Social Studies Instruction and
Reading Comprehension
Evidence from the Early Childhood Longitudinal Study
September 2020
By Adam Tyner and Sarah Kabourek
Foreword by Amber M. Northern and Michael J Petrilli
ABOUT THE FORDHAM INSTITUTE
The Thomas B. Fordham Institute promotes educational excellence for every child in America via
quality research, analysis, and commentary, as well as advocacy and exemplary charter school
authorizing in Ohio. It is affiliated with the Thomas B. Fordham Foundation, and this publication is
a joint project of the Foundation and the Institute. For further information, please visit our website
at www.fordhaminstitute.org. The Institute is neither connected with nor sponsored by Fordham
University.
SUGGESTED CITATION FOR THIS REPORT
Adam Tyner and Sarah Kabourek. Social Studies Instruction and Reading Comprehension: Evidence
from the Early Childhood Longitudinal Study. Washington D.C.: Thomas B. Fordham Institute
(September 2020). https://fordhaminstitute.org/national/resources/social-studies-instruction-and-
reading-comprehension.
ACKNOWLEDGMENTS
This report was made possible through the generous support of our sister organization, the Thomas B.
Fordham Foundation.
Sincere gratitude to Adam Tyner and Sarah Kabourek for their thoughtful approach to the study
and its implications. External reviewers Daniel Willingham, professor of psychology at the University
of Virginia, and Sco Latham, associate research scholar at the Education Research Section (ERS)
at Princeton University, provided useful feedback and advice on the dra report. David Grissmer
(University of Virginia) spoke to us about related research and Seth Gershenson (American University)
also offered input.
At Fordham, we extend thanks to Chester E. Finn, Jr. and Robert Pondiscio for reviewing dras, Victoria
McDougald for overseeing media relations, Olivia Piontek for handling funder communications,
and Pedro Enamorado for managing report production. Fordham research interns Tran Le, Trinady
Maddock, and Alice Tsai provided assistance at various stages in the process. Finally, we thank Dave
Williams for developing the report’s layout and design, Pamela Tatz for copyediting the report, and
Stockbyte from Getty Images for providing our cover image.
Contents
1 Foreword and Executive Summary
9 Section I: Introduction
11 Section II: The other “reading war”
14 Section III: Methodology
17 Section IV: How is time allocated in elementary
school classrooms?
24 Section V: How does instructional time relate to
reading improvement?
29 Section VI: Summary and implications
33 Appendix A: Technical appendix
36 Appendix B: Robustness check
38 Endnotes
1
Amber M. Northern and Michael J. Petrilli
Tell ‘em what you’re going to tell ‘em; then tell ‘em; then tell ‘em what you told ‘em. That’s
good advice from a lay preacher in 1908. Here goes.
We’re going to tell you what America needs to do if it is serious about wanting kids to
become beer readers. That’s this: Instead of devoting more class time to English language
arts (ELA), we should be teaching elementary school children more social studies—as in, rich
content about history, geography, and civics. That may seem counterintuitive, but that’s the
key takeaway of this groundbreaking study.
Mind you, we’re not the first to find that loads of time devoted to language arts instruction
does not improve student reading. But we are the first to find that literacy gains are more
likely to materialize when students spend more time learning social studies.
This novel result underscores the crucial insight that E.D. Hirsch set forth in 1987 in his path-
breaking book Cultural Literacy. Hirsch believed, as aptly summarized by journalist Eric Liu,
that “literacy is not just a maer of decoding the strings of leers that make up words or
the meaning of each word in sequence. It is a maer of decoding context: the surrounding
matrix of things referred to in the text and things implied by it.”
In other words, writers and speakers oen make assumptions about what their readers and
listeners know. They don’t take time to explain historical references and literary allusions
or to resolve ambiguities. When those assumptions are correct—when readers and writers
share a common body of knowledge—language comprehension comes much easier.
Think of it this way: Virtually all middle schoolers can “read” the words “Berlin Wall.” They
surely can sound out the leers. But only some middle schoolers will instantly recognize
the phrase and what it implies. They will know that Berlin was the capital of Germany
during World War II, aer which it was divided between the Western Allies and the Soviet
Union. They will know that the wall was built to keep East Germans from escaping to the
West. And they will know that those East Germans wanted to escape because of the lack of
Foreword and
Executive Summary
2
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
freedom and economic opportunity under Communist rule. So those students will be able to
comprehend a passage that alludes to the Berlin Wall, while their less-knowledgeable peers
will not.
Over the years, there’s been a small but growing body of literature showing that a focus
on academic content in the early grades—not generalized reading skills and strategies—
equips students with the background knowledge (à la the Berlin Wall) that they need
to comprehend all sorts of texts and make them truly literate.
1
Moreover, this focus on
knowledge has the potential to lessen socioeconomic and racial/ethnic gaps in literacy.
In fact, Hirsch made this case forcefully in his 1987 book:
To be culturally literate is to possess the basic information needed to thrive in the
modern world. The breadth of that information is great, extending over the major
domains of human activity from sports to science. It is by no means confined to
“culture” narrowly understood as an acquaintance with the arts. Nor is it confined to
one social class. Quite the contrary. Cultural literacy constitutes the only sure avenue
of opportunity for disadvantaged children, the only reliable way of combating the
social determinism that now condemns them to remain in the same social and
educational condition as their parents.
2
In other words, if we acknowledge that more affluent and white students oen have greater
access to knowledge-building opportunities and resources in the home (including, on
average, beer-educated parents), we can make education more equitable by teaching
knowledge-rich content to those without these same advantages.
Yet rather than spending the intervening three decades making sure that all students—but
especially poor students and students of color—got a strong dose of social studies, science,
and geography instruction in order to build this content knowledge and advance equity,
our schools mostly doubled down on the “skill” of reading. Instead of learning about the
world, students learn to “identify the main idea.” And as any serious analysis of reading
achievement can tell you, it hasn’t worked. It can’t. As Hirsch has tirelessly pointed out,
language comprehension is not a “skill” at all: content is comprehension.
Thankfully, there’s recently been renewed interest in ELA curricula that are serious about
building knowledge, especially in the early years. And evidence is finally starting to appear
that teaching “content-rich” instructional materials—which are organized coherently
around a topic to build student understanding—can improve students’ ability to read and
comprehend complex texts.
3
We would love to see every elementary school in the country adopt one of the handful of
such curricula that are commied to building content knowledge. But convincing districts
and schools to adopt particular instructional materials is a tricky thing, rife with questions
3
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
about who has the right to select it and how much autonomy teachers have to implement it.
Local politics oen prevent administrators from even trying.
We wondered whether there was another way—an alternate path to ensure that young
students are exposed to a healthy diet of thought-provoking content. What if they simply
spent more time in classrooms that devoted substantial aention to a wide array of subject
maer and knowledge-rich material? Is it possible that elementary school students whose
teachers spend more time on social studies, science, and the like end up becoming beer
readers? That’s the question that birthed this study.
To investigate, Fordham’s associate director of research Adam Tyner and early childhood
researcher Sarah Kabourek teamed up to beer understand how classroom time is currently
spent in U.S. elementary schools and how it might be beer utilized to promote literacy.
Tyner has authored high-profile reports on a variety of topics ranging from the impact of
end-of-course tests on student achievement, how participation in credit recovery courses
varies across states, and enrollment in gifted education in high-poverty schools. Kabourek’s
research focuses on preschool access, school finance, and equality of educational
opportunity. She provided methodological and technical assistance for the project.
Tyner and Kabourek plumbed nationally representative data from the federal Early
Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K: 2011), which samples
over 18,000 students in their kindergarten year and follows them through fih grade.
4
This
long view is important because knowledge and vocabulary accrue slowly. Their primary
analysis examines how much classroom time is spent on different subjects, whether students
who spend more time on certain subjects make greater progress in reading, and how these
effects differ by student characteristics. Their analytic models control for a host of student-,
teacher-, and school-level factors to further isolate the relationship between time-use and
reading growth.
Your own time will be well spent by reading the full report. But here’s a summary of a few of
its findings.
1) Elementary school students in the U.S. spend much more time
on ELA than on any other subject.
Figure ES-1 presents average time usage for grades 1 through 5. Elementary teachers report
that students spend more time on ELA than on any other subject, at two hours daily. Other
subjects receive far less instructional time: excluding math (at nearly an hour and a half per
day), students on average spend more time on literacy per day than on all other subjects
combined, including science (30 minutes), social studies (28 minutes), arts and music (23
minutes), physical education (PE; 19 minutes), and foreign language (3 minutes).
4
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
2) Students from less-affluent backgrounds, Hispanic students, and
those attending public schools (traditional and charter) spend
more classroom time on ELA than do other students.
Children from more affluent families spend slightly less time on ELA and math than
their less-affluent counterparts—which may be freeing up time for them to spend on
non-core subjects (such as art, music, and foreign language).
5
Students spend about
the same amount of time on social studies and science regardless of socioeconomic
status (see Finding 4 below for how the effect of that time varies across SES).
Differences in class time use for students of different racial/ethnic backgrounds are
negligible.
Private schools tend to spend less time on reading and math instruction and more
on other non-core subjects than do schools in other sectors, although few of these
differences are statistically significant, likely due to the smaller sample sizes.
Figure ES-1. Students spend an average of two hours per day on ELA
instruction.
