ArticlePDF Available

How to Make a Young Child Smarter: Evidence From the Database of Raising Intelligence

Authors:

Abstract and Figures

Can interventions meaningfully increase intelligence? If so, how? The Database of Raising Intelligence is a continuously updated compendium of randomized controlled trials that were designed to increase intelligence. In this article, the authors examine nearly every available intervention involving children from birth to kindergarten, using meta-analytic procedures when more than 3 studies tested similar methods and reviewing interventions when too few were available for meta-analysis. This yielded 4 meta-analyses on the effects of dietary supplementation to pregnant mothers and neonates, early educational interventions, interactive reading, and sending a child to preschool. All 4 meta-analyses yielded significant results: Supplementing infants with long-chain polyunsaturated fatty acids, enrolling children in early educational interventions, reading to children in an interactive manner, and sending children to preschool all raise the intelligence of young children. © The Author(s) 2013.
Content may be subject to copyright.
http://pps.sagepub.com/
Science
Perspectives on Psychological
http://pps.sagepub.com/content/8/1/25
The online version of this article can be found at:
DOI: 10.1177/1745691612462585
2013 8: 25Perspectives on Psychological Science
John Protzko, Joshua Aronson and Clancy Blair
How to Make a Young Child Smarter : Evidence From the Database of Raising Intelligence
Published by:
http://www.sagepublications.com
On behalf of:
Association For Psychological Science
can be found at:Perspectives on Psychological ScienceAdditional services and information for
http://pps.sagepub.com/cgi/alertsEmail Alerts:
http://pps.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
Perspectives on Psychological Science
8(1) 25 –40
© The Author(s) 2013
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1745691612462585
http://pps.sagepub.com
A good deal of research confirms what most people consider
self-evident: Intelligence matters for academic and life suc-
cess (Herrnstein & Murray, 1994; Nisbett et al., 2012). Accord-
ingly, many researchers and educators have attempted to
increase the intelligence of children—particularly children
from disadvantaged backgrounds. The question of the modifi-
ability of intelligence remains a contentious one, and findings
from these famous interventions are nuanced enough to sup-
port the conclusions of either side in the debate about the
nature and nurture of intelligence. Thus, effective interven-
tions are cited as evidence by “IQ environmentalists” to make
the case that intervention can raise IQ (e.g., Nisbett, 2009),
whereas “IQ nativists” cite the same results when making the
case that such gains are fleeting or even illusory (e.g., Herrn-
stein & Murray, 1994). We believe that a more complete con-
sideration of the best available data is required if we are to
determine whether interventions can meaningfully raise IQ
and, if they can, to consider how we might go about construct-
ing these interventions. To this end, we have compiled all
available and relevant high-quality studies into a central loca-
tion, the Database of Raising Intelligence (DORI).
Database of Raising Intelligence
The DORI is a continuously updated compendium of every
randomized controlled trial (RCT) designed to increase intel-
ligence across all age levels. To be included in the database, a
study has to include each of the following components: a
sample drawn from a general, nonclinical population; a pure
randomized controlled experimental design; a sustained inter-
vention; and a widely accepted, standardized measure of intel-
ligence as an outcome variable.
Our first criterion is that the participants must have been
drawn from the general population. Although data from clini-
cal populations can be informative, generalizing the effects of
interventions designed for clinical populations to nonclinical
populations is problematic. An intervention that helps some-
one overcome his or her disabilities or intellectual deficits
(e.g., Klingberg et al., 2005, on training working memory of
children with attention deficit/hyperactivity disorder) may not
have the same effects for members of a nonclinical population.
In addition, the amelioration of mental retardation may entail
different processes than the general increase of intelligence
(Spitz, 1986).
