School Readiness and Later Achievement

CORRECTED NOVEMBER 30, 2007; SEE LAST PAGE
Developmental Psychology
Copyright 2007 by the American Psychological Association
2007, Vol. 43, No. 6, 1428 –1446
0012-1649/07/$12.00
DOI: 10.1037/0012-1649.43.6.1428
School Readiness and Later Achievement
Greg J. Duncan
Chantelle J. Dowsett
Northwestern University
University of Texas at Austin
Amy Claessens
Katherine Magnuson
Northwestern University
University of Wisconsin–Madison
Aletha C. Huston
Pamela Klebanov
University of Texas at Austin
Princeton University
Linda S. Pagani
Leon Feinstein
Universite´ de Montre´al
University of London
Mimi Engel
Jeanne Brooks-Gunn
Northwestern University
Columbia University
Holly Sexton
Kathryn Duckworth
University of Michigan
University of London
Crista Japel
Universite´ de Que´bec a` Montre´al
Using 6 longitudinal data sets, the authors estimate links between three key elements of school
readiness—school-entry academic, attention, and socioemotional skills—and later school reading and
math achievement. In an effort to isolate the effects of these school-entry skills, the authors ensured that
most of their regression models control for cognitive, attention, and socioemotional skills measured prior
to school entry, as well as a host of family background measures. Across all 6 studies, the strongest
predictors of later achievement are school-entry math, reading, and attention skills. A meta-analysis of
the results shows that early math skills have the greatest predictive power, followed by reading and then
attention skills. By contrast, measures of socioemotional behaviors, including internalizing and exter-
nalizing problems and social skills, were generally insignificant predictors of later academic perfor-
mance, even among children with relatively high levels of problem behavior. Patterns of association were
similar for boys and girls and for children from high and low socioeconomic backgrounds.
Keywords: school readiness, socioemotional behaviors, attention, early academic skills
Supplemental materials: http://dx.doi.org/10.1037/[0012-1649.43.6.1428].supp
Greg J. Duncan, Amy Claessens, and Mimi Engel, School of Education
meetings of the Society for Research on Child Development, Atlanta,
and Social Policy, Northwestern University; Chantelle J. Dowsett and
Georgia, April 2005. We are grateful to the National Science
Aletha C. Huston, Department of Human Ecology, University of Texas at
Foundation-supported Center for the Analysis of Pathways from Child-
Austin; Katherine Magnuson, Department of Social Work, University of
hood to Adulthood (CAPCA; Grant 0322356) for research support. We
Wisconsin–Madison; Pamela Klebanov, Center for Research on Child
thank Larry Aber, Mark Appelbaum, Avshalom Caspi, David Cordray,
Wellbeing, Princeton University; Linda S. Pagani, Department of Psycho-
Herbert Ginsburg, David Grissmer, Mark Lipsey, Derek Neal, Cybele
education, Universite´ de Montre´al, Que´bec, Canada; Leon Feinstein and
Raver, Arnold Sameroff, Robert Siegler, Ross Thompson, Sandra Jo
Kathryn Duckworth, Department of Quantitative Social Science, Institute
Wilson, Nicholas Zill, and other members of CAPCA and the
of Education, University of London, London, England; Jeanne Brooks-
MacArthur Network on Families and the Economy for helpful com-
Gunn, Department of Pediatrics, Columbia University; Holly Sexton, Re-
ments.
seach Center for Group Dynamics, University of Michigan; Crista Japel,
Correspondence concerning this article should be addressed to Greg J.
De´partement d’e´ducation et formation spe´cialise´es, Universite´ de Que´bec
Duncan, School of Education and Social Policy, Northwestern University,
a` Montre´al, Que´bec, Canada.
2046
Sheridan
Road,
Evanston,
IL
60208.
E-mail:
greg-
A preliminary version of this article was presented at the biennial
duncan@northwestern.edu
1428