Note: The figure contains pooled averages of grades 1 through 5. The mean total instructional time is 302
minutes per day. Analytic sample includes 6,829 students. “Arts and music” includes art, music, dance, and
theater. Error bars represent 95 percent confidence intervals.
0
25
50
75
100
125
150
120
82
30
28
23
19
3
ELA Math Science Social
Studies
Arts and
Music
PE Foreign
Language
Subject
Minutes of Instruction per Day
5
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
3) Increased instructional time in social studies—but not in ELA—
is associated with improved reading ability.
Social studies is the only subject with a clear, positive, and statistically significant effect
on reading improvement.
6
On average, students who receive an additional thirty minutes
of social studies instruction per day (roughly equivalent to moving from the tenth to the
ninetieth percentile of social studies instructional time) in grades 1–5 outperform students
with less social studies time by 15 percent of a standard deviation on the fih-grade reading
assessment, even aer controlling for multiple measures of kindergarten reading ability and
a host of student, school, and teacher factors.
Figure ES-2 shows the effects of additional instructional time spent on each subject. Though
many elementary schools have lengthy reading blocks, oen every day, time spent on ELA is
not associated with reading improvement. Likewise, neither math instructional time nor time
spent on non-core subjects (including art, music, PE, and foreign languages) corresponds
to gains or losses in reading. Although also a content-rich subject, instructional time for
science has no relationship with reading development, either.
Figure ES-2. More instructional time devoted to social studies is
correlated with greater reading growth from first through fifth grade.
Note: Analytic sample includes 6,731 students. Effects are in standard deviations of fih-grade assessment
scores. For example, the first bar indicates that the effect of thirty minutes of additional ELA instruction is
associated with a 3 percent of a standard deviation increase in student reading progress from kindergarten to
fih grade. However, because the error bars overlap with the baseline (0 percent), this effect is not statistically
significantly different from zero. Contrast that with the 15 percent effect for social studies, where the error bars
do not overlap with the baseline, indicating that the result is statistically significantly different from zero. Error
bars represent 95 percent confidence intervals.
Percent of Standard Deviation Reading
Test Score Improvement for Thirty
Additional Minutes of Daily Instruction
3
6
3
-1
15
-40%
-20%
0%
20%
40%
ELA Math Non-Core Science
Social Studies
Subject
6
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
4) The students who benefit the most from additional social studies
time are girls and those from lower-income and/or non-English-
speaking homes.
Next, we examine how the effect of instructional time varies by students from
different income quartiles. The only significant difference, once again, is the
amount of time spent in social studies—not on ELA, math, science or other non-
core subjects.
7
Effects are consistently positive for students in the boom three SES
quartiles but nearly zero and statistically insignificant for students in the wealthiest
quartile. More specifically, students in the boom three quartiles have similar
positive effects from an additional thirty minutes of daily social studies instruction
during elementary school, corresponding to greater reading development between
17 and 21 percent of a standard deviation (Figure ES-3).
Figure ES-3. More instructional time in social studies is related to greater
reading growth from first through fifth grade for all students except those
in the top income quartile.
Note: Analytic sample includes 6,731 students. Indicators of socioeconomic status (for example, “affluent”)
reflect quartiles of the family income distribution. Effects are in standard deviations of fih-grade assessment
scores. Note that the 17 percent of a standard deviation effect for students in the “below average income
quartile” is only statistically significant at the 90 percent confidence level. Error bars represent 95 percent
confidence intervals.
Percent of Standard Deviation Reading
Test Score Improvement for Thirty
Additional Minutes of Daily Instruction
ELA Math Non-Core Science
Social Sciences
Subject
1
13
4
-5
21
2
8
6
-9
17
6
-4
2
-3
20
2
4
0
13
-3
-40%
-20%
0%
20%
40%
Highest
Income
Above Average
Income
Below Average
Income
Lowest
Income
7
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
Tyner and Kabourek offer three takeaways for policy and practice—with which we heartily
agree and which we summarize here. First and most obvious is that elementary schools
should make more room for high-quality instruction in history, civics, geography, and the
other knowledge-rich—and engaging—subjects that comprise social studies. Excessive
amounts of time spent on ELA appear not to yield the additional reading gains that well-
intentioned educators hoped for. Frankly, it is not clear to us why science—presumably also
a “content-rich” subject—did not demonstrate similar benefits to social studies. Perhaps
its discipline-specific vocabulary (“tier-three” words) is too specialized to impact literacy
broadly. What we do know is that there is rarely such a thing as wasted knowledge and
teachers should facilitate the gathering of riveting content for kids. David Coleman, one
of the architects of the Common Core standards and now College Board CEO, once aptly
referred to restoring “elementary teachers to their rightful role as guides to the world.”
That’s spot on.
Second, teachers should use the ELA block efficiently to build student knowledge. Of course
that could mean adopting one of the well-regarded knowledge-rich curricula now on the
market, but it doesn’t have to. Beefing up the literacy block with high-quality texts about
history, geography, and other social studies topics could go a long way—both in helping
young learners to become beer readers and in capturing their imaginations to increase
student engagement.
Third, policymakers and administrators should align reading assessments with curricular
content. Daniel Willingham, a cognitive psychologist at the University of Virginia, once wrote
this simple yet profound statement: “Reading tests are really knowledge tests in disguise.”
That means that we need to get serious not only about how we teach students knowledge
but also how we test it. It calls for a much more deliberate approach to how we sequence
particular content across grade levels and how we sample from it to inform new tests that
reward and prioritize knowledge development.
We add to the authors’ takeaways a few additional thoughts of our own. First, it’s oen the
small changes in education that can shi the tide. There’s no doubt that making bigger
waves—opening charter schools to compete with district schools, increasing the rigor of
state standards, replacing half or more of staff in failing schools—can rock a becalmed
boat. But it might also make faint-hearted passengers seasick. Throwing a smaller pebble
in—simply shrinking the massive ELA block to make more room for students to learn about
geography, history, civics, and the like—can make for a ripple effect.
And with lots of kids doing “remote learning” again this year, it is more important than
ever that kids be engaged. We need to give them more material—not less—that grabs their
imagination, piques their curiosity, inspires new interests, and—the icing on the cake—
improves how they read.
8
Social Studies Instruction and Reading Comprehension | Foreword and Executive Summary
Schools need to get creative about how to get young students engaged in building
knowledge while at home. Books are terrific (including read-alouds available on YouTube),
but documentaries and TV shows work, too.
Though surely well intended, the massive ELA block has become a barrier to progress in
reading, especially for poor kids and kids of color. The best elementary school teachers
direct students’ gaze not in the mirror, under the mistaken notion that children can only
be engaged by texts about their lives and experiences, but out the window where they can
encounter just a smidgeon of what our globe of awe and wonder has to offer.
We end by telling you what we told you: Spending more time in elementary school on the
“skill” of reading comprehension at the expense of teaching content may sound like a good
idea, but it actually works against the very outcomes we’re trying to achieve. So let’s stop
doing it!
9
sectionI: Introduction
Over thirty years ago, E.D. Hirsch advanced the argument that true literacy stems from
broad knowledge of the world. Without an understanding of art, music, history, science,
geography, and literature, students struggle with articles in the local newspaper, let alone
texts in college courses. They also earn lower scores on reading tests. It is background
knowledge, built in and outside the classroom and around the kitchen table, that enables
fluent reading comprehension. Unfortunately, many American elementary schools fail
to emphasize knowledge building, especially in the earliest grades. Instead, they tend to
treat reading comprehension as an independent “skill” and subjects like social studies and
science as domains to be studied later, in line with the common but false assumption that
children should “first learn to read and then read to learn.”
Curriculum enthusiasts have long argued that adopting content-rich language arts curricula
would help to remedy the knowledge deficit. Another approach is simply to devote more
classroom time to subjects like science and social studies. It’s an open question whether
more time spent on such content-rich subjects
in elementary school classrooms might positively
impact students’ reading progress over time.
With two-thirds of American fourth and eighth
graders failing to read proficiently—and far worse
outcomes for students from disadvantaged
backgrounds—improving the reading ability of
young students could hardly be a more urgent
priority for our elementary schools (Figure 1).
8
Over
decades of education reform, literacy levels have
barely budged, which raises questions about whether classroom time is being put to the
best use. If knowledge is a key to literacy, elementary classrooms must devote time to
building it. But do they?
It is background
knowledge, built in and
outside the classroom
and around the kitchen
table, that enables fluent
reading comprehension.
10
Social Studies Instruction and Reading Comprehension | I: Introduction
To help understand how classroom time is currently spent in U.S. elementary schools and
how it might be beer utilized, this study uses nationally representative data to address the
following question: How does the distribution of instructional time across subjects relate to
students’ reading growth?
More specifically, we address these three questions:
1. How does the amount of classroom time spent on different subjects—including English
language arts (ELA), math, science, social studies, and other subjects—differ across
elementary classrooms?
2. Do students in classrooms who spend more time on certain subjects make greater
progress in reading during their elementary school years than students who spend less
time in these subjects?
3. Do the effects of different uses of instructional time vary by key student factors such as
socioeconomic status, home language environment, and gender?
In Section II, we examine prior research on early literacy. Section III brings a brief discussion
of our methodology. Section IV presents key findings on the first research question, and
Section V includes results for the second and third questions. In Section VI, we discuss the
implications of our findings for education policy and practice.
Figure 1. Only a third of American students read proficiently.
