, behaviors, and general knowledge achievement
The first analysis began by describing the ways in which parents’ readiness beliefs,
transition practices, and children’s general knowledge achievement varied by socioeconomic
status and race/ethnicity. Parameter estimates for parent beliefs and behaviors followed a similar
pattern to those discussed in the structural models of reading and mathematics achievement.
Table 11 presents a summary of unstandardized and standardized parameter estimates.
It was hypothesized that parents’ school readiness beliefs would vary by socioeconomic
status and race/ethnicity. There was no statistically significant relationship between
socioeconomic status and parents’ school readiness beliefs. On average, African American
parents placed more importance on school readiness in comparison to European American
parents. Hispanic parents’ school readiness beliefs were not statistically different from European
American parents. The R² value for the latent factor representing parents’ readiness beliefs was
.003, indicating that only .3% of the variation was explained by socioeconomic status and
race/ethnicity.
It was also hypothesized that parents' transition practices would vary by socioeconomic
status and race/ethnicity. Parameter estimates indicate a statistically significant relationship
between socioeconomic status and parents’ transition practices, holding all else constant. In
general, more affluent parents reported engaging young children in more transition practices.
Holding socioeconomic status constant, there was a statistically significant relationship between
66

race/ethnicity and transition practices, but only for Hispanic parents. On average, Hispanic
parents reported engaging in fewer transition practices with their children in comparison to
European American parents from similar socioeconomic status backgrounds. There was no
statistically significant difference between the amount of transition practices reported by African
American and European American parents. The R² value for the latent factor representing
parents’ transition practices was .08, indicating that only 8% of the variation was explained by
socioeconomic status and race/ethnicity.
The next task was to describe children’s general knowledge achievement trajectories and
examine the ways in which performance varied by socioeconomic status and race/ethnicity. The
mean of the general knowledge intercept factor was 24.20 (SE = .134), p < .001. Interpretation
of this parameter estimate indicates that a European American student from an average
socioeconomic status home has a mean score of 24.2 on the fall kindergarten general knowledge
achievement assessment. The slope factor was 4.12 (SE = .042), p < .001. Thus, the average
increase in general knowledge achievement was 4.12 points at each data collection point for a
European American child from an average socioeconomic status household. The residual
covariance between the general knowledge intercept and slope was -2.66, p < .001.
Interpretation of this estimate indicates that a child who begins kindergarten with a high initial
level of general knowledge will have less rapid growth. Alternatively, a child who begins
kindergarten with a lower level of general knowledge will experience a more rapid rate.
It was hypothesized that children from affluent homes would begin kindergarten with
higher general knowledge scores. It was also hypothesized that European American children
begin kindergarten with higher general knowledge scores in comparison to African American
and Hispanic children. Parameter estimates of interest are presented in Table 11. There was a
67

significant positive association between socioeconomic status and intercept factor. In general,
children from affluent homes began kindergarten with higher general knowledge scores.
Interpretation of the unstandardized estimate would indicate that a one-unit change in
socioeconomic status would result in an increase of 3.28 points on the fall kindergarten general
knowledge assessment. There was also a significant negative relationship between race/ethnicity
and the intercept factor, holding socioeconomic status constant. On average, African American
and Hispanic children began kindergarten with lower general knowledge scores in comparison to
European American children. Interpretation of these estimates would indicate that on average,
African American children and Hispanic children from average socioeconomic status homes
scored 3.38 and 2.17 points lower, respectively, on the fall kindergarten general knowledge
assessment in comparison to European American children from average socioeconomic status
households.
It was also hypothesized that general knowledge growth would be greater for children
from more affluent homes. In addition, African American and Hispanic children would
experience lower rates of growth in general knowledge in comparison to European American
children. Socioeconomic status had no statistically significant relationship with the general
knowledge slope factor. There was a significant negative association between race/ethnicity and
the slope factor, but only for Hispanic children. Interpretation of the path coefficient would
indicate that on average, Hispanic children from an average socioeconomic status household
would have a rate of growth of 3.79 points in comparison to European American children from
similar socioeconomic backgrounds who had an average growth rate of 4.13 points at each data
collection point.

68

Based on fitted trajectories Figure 10 depicts the rate of general knowledge growth for
European American, African American, and Hispanic children, irrespective of socioeconomic
status, as they transition to elementary school. On average, European American children begin
kindergarten with higher general knowledge scores than African American children from similar
socioeconomic status backgrounds but experience similar rates of growth. The figure also
illustrates that Hispanic children begin kindergarten with higher average general knowledge
scores in comparison to African American children. Despite having a slightly more rapid rate of
growth than Hispanic children, African American children still have lower average general
knowledge scores by the end of first grade in comparison to Hispanic children.
Research questions 4-6: academic socialization and general knowledge achievement
The next task was to examine the direct effects of parents’ readiness beliefs on children’s
general knowledge achievement trajectories. It was hypothesized that parents’ readiness beliefs
would exert a positive influence on children’s general knowledge performance in kindergarten
and growth rate. However, this effect would decrease over time. Holding socioeconomic status
and race/ethnicity constant, there was a small significant positive relationship between parents’
readiness beliefs and the intercept factor. Parents’ readiness beliefs had no statistically
significant influence on the slope factor. This indicates that parents’ readiness beliefs had a
small positive effect on children’s general knowledge achievement at the beginning of
kindergarten but no influence over time. Interpretation of the unstandardized coefficients in
Table 11 suggests that a one unit increase in the importance parents place on school readiness
would result in a .40 point increase in children’s general knowledge assessment score in
kindergarten. Table 11 presents the standardized coefficients for the structural model of
academic socialization and children’s general knowledge achievement. Interpretation of the path

69

coefficient would indicate that a 1 SD increase in the importance parents place on school
readiness would result in a .04 SD increase in general knowledge achievement upon kindergarten
entry.
It was also hypothesized that parents’ transition practices would have a positive effect on
children’s academic general knowledge performance in kindergarten and over time. However,
the effect of parents’ transition practices would decrease as children transitioned into elementary
school. Holding socioeconomic status and race/ethnicity constant, there was a small significant
positive relationship between parents’ transition practices and the intercept but not the slope.
This indicates that parents’ transition practices had a small positive effect on children’s general
knowledge achievement at the beginning of kindergarten but no influence over time.
Interpretation of the unstandardized estimate in Table 11 suggests that a one unit increase in
transition practices would result in a 1.02 point increase on the general knowledge assessment
upon kindergarten entry. Interpretation of the standardized path coefficient in Table 14 indicates
that a 1 SD increase of parents’ transition practices would result in a .07 SD increase in general
knowledge achievement upon kindergarten entry. In contrast to the estimates from the reading
and mathematics models, transition practices exerted a stronger influence on children’s
beginning general knowledge achievement in comparison to the influence exerted by parental
beliefs. The overall model explained 32% of the variation in children’s beginning general
knowledge achievement and 1% of variation in growth.
The final task was to examine whether transition practices mediated the relation between
parents’ readiness beliefs and children’s academic performance in kindergarten and over time.
Again, parents’ readiness beliefs had a positive influence on transition practices. Formal tests of
mediation were conducted and confirmed that transition practices mediated the relation between

70

parents’ beliefs and children’s mathematics achievement upon kindergarten entry. Estimates are
presented in Table 12. Indirect effects of parent beliefs on children’s beginning general
knowledge performance, although minimal, were nonetheless significant but the direct effect of
parents’ beliefs on the intercept suggests only partial mediation. The unstandardized total effect
estimate indicates that a one unit increase in the importance parents place on school readiness
would result in a little more than a half of a point increase in beginning general knowledge
achievement. The standardized total effects of parents’ beliefs on children’s general knowledge
achievement upon kindergarten entry was β = .05, indicating that a 1 SD increase in parents’
readiness beliefs would result in a half of a standard deviation increase in children’s beginning
general knowledge achievement.

71

CHAPTER 5
DISCUSSION
Considerable attention has been focused on the persistent racial, ethnic, and
socioeconomic disparities in academic achievement during elementary school. Although
researchers associate socioeconomic status and parental level of education with school readiness
and subsequent academic success, parents play a pivotal role in supporting children’s early
learning and development through a process of socialization (Bempechat, 1992; Ginsberg et al.,
1992; Maccoby, 1992). The theory of academic socialization suggests that parents’ attitudes,
values, and beliefs about school influence parenting behaviors and in turn influence children’s
transitions to elementary school, all of which is shaped by one’s socioeconomic and cultural
contexts (Taylor et al., 2004). Although researchers have assumed parents’ school readiness
beliefs influence parenting practices, and ultimately children’s early academic achievement
(Barbarin et al., 2008; West et al., 1995), there are few studies which use mediating models to
support this claim (see Sy and Schulenberg, 2005 for an exception). If this relationship exists, it
seems plausible that differences in parents’ conceptions of school readiness beliefs and transition
practices contribute to the early academic achievement gap.
The principal goals of this study was to examine variation within parents’ readiness,
beliefs, transition practices, and children’s achievement trajectories while simultaneously
assessing the degree to which parents’ school readiness beliefs and transition practices
contributed to children’s academic achievement during the transition to elementary school. The
study began by describing the ways in which parents’ school readiness beliefs and transition
practices varied by race, ethnicity, and socioeconomic status. Children’s achievement
trajectories were estimated in reading, mathematics, and general knowledge, and differences

72

were described by race/ethnicity and socioeconomic status. The relationships between parents’
school readiness beliefs, transition practices, and children’s achievement trajectories were
examined in this study.
This chapter summarizes findings related to variability in parental beliefs and practices as
well as the relative influence of hypothesized predictors on the growth of reading, mathematics,
and general knowledge as children transition to formal schooling. In addition to providing an
overview of these findings, implications of results for theory, research, and practice are
discussed. Finally, the chapter concludes with a description of limitations of the study and
directions for future research.
Summary of Findings
Parental Beliefs and Practices
The first task of the study was to describe the ways in which parents’ readiness beliefs
and transition practices varied by race, ethnicity, and socioeconomic status. It was hypothesized
that socioeconomic status would be positively related to school readiness beliefs. In other words,
more affluent parents would place more importance on children mastering certain school
readiness skills and attributes, such as knowing the letters of the alphabet, counting,
communicating, and being calm. It was also hypothesized that race/ethnicity would be related to
school readiness beliefs, with European American parents placing greater emphasis on school
readiness in comparison to African American or Hispanic parents.
Results from each model indicated that parents’ readiness beliefs were not related to
socioeconomic status. Readiness beliefs, however, were related to parents’ race or ethnicity.
Findings indicated that African American parents, on average, placed a little more importance on