---

SCHOOL READINESS AND LATER ACHIEVEMENT
1429
Early childhood programs and policies that promote academic
attention, inhibit impulsive behavior, and relate appropriately to
skills have been gaining popularity among politicians and re-
adults and peers may be able to take advantage of the learning
searchers. For example, President George W. Bush (2002) en-
opportunities in the classroom, thus more easily mastering reading
dorsed Head Start reforms in 2002 that focus on building early
and math concepts taught in elementary school. For these reasons,
academic skills, observing that “on the first day of school, children
the skills children possess when entering school might result in
need to know letters and numbers. They need a strong vocabulary.
different achievement patterns in later life. If achievement at older
These are the building blocks of learning, and this nation must
ages is the product of a sequential process of skill acquisition, then
provide them” (p. 12). The National Research Council’s Commit-
strengthening skills prior to school entry might lead children to
tee on the Prevention of Reading Difficulties in Young Children
master more advanced skills at an earlier age and perhaps even
recommends providing environments that promote preliteracy
increase their ultimate level of achievement.
skills for all preschool children (Snow, Burns, & Griffin, 1998).
Although there are strong theoretical reasons to expect that
Similarly, the National Association for the Education of Young
individual differences in children’s early academic skills and be-
Children and the National Council of Teachers of Mathematics
havior are linked to subsequent behavior and achievement, sur-
(2002) issued a joint statement that advocated for high-quality
prisingly little rigorous research has been conducted to test this
mathematics education for children ages 3– 6.
hypothesis. Consequently, the purpose of this article is to assess as
Others, however, maintain that a broad constellation of behav-
precisely as possible, using six longitudinal, nonexperimental data
iors and skills enables children to learn in school. Asked to identify
sets, the association between skills and behaviors that emerge
factors associated with a difficult transition to school, kindergarten
during the preschool years and later academic achievement. As
teachers frequently mentioned weaknesses in academic skills,
with Robins’s (1978) classic study of adult antisocial behavior, our
problems with social skills, trouble following directions, and dif-
approach consists of comparable analyses of a number of different
ficulty with independent and group work (Rimm-Kaufman, Pianta,
longitudinal developmental studies.1 We are especially interested
& Cox, 2000). Researchers too have made this point. The National
in identifying academic, attention, and socioemotional skills and
Research Council and Institute on Medicine argued that “the
behaviors that may be learned or improved through experiences
elements of early intervention programs that enhance social and
prior to school entry. In the following sections, we draw from
emotional development are just as important as the components
developmental literature to identify key dimensions of school
that enhance linguistic and cognitive competence” (Shonkoff &
readiness and to derive theoretical predictions about how chil-
Phillips, 2000, pp. 398 –399).
dren’s school-entry skills and behaviors contribute to short- and
These two views have emerged in the current debate about what
long-term academic success.
constitutes school readiness and in particular about what skills
predict school achievement. Many early education programs, in-
Associations Between Early Skills and Later Achievement
cluding Head Start, are designed to enhance children’s physical,
intellectual, and social competencies on the grounds that each
Academic achievement is a cumulative process involving both
domain contributes to a child’s overall developmental competence
mastering new skills and improving already existing skills
and readiness for school. However, if early acquisition of specific
(Entwisle & Alexander, 1990; Pungello, Kuperschmidt, Burchinal,
academic skills or learning-enhancing behaviors forecasts later
& Patterson, 1996). Information about how children acquire read-
achievement, it may be beneficial to add domain-specific early
ing and math skills points to the importance of specific academic
skills to the definition of school readiness and to encourage inter-
skills but also indicates that more general cognitive skills, partic-
ventions aimed at promoting these skills prior to elementary
ularly oral language and conceptual ability, may be increasingly
school. Thus, understanding which skills are linked to children’s
important for later mastery of more complex reading and mathe-
academic achievement has important implications for early edu-
matical tasks. Basic oral language skills become critical for un-
cation programs.
derstanding texts as the level of difficulty of reading passages
We adopted a child-centered model of school transition, which
increases (NICHD Early Child Care Research Network, 2005b;
is nested within a broader ecological framework but focuses on the
Scarborough, 2001; Snow et al, 1998; Storch & Whitehurst, 2002;
set of individual skills and behaviors that children have acquired
Whitehurst & Lonigan, 1998). Likewise, mastery of foundational
prior to school entry (Rimm-Kaufman & Pianta, 2000). A child’s
concepts of numbers allows for a deeper understanding of more
individual characteristics contribute to the environments in which
complex mathematical problems and flexible problem-solving
the child interacts and the rate at which the child may learn new
techniques (Baroody, 2003; Ferrari & Sternberg, 1998; Hiebert &
skills; in turn, the child receives feedback from others in the
Wearne, 1996).
environment (Meisels, 1998). Thus, because they affect both the
Although children’s academic achievement is largely stable
child and the social environment, early academic skills and socio-
throughout childhood, children do demonstrate both transitory
emotional behaviors are linked to subsequent academic achieve-
fluctuations and fundamental shifts in their achievement trajecto-
ment because they provide the foundation for positive classroom
ries (Kowaleski-Jones & Duncan, 1998; Pungello et al., 1996).
adaptation (Cunha, Heckman, Lochner, & Masterov, 2006;
Nonexperimental data show that children’s achievement test
Entwisle, Alexander, & Olson, 2005).
For example, a child who enters kindergarten with rudimentary
1 Robins (1978) justified her approach as follows: “In the long run, the
academic skills may be poised to learn from formal reading and
best evidence for the truth of any observation lies in its replicability across
mathematics instruction, receive positive reinforcement from the
studies. The more the populations studied differ, the wider the historical
teacher, or be placed in a higher ability group that facilitates the
eras they span; the more the details of the methods vary, the more
acquisition of additional skills. Similarly, a child who can pay
convincing becomes that replication” (p. 611).