Our second criterion is that the authors must have used a
pure RCT. At the beginning of the study, every participant
must have had an equal chance of being assigned to an inter-
vention or a control group. We thus exclude studies that first
enrolled an experimental group and then later enrolled a
Corresponding Author:
John Protzko, New York University, 82 Washington Square East, New York,
NY 10003
E-mail: protzko@gmail.com
How to Make a Young Child Smarter:
Evidence From the Database of
Raising Intelligence
John Protzko, Joshua Aronson, and Clancy Blair
New York University
Abstract
Can interventions meaningfully increase intelligence? If so, how? The Database of Raising Intelligence is a continuously updated
compendium of randomized controlled trials that were designed to increase intelligence. In this article, the authors examine
nearly every available intervention involving children from birth to kindergarten, using meta-analytic procedures when more
than 3 studies tested similar methods and reviewing interventions when too few were available for meta-analysis. This yielded
4 meta-analyses on the effects of dietary supplementation to pregnant mothers and neonates, early educational interventions,
interactive reading, and sending a child to preschool. All 4 meta-analyses yielded significant results: Supplementing infants with
long-chain polyunsaturated fatty acids, enrolling children in early educational interventions, reading to children in an interactive
manner, and sending children to preschool all raise the intelligence of young children.
Keywords
intelligence, developmental psychology, early childhood, interventions, preschool
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
26 Protzko et al.
control group, used classroom or cluster randomized trials, or
compared an experimental group with a group selected after
the fact.
Our third criterion is that the intervention must be sustained
and not a “one-shot” treatment. Some authors alter the proce-
dures, instructions, or context of intelligence test administra-
tion to examine effects on IQ scores. Studies on “score
optimization,” which involves experimentally modifying the
behavior of intelligence testers, often yield significant effects
on the IQ scores participants obtained (Zigler & Butterfield,
1968); however, these effects are obtained as a result of the
one-time manipulation during the testing experience. Although
such interventions are informative, we exclude them from
DORI because their effects may reflect the role of extraneous
factors in the testing environment rather than genuine increases
in underlying intelligence.
Finally, only studies where the authors use widely accepted
standardized measures of intelligence are included in DORI.
Studies where authors incorporated tests of infant cognition
were included only if they demonstrated a substantial relation-
ship with measures of later cognitive ability. The two most
common tests of infant cognition are the Bayley Scales of
Infant and Toddler Development (Bayley, 1969) and the Fagan
Test of Infant Intelligence (FTII; Fagan & Shepard, 1987). The
Bayley test is used to measure early motor and developmental
behavior (e.g., copying the experimenter in arranging blocks
into a pyramid, pointing to the same item on a page that an
administrator had pointed to), but its scores have little or no
relationship to later cognitive ability (e.g., SAT) or to indices
of adult IQ (Columbo, 1993). The FTII, by contrast, is a mea-
sure of information processing (specifically, how long a child
takes to habituate to a stimulus); its scores significantly predict
SAT scores 16 years later (r = .59; Fagan, Holland, & Wheeler,
2007). DORI thus includes interventions in which experiment-
ers used the FTII but excludes those in which experimenters
used the Bayley test.
Interventions that influence academic achievement (e.g.,
grades and other metrics of performance in school), although
of obvious importance, do not qualify for inclusion because
such measures involve more than intelligence. Academic tests
ideally measure knowledge, intelligence, and motivation. The
individual effects of each of these inputs cannot be separated
fully.
Overview of Analyses
In this study, we examine a subset of DORI consisting of those
experiments involving young children (from the prenatal
period through 5 years of age). We coded all studies into effect
sizes based on the postintervention differences in intelligence
scores. In cases where no standard deviation data were avail-
able, we contacted the authors for the data. If the authors or the
data were unavailable, we imputed the standard deviations by
using the value from the standardization sample (most com-
monly 15 or 16). This represents a conservative approach, as
many studies use restricted samples, reducing the standard
deviation of the sample and increasing the effect sizes; using
the larger standard deviation of the standardization sample
deflates the postintervention effect sizes.
In our survey of both the published and unpublished litera-
ture, we found 74 interventions that were designed to raise the
intelligence of young children and met our inclusion criteria.
These 73 interventions yielded 181 effect sizes across 37,773
participants. However, only four types of interventions were
numerous enough to allow meta-analysis.
The first half of the article describes research on nutritional
supplement to mothers and children. Our meta-analysis exam-
ined the effects of providing long-chain polyunsaturated fatty
acids (LC-PUFA) supplements to pregnant or breast-feeding
mothers or directly to neonates and young children. We also
examined studies where the authors supplemented young chil-
dren with other nutrients.