73

school readiness skills and attributes in comparison to European American parents. There was
no difference between European American and Hispanic parents’ school readiness beliefs.
It was also hypothesized that parents use of transition practices would vary by
socioeconomic status and race/ethnicity. More specifically, socioeconomic status would be
positively related to transition practices and European American parents would report engaging
their young child in more transition practices in comparison to African American and Hispanic
parents. Results indicated that socioeconomic status positively predicted parental reports of
engaging children in transition practices prior to kindergarten. There was no difference between
the amount of transition practices employed by European American and African American
parents. That is, holding socioeconomic status constant, European American and African
American parents reported engaging young children in similar amounts of transition practices
prior to kindergarten. Hispanic parents, on the other hand, reported engaging their child in fewer
transition practices in comparison to European American parents from similar socioeconomic
backgrounds.
Overall, the findings presented thus far indicate that all parents place importance on
children exhibiting certain school readiness skills and attributes, but African Americans place
slightly more importance on school readiness in comparison to European American parents. In
terms of engaging children in transition practices prior to kindergarten, the findings indicated
that more affluent parents report engaging in greater amounts of transition practices. Irrespective
of socioeconomic status, Hispanic parents reported engaging their children in fewer transition
practices in comparison to European American parents.
Before discussing the structural relationships between parents’ readiness beliefs,
transition practices, and children’s achievement trajectories during the transition to elementary

74

school, it is important to note the consistency of the results within the existing literature and the
conceptual model of academic socialization. Although research examining parents’ readiness
beliefs is somewhat limited, there is evidence to indicate that all parents express concern for their
child’s readiness for school and transition to elementary school (Barbarin et al., 2008; Diamond
et al., 2000; Graue, 1993). Several studies have suggested that parents’ conceptions of school
readiness beliefs vary by parental level of education (West et al., 1995) or socioeconomic status
(Graue, 1993). Other studies have found that parental readiness beliefs do not vary by race
(Diamond et al., 2000) or socioeconomic status (Barbarin et al., 2008). Several studies
comparing parental and teacher conceptions of school readiness in a high-need preschool have
found that parents place more emphasis on academically-oriented school readiness skills in
contrast to teachers (Piotrkowski et al., 2000).
This study found that differences in parents’ readiness beliefs, although statistically
significant, was not very large and parents’ race and ethnicity did not account for much of the
variation in beliefs. This suggests that all parents believed that children’s acquisition of certain
readiness-related skills and attributes prior to kindergarten was important, which is consistent
with previous findings (Barbarin et al., 2008; Diamond et al., 2000; Graue, 1993). However, the
findings also suggest that African American parents placed a little more emphasis on school
readiness in comparison to European American parents which is in contrast to previous research
which found that conceptions of readiness did not differ by racial or ethnic background
(Diamond et al., 2000). The findings from this study do not indicate that parents’ conceptions of
school readiness are shaped to a large extent by parents’ race, ethnicity, or socioeconomic status.
It appears that other mechanisms may contribute to parents' conceptions of school readiness,

75

such as recollections of previous schooling experiences and parental involvement as suggested
by Barnett and Taylor (2009).
The concept of transition practices is fairly recent and has not received much attention in
the literature. The literature on home-based parental involvement during early childhood is more
extensive and generally focuses language and literacy in the home. Research suggests the
quality and quantity of language in the home is related to children’s early language and literacy
development (Aulls & Sollars, 2003; Hart & Risley, 1995). Parent-child interactive activities
such as telling stories, singing songs, and making books also encourage literacy development
(Boyce et al., 2010; Bryant et al., 1990; Maclean et al., 1987). Research, however, indicates that
differences in home-based parental involvement during early childhood are associated with
parents’ socioeconomic status, race, and ethnicity (Bouth & Crouter, 2008; Brooks-Gunn &
Markman, 2005; Hart & Risley, 1995; Heath, 1983; Phillips, 2011). Findings from this study
indicate that parental reports of engaging children in transition practices are positively related to
socioeconomic status which is consistent with previous research (Brooks-Gunn & Markman,
2005; Farkas & Hibel, 2008; Hart & Risley, 1995; Phillips, 2011; Reardon, 2011). This study
also finds that African American parents report engaging young children in a similar amount of
transition practices as European American parents from similar socioeconomic status
backgrounds. This is a noteworthy finding because much of the literature suggests that, on
average, African American parents engage their young children in fewer language and literacyrelated activities (Brooks-Gunn & Markman, 2005; Phillips, 2011) and education-related
activities (Farkas & Hibel, 2008; Phillips, 2011) in comparison to European American parents.
The theory of academic socialization suggests that parents’ beliefs, expectations, and
behaviors will be shaped by one’s socioeconomic and cultural contexts. Overall, findings from

76

this study offer support for this assumption, but mostly in terms of parents’ use of transition
practices. Although results indicated that African American parents’ placed a little more
importance on school readiness, very little of the variation in readiness beliefs was explained by
parents’ racial or ethnic heritage. On the other hand, parents’ transition practices varied by race,
ethnicity, and socioeconomic status. Affluent parents reported engaging their children in more
transition practices. Consistent with previous research which indicates that children from higher
socioeconomic status families experience more home-based education related experiences and
activities than children from lower socioeconomic status families (Reardon, 2011). Hispanic
parents, on average, reported engaging their children in fewer transition practices in comparison
to European American parents from similar socioeconomic backgrounds. The findings from this
study suggest that transition practices are shaped to a larger extent by one’s race, ethnicity, and
socioeconomic status in comparison to parents’ conceptions of school readiness.
The central questions investigated in this study addressed the degree to which parents’
conceptions of school readiness and transition practices contribute to children’s academic
achievement during the transition to elementary school as a means to better understand the
achievement gap and consider possible interventions aimed at maximizing home-based parental
involvement. Several hypotheses were under consideration. In this study, the predictive
relationship between academic socialization and children’s academic achievement was
estimated. The following sections will summarize these findings in the academic domains of
reading, mathematics, and general knowledge.
Reading Achievement
It was hypothesized that children’s beginning reading achievement and growth over time
would vary by socioeconomic status and race/ethnicity. It was hypothesized that parents’ school

77

readiness beliefs and transition practices would have a positive direct effect on children’s reading
achievement upon kindergarten entry and over time. Furthermore, it was hypothesized that
parents’ transition practices would mediate the relation between parents’ school readiness beliefs
and children’s reading achievement.
Results indicated that socioeconomic status was a positive predictor of children’s
beginning reading achievement and growth over time. In general, children from more affluent
homes began kindergarten with higher reading scores and experienced more rapid rates of
growth in reading achievement. Socioeconomic status was the strongest predictor of children’s
reading achievement and academic growth. Irrespective of socioeconomic status, African
American and Hispanic children began kindergarten with lower average scores in reading and
experienced slower average rates of growth in comparison to European American children.
These findings are consistent with previous research which indicates that race, ethnicity, and
socioeconomic status are related to children’s reading achievement in kindergarten (Farkas &
Hibel, 2008; Lee & Burkham, 2002; Rouse et al., 2005) and subsequent achievement (Fryer &
Levitt, 2004). An especially noteworthy finding based on fitted trajectories is that Hispanic
children began kindergarten with slightly lower average reading scores in comparison to African
American children from similar socioeconomic status backgrounds, but their average rate of
growth was rapid enough to have higher average reading scores by the end of first grade.
Overall, the model suggests that children who begin school with higher reading scores have more
rapid rates of growth, and children who begin school with lower reading achievement have
slower rates of growth. This phenomenon has been described in previous literature as the
Matthew Effect (Bast & Reitsma, 1997; Stanovich, 1986).

78

Parents’ school readiness beliefs positively predicted beginning reading achievement and
growth over time. Children whose parents placed more importance on school readiness had
higher average reading scores upon kindergarten entry and experienced more rapid rates of
growth. Comparison of standardized estimates indicates that parental readiness beliefs exerted
more of an effect on children’s beginning knowledge and half the effect on growth over time
than the effect exerted by race or ethnicity. Although socioeconomic status exerted the largest
effect on beginning reading achievement and growth, parents’ beliefs exerted almost half the
effect on children’s beginning reading scores and almost a third of the effect on growth over time
than the effect exerted by socioeconomic status.
Parents’ transition practices were also positively related to children’s beginning reading
achievement but had no significant association to reading growth rates. Parents’ readiness
beliefs exerted a positive influence on transition practices, and formal tests of mediation
indicated that transition practices partially mediated the relationship between parents’ school
readiness beliefs and children’s beginning reading achievement and growth over time. Parents
who placed more emphasis on school readiness skills and attributes reported engaging children
in more transition practices prior to kindergarten, which in turn had a modest positive total effect
on children’s beginning reading achievement. A one unit increase in the importance African
American and Hispanic parents place on school readiness would result in almost a two point
increase in beginning reading achievement which would close the gap between children of color
and their European American counterparts reading scores at school entry. Differences in rates of
growth would also decrease. It should be noted that parents’ readiness beliefs exerted a stronger
direct effect on reading achievement in kindergarten in comparison to the effect exerted by
transition practices. This finding is counterintuitive, as one would expect parental practices to

79

exert a greater effect on children’s academic achievement than parental beliefs. This is because
transition practices only partially mediated the relationship between parental beliefs and
children’s reading achievement.
Mathematics Achievement
Mathematics achievement trajectories were also under consideration in this study. Again
it was hypothesized that children’s beginning mathematics achievement and growth over time
would vary by socioeconomic status and race/ethnicity. Results indicated that socioeconomic
status was a positive predictor of children’s beginning mathematics achievement and growth.
Socioeconomic status was the strongest predictor of children’s mathematics achievement and
growth over time. In general, children from more affluent homes began kindergarten with higher
mathematics scores and experienced more rapid growth rates in mathematics through the first
grade. These findings are consistent with previous research, which indicates that socioeconomic
status is positively associated with children’s achievement in mathematics during early
elementary school (Bodovski & Farkas, 2007; Farkas & Hibel, 2008; Lee & Burkham, 2002).
Findings also indicated that European American children began kindergarten with higher
average mathematics scores and experienced more rapid average rates of growth in comparison
to African American and Hispanic children from similar socioeconomic backgrounds. Based on
fitted trajectories, Hispanic children, on average began kindergarten with slightly lower average
mathematics scores in comparison to African American children from similar socioeconomic
backgrounds, but Hispanic children’s rate of growth was rapid enough to close this gap and have
higher average mathematics scores than African American children by the end of first grade.
Similar to reading achievement, this model indicates that children who begin school with higher

80

mathematics scores have more rapid rates of growth, and children who begin school with lower
mathematics achievement will have a slower rate of growth.
Once more, it was hypothesized that parents’ school readiness beliefs and transition
practices would have a positive direct effect on children’s mathematics achievement upon
kindergarten entry and over time. It was also hypothesized that parents’ transition practices
would mediate the relation between parents’ school readiness beliefs and children’s mathematics
achievement. Results indicated that parents’ school readiness beliefs exerted a positive influence
on children’s beginning mathematics achievement and rates of growth. Parents’ beliefs exerted
more than half the effect on children’s beginning mathematics achievement in comparison to the
effect exerted by race and ethnicity and almost a third of the effect exerted by socioeconomic
status. Although the effect of parents’ beliefs on children’s mathematics achievement decreases
over time, it continues to exert a quarter of the effect exerted by socioeconomic status.
Parents’ transition practices were also positively related to children’s beginning
mathematics achievement but not to growth over time. Parents’ readiness beliefs exerted a
positive influence on transition practices, and formal tests of mediation supported the claim that
transition practices partially mediated the relationship between parents’ school readiness beliefs
and children’s beginning achievement and growth over time. Parents who placed more
importance on school readiness reported engaging children in more transition practices prior to
formal schooling, which in turn had a small positive effect on children’s beginning mathematics
achievement. Again parents’ readiness beliefs exerted a greater influence on children’s
beginning mathematics achievement in comparison to the effect of transition practices.
Transition practices partially mediated the relationship between parents’ school readiness beliefs
and children’s mathematics achievement.