---

1430
DUNCAN ET AL.
scores are related to prior cognitive functioning and the attainment
dence that high-quality programs for preschool children “at risk”
of basic skills in math and literacy such as number and letter
for school failure produce gains in cognitive and academic skills
recognition (Stevenson & Newman, 1986). In their meta-analysis,
and reduce behavior problems (Conduct Problems Prevention Re-
La Paro and Pianta (2000) found middle-range correlations in
search Group, 2002; Karoly, Kilburn, & Cannon, 2005; Love et al.,
cognitive/academic skills both from preschool to kindergarten
2003). Early educational interventions have also been found to
(.43) and from kindergarten to first or second grade (.48).
result in long-term reductions in special education services, grade
Attention-related skills such as task persistence and self-
retention, and increases in educational attainment (Campbell,
regulation are expected to increase the time during which children
Ramey, Pungello, Sparling, & Miller-Johnson, 2002; Lazar et al.,
are engaged and participating in academic endeavors. Research has
1982; Reynolds & Temple, 1998).
shown that signs of attention and impulsivity can be detected as
As is the case with nonexperimental studies, few intervention
early as age 2.5 but continue to develop until reaching relative
studies are designed to isolate the relative contributions of changes
stability between ages 6 and 8 (Olson, Sameroff, Kerr, Lopez, &
in achievement, attention, and behavior to later school achieve-
Wellman, 2005; Posner & Rothbart, 2000). Studies linking atten-
ment. A first problem is that behavioral interventions tend to
tion with later achievement are less common, but consistent evi-
measure behavioral but not achievement outcomes, whereas read-
dence suggests that the ability to control and sustain attention as
ing and math interventions tend to measure achievement but not
well as participate in classroom activities predicts achievement test
behavioral outcomes. Interesting exceptions are a small number of
scores and grades during preschool and the early elementary
experimental behavior-based interventions that tested for achieve-
grades (Alexander, Entwisle, & Dauber, 1993; Raver, Smith-
ment impacts (Coie & Krehbiel, 1984; Dolan et al., 1993). For
Donald, Hayes, & Jones, 2005). These attention skills, which are
example, a random-assignment evaluation of a behavioral inter-
conceptually distinct from other types of interpersonal behaviors,
vention targeting both aggressive and shy behaviors among first
are associated with later academic achievement, independent of
graders found short-run improvements in both teacher and peer
initial cognitive ability (McClelland, Morrison, & Holmes, 2000;
reports of aggressive and shy behavior but no crossover impacts on
Yen, Konold, & McDermott, 2004) and of prior reading ability and
reading achievement (Dolan et al., 1993; Kellam, Mayer, Rebok,
current vocabulary (Howse, Lange, Farran, & Boyles, 2003). Ex-
& Hawkins, 1998). Given evidence, albeit limited, that behavioral
amining attention separately from externalizing problems has clar-
interventions succeed at improving behavior but not achievement,
ified the role of each in achievement, suggesting that attention is
behavior would appear to play a limited role in academic success.
more predictive of later achievement than more general problem
A second problem is that many intervention programs target
behaviors (Barriga et al., 2002; Hinshaw, 1992; Konold & Pianta,
both children’s academic skills and their socioemotional behav-
2005; Ladd, Birch, & Buhs, 1999; Normandeau & Guay, 1998;
iors, rendering it impossible to assess their separate impacts
Trzesniewski, Moffitt, Caspi, Taylor, & Maughan, 2006).
through simple experimental contrasts. For example, the Fast
Children’s socioemotional skills and behaviors are also ex-
Track prevention program provided a number of services to chil-
pected to affect both individual learning and classroom dynamics.
dren who were identified as disruptive in kindergarten, including
Inadequate interpersonal skills promote child–teacher conflict and
direct tutoring in reading skills in first grade (Conduct Problems
social exclusion (Newcomb, Bukowski, & Pattee, 1993; Parker &
Prevention Research Group, 1992; 2002). It is possible to estimate
Asher, 1987), and these stressors may reduce children’s participa-
nonexperimental mediated models to determine whether program
tion in collaborative learning activities and adversely affect
effects are more likely to be due to children’s improved achieve-
achievement (Ladd et al., 1999; Pianta & Stuhlman, 2004). Cor-
ment, attention, or behavior skills (e.g., Reynolds, Ou, & Topitzes,
relational evidence linking problem behaviors to academic
2004). This is rarely done, however.
achievement is found in the Beginning School Study. First-grade
ratings on items describing a cheerful, outgoing temperament
The Present Study
(roughly the opposite of internalizing problems) predicted adult
educational attainment better than preschool or first-grade achieve-
This study builds on previous school readiness research in
ment scores (Entwisle et al., 2005). Other studies yield similar
several ways. First, the scope of the study is unprecedented. We
results. For example, children with consistently high levels of
estimated a carefully specified set of models with data from six
aggression from ages 2–9 were more likely than other children to
large-scale longitudinal studies, two of which were nationally
have achievement problems in third grade (NICHD Early Child
representative of U.S. children, whereas two were drawn from
Care Research Network, 2004).
multisite studies of U.S. children, with one each focusing on
children from Great Britain and Canada. Second, we included as
Experimental Evidence and Crossover Effects
predictors a wide representation of school readiness indicators
used in previous research and carefully distinguished between
Many nonexperimental studies find associations between early
related but conceptually distinct skills (e.g., oral language vs.
achievement, attention, and behavior and later achievement, yet
preliteracy skills, attention vs. externalizing problems) wherever
few of these studies are designed to determine which of these skills
possible. Third, we examined multiple dimensions of academic
can be modified prior to school entry or whether these changes
achievement outcomes, including grade completion and math and
predict later achievement. In theory, intervention research should
reading achievement as assessed by both teacher ratings and test
shed light on this gap by demonstrating ways to improve children’s
scores. Fourth, we implemented rigorous analytic methods that
skills and by testing whether improvements in early skills are
attempted to isolate the effects of school-entry academic, attention,
associated with better adjustment in the long term. Indeed, a small
and socioemotional skills by controlling for an extensive set of
number of experimental interventions provide encouraging evi-
prior child, family, and contextual influences that may have been

---

SCHOOL READINESS AND LATER ACHIEVEMENT
1431
related to children’s achievement. Finally, we assessed whether the
(Entwisle et al., 2005; Miech, Essex, & Goldsmith, 2001; Raver,
predictive power of school readiness components differs by gender
2004).
or socioeconomic status, which would indicate that some children
Despite differences in children’s behavior linked to gender and
are at heightened risk of low achievement.
family socioeconomic status, few studies have considered whether
We tested a number of hypotheses related to how school-entry
gender and socioeconomic status moderate the association be-
academic, attention, and socioemotional skills are associated with
tween these early skills and behaviors and subsequent achieve-
later school achievement. Developmental theory suggests that chil-
ment. We expected early academic skills, attention, and socioemo-
dren’s informal, intuitive knowledge of early language and math
tional behaviors to matter more for these subgroups, particularly
concepts plays an important role in the acquisition of more com-
when children enter school with very low levels of these skills.
plex skills formally taught in elementary school (Adams, Treiman,
To estimate the associations between early academic skills and
& Pressley, 1998; Baroody, 2003; Griffin, Case, & Capodilupo,
socioemotional behaviors and later school achievement, we sum-
1995; Tunmer & Nesdale, 1998). Theoretically, children’s atten-
marize results from a coordinated series of analyses across six
tion and socioemotional skills should also affect achievement
longitudinal data sets in two ways. First, we relate early academic,
because they influence children’s engagement in learning activities
attention, and socioemotional skills to later achievement in each of
and facilitate (or disrupt) classroom processes (Ladd, Birch, &
the six data sets and provide a basic summary of these results.
Buhs, 1999; Pianta & Stuhlman, 2004). However, some scholars
Second, we formally summarize the findings from these studies in
point out that it is important to distinguish between behaviors that
a meta-analysis, again focusing on the extent to which this collec-
are directly relevant for learning, such as attention, and those that
tion of early skills predicts later achievement.
may be correlated with attention but are less likely to be directly
linked with achievement, such as interpersonal skills and problem
Method
behavior (Alexander et al., 1993; Cooper & Farran, 1991; McClel-
In this section, we describe the data sets used in this study and
land et al., 2000; McWayne, Fantuzzo, & McDermott, 2004).
the common analytic procedures that were implemented across
Therefore, we expected early academic and attention-related skills
studies. Detailed information about the measures, descriptive sta-
to predict subsequent test scores and teacher achievement ratings,
tistics, and regression results from each study is presented in
and we expected attention skills to predict achievement more
Appendices A–F, which can be found online. As the goal of our
consistently than do socioemotional behaviors.
study was to relate early academic, attention, and socioemotional
In seeking a better understanding of the extent to which our
skills and behaviors to later achievement, each data set has mea-
broad set of early skills is associated with later achievement, it is
sures of these constructs, although there is variation across the
important to consider how outcomes are being measured. Although
studies with respect to when and how each skill or behavior is
test performance provides a key independent assessment of aca-
assessed.
demic achievement, teacher ratings also lend insight into chil-
Table 1 provides an overview of data sources and measures
dren’s everyday performance in the classroom. Teachers’ evalua-
available in each study. All six data sets provide measures of
tions are probably based on a broad picture of children’s
children’s academic skills as well as assessments of attention and
accomplishments, which include their academic skills as well as
socioemotional behaviors at about age 5 or 6. Because most
whether they complete assignments on time, work independently,
children enter elementary school at this age, we refer to the timing
get along with others, and show involvement in the learning
of these measures as school entry but alert the reader that the
agenda of the classroom. Moreover, previous research has found
precise timing varies considerably across studies. To facilitate
that children’s behavior can play a role that is equal to, if not
comparison of findings across studies, we standardized all mea-
greater than, prior cognitive ability in predicting teacher-rated
sures to have a mean of 0 and standard deviation of 1.
attainment or achievement (Lin, Lawrence, & Gorrell, 2003;
We measured achievement outcomes using teachers’ reports,
Schaefer & McDermott, 1999) and academic skills (National Cen-
test scores, and grade retention in early elementary school and, in
ter for Education Statistics, 1993). Consequently, we expected a
some studies, middle childhood. In terms of the timing of the
stronger relationship between school-entry socioemotional behav-
measurement of achievement outcomes, the children of the Na-
iors and subsequent teacher-rated achievement than with subse-
tional Longitudinal Survey of Youth (NLSY) measures are as-
quent test scores.
sessed as late as early adolescence, the National Institute of Child
Although many previous studies have examined the association
Health and Human Development Study of Early Child Care and
between early academic, attention, and socioemotional skills and
Youth Development (NICHD SECCYD) as late as fifth grade, and
subsequent achievement, few have systematically considered the
the 1970 British Birth Cohort Study (BCS) at age 10, whereas none
extent to which these associations differ by gender (Trzesniewski
of the other studies measures achievement beyond third grade. As
et al., 2006). On average, boys receive poorer grades and have
for measurement methods, two studies have both test-score-based
more difficulties related to school progress (e.g., grade retention,
and teacher reports of reading and mathematics achievement (the
special education, and drop out) than do girls (Dauber, Alexander,
Early Childhood Longitudinal Study–Kindergarten Cohort [ECLS-
& Entwisle, 1993; McCoy & Reynolds, 1999), and these differ-
K] and NICHD SECCYD).
ences are especially pronounced among low-income children (Hin-
We measured attention and socioemotional behaviors on the
shaw, 1992). Children from low-income families enter school with
basis of mothers’ reports, teachers’ reports, and observation. Table
lower mean academic skills, and the gap tends to increase during
1 provides an overview of the similarities and differences in
the school years (Lee & Burkam, 2002). These groups also have
measurement across the six studies. One of our data sets, the Infant
higher rates of problems with attention and externalizing behavior
Health and Development Program (IHDP), has observer reports of