Source: National Assessment for Educational Progress (NAEP) 2019.
0%
25%
50%
75%
100%
35%
21%
34%
19%
11
II: The other “reading war”
The so-called “reading wars” have seen bales and skirmishes for decades now. In
1955, Rudolf Flesch’s bestseller Why Johnny Can’t Read argued against the then-popular
“sight method” of teaching students to read, which deemphasized phonics. In the next
decade, Harvard’s Jeanne Chall referred to disagreements about whether to focus early
reading instruction on phonics as “the great debate.” In her definitive review of research on
literacy instruction from 1912 (!) to 1965, she concluded “that a code-emphasis method [i.e.,
phonics] . . . produces beer results.” In recent decades, evidence has continued to mount
that the best way to build a foundation for reading in the early years is the development of
phonemic awareness and decoding skills, which enable students to sound out words using
phonics. Yet the controversy has persisted.
9
Even as phonics bales rage in the realm of primary reading, another tussle has been with
us for ages regarding how best to develop the vital elements of reading ability that go
beyond decoding skills.
On one side, the typical solution to America’s abysmal elementary reading outcomes has
been the obvious one: Schools should spend more time on literacy instruction. Previous
research has shown that third-grade classrooms spend twice as many hours on language
arts as on math and twice as many on math as on social studies and science.
10
To improve
the reading proficiency of at-risk and underserved children, schools have invested ever-
more time in reading instruction, oen providing a “literacy block” that can stretch to two
hours—or more—per day.
11
Federal policy may have contributed to this trend by mandating
annual state testing in reading and math, which many educators and commentators blame
for narrowing the curriculum to the tested subjects.
12
On the other side, critics contend that long stretches of literacy instruction do more harm
than good because of how that time is spent. Education writer Natalie Wexler observes
that much of it is allocated to efforts to develop formal, generalized reading skills such
as “finding the main idea,” “determining the author’s perspective,” “summarizing,” and
“clarifying.”
13
12
Social Studies Instruction and Reading Comprehension | II: The other “reading war”
Yet a small but plucky army of cognitive psychologists, analysts, and educators has long cast
doubt on the view that reading is a discrete skill that can be mastered independently from
acquiring knowledge. To these contrarians, a focus on academic content—not generalized
reading skills and strategies—will equip students with the background knowledge they
need to comprehend all sorts of texts and make them truly literate. Both common sense
and substantial research affirm that students benefit from some time spent on reading
comprehension strategies, but the critics say their centrality to contemporary reading
instruction is crowding out subjects that would help students build knowledge. Moreover,
the status quo’s focus on more abstract reading skills may contribute to socioeconomic and
racial/ethnic gaps in literacy, as more affluent and White students tend to have more access
to knowledge-building resources in the home.
14
In the late 1990s, the National Reading Panel took up the banner of early phonemic
awareness, phonics, and decoding skills to achieve reading fluency. Its other pillars of
“scientifically-based reading” focused on vocabulary and comprehension, but the panel
did not explicitly address the extent to which these abilities are built on a foundation
of knowledge (for more, see “The link between
background knowledge and literacy”).
There’s lile doubt that background knowledge is
critical for a reader to make sense of a given text.
A key question, though, is whether spending more
time on content-rich subjects can improve reading
comprehension.
The present study puts this question to the test. It
builds on the intuition that growing students’ knowledge of the world through subjects
other than ELA may be more effective in developing literacy than additional “reading”
instruction. It also takes a longer view than most previous research on this topic: We follow
students’ classroom experiences and reading development over a six-year period, from
kindergarten through fih grade, capturing accumulating knowledge via assessments that
measure more sophisticated reading ability as students age.
15
The long view is important because knowledge and vocabulary accrue slowly. A third grader
may be able to sound out “Cincinnati,” but if she hasn’t studied some geography, she likely
won’t comprehend the word. Knowledge that helps build a broad vocabulary has a clear
effect on literacy.
16
Over multiple years, the students whose teachers spend more time on
content-rich learning in subjects such as social studies, science, and art may see a larger
accumulation of knowledge and greater improvement in reading.
There’s little doubt that
background knowledge
is critical for a reader
to make sense of a
giventext.
13
Social Studies Instruction and Reading Comprehension | II: The other “reading war”
The link between background knowledge and literacy
E.D. Hirsch, Jr., professor emeritus of education and humanities at the University of
Virginia, was one of the first to argue that American schools’ focus on reading skills
paradoxically undermines reading progress. Hirsch claimed that knowledge of the
relevant topics in a text—pertaining to literature, science, philosophy, history, arts,
culture, geography, and so on—does more to improve literacy than do generalized
reading strategies.
Experimental research reveals how background knowledge influences a reader’s
abilities. A number of studies show that students comprehend more, perform beer
at immediate recall, and acquire more additional information when they already
possess strong background knowledge of the topic.
17
Cognitive psychologist Daniel
Willingham explains the connection in his book The Reading Mind. “[W]riters
always omit a great deal of information needed to make sense of what they write,”
Willingham explains, so “[t]he reader must have the right information in memory to
make the inferences that bridge the meaning of what he reads.”
Scholars have run experiments on the effects of background knowledge in a variety
of topic areas, from learning the rules of chess and bridge to learning computer
programming and electronics. The best-known study showed that baseball-savvy
students strongly outperformed students with less knowledge of the sport in
comprehending a wrien description of a half-inning. Importantly, this was true
for high-knowledge students who were also generally poorer readers than their
low-knowledge counterparts, which demonstrates the significance of specific
background knowledge relative to general reading ability.
18
More recent experimental
studies show that providing background information improves test performance.
19
Particularly relevant to the present study, elementary students who were randomly
assigned to spend a greater share of ELA time on science and social studies content
performed beer in those subjects and slightly beer on literacy assessments.
20
Curricular evaluations have also demonstrated that students taught in Hirsch’s CKLA
program plus phonics improved in reading ability more quickly than similar students
not in the program.
21
Another study analyzed how kindergarten classroom contexts,
including the amount of language arts instruction, impacted students’ reading
and math skills, finding that students in classes that spent more time on reading
instruction did not improve relative to their peers who spent less time on it.
22
14
III: Methodology
Data and sample
We use nationally representative data from the federal Early Childhood Longitudinal Study,
Kindergarten Class of 2010–11 (ECLS-K: 2011), which samples over 18,000 students in their
kindergarten year and follows them through fih grade, collecting semiannual or annual
measures on students, their parents, teachers, and schools, including annual assessments
in reading and math. This longitudinal, nationally representative study design enables us to
assess reading progress associated with school experiences from kindergarten through fih
grade.
Our analytic sample includes all students for whom we have nonmissing observations on all
student-level measures (more below) and for whom we have teacher-reported classroom
time measures in at least three of the five years in which data were collected. We use
ECLS-K: 2011 analytic survey weights throughout the analysis to account for the sampling
design and to adjust for nonresponse. Our final analytic sample includes 6,829 students.
Measures
Reading assessments. Our key outcome is fih-grade reading performance, as measured
by an ECLS-K reading assessment, which is a traditional reading assessment including
questions that gauge basic reading skills (for example, print familiarity, leer recognition,
and sight vocabulary), vocabulary knowledge (including vocabulary in context), and reading
comprehension.
23
These assessments were administered by NCES in each round of data
collection, and item response theory (IRT) scale scores were calculated to enable comparison
of growth over time. The present study uses this fih-grade reading assessment as the final
outcome, while controlling for kindergarten reading scores, yielding a five-year measure of
reading development.
15
Social Studies Instruction and Reading Comprehension | III: Methodology
Time use. The key independent variable is a measure of classroom time use. Teacher surveys
ask teachers (1) how oen children in their class spend time in each subject area (ranging
from never to five days a week) and (2) how much time is spent on these lessons (ranging
from never to more than three hours).
24
Using these items, we generate a set of “time-use”
variables that estimate the amount of classroom time spent in specific subject areas. These
items are combined and transformed to generate a measure of time per day in each subject
(and summed to yield an indicator of the total amount of instructional time per day). Then,
we average the time-use variables across grades, yielding an estimate of student exposure
to different subjects in first through fih grade.
25
No data on specific curricula or teaching
practices are available for the analysis.
Student, teacher, and school controls. In regression models (described further below),
we control for a set of student-, teacher-, and school-level factors to further isolate
the relationship between time-use and reading growth. These include baseline scores
on two rounds of kindergarten reading assessments (fall and spring); family income;
home language; student age, race, disability status, and pre-kindergarten participation;
kindergarten measures of cognitive skills (science, math, and working memory) and non-
cognitive skills (executive function); and a host of teacher and school characteristics,
including school sector (private, public, magnet, or charter),
26
school locale (urban,
suburban, town, or rural), and average length of teacher tenure. We include a control
variable for “departmentalization” when schools employ specialized subject-area teachers,
as they may report instructional time differently than teachers who teach all of the core
subjects themselves.
27
Analysis
We first examine the distribution of time spent on ELA, math, science, social studies, and
other subjects on average across and within grade levels. We observe this distribution across
a series of student- and school-level indicators hypothesized to correlate with student
exposure to content-rich subject areas: student family income; student race/ethnicity;
whether the student aends a public, private, or charter/magnet school; and length of
teacher tenure.
Next, we address whether the amount of time spent in particular subjects generates greater
reading progress. We conduct regression analyses with fih-grade reading as the outcome
and time-use indicators by subject as the key independent variables, controlling for above,
as well as the total amount of reported time per day (to account for variation in the length of
the school day). These models are described further in Appendix A.