81

General Knowledge Achievement
This study also investigated children’s general knowledge achievement trajectories. It
was hypothesized that socioeconomic status as well as race and ethnicity would predict
children’s beginning general knowledge achievement and growth over time. Very few studies
have examined differences in children’s general knowledge using the ECLS-K data set (see
Farkas, 2008 for an exception) and no studies to my knowledge have examined growth in general
knowledge achievement. Results indicated that socioeconomic status was positively related to
children’s general knowledge achievement in kindergarten but not growth over time.
Socioeconomic status was the strongest predictor of children’s general knowledge achievement
at the onset of kindergarten. In general, children from more affluent homes began kindergarten
with higher general knowledge scores. Findings also indicated that, in general, European
American children began kindergarten with higher general knowledge scores and experienced
more rapid rates of growth in comparison to African American and Hispanic children,
irrespective of socioeconomic status. Socioeconomic status and racial/ethnic heritage were the
strongest predictors of children’s beginning general knowledge. These findings are consistent
with results presented by Farkas (2008). A new and particularly striking finding based on fitted
trajectories is that, on average, Hispanic children had higher general knowledge scores at the
beginning of kindergarten in comparison to African American children. Despite having a less
rapid rate of growth, by the end of first grade Hispanic children maintained higher average
general knowledge scores in comparison to African American children from similar
socioeconomic status backgrounds.
Overall, socioeconomic status, race, and ethnicity had little to no effect on children’s
general knowledge growth rates, suggesting that schools exert a stronger effect on children’s

82

general knowledge achievement over time. Children, irrespective of racial or ethnic heritage,
have similar rates of growth, however, European American children still have higher average
general knowledge scores in comparison to African American and Hispanic children from
similar socioeconomic backgrounds at the end of first grade. Finally, in contrast to the findings
for reading and mathematics growth, model estimates indicated that children who begin school
with higher general knowledge scores have slower rates of growth, while children who begin
school with lower levels of general knowledge achievement will have a faster rate of growth.
Again it was hypothesized that parents’ school readiness beliefs and transition practices
would have a positive direct effect on children’s general knowledge achievement and growth
over time. It was also hypothesized that parents’ transition practices would mediate the relation
between parents’ school readiness beliefs and children’s general knowledge achievement.
Parents’ school readiness beliefs positively predicted beginning general knowledge achievement.
Parents who placed more importance on school readiness had children with higher general
knowledge scores upon kindergarten entry. Parents’ readiness beliefs had no significant
association to growth rates in general knowledge.
Parents’ transition practices were also positively related to children’s beginning general
knowledge achievement but had no association to rates of growth. Contrary to the effects
reported in the reading and mathematics models, parents’ transition practices exerted a greater
influence on children’s beginning general knowledge achievement than the effect exerted by
parental readiness beliefs. Parents’ transition practices, however, only exerted one-fifth of the
influence exerted by socioeconomic status and barely a quarter of the influence exerted by race
and ethnicity on children’s beginning general knowledge achievement. Parents’ readiness beliefs
exerted a positive influence on transition practices, and formal tests of mediation indicated that

83

transition practices partially mediated the relationship between parents’ school readiness beliefs
and children’s beginning general knowledge achievement. Parents who placed more importance
on school readiness skills and attributes reported engaging children in more transition practices,
and transition practices had a small positive effect on children’s general knowledge achievement
at the beginning of kindergarten. Despite the positive influence that transition practices exert on
general knowledge achievement, children’s socioeconomic status, race, and ethnicity exerted a
much greater influence.
Academic Socialization and Achievement
Overall, parents’ school readiness beliefs and transition practices had a small-to-modest
effect on children’s reading, mathematics, and general knowledge on beginning achievement and
a small effect on growth in reading and mathematics achievement, thereby supporting the theory
of academic socialization. Consistent with previous research, socioeconomic status, race, and
ethnicity continued to be strong predictors of academic growth in reading and mathematics
during elementary school. As a whole, the findings demonstrated that children from higher
socioeconomic status homes began formal schooling with higher reading, mathematics, and
general knowledge scores. Socioeconomic status was the strongest predictor of beginning
reading, mathematics, and general knowledge achievement and reading and mathematics growth
over time. In general, European American children began kindergarten with higher reading,
mathematics, and general knowledge achievement and experienced more rapid rates of growth in
reading and mathematics compared to African American and Hispanic children from similar
socioeconomic backgrounds.
Based on fitted trajectories, the findings suggest that Hispanic children, in general had a
rate of growth in reading and mathematics similar to European American children from similar

84

socioeconomic backgrounds, in contrast to African American children who, on average, had less
rapid rates of growth in comparison to European American children. Although the achievement
gap between European American and Hispanic children did not decrease during the transition to
elementary school, it did not widen. While the gap between European American and African
American children from similar socioeconomic backgrounds appear to widen as children
transition to elementary school. The findings from fitted trajectories also suggest that Hispanic
children experience more rapid rates of growth in reading and mathematics in comparison to
African American children from similar socioeconomic backgrounds. These results are
consistent with previous research which demonstrates that the achievement gap in reading and
mathematics between Hispanic and European American children decreases as children progress
through elementary school while the gap between African American and European American
children widens (Reardon & Gallindo, 2009; Reardon, 2008).
Despite the role of socioeconomic status, race, and ethnicity on children’s early
achievement, academic socialization which was conceptualized as parents’ school readiness
beliefs and transition practices also contributed to children’s academic achievement. Parent’s
readiness beliefs exerted a small-to-moderate direct effect on children’s reading and mathematics
achievement upon kindergarten entry and continued to exert a small effect on growth over time.
This finding supports previous research which demonstrates that parents’ conceptions of school
readiness or expectations for mastery of developmental milestones are positively associated with
children’s early academic (Barbarin et al., 2008; Hess et al., 1984; Hess et al., 1980) and later
academic achievement (Hess et al., 1984).
The results from this study are consistent with previous literature which suggests that
home-based parental involvement during early childhood, as evidenced by interactive parent-

85

child engagement, is positively associated with children’s early academic outcomes (Boyce et
al., 2010; Bryant et al., 1990; Glazer, 1989; Hart & Risley, 1995; Maclean et al., 1987; Senechal
& Young, 2008; Sonnenschein et al., 1996). Although parents’ transition practices exerted a
positive influence on children’s early achievement, it exerted no influence on academic growth
over time. Furthermore, transition practices exerted less influence on children’s achievement in
reading and mathematics than parents’ readiness beliefs. Transition practices only partially
mediated the relationship between parental readiness beliefs and children’s achievement. What
can explain such a counterintuitive finding? One possible explanation is that items which
comprise parents’ transition practices measured the frequency of transition practices reported by
parents, not the duration of activity, or the quality of those interactions. According to Hart and
Risley (1995) the quality of language is a strong predictor of children’s language and literacy
development. It should be noted, however, that parents’ transition practices exerted a larger
influence on children’s general knowledge achievement at the beginning of kindergarten in
comparison to parents’ readiness beliefs. One plausible explanation for this finding relates to the
broad conceptualization of transition practices, which includes items that measure the frequency
that parents report engaging young children in a variety of interactive learning opportunities,
such as teaching the child about nature and building things together.
Limitations, Strengths, and Future Directions for Research
A major contribution of the current study is the finding that parents’ school readiness
beliefs exert a positive influence on transition practices, which in turn exert a positive influence
on children’s achievement. Few studies have used a mediating model to examine the
relationship between parental beliefs, involvement, and children’s achievement (see Sy and
Schulenberg, 2006 for an exception) and no study to my knowledge has used a path analysis

86

model to examine the relationship between parents’ conceptions of school readiness, practices,
and children’s achievement trajectories. Formal tests of mediation confirmed that parents’
transition practices partially mediated the relation between parents’ conceptions of school
readiness and children’s academic achievement in reading, mathematics, and general knowledge.
Therefore, results from this study support the theory of academic socialization which posits that
parental beliefs influence parenting practices, which in turn influences children’s transitions to
elementary school.
A major strength of this study was that it investigated children’s academic achievement
in reading, mathematics, and general knowledge across time. In general, previous research has
examined achievement gaps using differences in mean scores, differences in standardized scores,
and by so-called metric-free gap measures (Reardon & Robinson, 2007). These approaches use
cross-sectional data, to show differences in children’s academic achievement, measured at
specific time points, such as first grade, third grade, and fifth grade. This study, in contrast, used
a longitudinal data analysis approach, which allows the estimation of individual growth
trajectories while including predictors of change. The findings presented from this study provide
a more nuanced description of differences in academic achievement trajectories between
European American, African American, and Hispanic children during the transition to formal
schooling. It should be noted that longitudinal analyses often suffer from problems with missing
data. Although missing data was handled with a combination of a full information likelihood
method and pairwise present method, the potential for bias is always present when large amounts
of data are missing for variables of interest.
Few studies have considered the ways in which parenting practices mediate the
relationship between parental beliefs and children’s early academic outcomes. It is a common

87

assumption that parents’ school readiness beliefs will influence the ways that parents engage
children in specific activities as a means to support children’s readiness for school and successful
transition to elementary school. Findings from this study validate the assumption that parents’
beliefs influence their behaviors which in turn influence children’s academic outcomes. This
study investigated the relationship between parents’ conceptions of school readiness, transition
practices, and children’s academic growth using structural equation modeling, a sophisticated
multivariate analysis technique that allowed for the simultaneous investigation of variability in
predictors and the relative impact of hypothesized predictors on outcomes of interest. The
specific type of structural equation modeling employed was latent growth curve modeling, which
was used to model change across time using longitudinal data. This approach is a
methodological strength of the study and contributes to the achievement gap literature by
providing a more detailed and nuanced picture of the mean differences in academic growth
between European American, African American, and Hispanic children during the transition to
elementary school.
Given the correlational nature of this study, it is not clear whether the relationship
between parents’ school readiness beliefs, transition practices, and children’s achievement is
causal. Although caution is needed in interpreting these results given that they are correlational,
one implication of the findings is that parental conceptions of school readiness positively
influence parenting practices. This provides evidence that we must expand our understanding of
home-based parental involvement during early childhood and highlights potential targets for
interventions which maximize the important role of parents as their child’s first teacher. Thus,
these findings suggest that providing parents with information about school readiness and ways
to support children’s early learning and development may help children successfully transition to