---

1432
DUNCAN ET AL.
Test:
word
and
Math
and
10
of
.96)
arithmetic,
algebra,
statistics
use
5
of
fractions,
Test:
age
Bristol
Maternal
and
e
Inattention
Reading
(
g
sequencing,
of
A
intelligence
.67):
BCS
Rules
value,
.93)
Picture
Scale:
Self-completion,
recognition,
understanding
syntax,
comprehension,
retention
Test:
place
measurement,
geometry,
(
Vocabulary
Verbal
(
Report
Edinburgh
University
English
Rutter
(
and
.72
(
.94):
Split-
.85
3
A
etc.
Test:
etc.
French,
oral
and
Test:
sequence
Report
seems
Report
(
division,
kindergarten
week
concentrates,
Grade
of
of
Vocabulary
Forms
.66
respectively;
1
Use/
counting,
decimals
adaptation:
B,
at
Child
Report
sequence,
subtraction,
senior
number
number
Child
Teacher
impulsive,
Teacher
Knowledge
Picture
and
Knowledge
attentively,
MLEPS
Skills:
and
(PPVT),
reliability
French
A
.82):
.90):
understanding
characteristics
communication
Teacher
Number
addition,
multiplication,
fractions,
Test
B,
half
for
test–retest
Informal
knowledge,
addition,
listens
agitated,
Verbal
Number
Junior
Peabody
Number
Attention:
Hyperactive:
8
.90s)
Scale
Broad
5
Verbal
Profile,
(
age
Broad
(
.94)
Child
Preschool
of
of
(
Age
Primary
Maternal
(
IHDP
Intelligence
and
of
(WPPSI):
IQ
Behavior
Attention
problems
.61):
Report
Behaviors
Tests
Achievement—
Revised:
Reading
.90s)
Tests
Achievement—
Revised:
Math
Woodcock–Johnson
Woodcock–Johnson
Wechsler
Achenbach
5
Fall
.84)
Scale–
Grade
Psycho-
Reading:
Report
Teacher
Ratings
Report
.84)
Checklist
.93):
Word
Ratings
Letter-Word
(
Socioemotional
Battery—
4.5
Performance
Applied
Applied
(
Performance
(
(WJ-R)
Attention
Skills
Skills
Language
and
SECCYD
Teacher
Teacher
Age
Attention
Behavior
Impulsivity
Math:
Reading:
Expressive
Math:
Kindergarten
Continuous
Educational
Revised
Letter-Word,
Attack
Scale:
problems
Scale:
Identification
3:
communication
problems
Task:
Child
(CBCL),
problems
of
Report
Task:
NICHD
Woodcock–Johnson
Academic
WJ-R
Academic
WJ-R
Preschool
WJ-R
Continuous
(I)
(II)
Attention
and
and
and
words
13–14
Tests
counting,
(difficulty
letters,
Application
5–6
(test–retest
number
fractions,
algebra
(test/retest
etc.):
Report
Achievement,
age
Reading
names,
single
Individual
Age
loud
Math:
mathematical
.74)
Reading
Math
NLSY
Achievement
(PIAT)
Recognition:
Matching
naming
reading
out
.89)
of
concepts,
recognition,
multiplication,
division,
advanced
geometry
Recognition
concentrating,
restless,
Maternal
Peabody
PIAT
PIAT
PIAT
Hyperactivity
Preschool
(
and
3
Scale
Math
Scale
Reading
Math
Report
response
ending
Entry
Word
subscales:
size
Grade
Test
subscales:
Teacher
Test
Teacher
subscales:
Learning
Test
Test
ordinality
item
and
to
Rating
Rating
(IRT)
kindergarten
Teacher
.95):
.94):
recognition,
“advanced”
of
“early”
sounds
“early”
School
relative
ECLS-K
.89):
Reading
theory
“advanced”
Extrapolation,
Evaluation
(
Report
IRT
subscales:
Multiplication/division,
Place-value,
problems
(
Report
Fall
IRT
Letter
beginning
word
IRT
Counting,
and
Achievement
Academic
Achievement
Academic
Achievement
Achievement
Approaches
Outcomes,
of
skills
skills
problems
skills
entry
1
Measure
Measures
ability
Reading
Language/verbal
Math
Attention
Attention
School
Reading
Math
Achievement
Attention
Table
Study
Outcomes