16
Social Studies Instruction and Reading Comprehension | III: Methodology
Finally, we examine whether the effects in time use vary by certain student factors, including
family socioeconomic status, home language environment, and gender. Again, see Appendix
A for more.
Because the amount of classroom time devoted to different subjects is not randomly
assigned and may correlate with other factors for which we are unable to control—such
as teacher quality—it is possible that one or more of these factors is driving a correlation
between class time usage and student learning. Still, a finding that different use of class
time is correlated with greater improvement in reading aer controlling for various potential
confounders could imply a causal effect or, at the very least, suggest how beer teachers
employ instructional time.
17
We first look at how instructional time is spent in elementary school classrooms across
different subjects, schools, teachers, students, and grade levels.
Finding 1: Elementary school students spend much more time on ELA than
on any other subject.
Figure 2 presents average time usage for grades 1 through 5. Elementary teachers report
that students spend more time on ELA than on any other subject, at two hours daily. Math
is second, at nearly an hour and a half per day. Other subjects receive far less instructional
time: Excluding math, students on average spend more time on literacy than on all other
subjects combined, including science (30 minutes), social studies (28 minutes), arts and
music (23 minutes),
28
physical education (PE) (19
minutes), and foreign language (3 minutes).
Said another way, ELA instruction accounts for 39
percent of instructional time in U.S. elementary
schools (Figure 3), which average about five hours
(302 minutes) of total instructional time per day.
More than a quarter (27 percent) of that time is
spent on math, while science and social studies each
occupy about 10 percent. In Figure 3 and thereaer,
we combine the smaller subjects (arts and music, PE, and foreign language) into a “non-
core” category,
29
which accounts for 15 percent of instructional time.
IV: How is time allocated
in elementary school
classrooms?
Excluding math, students
on average spend more
time on literacy than
on all other subjects
combined...
18
Social Studies Instruction and Reading Comprehension | IV: How is time allocated in elementary school classrooms?
Figure 3. Of the total instructional time, 39 percent is spent on ELA, while
barely 20 percent is spent on science and social studies combined.
Note: The figure contains pooled averages of grades 1 through 5. The mean total instructional time is 302
minutes per day. “Non-core” includes art, music, theater, dance, physical education, and foreign language.
Analytic sample includes 6,829 students.
Math
ELA
Non-Core
Science
Social Studies
10%
15
%
39%
27%
9%
Figure 2. Students spend an average of two hours per day on ELA
instruction.
Note: The figure contains pooled averages of grades 1 through 5. The mean total instructional time is 302
minutes per day. Analytic sample includes 6,829 students. “Arts and music” includes art, music, dance, and
theater. Error bars represent 95 percent confidence intervals.
0
25
50
75
100
125
150
120
82
30
28
23
19
3
ELA Math Science Social
Studies
Arts and
Music
PE Foreign
Language
Subject
Minutes of Instruction per Day
19
Social Studies Instruction and Reading Comprehension | IV: How is time allocated in elementary school classrooms?
Still, considerable variation exists across students in how much instructional time they
receive in each subject (Table 1). For example, although students average 120 minutes per
day of ELA instructional time from first to fih grade (column 4), 10 percent of students
receive 93 minutes or less (column 2), while the 10 percent at the other extreme receive 147
minutes or more (column 6).
Other subjects exhibit even more variation. Students at the 90th percentile of social studies
and science instructional time (column 6), for example, spend nearly three times as much
time (43-45 minutes) on those subjects as students in the 10th percentile (15-16 minutes).
Non-core subjects show a similar paern, where students in the 90th percentile receive
68 minutes of instruction, an 183 percent increase over students in the 10th percentile (24
minutes) and a 62 percent increase over the median student (42 minutes). Math instruction
ranges from 63 minutes at the 10th percentile to 102 minutes at the 90th percentile.
Table 1. There is substantial variation in the distribution of instructional
time by subject.
Note: Table contains pooled averages of grades 1 through 5. “Non-core” includes art, music, theater,
dance, physical education, and foreign language. Analytic sample includes 6,829 students.
Subject
(1)
10th
Percentile
(2)
25th
Percentile
(3)
Median
(4)
75th
Percentile
(5)
90th
Percentile
(6)
ELA 93 105 120 135 147
Math 63 73 81 90 102
Non-Core 24 31 42 54 68
Science 16 22 29 37 45
Social Studies 15 21 28 35 43
All 252 275 302 328 352
20
Social Studies Instruction and Reading Comprehension | IV: How is time allocated in elementary school classrooms?
Finding 2: In the later elementary grades, somewhat more instructional time
is allocated to science and social studies and somewhat less to ELA and
non-core subjects.
Although elementary students spend more time in ELA relative to any other subject,
its allocation tends to diminish as students age. As we see in Figure 4, the peak of ELA
instruction occurs in first grade at 129 minutes; by grade 5, however, students spend an
average of 106 minutes per day in ELA.
Reported math instruction comprises 68 minutes per day in kindergarten, peaking at 85
minutes in third grade and then falling back to 75 minutes by fih grade. As students move
into the later elementary grades, they spend a lile more than a half hour on science or
social studies, which get even shorter shri in grades K–2 (25 minutes or less on each subject
per day). Time spent on other non-core subjects falls from an average of 64 minutes per day
in kindergarten to just 39 minutes per day in grade 5.
Figure 4. As elementary students progress through the grades, more time
is allocated to science and social studies, while less is allocated to ELA
and non-core subjects.
Note: “Non-core” courses include art, music, theater, dance, physical education, and foreign language. Analytic
sample includes 5,449 students. The overall mean total instructional time by grade (not shown) ranges from 293
to 303 minutes in kindergarten to fih grade. 22 percent aend “half-day” kindergarten yet receive more than
half (63 percent) of the instructional time that full-day kindergarteners receive, at 203 and 322 total minutes,
respectively. Error bars represent 95 percent confidence intervals.
Science Social StudiesMathELA Non-Core
K1 23
45
Grade
0
25
50
75
100
125
150
Minutes per Day
21
Social Studies Instruction and Reading Comprehension | IV: How is time allocated in elementary school classrooms?
Finding 3: Students from less affluent backgrounds, Hispanic students,
and those attending public schools (traditional and charter) spend more
classroom time on ELA than do other students.
Figure 5 shows how classroom time is utilized for different subjects relative to family
income.
30
Although the general paern remains the same, those from more affluent families
spend slightly less time on ELA and math than their less affluent counterparts—which may
be freeing up time for them to spend on non-core subjects. Students spend about the same
amount of time on social studies and science regardless of socioeconomic status.
Figure 5. Compared to their less affluent peers, students from more
affluent families spend slightly less classroom time on ELA and math and
slightly more on non-core subjects.
Note: The figure shows pooled averages of grades 1–5. Indicators of socioeconomic status (for example,
“highest income”) reflect quartiles of the family income distribution. “Non-core” courses include art, music,
theater, dance, physical education, and foreign language. Analytic sample includes 6,829 students. The total
mean instructional time per day (not shown) by income is as follows: 301 minutes (lowest income), 300 minutes
(below average income), 297 minutes (above average income), and 301 minutes (highest income). Error bars
represent 95 percent confidence intervals.
116
78
49
30
29
119
80
45
29
28
122
84
43
29
28
122
84
41
30
29
0
25
50
75
100
125
150
ELA Math Non-Core Science Social Studies
Subject
Minutes of Instruction per Day
Highest
Income
Above Average
Income
Below Average
Income
Lowest
Income
22
Social Studies Instruction and Reading Comprehension | IV: How is time allocated in elementary school classrooms?
Differences in class time use for students of different racial/ethnic backgrounds are
negligible, but Hispanic students do spend slightly more time on ELA than Black or White
students (Figure 6), although the differences are not statistically significant.
31
Figure 6. There are no substantial differences in instructional time usage
by race/ethnicity.
Note: The above figure shows pooled averages of grades 1–5. “Non-core” courses include art, music, theater,
dance, physical education, and foreign language. Analytic sample includes 590 Black students, 1,527 Hispanic
students, and 3,813 White students. The mean total instructional time per day (not shown) is 305 minutes for
Black students, 309 minutes for Hispanic students, and 299 minutes for White students. Error bars represent 95
percent confidence intervals.
Minutes of Instruction per Day
Black Hispanic White
117
82
41
34
33
122
84
46
31
28
119
80
45
29
28
0
25
50
75
100
125
150
ELA Math Non-Core Science
Social Studies
Subject
23
Social Studies Instruction and Reading Comprehension | IV: How is time allocated in elementary school classrooms?
Figure 7. Private schools spend much less time on reading and math and
more time on non-core subjects than schools in other sectors.
Note: The figure shows data for grades 1–5. Charter and magnet schools have a common code in the public
use ECLS-K data. Analytic sample includes 3,019 public school students, 507 private school students, and 106
magnet and charter school students. “Non-core” courses include art, music, theater, dance, physical education,
and foreign language. The mean total instructional time per day (not shown) is 303 minutes for traditional
public school students, 290 minutes for students in private schools, and 307 minutes for students in magnet
and charter schools. Error bars represent 95 percent confidence intervals.