88

elementary school. Therefore, experimental and quasi-experimental studies are needed to
evaluate the effectiveness of intervention programs which inform parents about school readiness
issues and transition practices as a means to increase children’s academic achievement during the
transition to elementary school.
Despite the advantages of using existing large-scale data sets, such as the potential to
increase external validity of the study due to the large nationally representative sample, there are
several disadvantages of this approach which limit the findings of this study. For one, these
studies are generally correlational and may fail to include confounding variables. Another issue
pertains to measurement error, or the use of imprecise measures, which may underestimate the
influence of a predictor. For example, data were not included that measure the quality of
parents’ transition practices, and research indicates that the quality of parent-child interactions
during early childhood is a strong predictor of children’s vocabulary development (Hart &
Risley, 1995; Landry & Smith, 2008; Landry, Smith, & Swank, 2003). In future studies, it
would be worthwhile to investigate findings reported here with more refined measures of
parenting practices, collected through observation by trained researchers.
Another limitation of this study relating to the use of large-scale data concerns its use of
self-reported data. While the outcome data of interest for this study consisted of direct-child
assessments, the predictors examined all relied on self-reported data from parents. Parent
responses may be compromised by their inability to accurately report the frequency of transition
practices. There may also be issues of social desirability when reporting the frequency of
engaging in transition practices. In other words, parents may provide socially desirable
responses to questions about school readiness. It is hard to imagine a parent suggesting that it is
not important for a child to be calm or be able to communicate prior to beginning kindergarten.

89

Also, survey questions measuring parents’ school readiness beliefs may not adequately address
the skills and behaviors parents believe are necessary upon kindergarten entry. Perhaps there are
a group of parents who strongly believe a child should be able to identify sounds in words or do
simple addition and subtraction before entering kindergarten. If so, do these parents engage
children in different ways that support the acquisition of those skills?
Given the wide variability within the groups examined in this study, interpretation based
on broad group definitions should be tentative. In particular, recent scholars have estimated that
results from grouped analyses of people identified as Hispanic may not always be appropriate
(Reardon & Gallindo, 2009). Future studies could examine variability in academic achievement
trajectories for Hispanic subpopulations based upon parental country of origin or generational
status.
The findings of the present study suggest several fruitful avenues of future work.
Measures of children’s social and emotional development as outcomes of interest were not
considered here, but present a compelling expansion to the present work. Recent scholars have
argued that children’s approaches toward learning, which include persistence, emotional
regulation, attentiveness, and flexibility (Fantuzzo et al., 2007; McWayne, Fantuzzo, &
McDermott, 2004) may lay the foundation of basic academic skills that facilitate later acquisition
of more advanced academic skills (Cunha, Heckman, Lochner, & Masterov, 2006). A recent
study using data from the ECLS-K study has demonstrated that children’s approaches toward
learning are positively related to children’s academic growth trajectories (Li-Grining et al.,
2010). Parents’ conceptions of school readiness and transition practices may predict children’s
social and emotional development during the transition to elementary school.

90

A particularly interesting finding, based on fitted trajectories, was the observed
differences between Hispanic and African American children’s academic achievement growth
trajectories. Future research could explore these observed differences, specifically between
children attending similar schools. Subsequent studies could include additional family and
school characteristics as predictors of children’s achievement trajectories in order to identify
additional sources of variability between European American, African American, and Hispanic
children’s academic achievement as they progress in elementary school.
Conclusion
This study investigated the relationship between parents’ conceptions of school readiness,
transition practices, and children’s academic growth using latent growth curve modeling. The
findings demonstrated that on average, children from more affluent homes began school with
higher levels of achievement and experienced more rapid rates of growth in reading,
mathematics, and general knowledge in comparison to less affluent children. The results also
illustrated that European American children in general began kindergarten with higher levels of
academic achievement and experienced more rapid rates of academic growth in comparison to
African American and Hispanic children from similar socioeconomic backgrounds. Moreover,
fitted trajectories illustrated that although Hispanic children may begin kindergarten with lower
average achievement in reading and mathematics compared to African American children,
Hispanic children have more rapid rates of academic growth and have higher average scores the
end of first grade. Therefore, this study highlights the importance of considering individual
variation in children’s achievement trajectories over time, particularly during important
developmental transitions.

91

The primary substantive contribution of this study is that the findings broaden our
understanding of home-based parental involvement during early childhood by illuminating the
positive relationship between parents’ school readiness beliefs, parenting practices, and
children’s early academic achievement. The findings suggest that parents who placed more
importance on school readiness reported engaging children in more transition practices. Parents’
school readiness beliefs were directly and indirectly related to children’s beginning academic
achievement. Considering parents’ beliefs as an indirect predictor of children’s academic
achievement provides a more nuanced understanding of one way in which parents influence
children’s early education. The findings provide support for the theory of academic socialization
posited by Taylor et al. (2004) which suggests parents’ beliefs and values about schooling
influence parenting behaviors, which in turn, influence children’s transition to formal schooling.
This study also examined differences in parents’ conceptions of school readiness.
Although African American parents placed a little more importance on school readiness in
comparison to European American parents, in general, the findings suggest that all parents place
importance on children displaying certain school readiness skills and attributes regardless of
racial, ethnic, or socioeconomic background. Although parents’ readiness beliefs had a positive
direct and indirect effect on children’s early academic achievement and growth over time, there
were no major differences in parents’ conceptions of school readiness by race, ethnicity, or
socioeconomic status. Therefore, the findings from this study do not indicate that gaps in
children’s early academic achievement are attributed to differences in parents’ conceptions of
school readiness.
This study has important practical implications for potential interventions aimed at
reducing the early achievement gap. If parents’ beliefs about school readiness positively

92

influence parenting practices, then possible interventions can look toward informing parents
about school readiness even before the birth of their children and throughout their children's
early developmental years. Having a broader understanding of the types of knowledge, skills,
and attributes that contribute to a child’s successful transition to elementary school in
combination with a deeper understanding of developmentally appropriate practices which
support children’s early learning and development empowers parents to maximize their child’s
potential during early childhood.
Although parents have the potential to play an important role, the findings from this study
are consistent with previous research which indicates that socioeconomic status exerts a strong
positive influence on children’s early academic achievement. Parental beliefs and transition
practices were not able to overcome the cumulative advantage bestowed upon children from
affluent families with highly educated parents. In the current policy context, there is an
increasing awareness on the role of parents in supporting children’s early learning and
development, and a growing consensus for policy to shift more attention toward early
interventions (Heckman, 2011). These early interventions should provide culturally sensitive
encouragement and support for parents by informing parents about school readiness issues and
developmentally appropriate ways to engage young children in order to support their transition to
elementary school.

93

APPENDICES

94

.

Readiness
Beliefs

Kindergarten
achievement
Socioeconomic Status
Race and Ethnicity
Achievement
trajectory

Transition
Practices

Figure 1. Conceptual model representing the influence of academic socialization
on children’s achievement.

95

Alphabet

1

Count
Readiness
Beliefs

Draw
Communicate
Calm
Share

1

Latent
Intercept

1
1
0

Read

Latent
Latent
Slope
Slope
1

Stories

Fall
Kindergarten
Reading
Spring
Kindergarten
Reading

1
3

Spring
First Grade
Reading

Song
Art

Transition
Practices

Games
Build
Nature
Sport

Figure 2. Measurement model of academic socialization and reading achievement. Final
measurement model allows importance alphabet and importance count error terms to covary;
not illustrated in path diagram.

96

Alphabet

1

Count
Readiness
Beliefs

Draw
Communicate
Calm
Share

1

Latent
Intercept

1
1
0

Read

Latent
Slope

1

Stories

1
3

Fall
Kindergarten
Mathematics
Spring
Kindergarten
Mathematics
Spring
First Grade
Mathematics

Song
Art

Transition
Practices

Games
Build
Nature
Sport
Figure 3. Measurement model of academic socialization and mathematics achievement. Final
measurement model allows importance alphabet and importance count error terms to covary;
not illustrated in path diagram.

97

Alphabet

1

Count
Readiness
Beliefs

Draw
Communicate
Calm
Share

Latent
Intercept

1
1
1
0

Read

Latent
Slope
1

Stories

Fall
Kindergarten
General
Spring
Kindergarten
General

1

3

Spring
First Grade
General

Song
Art

Transition
Practices

Games
Build
Nature
Sport

Figure 4. Measurement model of academic socialization and general knowledge achievement.
Final measurement model allows importance alphabet and importance count error terms to
covary; not illustrated in path diagram.

98

Alphabet

1

Count
Readiness
Beliefs

Draw
Communicate
Calm
Share

1

Latent
Intercept

1
1
0

Read

Latent
Slope
1

Stories

Fall
Kindergarten
Reading
Spring
Kindergarten
Reading

1
3

Spring
First Grade
Reading

Song
Art

Transition
Practices

Games
Build
Nature
Sport

Figure 5. Structural model of academic socialization as a predictor of reading achievement.
Socioeconomic status and race/ethnicity (African American and Hispanic dummy coded with
European American as reference category) included in full structural model as covariates; not
illustrated in path diagram.

99

Alphabet

1

Count
Readiness
Beliefs

Draw
Communicate
Calm
Share

1

Latent
Intercept

1
1
0

Read

Latent
Slope
1

Stories

Fall
Kindergarten
Mathematics
Spring
Kindergarten
Mathematics

1
3

Spring
First Grade
Mathematics

Song
Art

Transition
Practices

Games
Build
Nature
Sport
Figure 6. Structural model of academic socialization as a predictor of mathematics
achievement. Socioeconomic status and race/ethnicity (African American and Hispanic
dummy coded with European American as reference category) included in full structural
model as covariates; not illustrated in path diagram.
100

Alphabet

1

Count
Readiness
Beliefs

Draw
Communicate
Calm
Share

Latent
Intercept

1
1
1
0

Read

Latent
Slope
1

Stories

Fall
Kindergarten
General
Spring
Kindergarten
General

1
3

Spring
First Grade
General

Song
Art

Transition
Practices

Games
Build
Nature
Sport
Figure 7. Structural model of academic socialization as a predictor general knowledge
achievement. Socioeconomic status and race/ethnicity (African American and Hispanic
dummy coded with European American as reference category) included in full structural
model as covariates; not illustrated in path diagram.