---

SCHOOL READINESS AND LATER ACHIEVEMENT
1433
of
is
(
the
.82)
.46)
Study
Survey
(
Study
Report
seems
(
Report
(
Cube
Personal
Preschool
5
others,
.95)
(
etc.
(
during
Externalizing:
Internalizing:
etc.
months
Vocabulary
Interviewer
Language
child’s
Cohort
Age
42
(
&
Copying
test
of
bullies
worries,
designs
Maternal
.79)
months:
months:
Scale,
Maternal
Scale,
Birth
Age
Wechsler
22
.93),
22
Longitudinal
5
Child
disobedient,
.72):
Child
miserable,
.54):
(
.83)
stacking,
designs
ratings
cooperation
assessment
development
Development
Rutter
Rutter
Counting
Speaking
Copying
Age
Cooperation:
Age
age
British
Human
National
(
.93).
etc.
and
.72
.85
and
includes
etc.
Report
A
Split-
SY
(
(
Child
colors,
Report
helpful,
and
Test:
sequence,
alth
NL
senior
fights,
etc.
others,
senior
Report
often,
is
.66
week
He
(IHDP)
2.
to
Vocabulary
Forms
respectively;
1
and
Teacher
shapes,
Child
and
adaptation:
B,
at
cries
number
and
Teacher
Child
number
problems
kindergarten
Child
1
kindergarten
others,
Teacher
.92):
Picture
and
Knowledge
Junior
(PPVT),
reliability
of
Program
.80):
Junior
French
A
bullies
.72):
worries,
sympathetic
(
Test
B,
half
for
test–retest
Informal
knowledge,
counting,
addition
behavior
Tables
Aggression:
Anxious/Depressed:
Prosocial:
Peabody
Number
in
Institute
Overall
IQ
.94):
Development
5
3
problems
Report
National
and
(
consistently
.94):
Report:
Age
Age
Overall
observer’s
Health
Test
behavior
(
Maternal
Test
report
Stanford-Binet
CBCL,
Attention
knowledge.
Maternal
measures
Infant
index
Fall
(
general
Profile
(
Teacher
System
of
.93):
numbers,
.86)
measure.
Teacher
Maternal
4.5
of
Teacher
3
Skills
laboratory
test
numerals
(
of
Rating
.93)
Scale
Behavior
Cooperation,
self-control
Age
Fall
Age
letters,
toy
Ability
entry
Aggressive
Internalizing
Fall
Basic
Developmental
expressive
Externalizing,
Impulsivity
Roman
Kindergarten
Skills
.93):
Child
behavior
of
Report
.93):
Kindergarten
Report
(SSRS):
assertion,
Kindergarten
(colors,
etc.,
Language
(vocabulary
comprehension,
.93;
language,
internalizing:
Report
task:
school
CBCL,
CBCL:
Social
Bracken
Reynell
CBCL,
Forbidden
a
Cognitive
Drawing.
of
of
WJ-R
Achenbach
5
sad):
.59);
.93):
includes
4.5
Profile
(cheats,
age
Test—
ratings
5–6
strong
Report
Report
Report
3–4
(
Report
age
etc.):
and
etc.):
(PPVT-R;
reliability
(
and
Study.
Strong
cooperation
the
Age
unloved,
Age
Picture
(ECLS-K)
.93)
Test,
Head
Antisocial
(stubborn,
temper,
Maternal
bullies,
Maternal
(feels
Maternal
Vocabulary
Revised
split-half
.80)
Maternal
Interviewer
child’s
during
assessment
includes
(I)
(II)
Anxious/depressed
Peabody
Compliance
Sociability
(
Cohort
Preschool
IQ
Designs
(
SECCYD)
Report
Skills
(
Report
Copying
Problems
Teacher
Problems
Teacher
(NICHD
Performance
kindergarten
Teacher
of
.90):
Control
Teacher
.89):
Study-Kindergarten
Drawing,
.80):
Fall
Interpersonal
(
Report
Self
.79):
(
Report
(WPPSI)
Externalizing
Internalizing
(I)
(II)
Longitudinal-Experimental
Figure
Development
Longitudinal
)
Youth
Intelligence
Human
Montreal
5
of
skills
and
behaviors
problems
problems
controls
age
Childhood
entry
Care
Scale
(
continued
MLEPS
skills
child
1
cognitive/
attention
Early
Child
includes
School
Prior
achievement
skills/socioemotional
behavior
Primary
Externalizing
Internalizing
Social
Prior
Prior
Youth;
Table
Socioemotional
Note.
Early
and
(BCS)
of