Minutes of Instruction per Day
Magnet/Charter Private Traditional Public
117
80
43
32
32
102
68
51
31
31
122
82
42
28
27
0
25
50
75
100
125
150
175
ELA Math Non-Core Science Social Studies
Subject
Figure 7 shows how instructional time is spent in different school sectors, including
traditional public schools, private schools, and (combined) magnet and charter schools.
32
Private schools tend to spend less time on reading and math instruction and more on other
non-core subjects than do schools in other sectors, although few of these differences are
statistically significant, likely due to the smaller sample sizes. On the other hand, magnet
and charter schools spend a bit more time on social studies and science than do other
school types.
We also examined differences in instructional time by years of teacher experience (not
shown). Teachers with longer tenure (above the median of 13.5 years) tend to spend a few
more minutes per day on ELA than their less-experienced colleagues, but otherwise these
differences are negligible.
24
This section examines the potential impacts on reading improvement of how time
is spent in the classroom. First, we examine potential effects of instructional time on all
students, and then we report effects by student socioeconomic status, early reading ability,
gender, and home language.
Effects on all students
Finding 4: Increased instructional time in social studies—but not in ELA—is
associated with improved reading ability.
Figure 8 shows the effects of additional instructional time spent on each subject.
33
Contrary
to the practices of many schools, time spent on ELA—the subject that would appear most
relevant to the outcome we’re measuring—is not associated with reading improvement.
Likewise, neither math instructional time nor time spent on non-core subjects (including
art, music, and other non-core subjects) corresponds to gains or losses in reading. Although
presumably also a content-rich subject, instructional time for science has no relationship
with reading development either.
In fact, social studies is the only subject with a clear, positive, and statistically significant
effect on reading improvement. On average, students who receive an additional 30 minutes
of social studies instruction per day (roughly equivalent to moving from the 10th to the
90th percentile of social studies instructional time) in grades 1–5 outperform students with
less social studies time by 15 percent of a standard deviation on the fih-grade reading
assessment, even aer controlling for multiple measures of kindergarten reading ability and
a host of student, school, and teacher factors.
34
V: How does instructional
time relate to reading
improvement?
25
Social Studies Instruction and Reading Comprehension | V: How does instructional time relate to reading improvement?
Effects by student characteristics
As shown above, increased instructional time in social studies is correlated with increased
reading ability during elementary school. But does this effect vary by student characteristics?
We tackle that question next.
Finding 5: The students who benefit the most from additional social studies
time are girls and those from lower-income and/or non-English-speaking
homes.
First, we examine how the effect of instructional time varies by students whose families are
in different income quartiles (Figure 9). The only significant difference by socioeconomic
status (SES) is the amount of time spent in social studies. Effects are consistently positive for
students in the boom three SES quartiles but nearly zero and statistically insignificant for
students in the wealthiest quartile. More specifically, students in the boom three quartiles
Figure 8. More instructional time devoted to social studies is correlated
with greater reading growth from first through fifth grade.
Note: Analytic sample includes 6,731 students. Effects are in standard deviations of fih-grade assessment
scores. For example, the first bar indicates that the effect of thirty minutes of additional ELA daily instruction is
associated with a 3 percent of a standard deviation increase in student reading progress from kindergarten to
fih grade. However, because the error bars overlap with the baseline (0 percent), this effect is not statistically
significantly different from zero. Contrast that with the 15 percent effect for social studies, where the error bars
do not overlap with the baseline, indicating that the result is statistically significantly different from zero. Error
bars represent 95 percent confidence intervals.
Percent of Standard Deviation Reading
Test Score Improvement for Thirty
Additional Minutes of Daily Instruction
3
6
3
-1
15
-40%
-20%
0%
20%
40%
ELA Math Non-Core Science
Social Studies
Subject
26
Social Studies Instruction and Reading Comprehension | V: How does instructional time relate to reading improvement?
have similar positive effects from an additional 30 minutes of daily social studies instruction
during elementary school, corresponding to greater reading development between 17 and 21
percent of a standard deviation.
There are generally no statistically significant differences for students based on the amount
of ELA, math, non-core, or science instructional time, regardless of SES (the one exception
is for students in the most affluent quartile, for whom additional science instruction is
positive).
Figure 9. More instructional time in social studies is related to greater
reading growth from first through fifth grade for all students except those
in the top income quartile.
Note: Analytic sample includes 6,731 students. Indicators of socioeconomic status (for example, “affluent”)
reflect quartiles of the family income distribution. Effects are in standard deviations of fih-grade assessment
scores. Note that the 17 percent of a standard deviation effect for students in the “below average income
quartile” is only statistically significant at the 90 percent confidence level. Error bars represent 95 percent
confidence intervals.
Percent of Standard Deviation Reading
Test Score Improvement for Thirty
Additional Minutes of Daily Instruction
ELA Math Non-Core Science
Social Sciences
Subject
1
13
4
-5
21
2
8
6
-9
17
6
-4
2
-3
20
2
4
0
13
-3
-40%
-20%
0%
20%
40%
Highest
Income
Above Average
Income
Below Average
Income
Lowest
Income
27
Social Studies Instruction and Reading Comprehension | V: How does instructional time relate to reading improvement?
Figure 10 shows the effects of instructional time for students by gender. Additional social
studies instruction for boys and girls provides similar benefit, but the effect for girls is a bit
stronger. There are no statistically significant effects of instructional time in other subjects
for students of either gender.
Figure 10. For both boys and girls, additional social studies time is
associated with greater progress in reading.
Note: Analytic sample includes 6,731 students. Effects are in standard deviations of fih-grade assessment
scores. Error bars represent 95 percent confidence intervals.
Percent of Standard Deviation Reading
Test Score Improvement for Thirty
Additional Minutes of Daily Instruction
Girls Boys
4
4
5
-3
18
3
7
5
2
12
-40%
-20%
0%
20%
40%
ELA Math Non-Core Science
Social Studies
Subject
Finally, we examine the effects of instructional time by home language. Students from
homes in which English is not the primary language see larger effects from social studies
instructional time than do students from homes where it is (Figure 11). For students from
non-English-speaking homes, an additional 30 minutes of social studies time per day
during elementary school corresponds to about a quarter of a standard deviation increase
in reading ability. For students from primarily English-speaking families, that same 30
additional minutes corresponds to an improvement in reading of about 12 percent of
a standard deviation (statistically significant only at the 90 percent confidence level).
Interestingly, the effects of additional ELA instructional time are small and statistically
insignificant for both groups, although the estimated effect on students from non-English-
speaking homes is about double that for other students (7 versus 3 percent of a standard
deviation).
28
Social Studies Instruction and Reading Comprehension | V: How does instructional time relate to reading improvement?
Figure 11. Additional social studies time is tied to greater progress in
reading regardless of students’ home language, although the effect is
stronger for students in non-English-speaking homes.
Note: Analytic sample includes 6,731 students. Effects are in standard deviations of fih-grade assessment
scores. Error bars represent 95 percent confidence intervals.
Percent of Standard Deviation Reading
Test Score Improvement for Thirty
Additional Minutes of Daily Instruction
English Non-English
3
5
6
3
12
7
11
3
-7
25
-40%
-20%
0%
20%
40%
ELA Math Non-Core Science
Social Studies
Subject
29
VI: Summary and
implications
Here’s what we’ve learned: Devoting more instructional time to social studies is
associated with increased reading ability over the course of five years of elementary school.
This is true for all but the most affluent students. Female and low-income students, as
well as those from non-English-speaking families, are especially likely to benefit. On the
other hand, increased instructional time in math, non-core, science, and—crucially—ELA is
generally not associated with more reading improvement.
Spread out over first through fih grade, these impacts are modest but suggest that, at the
margin, students are not geing additional benefit from lengthy periods of ELA instruction.
We draw three lessons from these findings.
1) Elementary schools should make more room for high-quality social
studies instruction.
In the average elementary classroom, students spend two hours on language arts, 28
minutes on social studies, and even less time in art, music, or foreign language, despite
the fact that additional social studies time is associated with improved reading scores. Our
findings imply that shiing twenty, thirty, or even forty minutes away from less effective
ELA activities (such as practicing comprehension skills) and reinvesting that time to learn
more about geography, history, civics, and the like will improve students’ reading ability.
Just as important, additional social studies time will probably also help students develop
the strong knowledge base needed for a successful transition to middle school. That’s
why organizations like the Council of Chief State School Officers recommend elementary
classrooms dedicate at least 45 minutes to social studies each day.
35
The link between social studies and reading may stem from the way that social studies
instruction can help build systematic knowledge and vocabulary in multiple domains, which
are broadly applicable and transferable to other topics. Social studies can help students
understand history, current events, family and social relationships, and common narratives;
whereas, reading passages that putatively cover other subjects, such as literature or drama,
may assume the reader already has a grasp of such knowledge. Although many have oen
30
Social Studies Instruction and Reading Comprehension | VI: Summary and implications
considered science to be another content-rich subject,
36
the inherent vocabulary in science
classes may be too specialized to impact literacy broadly.
Unfortunately, social studies has long been neglected in American primary schools.
Elementary teachers are oen taught that students should “first learn to read, so they can
read to learn,” even though youngsters can learn a lot about the world before they can
decode (who hasn’t seen animated kindergarteners wrapped in suspense as their teachers
share aloud well-wrien historical and mythical narratives?). When schools do teach social
studies, it is oen so watered down or devoid of controversy that it neither builds knowledge
nor captures student interest.