101

Table 1
Summary Statistics and Missing Rates for Model Variables
Variable
Reading
Fall kindergarten
Spring kindergarten
Spring first grade
Mathematics
Fall kindergarten
Spring kindergarten
Spring first grade
General knowledge
Fall kindergarten
Spring kindergarten
Spring first grade
Covariates
Socioeconomic status
Race/Ethnicity
European American
African American
Hispanic
Parent Belief indicators
Importance count
Not important
Somewhat important
Important
Very important
Essential
Importance Share
Not important
Somewhat important
Important
Very important
Essential
Importance Draw
Not important
Somewhat important
Important
Very important
Essential
Importance Calm
Not important
Somewhat important
Important

M

SD

N

Missing rate (%)

26.90
37.93
66.87

9.53
12.75
20.58

12,261
12,556
12,857

7
4
2

20.97
30.85
53.95

8.61
11.36
15.99

13,061
13,090
13,102

1
1
1

22.23
27.10
34.28

7.44
7.77
7.58

12,238
12,541
12,850

7
4
2

-.038

.78

12,816
13,133

2
0

12,452

5

12,458

5

12,458

5

12,456

5

.62
.17
.21

.01
.07
.32
.43
.17
.00
.00
.05
.61
.34
.00
.02
.24
.52
.22
.00
.01
.15
102

Table 1 (cont’d)

Very important
Essential
Importance Alphabet
Not important
Somewhat important
Important
Very important
Essential
Importance Communicate
Not important
Somewhat important
Important
Very important
Essential
Transition practices
Frequency Read
not at all
once or twice a week
3 to 6 times a week
everyday
Frequency tell stories
not at all
once or twice a week
3 to 6 times a week
everyday
Frequency sing songs
not at all
once or twice a week
3 to 6 times a week
everyday
Frequency play games
not at all
once or twice a week
3 to 6 times a week
everyday
Frequency make art
not at all
once or twice a week
3 to 6 times a week
everyday
Frequency teach about nature
not at all
once or twice a week
3 to 6 times a week
everyday

.59
.25
12,457

5

12,459

5

12,453

5

12,449

5

12,452

5

12,454

5

12,447

5

12,443

5

.01
.05
.25
.50
.19
.00
.00
.06
.58
.36

.01
.19
.35
.45
.08
.36
.31
.25
.05
.23
.28
.44
.04
.35
.39
.22
.07
.41
.32
20
.20
.48
.22
.10
103

Table 1 (cont’d)

Frequency build
not at all
once or twice a week
3 to 6 times a week
everyday
Frequency play sports
not at all
once or twice a week
3 to 6 times a week
everyday
Note. N varies due to missing data.

12,444

5

12,449

5

.18
.44
.25
.13
.10
.36
.33
.20

104

Table 2
Correlations Among Parents’ Readiness Beliefs, Transition Practices, and Reading Achievement Outcomes
1
Child outcomes
1. Fall K
2. Spring K
3. Spring 1st
Parent beliefs
4. Count
5. Share
6. Draw
7. Calm
8. Alphabet
9. Communicate
Practices
10. Read

2

3

.82
.67

.77

-

.07
.09
.07
.01ᵃ
.08
.11

.07
.10
.07
.03
.07
.12

.03
.11
.05
.03
.04
.12

4

5

6

7

8

9

.36
.46
.36
.63
.24

.47
.47
.36
.50

.43
.50
.37

.44
.41

.33

-

.11
.05
-.01ᵃ
11. Story
.07
.07
.08
.07
.06
.07
12. Sing songs
.03
.05
.05
.07
.12
.08
13. Play games
.06
.05
.04
.05
.09
.06
14. Art project
.04
.04
.05
.05
.07
.09
15. Teach nature
.12
.12
.12
.12
.06
-.01ᵃ
16. Build
.03
.05
.03
-.01ᵃ -.01ᵃ -.01ᵃ
17. Play sports
-.04
-.03
-.02
.05
.07
.02
Note. Polychoric correlations estimated between ordinal variables.
.22

.21

.20

.00ᵃ
.01ᵃ
.04
.04
.03
.01ᵃ

-.02ᵃ
.07
.07
.07
.03
.00ᵃ

10

.09

.43
.21
.28
.27
.28

12

13

14

.27
.33
.28
.30

.26 .26 .34 .26 .30 .30

15

16

17

-

.08
.12
.07
.08
.12

11

-

.01ᵃ
.00ᵃ
.01ᵃ .25 .29 .20 .36 .32 .32 .03
.03
.06
.20 .24 .24 .32 .27 .26 .32 Polyserial correlations estimated between ordinal and continuous

variables. ᵃ Coefficient is not significant at the p < .05 level.

105

Table 3
Correlations Among Parents’ Readiness Beliefs, Transition Practices, and Mathematics Achievement Outcomes
1
Child outcomes
1. Fall K
2. Spring K
3. Spring 1st
Parent beliefs
4. Count

2

3

.82
.71

.78

4

5

6

7

8

9

10

-

.03
.13
.10
.00ᵃ
.02
.13

.01
.13
.06
.01ᵃ

.01ᵃ
.12
.05
.00ᵃ

.36
.47
.36

.47
.47

.43

-.01ᵃ
.11

.63
.24

.36
.50

.50
.37

.44
.41

.18

.17

.11

.05

.00ᵃ

.05

.06

.06

.07

.08

.43

12. Sing songs
13. Play games
14. Art project
15. Teach nature

.04
.08
.06
.14

.02
.08
.06
.14

.01ᵃ
.05
.05
.13

.12
.09
.07
.12

.08
.06
.09
.06

.01ᵃ
.04
.04
.03
.01ᵃ

.07
.07
.03
.00ᵃ

.12
.07
.08
.12

.21
.28
.27
.28

16. Build

.02

14

15

16

17

-

.07

-.02ᵃ
.07

.09

11. Story

-.01ᵃ
.07

13

.33

.20

12

-

.01ᵃ
.13

11

5. Share
6. Draw
7. Calm
8. Alphabet
9. Communicate
Practices
10. Read

.07
.05
.05
-.01ᵃ
.03

.27 .33 .26 .28 .26 .34 .30 .26 .30 .30

-

.01ᵃ
.01ᵃ .25 .29 .20 .36 .32 .32
.01ᵃ
17. Play sports
.02
.05
.06
.02
.03
.03
.06 .20 .24 .24 .32 .27 .26 .32 .02ᵃ
.01ᵃ
Note. Polychoric correlations estimated between ordinal variables. Polyserial correlations estimated between ordinal and continuous
.03

.02

.05

.03

outcomes. ᵃ Coefficient is not significant at the p < .05 level.

106

Table 4
Correlations Among Parents’ Readiness Beliefs, Transition Practices, and General Knowledge Achievement Outcomes
1

2

3

.86
.79

.84

-

-.08
.10
.03
.04
-.10
.13

-.09
.12
.04
-.03
-.08
.14

.25

12. Sing songs
13. Play games
14. Art project
15. Teach nature

.08
.02
.05
.09
.21

16. Build

.03

Child outcomes
1. Fall K
2. Spring K
3. Spring 1st
Parent beliefs
4. Count
5. Share
6. Draw
7. Calm
8. Alphabet
9. Communicate
Practices
10. Read
11. Story

4

5

6

7

8

-.09
.14
.02
-.04
-.09
.14

.36
.48
.36
.63
.24

.46
.47
.36
.50

.43
.50
.37

.44
.41

.33

.26

.25

.11

.05

.00ᵃ

.09
.03
.05
.09
.22

.09
.02
.05
.09
.21

-.01ᵃ
.07
.07
.05
.05
-.01ᵃ
.03

.06
.12
.09
.07
.12

.07
.08
.06
.09
.06

.01ᵃ
.04
.04
.03
.01ᵃ

-.02ᵃ
.07
.07
.07
.03
.00ᵃ

9

10

.09

-

.08
.12
.07
.08
.12

.43
.21
.28
.27
.28

11

12

13

14

.27 .33 .26 .28 .26 .34 .30 .26 .30 .30

15

16

17

-

.01ᵃ
.01ᵃ
.01ᵃ .25 .29 .20 .36 .32 .32 17. Play sports
-.02
.03
.06
.20 .24 .24 .32 .27 .26 .32 .05
.07 .02 .03
.00ᵃ
.01ᵃ
Note. Polychoric correlations estimated between ordinal variables. Polyserial correlations estimated between ordinal and continuous
.03

.05

.05

.03

variables. ᵃ Coefficient is not significant at the p < .05 level.

107

Table 5
Summary of Model Fit Indices for all Measurement and Structural Models Under Study
χ²

df

Reading
Measurement model
552.58
30
Structural model
578.73
36
Mathematics
Measurement model
517.50
30
Structural model
503.08
37
General knowledge
Measurement model
682.77
26
Structural model
506.04
32
Note. N varies due to missing values.

p<

CFI

TLI

RMSEA

N

.001
.001

.96
.96

.97
.96

.036
.034

13,108
12,801

.001
.001

.96
.96

.96
.96

.035
.031

13,128
12,812

.001
.001

.92
.95

.93
.95

.044
.034

13,118
12,806

108

Table 6
Factor Loadings of Indicators for Structural Model of Academic Socialization and Reading
Achievement
Factors/Items
Readiness beliefs
Importance alphabetᵃ
Importance count
Importance draw
Importance communicate
Importance calm
Importance share
Transition practices
Frequency readᵃ
Frequency tell stories
Frequency sing songs
Frequency make art
Frequency play games
Frequency build together
Frequency teach nature
Frequency play sports
Latent Intercept

Unstandardized Standardized
factor loadings factor loadings

R²

1.00

.66

.44

.90***
1.06***
.92***
1.00***
1.07***

.60
.70
.61
.66
.70

.35
.49
.37
.43
.50

1.00

.52

.27

1.15***
.90***
1.06***
1.17***
1.05***
1.03***
.96***

.58
.46
.55
.60
.55
.54
.50

.33
.21
.30
.36
.30
.29
.25

Fall kindergarten readingᵇ

1.00

.98

Spring kindergarten readingᵇ

1.00

.73

Spring first grade readingᵇ
Latent Slope

1.00

.46

Fall kindergarten readingᵇ

0

0

Spring kindergarten readingᵇ

1.00

.34

Spring first grade readingᵇ

3.00

.64

Note. ᵃ Unstandardized factor loading was fixed to equal 1.0 and was not

ᵇ Unstandardized factor loadings were fixed to
estimate growth curve. Model fit statistics, X ² (36, N = 12,801) = 578.73,
RMSEA = .034, CFI =.96, TLI = .96. R² value represents the amount of
variation for the observed variable explained by the latent factor. R² fall
kindergarten reading = .96; R² spring kindergarten reading = .81; R² spring
tested for significance.

kindergarten reading = .83. * p < .05, ** p < .01, *** p < .001.