---

1434
DUNCAN ET AL.
attention, another (NICHD SECCYD) has both test-based and
age 5– 6. Consequently, our sample comprises children who were
teacher-rated measures of attention, and three (NLSY, IHDP, and
age 5 or 6 in 1986, 1988, 1990, or 1992. The age 13–14 achieve-
BCS) have parent rather than teacher reports of socioemotional
ment and behavior of these children were assessed in the respec-
behaviors. In addition, two of the studies (NICHD SECCYD and
tive 1994, 1996, 1998, and 2000 interviews.
the Montreal Longitudinal-Experimental Preschool Study
School readiness measures, including math and reading test
[MLEPS]) measure both attention skills and problems, whereas
scores (Peabody Individual Achievement Test; Dunn & Mark-
three (NLSY, IHDP, and BCS) have measures of attention prob-
wardt, 1970) and maternal reports of children’s behavior problems
lems but not skills, and one study (ECLS-K) has a measure of
(adapted from the Achenbach Behavior Problems Checklist;
attention skills but not attention problems. In addition, with one
Baker, Keck, Mott, & Quinlan, 1993) were collected at age 5 or
exception, all of the studies provide measures of academic, atten-
age 6. Academic achievement outcome measures were collected
tion, and socioemotional skills prior to the point of school entry,
biennially for children between the ages of 5 and 14. In addition,
which we used as key control variables in our analyses.
key control variables include children’s receptive vocabulary (Pea-
body Picture Vocabulary Test—Revised; Dunn & Dunn, 1981)
The Studies and Samples
and children’s temperament (compliance and sociability) at age 3
or 4. Additional family- and child-level control variables are
The Early Childhood Longitudinal Study–Kindergarten Cohort
described in Appendix B.
(ECLS-K).
The ECLS-K follows a nationally representative sam-
The NICHD Study of Early Child Care and Youth Development
ple of 21,260 children who were in kindergarten in 1998 –1999.
(SECCYD).
Longitudinal data from the NICHD SECCYD are
We used data from kindergarten, first grade, and third grade. Data
drawn from a multisite study of births in 1991 (NICHD Early
were collected from multiple sources, including direct achieve-
Child Care Research Network, 2005a). Participants were recruited
ment tests of children and surveys of parents, teachers, and school
from hospitals located at 10 sites across the United States. During
administrators (see Table 1; National Center for Education Statis-
24-hr sampling periods, 5,265 new mothers met the selection
tics, 2001).
criteria and agreed to be contacted after returning home from the
Achievement tests were administered in the fall of kindergarten
hospital. At 1 month of age, 1,364 healthy newborns were enrolled
and in the spring of kindergarten, first grade, and third grade. We
in the study. Although it is not nationally representative, the study
used teacher reports of children’s “approaches to learning” (which
sample closely matches national and census tract records with
measure both attention skills and achievement motivation) and
respect to demographic variables such as ethnicity and household
socioemotional behaviors, including internalizing and externaliz-
income. The majority of children in the sample are White, 12% are
ing problems, self-control with peers, and interpersonal skills,
African American, and 11% are Hispanic or of another ethnicity.
collected in the fall and spring of kindergarten.
About 30% of mothers had a high school education or less, and
The battery of achievement tests given as part of the ECLS-K
14% were single parents (NICHD Early Child Care Research
kindergarten and first-grade assessments covered three subject
Network, 1997). The analysis sample had valid data on the
areas: language and literacy, mathematical thinking, and general
achievement outcome measures and at least three sources of in-
knowledge. For third grade, the achievement tests included math-
formation on the key independent variables (approximately 981 at
ematics, reading, and science. We used item response theory
first grade, 928 at third grade, and 907 at fifth grade).
scores for the first two of these as key dependent variables. These
School readiness measures, including achievement tests and
third-grade assessments required students to complete workbooks
attention/impulsivity tasks, were administered in a controlled lab-
and open-ended mathematics problems. As detailed in Appendix
oratory setting at age 4.5, and attention problems, aggression,
A, a host of family- and some child-level controls are available in
internalizing behavior, and social skills were measured by teacher
the data.
report in the fall of the kindergarten year. Outcomes at first, third,
The children of the National Longitudinal Survey of Youth
and fifth grades include achievement in math and reading accord-
(NLSY).
The NLSY is a multistage stratified random sample of
ing to teacher ratings and Woodcock–Johnson Tests of Achieve-
12,686 individuals age 14 to 21 in 1979 (Center for Human
ment—Revised test scores (Woodcock & Johnson, 1990; see Table
Resource Research, 2004). Black, Hispanic, and low-income youth
1). Key control variables at age 3 include children’s cognitive
were overrepresented in the sample. Annual (through 1994) and
ability, language skills, impulsivity, externalizing problems, and
biennial (between 1994 and 2000) interviews with sample mem-
internalizing problems. The NICHD SECCYD also collects infor-
bers and very low cumulative attrition in the study contribute to the
mation from infancy about children’s early environments, includ-
quality of the study’s data.
ing child-care type and quality, home environment, and parenting;
Beginning in 1986, the children born to NLSY female partici-
these and other child- and family-level covariates are described in
pants were tracked through biennial mother interview supplements
Appendix C.
and direct child assessments. Given the nature of the sample, it is
The Infant Health and Development Program (IHDP).
The
important to note that early cohorts of the child sample were born
IHDP is an eight-site randomized clinical trial designed to evaluate
disproportionately to young mothers. With each additional cohort,
the efficacy of a comprehensive early-intervention program for
the children become more representative of all children, and NLSY
low birth weight (LBW) premature infants. Infants weighing 2,500
children younger than age 14 in 2000 share many demographic
g (5.51 lb) or less at birth were screened for eligibility if their
characteristics of their broader set of age mates.
postconceptional age between January and October 1985 was 37
The sample used in the present analysis consists of 1,756 chil-
weeks or less and if they were born in one of eight participating
dren whose academic achievement was tracked from age 7– 8 to
medical institutions. A total of 985 infants was randomly assigned
age 13–14 and whose achievement and behavior was assessed at
either to a medical follow-up only or to a comprehensive early