The second half of the article describes research on envi-
ronmental changes. In our second meta-analysis, we evaluated
early educational interventions conducted before children
began preschool. These include enrichment of the child’s envi-
ronment through home visits, parent training, special child
development centers, or a combination of the three. In our
third meta-analysis, we examined the effects of engaging
young children in an interactive reading intervention at home.
With our fourth meta-analysis, we examined the effects of
sending a young child to preschool. Throughout the article, we
also review the remaining interventions with too few replica-
tions to enable meta-analyses. All statistical tests and analyses
are in the Appendix.
Nutritional Supplements
We first analyzed the effects of several nutritional supple-
ments given to expecting mothers, new mothers, and their
children in the hopes of raising the children’s intelligence. We
discuss each of these interventions, starting with those we are
most confident can raise intelligence, as well as possible
causal mechanisms. Readers should note that the causal mech-
anisms discussed in this article are not exhaustive; we direct
interested readers to the relevant literature for a more thorough
treatment of the biochemical interactions of supplements and
their effects on human cognitive function.
LC-PUFA supplements and intelligence: A
meta-analysis
Using correlational studies, researchers have found that breast-
fed children are more intelligent than their bottle-fed counter-
parts (Anderson, Johnstone, & Remley, 1999). Because many
infant formulae do not contain breast milk’s LC-PUFA, which
is essential for nerve development, researchers interested in
raising intelligence began studying the effects of supplement-
ing formula with LC-PUFA.1
Our search of the literature (see Appendix) finally yielded
10 effect sizes across 844 participants (see Table 1). In these
studies, mothers’ diets were supplemented with over 1,000 mg
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 27
of LC-PUFA, specifically docosahexaenoic acid (DHA), per
day; infants who were supplemented received formula that
ranged from 0.2% to 0.5% LC-PUFA.
We found that supplementing either a pregnant mother
or supplementing infant formula with LC-PUFA raises a
young child’s IQ by more than 3.5 points2 (g = 0.236, 95%
CIWeighted Variance = .043 to .429); the benefits to LC-PUFA sup-
plementation does not appear when the children are tested in
infancy (g = −0.064).
LC-PUFA plus arachidonic acid supplements
and intelligence
Two studies from the LC-PUFA supplementation literature
merit closer attention, as they included a condition in which
infants’ diets were supplemented with not only DHA but also
arachidonic acid (ARA),3 a second, nonessential fatty acid.
Although the two interventions introduced supplements to
neonates’ diets during the same time period (within 5 days
after birth), they differed in duration. Whereas neonates in one
study received supplements for only 3 months, those in the
other continued on DHA- and ARA-supplemented diets for 12
months. It is perhaps surprising, then, that the shorter interven-
tion proved more effective; toddlers who had received the
supplements scored 6.5 IQ points higher than their peers in the
control group who had not received dietary supplements (g =
0.54, p < .001; Birch et al., 2007). Participants in the longer
intervention showed no such benefit (g = −0.142; Auestad
et al., 2003). A simple average of the two effect sizes yields a
small effect (g = 0.199); a weighted average yields an insig-
nificant effect (g = 0.092). This pattern of results does not pro-
vide sufficient evidence for us to conclude that supplementing
neonates’ diets with both DHA and ARA will contribute to
higher IQs in young childhood.
Our meta-analysis indicates that supplementing pregnant
mothers’ or neonates’ diets with LC-PUFA raises a young
child’s IQ. There is a large literature explaining how such sup-
plementation works. LC-PUFA are considered essential fatty
acids because they provide the building blocks for nerve cell
development that the body cannot produce on its own (Kurlak
& Stephenson, 1999). Lipids (of which LC-PUFA are one
class) make up the majority of dry matter in the brain, and dur-
ing periods of deprivation, they are preferentially depleted
from all of the body’s organs but the brain (Salem, Kim, &
Yergey, 1986).