109

Table 7
Structural Model Estimates of Academic Socialization as a Predictor of Reading Achievement
Latent factor
1
2
3
4
5
6
7
1. Intercept
1.87*** / .13
.49* / .03
4.13*** / .34 -1.75*** / -.07 -2.17*** / -.09
2. Slope
.48*** / .07
.13 / .02
1.42*** / .25 -1.47*** / -.12 -.60*** / -.06
3. Beliefs
-.02 / .02
.09** / .05
-.01 / .01
4. Practices
.12*** / .15
.12*** / .17
-.03 / .02
-.15*** / -.11
Covariates
5. SES
6. African American
7. Hispanic
Note. The first estimate is unstandardized, and the second is standardized. Columns represent exogenous factors (predictors) and
rows represent endogenous factors (outcomes). Model fit statistics, X ² (36, N = 12,801) = 578.73, RMSEA = .034, CFI =.96, TLI
= .96. Mean intercept = 27.08, Mean slope = 13.78. R² intercept =.174; R² slope =.105; R² beliefs =.003; R² behavior =.075. * p < .05,
** p < .01, *** p < .001.

110

60
55

IRT Reading Score

50
45
40
35
30
25
20
15
10
5
0

Fall K

European American

Spring K
Time of Data Collection
African American

Spring 1st

Hispanic

Figure 8. Reading achievement trajectories by race and ethnicity conditional on socioeconomic
status

111

Table 8
Factor Loadings of Indicators for Structural Model of Academic Socialization and
Mathematics Achievement
Factors/Items
Readiness beliefs
Importance alphabetᵃ
Importance count
Importance draw
Importance communicate
Importance calm
Importance share
Transition practices
Frequency readᵃ
Frequency tell stories
Frequency sing songs
Frequency make art
Frequency play games
Frequency build together
Frequency teach nature
Frequency play sports
Latent Intercept

Unstandardized Standardized
factor loadings factor loadings

R²

1.00

.65

.43

.90***
1.07***
.93***
1.01***
1.09***

.59
.70
.61
.66
.71

.35
.49
.37
.44
.50

1.00

.52

.27

1.12***
.89***
1.07***
1.17***
1.07***
1.04***
.96***

.58
.46
.55
.61
.55
.54
.50

.33
.21
.30
.36
.30
.29
.25

Fall kindergarten mathematicsᵇ

1.00

.98

Spring kindergarten mathematicsᵇ

1.00

.75

Spring first grade mathematicsᵇ
Latent Slope

1.00

.53

Fall kindergarten mathematicsᵇ

0

0

Spring kindergarten mathematicsᵇ

1.00

.29

Spring first grade mathematicsᵇ

3.00

.62

Note. ᵃ Unstandardized factor loading was fixed to equal 1.0 and was not tested
for significance.

ᵇ Unstandardized factor loadings were fixed to estimate growth

curve. R² value represents the amount of variation for the observed variable
explained by the latent factor. Model fit statistics, X ² (37, N = 12,812) = 503.08,
RMSEA = .031, CFI = .96, TLI = .96. R² fall kindergarten mathematics =.96; R²
Spring kindergarten mathematics =.78; R² Spring first grade mathematics = .88.
* p < .05, ** p < .01, *** p < .001.

112

Table 9
Structural Model Estimates of Academic Socialization as a Predictor of Mathematics Achievement
Latent factor
1
2
3
4
5
6
7
1. Intercept
1.34*** / .10
.47* / .03
3.93*** / .36
-3.38*** / -.15 -3.75*** / -.17
2. Slope
.30*** / .06
-.04 / .01
.870*** / .21
-1.68*** / -.19
-.50** / -.06
3. Beliefs
-.02 / -.02
.09** / .05
-.01 / -.01
4. Practices
.12*** / .15
.12*** / .17
-.03 / -.02
-.15*** / -.11
Covariates
5. SES
6. African
American
7. Hispanic
Note. The first estimate is unstandardized, and the second is standardized. Columns represent exogenous factors (predictors) and
rows represent endogenous factors (outcomes). Model fit statistics, X ² (36, N = 12,812) = 503.08, RMSEA = .031, CFI = .96, TLI
= .96. Mean intercept = 22.31 Mean slope = 11.41. R² intercept = .24; R² slope = .10; R² beliefs = .003; R² behavior = .075. * p < .05,
** p < .01, *** p < .001.

113

50

MATHEMATICS IRT SCORE

45
40
35
30
25
20
15
10
5
0

Fall K

European American

Spring K
Time of Data Collection
African American

Spring 1st

Hispanic

Figure 9. Mathematics achievement trajectories by race/ethnicity conditional on socioeconomic
status

114

Table 10
Factor Loadings of Indicators for Structural Model of Academic Socialization and
General Knowledge Achievement
Factors/Items
Readiness beliefs
Importance alphabetᵃ
Importance count
Importance draw
Importance communicate
Importance calm
Importance share
Transition practices
Frequency readᵃ
Frequency tell stories
Frequency sing songs
Frequency make art
Frequency play games
Frequency build together
Frequency teach nature
Frequency play sports
Latent Intercept

Unstandardized Standardized
factor loadings factor loadings

R²

1.00

.65

.42

.90***
1.09***
.95***
1.03***
1.10***

.58
.70
.61
.66
.71

.34
.49
.37
.44
.50

1.00

.52

.27

1.10***
.88***
1.06***
1.15***
1.04***
1.04***
.94***

.58
.46
.55
.60
.55
.54
.49

.33
.21
.31
.36
.30
.29
.24

Fall kindergarten generalᵇ

1.00

.96

Spring kindergarten generalᵇ

1.00

.93

Spring first grade generalᵇ
Latent Slope

1.00

.96

Fall kindergarten generalᵇ

0

0

Spring kindergarten generalᵇ

1.00

.19

Spring first grade generalᵇ

3.00

.58

Note. ᵃ Unstandardized factor loading was fixed to equal 1.0 and was not tested

ᵇ Unstandardized factor loadings were fixed to estimate
growth curve. Model fit statistics, X ² (32, N = 12,806) = 506.04, RMSEA =
.034, CFI = .95, TLI = .95. R² value represents the amount of variation for the
observed variable explained by the latent factor. R² fall kindergarten general = .93;
R² spring kindergarten general = .83; R² spring first grade general = .99. * p <

for significance.

.05, ** p < .01, *** p < .001.

115

Table 11
Structural Model Estimates of Academic Socialization as a Predictor of General Knowledge Achievement
Latent factor
1
2
3
4
5
6
7
1. Intercept
.40** / .04
1.02*** /.07
3.28*** / .35 -5.58*** / -.29 -3.72*** / -.21
2. Slope
.02 / .01
-.05 / -.02
.01 / .01
.04 / .01
-.34*** / -.10
3. Beliefs
-.02 / .02
.09** / .05
-.01/ -.01
4. Practices
.12*** / .15
.12*** / .17
-.03 / -.02
-.15*** / -.11
Covariates
5. SES
6. African American
7. Hispanic
Note. The first estimate is unstandardized, and the second is standardized. Columns represent exogenous factors (predictors) and rows
represent endogenous factors (outcomes). Model fit statistics, X ² (32, N = 12,806) = 506.04, RMSEA = .034, CFI = .95, TLI = .95.
Mean intercept = 24.20, Mean slope = 4.12. R² intercept = .324; R² slope = .01; R² beliefs = .003; R² behavior = .076. * p < .05, ** p < .01,
*** p < .001.

116

General Knowledge IRT Score

35
30
25
20
15
10
5
0

Fall K

European American

Spring K
Time of Data Collection
African American

Spring 1st

Hispanic

Figure 10. General knowledge achievement trajectories by race and ethnicity conditional on
socioeconomic status

117

Table 12
Tests of Mediation for Structural Equation Models from Parents’ Beliefs to Achievement Outcomes
Total effect
Unstandardized Standardized
(SE)

Direct effect
Unstandardized Standardized
(SE)

Indirect effect
Standardized
Unstandardized
(SE)

Path
Reading
Belief
intercept
1.928 (.17)***
.135
1.869 (.17)***
.131
.059 (.03)*
.004
Belief
slope
.494 (.11)***
.074
.479 (.11)***
.072
.015 (.018)
.002
Mathematics
Belief
intercept
1.398 (.14)***
.107
1.34 (.14)***
.103
.057 (.02)**
.004
Belief
slope
.274 (.093)**
.054
.280 (.09)**
.055
-.005 (.01)
-.001
General knowledge
Belief
intercept
.518 (.134)***
.046
.391 (.134)**
.035
.127 (.024)***
.011
Belief
slope
.010 (.035)
.005
.016 (.036)
.007
-.006 (.007)
-.003
Note. Effect sizes are standardized regression coefficients and can be interpreted as the standard deviation change in the
outcome variable associated with a change of 1 SD in parent beliefs and behaviors (Hedges, 2008). Indirect effects are based
on a series of mediators; the resulting coefficients are quite small even though each of the contributing path coefficients are
sometimes in the medium range. * p < .05, ** p < .01, *** p < .001.

118

REFERENCES

119

REFERENCES

Ackerman, D. J., & Barnett, W. S. (2005). Prepared for kindergarten: What does readiness
mean? Retrieved from National Institute for Early Education Research website:
http://nieer.org/resources/policyreports/report5.pdf
Alexander, K., Entwisle, D., & Bedinger, S. (1994). When expectations work: Race and
socioeconomic differences in school performance. Social Psychology Quarterly, 57(4),
283–299.
Aulls, M. W., & Sollars, V. (2003). The differential influence of the home environment on the
reading ability of children entering grade 1. Reading Improvement, 40, 164–178.
Barbarin, O. A., Early, D., Clifford, R., Bryant, D., Frome, P., Burchinal, M., et al. (2008).
Parental conceptions of school readiness: Relations to ethnicity, socioeconomic status,
and children's skills. Early Education and Development, 19(5), 671–701.
Barnett, M. A., & Taylor, L. C. (2009). Parental recollections of school experiences and current
kindergarten transition practices. Journal of Applied Developmental Psychology, 30(2),
140–148.
Bast, J., & Reitsma, P. (1997). Matthew effects in reading: A comparison of latent growth curve
models and simplex models with structured means. Multivariate Behavioral Research,
32, 135–167.
Bempechat, J. (1992). The role of parental involvement in children's academic achievement. The
School Community Journal 2(2), 31–41.
Bodovski, K., & Farkas, G. (2007). Mathematics growth in early elementary school: The roles of
beginning knowledge, student engagement, and instruction. Elementary School Journal,
108(2), 115–130.
Booth, A., & Crouter, A. C. (Eds.). (2008). Disparities in school readiness: How families
contribute to transitions into school. New York, NY: Lawrence Erlbaum Associates.
Boyce, L. K., Innocenti, M. S., Roggman, L. A., Norman, V. K., & Ortiz, E. (2010). Telling
stories and making books: Evidence for an intervention to help parents in migrant Head
Start families support their children's language and literacy. Early Education and
Development, 21(3), 343–371.
Bronfenbrenner, U. (1986). Ecology of the family as a context for human development: Research
perspectives. Developmental Psychology, 22, 723–742.
Brooks-Gunn, J., & Markman, L. B. (2005). The contribution of parenting to ethnic and racial
gaps in school readiness. The Future of Children, 15(1), 139–168.
120

Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen
& J. S. Long (Eds.), Testing structural equation models (pp. 136–162). Newbury Park,
CA: Sage.
Bryant, P. E., Maclean, M., Bradley, L., & Crossland, J. (1990). Rhyme and alliterations,
phoneme detection, and learning to read. Developmental Psychology 26, 429–438.
Byrne, B. M. (2011). Structural equation modeling with Mplus. New York, NY: Routledge.
Chen, H., & Lan, W. (1998). Adolescents' perceptions of their parents' academic expectations:
Comparison, of American, Chinese-American, and Chinese high school students.
Adolescence, 33, 385–390.
Claessens, A., Duncan, G., & Engel, M. (2009). Kindergarten skills and fifth-grade achievement:
Evidence form ECLS-K. Economics of Education Review, 28, 415–427.
Coley, R. (2002). An uneven start: Indicators of inequality in school readiness. Retrieved from
Educational Testing Service website:
http://www.ets.org/Media/Research/pdf/PICUNEVENSTART.pdf
Cooper, C. E., Crosnoe, R., Suizzo, M., & Pituch, K. A. (2010). Poverty, race, and parental
involvement during the transition to elementary school. Journal of family Issues, 31(7),
859–883.
Crinic, K., & Lamberty, G. (1994). Reconsidering school readiness: Conceptual and applied
perspectives. Early Education and Development, 5, 91–105.
Cunha, F., Heckman, J., Lochner, L., & Masterov, D. (2006). Interpreting the evidence on life
cycle skill formation. In E. A. Hanushek & F. Welch (Eds.), Handbook of the economics
of education (pp. 697–812). Amsterdam, the Netherlands: North-Holland.
Davis-Kean, P. (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.
de Jong, P. F., & Leseman, P. P. M. (2001). Lasting effects of home litearcy on reading
achievement in school. Journal of School Psychology, 39, 389–414.
Diamond, K. E., Reagan, A. J., & Bandyk, J. E. (2000). Parents' conceptions of kindergarten
readiness: Relationships with race, ethnicity, and development. The Journal of
Educational Research, 94, 93–100.
Dilworth-Bart, J. E., Khurshid, A., & Vandell, D. L. (2007). Do maternal stress and home
environment mediate the relation between early income-to-need and 54-months attention
ability. Infant and Child Development, 16, 525–552.
121

Downey, D. B., Von Hippel, P. T., & Broh, B. A. (2004). Are schools the great equalizer?
Cognitive inequality during the summer months and the school year. American
Sociological Review, 69(5), 613–635.
Duncan, G., & Magnuson, K. (2005). Can family socioeconomic resources account for racial and
ethnic test score gaps? The Future of Children, 15(1), 35–54.
Duncan, T. E., Duncan, S. C., & Strycker, L. A. (2010). An introduction to latent variable
growth curve modeling (2nd ed.). New York, NY: Psychology Press.
Entwisle, D. R., & Alexander, K. L. (1992). Summer setback: Race, poverty, school composition
and mathematics achievement in the first two years of school. American Sociological
Review, 57(1), 72–84.
Entwisle, D. R., & Alexander, K. L. (1994). Winter setback: The racial composition of schools
and learning to read. American Sociological Review, 59(3), 446–460.
Entwisle, D. R., & Alexander, K. L. (1996). Family type and children's growth in reading and
math over the primary grades. Journal of Marriage and Family, 58, 381–355.
Entwisle, D. R., & Alexander, K. L. (1999). Early schooling and social stratification. In R. C.
Pianta & M. J. Cox (Eds.), The transition to kindergarten. Baltimore, MD: Paul H.
Brookes Publishing Company.
Entwisle, D. R., Alexander, K. L., & Olsen, L. S. (2005). First Grade and Educational
Attainment by Age 22: A New Story. American Journal of Sociology, 110(5), 1458–
1502.
Epstein, J. L., & Connor, L. J. (1992). School and Family Partnerships. [ERIC Document
Reproduction Service No. ED347638]. Practitioner, 18(4), 1–8.
Fantuzzo, J., Bulotsky-Shearer, R., McDermott, P., McWayne, C., Frye, D., & Perlman, S.
(2007). Investigation of dimensions of social-emotional classroom behavior and school
readiness for low-income urban preschool children. School Psychology Review, 36, 44–
62.
Farkas, G., & Hibel, J. (2008). Being unready for school: Factors affecting risk and resilience. In
A. Booth & A. Crouter (Eds.), Disparities in School Readiness (pp. 3–30). New York,
NY: Lawrence Erlbaum Associates.
Flora, D.B. & Curran, P.J. (2004). An empirical evaluation of alternative methods of estimation
for confirmatory factor analysis with ordinal data. Psychological Methods, 9, 466–491.

122

Foster, M. A., Lambert, R., Abbott-Shim, M., McCarty, F., & Franze, S. (2005). A model of
home learning environment and social risk factors in relation to children's emergent
literacy and social outcomes. Early Childhood Research Quarterly, 20, 13–36.
Fryer, R. G., & Levitt, S. D. (2004). Understanding the black-white test score gap in the first two
years of school. The Review of Economics and Statistics, 86(2), 447–464.
Fryer, R. G., & Levitt, S. D. (2005). The black-white test score gap through third grade
(Working paper No. w11049). American Law and Economics Review, 8(2), 249–281.
Retrieved from http://www.nber.org/papers/w11049.pdf
Ginsberg, H., Bempechat, J., & Chung, E. (1992). Parent influences on children's mathematics.
In T. Stitcht (Ed.), The integenerational transfer of cognitive skills. Norwood, NJ: Ablex.
Glazer, S. M. (1989). Oral language and litearcy In D. S. Strickland & L. M. Morrow (Eds.),
Emerging literacy: Young children learn to read and write (pp. 16–26). Newark, DE:
International Reading Association.
Glick, J. E., & White, M. J. (2004). Post-secondary school participation of immigrant and native
youth: The role of familial resources and educational expectations. Social Science
Research, 33, 272–299.
Goldenberg, C., Gallimore, R., Reese, L., & Garnier, H. (2001). Cause or effect? A longitudinal
study of immigrant Latino parents' aspirations and expectations, and their children's
school performance. American Educational Research Journal, 38(3), 547–582.
Goodnow, J., & Collins, W. (1990). Development according to parents: The nature, sources, and
consequences of parents' ideas. East Sussex, United Kingdom: Lawrence Erlbaum and
Associates, Ltd., Publishers.
Goodnow, J. J. (1998). Parents' ideas, actions, and feelings: Models and methods from
developmental and social psychology. Child Development, 59, 286–320.
Goyette, K., & Xie, Y. (1999). Educational expectations of Asian American youths:
Determinants and ethnic differences. Sociology of Education, 72, 22–36.
Graue, M. E. (1993). Ready for What? Constructing meanings of readiness for kindergarten.
Albany, NY: State University of New York Press.
Grolnick, W.S. & Slowiaczek, M.L. (1994). Parent's involvement in children's schooling: A
multidimensional conceptualization and motivational model. Child Development, 65,
237–252.
Haines, A. H., Fowler, S. A., Schwartz, I. S., Kottwitz, E., & Rosenhoetter, S. (1989). A
comparison of preschool and kindergarten teacher expectations for school readiness.
Early Childhood Research Quarterly, 4, 75–88.
123

Hair, E., Halle, T., Terry-Humen, E., Lavelle, B., & Calkins, J. (2006). Children's school
readiness in the ECLS-K: Predictions to academic, health, and social outcomes in first
grade. Early Childhood Research Quarterly, 21, 431–454.
Halle, T., Kurtz-Costes, B., & Mahoney, J. (1997). Family influences on school achievement in
low-income, African American children. Journal of Educational Psychology, 89, 527–
537.
Hancock, G. R., & Mueller, R. O. (2010). Structural equation modeling. In G. A. Hancock & R.
O. Mueller (Eds.), The reviewer's guide to quantitative methods in the social sciences.
New York Routledge.
Harradine, C. C., & Clifford, R. M. (April, 1996). When are children ready for kindergarten?
Views from families, kindergarten teachers, and child care providers. Paper presented at
the American Education Research Association, New York, NY. Retrieved from
http://www.eric.ed.gov/PDFS/ED399044.pdf
Hart, B., & Risley, T. (1995). Meaningful differences in everyday experiences of young
American children. Baltimore, MD: Paul H. Brookes Publishing Co.
Heath, S. B. (1983). Ways with words. New York, NY: Cambridge University.
Heckman, J. (2011). The American family in black and white: A post-racial strategy for
improving skills to promote equality. Daedalus, 140(2),1–20.
Hedges, L. V. (2008). What are effect sizes and why do we need them? Child Development
Perspectives 2(3), 167–171.
Hess, R. D., Holloway, S. D., Dickson, W. P., & Price, G. G. (1984). Maternal variables as
predictors of children's school readiness and later achievement in vocabulary and
mathematics in sixth grade. Child Development, 55(5), 1902–1912.
Hess, R. D., Kahiwagi, K., Azuma, H., Price, G. G., & Dickson, W. P. (1980). Maternal
expectations for mastery of developmental tasks in Japan and the United States.
International Journal of Psychology, 15, 259–271.
Hill, N. E. (2001). Parenting and academic socialization as they relate to school readiness: The
roles of ethnicity and family income. Journal of Educational Psychology, 93(4), 686–
697.
Hill, N.E., & Craft, S.A. (2003). Parent-school involvement and school performance: Mediated
pathways among socioeconomically comparable African-American and Euro-American
families. Journal of Educational Psychology, 95, 74–83.