---

SCHOOL READINESS AND LATER ACHIEVEMENT
1435
childhood intervention group immediately following hospital dis-
behavioral development, including physically aggressive, anxious,
charge.
depressive, hyperactive, inattentive, and prosocial behavior. Third-
Infants in both the comprehensive early childhood intervention
grade assessments include a group-administered math test and
and medical follow-up only groups participated in a pediatric
teacher ratings of children’s French language skills (see Table 1).
follow-up program of periodic medical, developmental, and famil-
Key control variables include number knowledge and vocabulary
ial assessments from 40 weeks of conceptional age (when they
measured on entry into junior kindergarten (age 4) for Cohort 1
would have been born if they had been full term) to 36 months of
and on entry into senior kindergarten (age 5) for Cohort 2. Addi-
age corrected for prematurity. The intervention program, lasting
tional family- and child-level control variables are detailed in
from hospital discharge until 36 months, consisted of home visits,
Appendix E.
child-care services, and parent group meetings. A coordinated
The 1970 British Birth Cohort Study (BCS).
The U.K. 1970
educational curriculum of learning games and activities was used
BCS, a nationally representative longitudinal study, has followed
both during home visits and at the center.
into adulthood a cohort of children born in Great Britain during 1
The primary analysis group consisted of 985 infants. Of these
week in 1970 (Bynner, Ferri, & Shepherd, 1997). The birth sample
985 infants, cognitive assessments are available for 843 children at
of 17,196 infants was approximately 97% of the target birth
age 3, 745 children at age 5, and 787 children at age 8. In addition,
population. Attrition has reduced the original sample to 11,200
76 children who were born at an extremely low birth weight
participants. Nevertheless, the representativeness of the original
(ELBW; 1,000 g [3.27 lb] or less) were excluded from the sample
birth cohort has largely been maintained, although the current
because ELBW children differ markedly from other LBW children
sample is disproportionately female and highly educated (Ferri &
in cognitive and behavioral functioning (Klebanov, Brooks-Gunn,
Smith, 2003). Missing data on key variables reduce the sample size
& McCormick, 1994a, 1994b). Thus, this study focuses on a
for most analyses to between 9,000 and 10,000 cases.
subsample of 690 children who were not born ELBW and for
At each wave, cohort members were given a battery of tests of
whom cognitive assessment and family background data were
intellectual and behavioral development (see Table 1). School
available.
readiness measures include vocabulary and copying skills tests,
Data come from a variety of sources: questionnaires, home
and maternal reports of attention, externalizing behavior, and
visits, and laboratory tests (see Table 1). School readiness mea-
internalizing behavior were collected when the children were 5
sures include preschool performance and verbal test scores, paren-
years of age. Reading and mathematics achievement tests were
tal reports of children’s mental health and aggressive behavior, and
administered at age 10. Key control variables include measures of
observer reports of children’s attention and task persistence. We
basic skills and behavior at ages 22 and 42 months for a 10%
assessed reading and math achievement using the Woodcock–
subsample of the data. Additional family- and child-level controls
Johnson Tests of Achievement—Revised broad reading and math
are described in Appendix F.
tests and the Wechsler Intelligence Scale for Children—Third
Edition (Wechsler, 1991) performance and verbal tests at 8 years
Analysis Plan
of age. Key control variables include cognitive ability, sustained
We begin our analysis by estimating a similar set of regression
attention, and behavior problems at age 3. Additional family- and
models across all six studies, in which school-entry academic,
child-level control variables are described in Appendix D.
attention, and socioemotional skills are related to later academic
The Montreal Longitudinal-Experimental Preschool Study
achievement. For example, in ECLS-K data, the school-entry skills
(MLEPS).
The MLEPS comprises several consecutive cohorts
and behaviors are measured in the fall of kindergarten (referred to
launched from 1997 to 2000. The original sample of 4- and
hereafter as FK), whereas math and reading achievement are
5-year-old children (N
1,928), representing one third of its
measured in the spring of third grade (referred to hereafter as 3rd).
population base, was obtained after a multilevel consent process
The resulting equation is as follows:
involving school board administrators, local school committees,
parents, and teachers. Given that its final cohort (2000) does not
ACHi3rd
a1
1ACADiFK
2ATTNiFK
3SEiFK
meet all the data requirements for the research objective examined
here, we limited ourselves to the sample of children beginning
1FAMi
2CHILDi
eit,
(1)
kindergarten in the fall of 1998 and the fall of 1999.
where ACH
is the math or reading2 achievement of child i in
i3rd
Incomplete data reduced the sample from 1,369 to 767 children.
the spring of third grade; ACAD
is the collection of math,
iFK
Students in the final sample had a valid value on any of the four
reading, and general knowledge skills that child i has acquired at
outcome measures of interest (first- and third-grade achievement
school entry, assessed by achievement tests in the fall of the
measures) and on at least four of the six socioemotional measures.
kindergarten year; ATTN
is a teacher-reported measure of
iFK
Of the 767 participants in the final sample, 439 began kindergarten
attention; SE
is the collection of socioemotional skills that child
iFK
in 1998 and 328 began kindergarten in the fall of 1999. Addition-
i’s teacher reports; FAM and CHILD are sets of family back-
i
i
ally, for 350 of the 767 students, initial data were collected during
ground and child characteristics, respectively, included in analyses
the fall of junior kindergarten (332 who began junior kindergarten
to control for individual differences that might influence child
in 1997 and 18 who began junior kindergarten in 1998).
achievement before and after school entry; a is a constant; and e
1
it
Initial and follow-up data were collected from multiple sources,
is a stochastic error term.
including direct cognitive assessments of children and surveys of
parents and teachers. Early academic assessments include individ-
ually administered number knowledge and receptive vocabulary
2 We use reading as shorthand for the set of reading, language, and
tests at the end of senior kindergarten. Teachers rated children’s
verbal ability skills measured in our data sets.