37
Given the positive impacts of today’s oen mediocre social
studies instruction, imagine how students might benefit if we delivered even more engaging
instruction in history, geography, and civics. Moreover, because the effects of social studies
instruction are strongest for students in lower socioeconomic classes, more time in the
subject may help close reading gaps.
2) Teachers should use the literacy block efficiently to build student
knowledge.
Whether or not schools trim their sprawling literacy blocks to make room for more instruction
in social studies and other content-rich subjects, they can still make beer use of that time
by infusing it with content-rich texts and topics.
Curriculum advocates are increasingly aware of the importance of building knowledge. In
fact, EdReports, a nonprofit group that provides reviews of K–12 instructional materials,
includes in its evaluations a measure of knowledge richness (“building knowledge with
texts, vocabulary, and tasks”). For example, it gives high marks to the ELA curriculum Wit
& Wisdom, noting that it “provide[s] ample opportunities for students to build knowledge
through content-rich, integrated reading, writing, speaking, listening, and language
experiences.”
38
Encouraging schools to use such curricula could be transformative to the
reading block.
Regardless of whether a school embraces a specifically packaged curriculum, teachers can
still play a crucial role in beefing up the literacy block. The notion that students’ choice of
reading material boosts intrinsic motivation is well grounded in research,
39
but that doesn’t
mean that students don’t benefit from teacher guidance. Allowing them to choose between
a handful of teacher-selected, high-quality texts on various topics fosters buy-in and may
inspire new interests. One way that teachers curate is through the use of “text sets,” which
facilitate deep learning around specific topics.
40
31
Social Studies Instruction and Reading Comprehension | VI: Summary and implications
3) Policymakers and administrators should align reading assessments with
curricular content.
That improving students’ reading ability stems more from exposure to additional social
studies than to more English language arts suggests a rethinking of how we measure
literacy. Reading assessments—including the ECLS-K assessment used in this report—tend
to focus on abstract reading skills. Such skills-centric assessments nonetheless require
background knowledge, as they include reading passages to assess comprehension. But
because those passages may relate to any conceivable topic, they can incentivize classroom
teachers to focus on test-prep and reading strategies. In turn, districts oen choose curricula
with the same orientation that prizes skills over rich content.
What we need, in the words of Hirsch and Pondiscio, is a “domain-specific approach to
language arts.” In other words, “rather than idle away precious hours on trivial stories or
randomly chosen nonfiction, reading, writing, and listening instruction would be built into
the study of ancient civilizations in first grade, for example, Greek mythology in second, or
the human body in third.”
41
In this way, reading tests would subsequently include texts from a variety of genres on these
same topics. Cue Hirsch and Pondiscio: “Teachers would still teach to the test, emphasizing
domain-specific knowledge (because it might be on the test), but no one would object,
because it would help students not only pass the current year’s test but build the broad
background knowledge that enables them to become stronger readers in general.”
42
Of
course, “no one” objecting is a high bar, but we can imagine that many teachers might
embrace an assessment that reflects some of the particular texts and content topics they’ve
covered during the year.
More specifically, this new generation of reading assessments would, in the earliest grades,
use rich texts to target decoding and basic comprehension skills. Then, as students master
the basics, the focus would shi to broadening vocabulary, reading comprehension, and
writing about specific topics, including science, art, geography, and history.
Louisiana is taking a step in this direction by pioneering a set of reading assessments that
align with the state’s social studies curricula. For now, it is limited to a few districts and
charter networks in the Bayou state—but, if fruitful, such an approach could have much
broader impacts for literacy across the nation.
43
32
Social Studies Instruction and Reading Comprehension | VI: Summary and implications
We are not the first to find that the expansive time devoted to language arts instruction does
not improve student reading.
44
But we are the first to find that literacy gains are more apt to
materialize when students spend more time in social studies. Diminishing or transforming
the ELA block, then, should move to the top of the literacy agenda. Yes, we should keep
fighting for high-quality phonics-based instruction in the earliest grades, but that must
be complemented by a diet rich in content that grabs the hearts and imaginations of our
youngest learners.
33
Appendix A:
Technical appendix
Time-use measure
The ECLS-K teacher surveys have been used in numerous studies to test the relationship
between classroom content and student outcomes.
45
Surveys like these are subject to
some degree of bias from both nonresponse and respondent bias. ECLS-K corrects for
nonresponse bias through the adjustment of sample weights postsurvey, and the weighted
response rates do not reflect nonresponse bias.
46
Overall, response rates were high among teachers across questionnaires. For example,
in fih grade, teacher response rates average approximately 82 percent of the weighted
sample. In order to ensure that unusual response paerns did not drive results, we
eliminated teacher responses where the daily instructional time averaged less than two
hours or more than eight hours and include only students for whom a majority of years
of data are valid and within this range. Respondent bias does not affect sample size but
does generate measurement error, which aenuates results and could possibly affect their
interpretation.
Analyses
In Section V, we present the results of regression analyses linking instructional time and
student reading development. Although these models are described briefly in Section III, we
discuss them in detail below.
The models with results presented in Section V include fih-grade reading scores as the
outcome variable and instructional time by subject as the key independent variables. We
control for fall and spring kindergarten reading assessments (and an interaction between
those assessment scores); kindergarten assessments in math, science, and in executive
functioning; total instructional time; and a range of student, school, and teacher factors.
Student socioeconomic status is based on the ECLS-K variable x12sesl, which is a continuous
measure of family socioeconomic status. The SES indicator in the ECLS-K: 2011 reflects
the socioeconomic status of the household at the time of data collection and includes
34
Social Studies Instruction and Reading Comprehension | Appendix A: Technical appendix
five components: parent/guardian 1’s education, parent/guardian 2’s education, parent/
guardian 1’s occupational prestige score, parent/guardian 2’s occupational prestige score,
and household income.
47
The main models, presented in Section V, take the following form:
(1) Y
s
= β
0
+ β
1
InstructionalTime
s
+ β
2
KinderReading
s
+ β
3
TotalInstructionalTime
s
+
β
4
StudentDemo
s
+ β
5
School
st
+ β
6
Teacher
st
+ ε
st
where Y
s
is fih-grade reading performance for student s. InstructionalTime
s
is a vector of
classroom time-use variables including ELA, math, social studies, science, and a pooled
measure of the time spent on other “non-core” subjects including art, music, dance, theater,
PE, and foreign language. KinderReading
s
is a vector of (lagged) kindergarten reading
measures for student s, including a fall reading assessment, a spring reading assessment,
and an interaction term for the two kindergarten assessments. TotalInstructionalTime
s
represents the total amount of instructional time, which is a sum of the InstructionalTime
s
measures for each student. StudentDemo
s
is a vector of student-level control variables,
including the SES variable described above; age at kindergarten matriculation; race;
disability status; and dummies representing whether the student’s home language is
English and whether the student aended a pre-kindergarten education program. School
is a vector of school-level factors including whether the school is private, traditional public,
or public charter or magnet in each school year and predominant urbanicity of the schools
the student aended during elementary school (urban, suburban town, or rural). Teacher
s
is a vector of teacher factors including the average length of students’ teacher’s experience
from grades 1–5 and dummies representing whether the student’s teacher concentrates in a
specific subject in grades 3, 4, or 5. The coefficient on ß
1
for each academic subject provides
the average effects of additional instructional time on student reading development.
Heterogeneous effects
To examine heterogeneity in the observed effects, we run additional analyses to test
differences by SES, gender, and English-language status. In these models, we transform the
continuous SES measure into a quartile categorical variable, where 1 is the lowest quartile
group and 4 is the highest. Gender remains a binary variable, where 1 equals female. Our
indicator of home language is a binary variable where 1 indicates that a language other
than English is spoken at home (gleaned from the kindergarten parent survey). We run the
following model for these analyses:
(2) Y
s
= β
0
+ β
1
InstructionalTime
s
+ β
2
SES
s
+ β
3
InstructionalTime*SES
s
+ β
3
KinderReading
s
+ β
4
TotalInstructionalTime
s
+ β
5
StudentDemo
s
+ β
6
School
st
+ β
7
Teacher
st
+ ε
st
35
Social Studies Instruction and Reading Comprehension | Appendix A: Technical appendix
(3) Y
s
= β
0
+ β
1
InstructionalTime
s
+ β
2
Gender
s
+ β
3
InstructionalTime*Genders
s
+
β
3
KinderReading
s
+ β
4
TotalInstructionalTime
s
+ β
5
StudentDemo
s
+ β
6
School
st
+
β
7
Teacher
st
+ ε
st
(4) Y
s
= β
0
+ β
1
InstructionalTime
s
+ β
2
HomeLang
s
+ β
3
InstructionalTime*HomeLang
s
+
β
3
KinderReading
s
+ β
4
TotalInstructionalTime
s
+ β
5
StudentDemo
s
+ β
6
School
st
+
β
7
Teacher
st
+ ε
st
All coefficients remain the same as the main model, except SES, gender, and home
language are in turn removed from the StudentDemo
s
vector and interacted with the main
variables of interest, InstructionalTime
s
.