124

Hill, N. E., & Taylor, L. C. (2004). Parental school involvement and children's academic
achievement: Pragmatics and issus. Current Directions in Psychological Science, 13(4),
161–164.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.
Jencks, C., & Phillips, M. (1998). The Black-White test score gap: An Introduction. In C. Jencks
& M. Phillips (Eds.), The Black-White test score gap. Washington DC: Brookings
Institution Press.
Keels, M. (2009). Ethnic group differences in early had start parent's parenting beliefs and
practices and links to children's early cognitive development. Early Childhood Research
Quarterly, 24, 381–397.
Kinlaw, C. R., Kurtz-Costes, B., & Goldman-Fraser, J. (2001). Mothers' achievement beliefs and
behaviors and their children's school readiness: A cultural comparison. Applied
Developmental Psychology, 22, 493–506.
Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New
York, NY: Guilford Press.
Kohn, M. L. (1969). Class and conformity: A study in values. Homewood, IL: Dorsey.
Lamb Parker, F., Boak, A. Y., Griffin, K. W., Ripple, C., & Peavy, L. (1999). Parent-child
relationship, home learning environment, and school readiness. School Psychology
Review, 28(3), 413–425.
Landry, S. H., & Smith, K. E. (2008). Family processes that support school readiness: Specific
behaviors and contextual conditions that set this process in motion. In A. Booth & A. C.
Crouter (Eds.), Disparities in School Readiness. New York, NY: Lawrence Erlbaum
Associates.
Landry, S. H., Smith, K. E., & Swank, P. R. (2003). The importance of parenting during early
childhood for school-age development. [Proceedings Paper]. Developmental
Neuropsychology, 24(2-3), 559–591.
Lareau, A. (2003). Unequal childhoods: Class, race, and family life. Berkeley, CA: University of
California Press.
Lee, V., & Burkham, D. (2002). Inequality at the starting gate. Washington DC: Economic
Policy Institute.
Levine, R. A. (1988). Human parental care: Universal goals, cultural strategies, individual
behavior. In R. A. LeVine, P. M. Miller & M. M. West (Eds.), Parental behavior in
diverse societies. San Francisco, CA: Jossey-Bass.
125

Li-Grining, C. P., Votruba-Drzal, E., Maldonado-Carreno, C., & Haas, K. (2010). Children's
early approaches to learning and academic trajectories through fifth grade.
Developmental Psychology, 46(5), 1062–1077.
Maccoby, E. E. (1992). The role of parents in the socialization of children: An historical
overview. Developmental Psychology, 28, 1006–1017.
Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the family: Parent-child
interaction. In P. H. Mussen (Ed.), Handbook of Child Psychology (pp. 1–101). New
York: Wiley.
Maclean, M., Bryant, P. E., & Bradley, L. (1987). Rhymes, nursery rhymes, and reading in early
childhood. Merrill-Palmer Quarterly, 33, 255–281.
McWayne, C., Fantuzzo, J., & McDermott, P. (2004). Preschool competency in context: An
investigation of the unique contribution of child competencies to early academic success.
Developmental Psychology, 40, 633–645.
Miller, S. (1988). Parents' beliefs about children's cognitive development. Child Development,
59, 259–285.
Milne, A., & Plourde, L. A. (2006). Factors of a low-SES household: What aids academic
achievement. Journal of Instructional Psychology, 33, 183–193.
Mistry, R. S., Benner, A. D., Biesanz, J. C., Clark, S. L., & Howes, C. (2010). Family and social
risk, and parental investments during the early childhood years and predictors of lowincome children's school readiness outcomes. Early Childhood Research Quarterly, 25,
432–449.
Mueller, R. O., & Hancock, G. R. (2010). Structural equation modeling. In G. A. Hancock & R.
O. Mueller (Eds.), The reviewer's guide to quantitative methods in the social sciences.
New York Routledge.
Murphey, D. (1992). Constructing the child: Relations between parents' beliefs and child
outcomes. Developmental Review, 12, 199–232.
Muthén, B. O. (1984). A general structural equation model with dichotomous, ordered
categorical, and continuous latent variable indicators. Psychometrika, 49, 115–132.
Muthén, B. O., & Asparouhov, T. (2002). Latent variable analysis with categorical outcomes:
Multiple-group and growth modeling in Mplus. Retrieved from
www.statmodel.com/download/webnotes/CatMGLong.pdf
Muthén, L. K., & Muthén, B. O. (1998-2007). Mplus User’s Guide. (5th ed.). Los Angeles, CA:
Muthén & Muthén.
126

National Center for Education Statistics. (2000). User's manual for the ECLS-K first grade
public-use data files and electronic code book. Washington, DC: U.S. Department of
Education, Office of Educational Research and Improvement.
Newman, D. (2003). Longitudinal modeling with randomly and systematically missing data: A
simulation of ad hoc, maximum likelihood, and multiple imputation techniques.
Organizational Research Methods, 6(3), 328–362.
Okagaki, L., & Frensch, P. A. (1998). Parenting and children's school achievement: A
multiethnic perspective. American Educational Research Journal, 35(1), 123–144.
Okagaki, L., & Sternberg, R. (1993). Parental beliefs and children's school performance. Child
Development, 64, 36–56.
Olds, D. L., & Kitzman, H. (1993). Review of research on home visiting for pregnant women
and parents of young children. The Future of Children, 3(3), 53–92.
Olds, D. L., Kitzman, H., Hanks, C., Cole, R., Anson, E., Sidora-Arcoleo, K., et al. (2007).
Effects of nurse home visiting on maternal and child functioning: Age-9 follow-up of a
randomized trial. Pediatrics, 120(4), 832-845.
Orr, A. J. (2003). Black-white differences in achievement: The importance of wealth. Sociology
of Education, 76, 281–304.
Pearce, R. R. (2006). Effects of cultural and social structural factors on the achievement of white
and Chinese American students at school transition points. American Educational
Research Journal, 43(1), 75–101.
Phillips, M. (2011). Parenting, time use, and disparities in academic outcomes. In G. Duncan &
R. Murnane (Eds.), Whither Opportunity? (pp. 207–228). New York, NY: Russell Sage
Foundation.
Phillips, M., Brooks-Gunn, J., Duncan, G., Klebanov, P. K., & Crane, J. (1998). Family
background, parenting practices, and the Black-White test score gap. In C. Jencks & M.
Philips (Eds.), The Black-White test score gap. Washington, DC: Brookings Institution
Press.
Pianta, R. C., Cox, M. J., & Snow, K. L. (2007). School Readiness & the Transition to
Kindergarten in the Era of Accountability. Baltimore, MD: Paul H. Brookes Publishing
Co.
Pianta, R.C. & Walsh, D.J. (1996). High-risk children in schools: Constructing sustaining
relationships. New York, NY: Routledge.

127

Piotrkowski, C. S. (2004). A community-based approach to school readiness in Head Start. In E.
Zigler & S. J. Styfco (Eds.), The Head Start debates. Baltimore, MD: Paul H. Brookes
Publishing Co.
Piotrkowski, C. S., Botsko, M., & Mathews, E. (2000). Parents' and teachers' beliefs about
children's school readiness in a high-need community. Early Childhood Research
Quarterly, 15, 537–558.
Raykov, T., & Marcoulides, G. A. (2006). A first course in structural equation modeling (2nd
ed.). Mahwah, NJ: Lawrence Erlbaum Associates, Inc., Publishers.
Reardon, S. (2011). The widening academic achievement gap between the rich and the poor:
New evidence and possible explanations. In G. Duncan & R. Murname (Eds.), Whither
Opportunity? (pp. 507–522). New York, NY: Russell Sage Foundation.
Reardon, S., & Gallindo, C. (2009). The Hispanic-White achievement gap in math and reading in
the elementary grades. American Educational Research Journal, 46(3), 853–891.
Reardon, S. F., & Robinson, J. P. (2008). Patterns and trends in racial/ethnic and socioeconomic
academic achievement gaps. In H. Ladd & E. Fiske (Eds.), Handbook of Research in
Education Finance and Policy (pp. 497–516). New York, NY: Routledge.
Reynolds, A. J., & Gill, S. (1994). The role of parental perspectives in the school adjustment of
inner-city black children. Journal of Youth and Adolescence, 23(6), 671–694.
Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. New
York, NY: Oxford University Press.
Rosenthal, D. A., & Gold, R. (1989). A comparison of Vietnamese-Australian and AngloAustralian mothers' beliefs about intellectual development. International Journal of
Psychology, 24, 179–193.
Rouse, C., Brooks-Gunn, J., & Mclanahan, S. (2005). School readiness: Closing racial and ethnic
gaps. The Future of Children, 15(1), 5–14.
Schweinhart, L. J., Montie, J., Xiang, Z., Barnett, W. S., Belfield, C. R., & Nores, M. (2005).
Lifetime effects: The High/Scope Perry preschool study through age 40. Ypsilanti, MI:
High/Scope Press.
Scott-Jones, D. (1995). Parent-child interactions and school achievement. In B. A. Ryan, G. R.
Adams & T. P. Gullotta (Eds.), The family-school connection: Theory, research, &
practice (pp. 75–107). Thousand Oaks, CA: Sage.
Senechal, M., & Young, L. (2008). The effect of family literacy interventions on children's
acquisition of reading from kindergarten to grade 3: A meta-analytic review. Review of
Education 78(4), 880–907.
128

Singer, J. D., & Willet, J. B. (2003). Applied longitudinal data analysis: Modeling change and
event occurence. New York, NY: Oxford University Press.
Smith-Maddox, R. (1999). The social networks and resources of African American eighth
graders: Evidence from the National Educational Longitudinal Study of 1988.
Adolescence, 34, 169–183.
Smith, J. R., Brooks-Gunn, J., & Klebanov, P. K. (1997). Consequences of living in poverty for
young children's cognitive and verbal ability and early school achievement. In G. Duncan
& J. Brooks-Gunn (Eds.), Consequences of Growing Up Poor (pp. 132–188). New York,
NY: Russell Sage Foundation.
Snow, K. (2006). Measuring school readiness: Conceptual and practical considerations. Early
Education and Development, 17, 7–41.
Sonnenschein, S., Brody, G., & Musterman, K. (1996). The influence of family beliefs and
practices on children's early reading development. In L. Baker, P. Afferbach & D.
Reinking (Eds.), Developing engaged readers in school and home communities (pp. 3–
20). Mahway, NJ: Lawrence Erlbaum Associates, Publishers.
Stanovich, K. E. (1986). Matthew effects in reading: Some consequences of individual
differences in the acquisition of literacy. Reading Research Quarterly, 21(4), 360–407.
StataCorp. (2009). Stata: Release 11. Statistical Software. College Station, TX: StataCorp LP.
Stormshak, E. A., Kaminski, R. A., & Goodman, M. R. (2002). Enhancing the parenting skills of
Head Start families during the transition to kindergarten. Prevention Science, 3(3), 223–
234.
Suizzo, M. & Stapleton, L. (2007): Home‐based parental involvement in young children’s
education: Examining the effects of maternal education across U.S. ethnic groups.
Educational Psychology: An International Journal of Experimental Educational
Psychology, 27(4), 533–556.
Sy, S. R., & Schulenberg, J. E. (2005). Parent beliefs and children's achievement trajectories
during the transition to school in Asian American and European American families.
International Journal of Behavioral Development, 29(6), 505–515.
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.
Votruba-Drzal, E. (2006). Economic disparities in middle childhood development: Does income
matter? Developmental Psychology, 42, 1154–1167.
129

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes.
Cambridge, MA: Harvard University Press.
Wentzel, K. R. (1998). Parents' aspirations for children's educational attainments: Relations to
parental beliefs and social address variables. Merrill-Palmer Quarterly, 44, 20–37.
Wertsch, J. V. (1979). From social interaction to higher psychological processes: A clarification
and application to Vygotsky's theory. Human Development, 22, 1–22.
West, J., Germino-Hausken, E., & Collins, M. (1995). Readiness for kindergarten: Parent and
teacher beliefs (Statistics in brief). Washington, DC: National Center for Education
Statistics.
Zill, N., & West, J. (2001). Entering kindergarten: A portrait of American children when they
begin school: Findings from the Condition of Education 2000. Washington, DC: National
Center for Education Statistics.

130