---

1436
DUNCAN ET AL.
Our interest is in estimating
,
, and
, which, if correctly
power of attention and socioemotional skills. This would occur if
1
2
3
modeled, can be interpreted as the impact of school-entry aca-
one of the ways in which attention and socioemotional skills
demic, attention, and socioemotional skills on subsequent achieve-
affected later achievement were to raise children’s school entry
ment. A key challenge in this approach is ensuring that we have
academic skills. We investigate this possibility by estimating ver-
accounted for the possibility of omitted variable bias, which is
sions of Equation 2 that omit ACAD
.
iFK
likely to arise if unobserved family or child characteristics are
In the second step of our analysis, we use meta-analytic tech-
correlated with both children’s school entry skills and their later
niques to summarize coefficients obtained from our six studies’
achievement. Our principal strategy for securing unbiased estima-
estimates of Equation 2 and seek to determine whether particular
tion of
,
, and
is to estimate a model of the form of
1
2
3
study characteristics are associated with larger (or smaller) coef-
Equation 1 that includes as many prior measures of relevant child
ficients. More specifically, the meta-analysis treats the standard-
and family characteristics as possible.
ized regression coefficients from Equation 2 as observations in a
All of the studies contain important measures of child and
regression predicting academic achievement measured as late in
family characteristics that may be confounded with children’s
childhood as possible. Independent variables in the meta-analytic
achievement, attention, and behavior. Although the specific set of
regression include (a) the type of school-entry measure,5 (b)
covariates varies across studies, most studies include measures of
elapsed time (scaled in years) between measurement of school-
the child’s race and ethnicity, maternal education, family structure,
and family income or economic well-being. In some studies,
entry characteristics and the outcome, (c) whether the outcome is
measures of child health, maternal depressive symptoms, parent-
math or reading achievement, and (d) whether the outcome is
ing, and quality of the home environment, as well as children’s
based on a test or a teacher report. In keeping with standard
participation in early child care and education during early child-
meta-analytic practices, we weighted each regression coefficient
hood were also included as controls.3 Details about the specific
observation by the inverse of its variance (Hedges & Olkin, 1985).
controls used in each study are provided in the appendices, and a
complete list of covariates for each study can be found in Tables
Results
A6, B6, C6, D5, E6, and F5.
Our analysis was designed to examine the relations between
Regression Results
early skills and later achievement, irrespective of the characteris-
tics of the classroom/school the child attends. The ECLS-K is an
To consider whether school entry6 achievement, attention,
exception, owing to a sample design that selects an average of 4
and socioemotional skills are predictive of subsequent achieve-
students per kindergarten classroom to be enrolled in the study. We
ment, we first estimated a comparable set of regressions (Equa-
took advantage of this classroom clustering by adjusting our
tion 2) across all of the studies. For each study, reading and
ECLS-K estimates for classroom fixed effects. Thus, all of the
math outcomes measured as late in the data set as possible were
variation used in the regression stems from within-classroom dif-
regressed on school-entry achievement, attention, and socio-
ferences, which holds constant school and classroom characteris-
tics.
emotional behaviors, with controls for important family and
Of course, we cannot be certain that even a comprehensive set
child characteristics also included in the regression. In all but
of control variables captures all of the important confounds, which
two cases, our regressions include measures of both cognitive
leaves open the possibility that this approach will still produce
ability and either attention or socioemotional behaviors.
biased estimates of
,
, and
. For example, an obvious bias
1
2
3
of this sort would arise if scores on a kindergarten mathematics test
3 Our list of child and family control variables is more extensive than in
reflected both math skills and underlying cognitive ability.
most developmental studies. In selecting these variables, we were careful
To further reduce the possibility of biases, we include measures
to include only variables measured prior to or concurrently with our
of a child’s attention and socioemotional behaviors and either
school-entry measures of achievement and behavior. We were also mindful
cognitive ability or preacademic skills assessed prior to school-
that added controls might introduce multicolinearity into our regression
entry, which are available in all but two studies (ECLS-K and
estimation, but there was no indication that this might be the case. And
MLEPS).4 With these prior measures, our model becomes
finally, our appendix tables compare models run with and without our child
and family controls and show that the results of our analyses depend little
ACH
on adjustments for these factors; concurrent controls for the other achieve-
i3rd
a1
1ACADiFK
2ATTNiFK
3SEiFK
ment and behavioral measures matter much more.
4
4ACADiPre-FK
5ATTNiPre-FK
6SEiPre-FK
The MLEPS provides preschool cognitive measures but not attention
or socioemotional behaviors (see Table 1).
1FAMi
2CHILDi
eit.
(2)
5 The decision of which type of measure should serve as the omitted
dummy variable category is noteworthy, because the coefficients on the
ACAD
, ATTN
, and SE
refer to child i’s respec-
iPre-FK
iPre-FK
iPre-K
included measure categories represent differences from the omitted cate-
tive achievement, attention, and socioemotional behavior prior to
gory. We selected internalizing behavior problem coefficients as the omit-
school entry, respectively. This constitutes a particularly powerful
ted category because the simple average of their regression coefficients
version of Equation 1, because controlling for the child’s cognitive
was very close to zero (–. 01 for reading outcomes and –. 01 for math
and behavioral skills before school entry should reduce, if not
outcomes).
eliminate, omitted-variable bias in
,
, and
.
6
1
2
3
We remind the reader we use the term school entry somewhat loosely.
One concern about Equation 2 is that by controlling for school
It refers to age 5 in four cases, age 5– 6 in one case, and the fall of the
entry achievement, we might reduce the deserved explanatory
kindergarten year in only one case.

---

SCHOOL READINESS AND LATER ACHIEVEMENT
1437
Standardized regression coefficients and standard errors from
outcomes measured as late in childhood as possible. These results are
models predicting achievement from the school-entry academic,
shown in the second column of Table 3. The second meta-analytic
attention, and socioemotional behaviors are presented in Table
regression is based on the 228 coefficients taken from regressions with
2. Complete regression results using all available reading and
outcomes measured at all possible points in a given study. These coeffi-
math outcomes are presented in appendix tables and are sum-
cients are shown in the appendix tables and produce the results shown in
marized below in our meta-analysis.
the third, fourth, and fifth columns of Table 3.
As expected, the regression results indicate that school-entry read-
A clear conclusion from the first meta-analytic regression is
ing and math skills are almost always statistically significant predic-
that only three of the school-entry skill categories predict
tors of later reading and math achievement, with standardized coef-
subsequent reading and math achievement: reading/language, math,
ficients ranging from .05 to .53. Not surprisingly, school-entry reading
and attention. Moreover, rudimentary mathematics skills appear to
skills predict subsequent reading achievement better than subsequent
matter the most, with an average standardized coefficient of .33.9 The
math achievement, just as early math skills are more predictive of later
association of reading skill with later achievement was less than half
math than reading achievement.
as large (.13), and, at .07, the average standardized coefficients on the
In the case of attention skills and attention problems, coeffi-
attention-related measures was less than one quarter the size of the
cients are usually smaller than those for math skills, but they are
mean math-skills coefficient. As expected from Table 2, the meta-
statistically significant for more than half of the coefficients. In
analysis results confirm that behavior problems and social skills are
contrast, coefficients for socioemotional behaviors— externalizing
not associated with later achievement, holding constant achievement
and internalizing behavior problems and social skills—rarely pass
as well as child and family characteristics. Indeed, none had a stan-
the threshold of statistical significance.
dardized coefficient that averaged more than .01 in absolute value.
This general pattern—relatively strong prediction from
Turning to the other coefficients listed in the second column
school-entry reading and math skills, moderate predictive
of Table 3, one can see that the school-entry skills coefficients
power for attention skills, and few to no statistically significant
decreased a little (.010 per year) with each additional year
coefficients on socioemotional behaviors—is also found for
between school entry and the point of assessment of the math or
reading and math achievement measured at earlier points in the
reading outcome. As for whether teacher-report outcomes or
studies and in logistic regressions in which grade retention is
direct skill assessments are more likely to be predicted by early
the dependent variable (results shown in Tables