36
Appendix B:
Robustness check
An advantage of this study is its relatively long horizon, which allows us to analyze
how more time in content-rich subjects may accumulate over five years of elementary
school to improve reading outcomes. Still, a critical reader may wonder whether this long-
term correlation is driven by beer readers needing—and receiving—less ELA instruction,
resulting in more time for instruction in social studies and other subjects. Because our
analysis controls for multiple measures of kindergarten reading ability, it cannot be the case
that differences in initial reading ability are driving the correlation. But what if students who
accelerate fastest in reading in the early grades are subsequently instructed less in ELA and
more in social studies in the later grades? If this were true, it would explain the correlation
between reading improvement and social studies instruction but would suggest that the
arrow of causality runs in the opposite direction. In other words, it would imply that social
studies instruction is not influencing student reading abilities. This appendix presents the
results of this robustness check.
Specifically, we test this question: Do students who make the most progress in reading
in early grades receive less ELA instruction—and more social studies instruction—in later
grades?
We find that students who make substantial progress in reading in the early grades receive
very similar amounts of reading and social studies instruction as students who make very
lile progress. Table B1 shows that, compared to third-grade students in the 1st percentile
of prior reading growth, third-grade students in the 99th percentile of prior reading growth
have very similar amounts of instructional time in ELA and social studies, receiving just six
fewer minutes in ELA and less than one additional minute in social studies. Neither of these
differences is statistically significant (the final two rows of the table replicate the analysis for
kindergarten through fourth grade, where the differences are even smaller).
Because instructional time is measured by teacher responses to surveys about their
classrooms in general, it makes sense that student progress is not a good predictor of
instructional time in later grades. Given these similarities in later instructional time for
37
Social Studies Instruction and Reading Comprehension | Appendix B: Robustness check
students at the furthest extremes of the growth spectrum, we can be confident that the
correlations between social studies instruction and reading improvement identified in the
main report are not driven by this potentially confounding paern.
Table B1. Student reading growth in earlier grades is not a good predictor
of instructional time in later grades.
Note: Analytic sample includes 6,115 to 6,350 students, depending on the grade level. Results are from
regression models estimating effects of reading growth in earlier grades on instructional time in later grades,
controlling for total instructional time in the later grades. None of the differences are statistically significant.
Students in 1st
percentile growth in
early grades
(1)
Students in 99th
percentile growth in
early grades
(2)
Difference
(3)
ELA (3rd grade) 121.3 minutes 114.9 minutes −6.4
Social studies (3rd grade) 26.7 minutes 27.6 minutes +0.9
ELA (4th grade) 115.8 minutes 116.0 minutes +0.2
Social studies (4th grade) 31.2 minutes 32.5 minutes +1.3
38
Endnotes
1 Donna R. Recht and Lauren Leslie, “Effect of Prior Knowledge on Good and Poor Readers’ Memory of
Text,” Journal of Educational Psychology 80, no.1 (1988): 16–20, doi:10.1037/0022-0663.80.1.16; Katherine
A. Rawson and Walter Kintsch, “Exploring Encoding and Retrieval Effects of Background Information
on Text Memory,” Discourse Processes 38, no. 3 (2004): 323–44, doi:10.1207/ s15326950dp3803_3; Carol
McDonald Connor et al., “Acquiring Science and Social Studies Knowledge in Kindergarten Through
Fourth Grade: Conceptualization, Design, Implementation, and Efficacy Testing of Content-Area
Literacy Instruction (CALI),” Journal of Educational Psychology 109, no. 3 (2017): 301–20, doi:10.1037/
edu0000128.
2 Eric Donald Hirsch Jr., Joseph F. Ke, and James S. Trefil. Cultural Literacy: What Every American Needs
to Know (Boston: Houghton Mifflin, 1987), xiii.
3 See endnote 21.
4 The final analytic sample includes 6,829 students.
5 “Core” refers to ELA, math, science, and social studies, and “non-core” refers to arts and music,
physical education (PE), and foreign language. These descriptors are not meant to connote greater or
lesser importance of certain courses over others.
6 Because the amount of classroom time devoted to different subjects is not randomly assigned and
may correlate with other factors for which we are unable to control—such as teacher quality—it is
possible that one or more of these factors is driving a correlation between class time usage and
student learning. Still, a finding that different use of class time is correlated with greater improvement
in reading aer controlling for various potential confounders could imply a causal effect or, at the very
least, suggest how beer teachers employ instructional time.
7 The one exception is for students in the most affluent quartile, for whom additional science
instruction is positive.
8 National Assessment of Educational Progress, 2019.
9 For a discussion of contemporary controversies around the use of direct phonics instruction, see Emily
Hanford, “Hard Words: Why aren’t kids being taught to read?,” American Public Media, September
10, 2018, https://www.apmreports.org/story/2018/09/10/hard-words-why-american-kids-arent-being-
taught-to-read.
10 Kathleen M. Hoyer and Dinah Sparks, “Instructional Time for Third- and Eighth-Graders in Public
and Private Schools: School Year 2011–12,” Statistics in Brief (Washington, D.C.: National Center for
Education Statistics, 2017), https://nces.ed.gov/pubs2017/2017076.pdf.
11 American schools may spend more time on literacy than any OECD country. As we report in
Section IV, American elementary classrooms spend 40 percent of instructional time on literacy. The
Organization for Economic Cooperation and Development (OECD) does not report this outcome for
the U.S., but their studies indicate that the countries with the greatest proportion of time spent on
literacy are France (37 percent) and Mexico (35 percent), both of which spend a smaller proportion
of instructional time on literacy than the U.S. Many OECD countries seem to spend far less time
on elementary literacy instruction than the U.S., including Canada (27 percent), Japan (24 percent),
Finland (24 percent), and Germany (20 percent). “Chart D1.2a. Instruction time per subject in primary
39
Social Studies Instruction and Reading Comprehension | Endnotes
education (2014),” Education at a Glance 2014: OECD Indicators, OECD Publishing, https://doi.
org/10.1787/eag-2014-en.
12 See Erik W. Robelen, “Most Teachers See the Curriculum Narrowing, Survey Finds,” Curriculum
Maers, Education Week, December 8, 2011, http://blogs.edweek.org/edweek/curriculum/2011/12/
most_teachers_see_the_curricul.html. Although federal accountability policies, beginning with the
No Child Le Behind Act of 2001, may have contributed to the emphasis on reading over potentially
more content-rich subjects such as science and social studies, the disproportionate amount of time
spent on reading instruction predates these changes to federal law. A study from the late 1970s shows
that in kindergarten through third grade, more instructional time was spent on reading than on math,
social studies, and science combined. In fourth through sixth grade, the distribution of instructional
time was more evenly distributed across core subjects, but more than seventy minutes per day
was still spent on reading in those grades. Iris Weiss, Report of the 1977 National Survey of Science,
Mathematics, and Social Studies Education (Research Triangle Park, NC: Center for Educational
Research and Evaluation, 1978), https://files.eric.ed.gov/fulltext/ED152566.pdf.
13 Natalie Wexler, The Knowledge Gap: The Hidden Cause of America’s Broken Education System–and
How to Fix It (New York, NY: Avery, 2019).
14 The What Works Clearinghouse Practice Guide on reading comprehension in grades K to 3 gives
explicit teaching of comprehension strategies a “strong” evidence rating. Timothy Shanahan, Kim
Callison, Christine Carriere, Nell Duke, P. David Pearson, Christopher Schatschneider, and Joseph
Torgesen, Improving reading comprehension in kindergarten through 3rd grade: A practice guide
(NCEE 2010-4038), Washington, D.C.: National Center for Education Evaluation and Regional
Assistance, Institute of Education Sciences, U.S. Department of Education, whatworks.ed.gov/
publications/practiceguides.
15 In later grades, assessments generally become more sophisticated, shiing from low-level
“decoding” skills to passages that require greater context knowledge. See, Chris Schatschneider
et al., A Multivariate Study of Individual Differences in Performance on the Reading Portion of the
Florida Comprehensive Assessment Test: A Brief Report (Tallahassee, FL: Florida Center for Reading
Research, 2004), hps://eric.ed.gov/?id=ED495465.
16 Lowry Hemphill and Terrance Tivnan, “The Importance of Early Vocabulary for Literacy Achievement
in High-Poverty Schools,” Journal of Education for Students Placed at Risk 13, no.4 (2008): 426–51,
doi:10.1080/10824660802427710.
17 For a review of these studies, see Donna R. Recht and Lauren Leslie, “Effect of Prior Knowledge on
Good and Poor Readers’ Memory of Text,” Journal of Educational Psychology 80, no.1 (1988): 16–20,
doi:10.1037/0022-0663.80.1.16.
18 Recht and Leslie, “Effect of Prior Knowledge,” 16–20.
19 Katherine A.Rawson and Walter Kintsch, “Exploring Encoding and Retrieval Effects of Background
Information on Text Memory,” Discourse Processes 38, no.3 (2004): 323–44, doi:10.1207/
s15326950dp3803_3.
20 Carol McDonald Connor et al., “Acquiring Science and Social Studies Knowledge in Kindergarten
Through Fourth Grade: Conceptualization, Design, Implementation, and Efficacy Testing of Content-
Area Literacy Instruction (CALI),” Journal of Educational Psychology 109, no.3 (2017): 301–20, doi:10.1037/
edu0000128.
21 For example, a study of the CKLA curriculum in New York City showed that CKLA students had more
than double the fall-to-spring reading growth of a matched sample of students in the city and
outperformed the control group over three years, as well. Additionally, researchers from Florida State
University are currently conducting a five-year, $3.2 million Institute of Education Sciences–funded
study on the effectiveness of the CKLA curriculum for students in kindergarten through second grade.
See https://ies.ed.gov/funding/grantsearch/details.asp?ID=1791. Core Knowledge, “The NYC Core
Knowledge Early Literacy Pilot: K–Grade 2 Results,” New York City Department of Education, 2012,
https://www.coreknowledge.org/wp-content/uploads/2016/12/CK-Early-Literacy-Pilot-3-12-121.pdf.
See also Robert Bradley, “Evaluation of the Core Knowledge Program in Arkansas” (Charloesville,
40
Social Studies Instruction and Reading Comprehension | Endnotes
VA: 2005), https://www.coreknowledge.org/wp-content/uploads/2016/12/CK_Eval_Rpt_AR2005.
pdf; Susan Sonnenschein, Linda Baker, and Adia Garre, “An Analysis of Academic Progress of
Children Participating in the Core Knowledge Preschool Program in Baltimore County Head Start
Centers,” Core Knowledge, 2005, https://www.coreknowledge.org/wp-content/uploads/2016/12/
Baltimore_Report.pdf; John Wedman and Alex Waigandt, “Core Knowledge Curriculum and School
Performance: A National Study,” Core Knowledge, 2004, https://www.coreknowledge.org/wp-content/
uploads/2016/12/CK_National_Study_2004.pdf; Herbert J. Walberg and Joseph Meyer, “The Effects of
Core Knowledge on State Test Achievement in North Carolina,” Core Knowledge, 2004, https://www.
coreknowledge.org/wp-content/uploads/2016/12/Walberg_phase3.pdf; and Martha A. Mac Iver, Sam
Stringfield, and Barbara McHugh, Core Knowledge Curriculum: Five-Year Analysis of Implementation
and Effects in Five Maryland Schools (Baltimore, MD: Center for Research on the Education of
Students Placed At Risk, 2000), https://www.coreknowledge.org/wp-content/uploads/2016/12/
FiveYearEffects_Maryland_2000.pdf.
22 Katherine A. Magnuson, Christopher Ruhm, and Jane Waldfogel, “The Persistence of Preschool
Effects: Do Subsequent Classroom Experiences Maer?,” Early Childhood Research Quarterly 22, no.1
(2007): 18–38.
23 The most comprehensive information available on the content of the assessments comes from the
ECLS-K psychometric reports. These reports indicate that “the ECLS-K:2011 reading framework
was modeled aer the NAEP 2011 reading framework.” Regarding the content of the assessments,
these reports indicate that “the number of informational texts and their level of sophistication
increases gradually in the ECLS-K:2011 testing baery, such that in second, third, fourth, and fih
grades, approximately two-thirds of the passages are literary texts and one-third of the passages
are informational texts.” Michelle Najarian et al., Early Childhood Longitudinal Study, Kindergarten
Class of 2010–11: Third-Grade, Fourth-Grade, and Fih-Grade Psychometric Report, NCES 2020-123
(Washington, D.C.: U.S. Department of Education, National Center for Education Statistics, 2020),
hps://nces.ed.gov/pubs2020/2020123.pdf. See also Karen Tourangeau et al., Early Childhood
Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K: 2011): User’s Manual for the ECLS-K:2011
Kindergarten–Fih Grade Data File and Electronic Codebook, Public Version, NCES 2019-051
(Washington, D.C.: U.S. Department of Education, National Center for Education Statistics, 2019),
https://nces.ed.gov/pubs2019/2019051.pdf.
24 The exact wording of the first question is as follows: “How oen does the typical child in your class
usually work on lessons or projects in the following general subject areas, whether as a whole class, in
small groups, or in individualized arrangements?” For each subject (ELA, math, social studies, science,
music, art, physical education, dance, theater, and foreign language), the teacher indicates one of
the following options: “never,” “less than once a week,” “1 day a week,” “2 days a week,” “3 days a
week,” “4 days a week,” or “5 days a week.” The exact wording of the second question is as follows:
“On the days children work in these areas, how much time does the typical child in your class usually
work on lessons or projects in the following general subject areas?” For each subject (see above), the
teacher indicates one of the following options: “not applicable/never,” “less than ½ an hour a day,”
“½ hour to less than 1 hour,” “1 to less than 1 ½ hours,” “1 ½ to less than 2 hours,” “2 to less than 2 ½
hours,” “2 ½ hours to less than 3 hours,” or “3 hours or more.” We combine responses for art, dance/
creative movement, and theater/creative dramatics into one scale measure of “arts and music.” Our
construction of these variables follows Mimi Engel, Amy Claessens, and Maida A. Finch, “Teaching
students what they already know? The (Mis)Alignment Between Mathematics Instructional Content
and Student Knowledge in Kindergarten,” Educational Evaluation and Policy Analysis 35, no. 2 (2013):
157–78, doi:10.3102/0162373712461850.
25 Although information about how instructional time is used can help us understand how much time is
spent on different subjects in different types of schools, no information is available on the content of
the subjects, curricula, or how instructional practice may emphasize different forms of content.
41
Social Studies Instruction and Reading Comprehension | Endnotes
26 The ECLS-K does not distinguish between charter schools and magnet schools in the public-use files
analyzed for this report.
27 This variable indicates whether there were some core subjects that a teacher did not teach. Although
some teachers do not teach all subjects, they are asked to report overall instructional time use for
the classroom, not just the instructional time for which they are personally responsible. Because
these teachers may report instructional time differently than teachers who teach all the core subjects
themselves, this control variable accounts for some of these potential differences in response
paerns.
28 The ECLS-K survey solicits information about art, music, dance, and theater, which we combine here.
29 Throughout this report, we use the term “core” to refer to ELA, math, science, and social studies and
the term “non-core” to refer to arts and music, physical education (PE), and foreign language. These
descriptors are not meant to connote greater or lesser importance of certain courses over others.
30 Economic quartiles are for the families of the sample of students, which is representative of the
national student population but not the broader population.
31 Additional ELA time for Hispanic students is likely driven by the fact that 53 percent of Hispanic
students come from homes where English is not the primary language and therefore may be
receiving supplemental ELA instruction, whereas the percentages of students coming from
non-English-speaking homes is smaller for Black and White students (4 percent and 2 percent,
respectively).
32 Charter and magnet schools have a common code in the public-use ECLS-K data.
33 Because there may be increasing or diminishing returns to additional instructional time for a given
subject, we also examined nonlinear relationships between instructional time use and reading
development by including quadratic transformation of each subject variable, but none of those terms
were statistically significant.
34 A discerning reader may wonder whether this correlation over five years is driven by beer readers
needing—and receiving—less ELA instruction, resulting in more time for instruction in social studies
and other subjects. We address this potential threat to the validity of our results in Appendix B:
Robustness check.
35 “The Marginalization of Social Studies,” Council of Chief State School Officers, 2018, hps://ccsso.org/
sites/default/files/2018-11/Elementary%20SS%20Brief%2045%20Minute%20Version_0.pdf.
36 For example, Wexler cites a teacher concluding from a lecture by psychology professor Dan
Willingham that “pulling kids out of science and social studies to give them extra help in reading skills
is actually ‘the worst thing you can do’” (p. 57).
37 For example, Wexler bemoans the “standard ‘me and my community’ social studies curriculum” of
elementary schools, and Fordham’s Chester Finn has targeted the academic guidelines promulgated
by the National Council of Social Studies as “avowedly, even proudly, devoid of all content.” See
Wexler, The Knowledge Gap, 18, and Chester E. Finn, Jr., “Social studies and poison gas,” Flypaper
(blog), September 18, 2013, https://fordhaminstitute.org/national/commentary/social-studies-and-
poison-gas.
38 Also receiving high marks from EdReports on this measure is Hirsch’s open-source Core Knowledge
Language Arts curriculum. “Wit & Wisdom (2016): Series Overview,” EdReports, accessed July 27, 2020,
https://www.edreports.org/reports/overview/wit-wisdom-2016.
39 Alexandra Usher and Nancy Kober, Summary: Student Motivation—An Overlooked Piece of School
Reform (Washington, D.C.: Center on Education Policy, 2012).
40 Find more about text sets from Achieve the Core, Newsela, and ReadWorks at the following websites,
respectively: https://achievethecore.org, https://newsela.com, and https://www.readworks.org.
41 E.D. Hirsch and Robert Pondiscio, “There’s no such thing as a reading test,” The American Prospect,
June 14, 2010, https://prospect.org/special-report/thing-reading-test.
42 Ibid.
42
Social Studies Instruction and Reading Comprehension | Endnotes
43 John C. White, “States don’t measure what kids actually know. That needs to change,” The Hill, April 3,
2018, hps://thehill.com/opinion/education/381285-states-dont-measure-what-kids-actually-know-
that-needs-to-change.
44 Magnuson et al., “The Persistence of Preschool Effects.”
45 Daphna Bassok, Sco Latham, and Anna Rorem, “Is Kindergarten the New First Grade?” AERA
Open 2, no.1 (2016), doi:10.1177/2332858415616358; Engel et al., “Teaching Students what They Already
Know?”; and Vi-Nhuan Le, Diana Schaack, Kristen Neishi, Marc W. Hernandez, and Rolf Blank,
“Advanced Content Coverage at Kindergarten: Are There Trade-Offs Between Academic Achievement
and Social-Emotional Skills?” American Educational Research Journal 56, no.4 (2019): 1254–80.
46 Tourangeau et al., Early Childhood Longitudinal Study.
47 Ibid.