Other researchers have found that supplementing the diets
of children diagnosed with attention deficit/hyperactivity dis-
order with LC-PUFA can reduce attention problems, impulse
problems, and inhibitory control problems (Stevens et al.,
2003), all of which are mediated by the prefrontal cortex (Bar-
kley, 1997). Therefore, it is probable that LC-PUFA supple-
mentation raises IQ by providing critical resources for
synaptogenesis, which neonates’ brains then allocate to the
development of the prefrontal cortex. Partial confirmation of
this hypothesis comes from a study of brain activation after 8
weeks of LC-PUFA supplementation; children who had
received these supplements demonstrated more baseline acti-
vation in the left dorsolateral prefrontal cortex than did those
who had not (McNamara et al., 2009).
Overall, the research suggests that supplementing pregnant
women’s, breast-feeding women’s, and neonates’ diets with
LC-PUFA increases young children’s IQ; our analysis here
confirms this finding.
Other nutrients
In additional interventions, researchers supplemented moth-
ers’ or young children’s diets with iron; thiamine, ascorbic
acid, and B-complex vitamins; multivitamins, or zinc.
Iron. Authors of several studies provided pregnant women
and young children with iron supplements in the hopes of
Table 1. Studies Used in the Meta-Analysis on the Effects of LC-PUFA Supplementation on IQ.
Study Test gDuration (years)
Judge, Harel, & Lammi-Keefe, 2007 Fagan −0.48 0.3
O’Connor et al., 2003; LC-PUFA from egg Fagan 0.07 0.59
O’Connor et al., 2003; LC-PUFA from fish Fagan 0.08 0.59
Auestad et al., 2001; LC-PUFA from fish Fagan 0.05 0.64
Auestad et al., 2001; LC-PUFA from egg Fagan 0.15 0.64
Werkman & Carlson, 1996 Fagan 0 0.75
Average (mixed) −0.06
Birch et al., 2007 WPPSI–R 0.37 0.33
Dunstan, Simmer, Dixon, & Prescott, 2008 PPVT 0.4 0.56
Helland, Smith, Saarem, Saugstad, & Drevon, 2003 K-ABC 0.44 1
Auestad et al., 2003 S-B L-M −0.29 1
Average (Mixed) 0.24
Note. LC-PU FA = long-chain polyunsaturated fatty acids; Fagan = Fagan Test of Infant Intelligence; WPPSI– R
= Wechsler Preschool and Primary Scale of Intelligence , Revised; PPVT = Peabody Picture Vocabulary Test;
K-ABC = Kaufman Assessment Battery for Children ; S-B L-M = Stanford-Binet Intelligence Scale, Form L-M.
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
28 Protzko et al.
increasing these children’s intelligence. In one, researchers
randomly assigned pregnant women to receive either 20 mg of
iron supplements or a placebo daily for the final 20 weeks of
their pregnancies. The supplementing of these pregnant wom-
en’s diets with iron had no effect on the children’s IQ by the
time they had reached the age of 4 (Zhou, Gibson, Crowther,
Baghurst, & Makrides, 2006). A large-scale RCT in Chile, in
which infants were randomly assigned to receive either iron-
supplemented formula or unsupplemented formula for the sec-
ond 6 months of life, produced similarly disappointing results
(Lozoff et al., 2003).
Although providing iron supplements to pregnant mothers
and infants may not boost young children’s IQ, introducing
them later in a child’s life might. In one study, researchers ran-
domly assigned 35 Indian children (M = 5.5 years old) to
receive either 20 mg of iron and 0.1 mg of folic acid or a pla-
cebo daily for 60 days. Although the experimental group had a
preexisting 10-point IQ advantage over the control group prior
to receiving the iron and folic acid supplements, supplementa-
tion boosted its members’ IQ, widening that gap to 23 points
(g = 2.205). In light of these results, it seems that iron supple-
ments may need to be administered during a specific period of
development if they are to have salutary effects on young chil-
dren’s IQ.
A possible mechanism for iron’s contributions to a young
child’s intelligence can be found in the reversibility of iron-
deficiency anemia’s detrimental cognitive effects. Children
who fail to consume enough iron during their early develop-
ment often develop iron-deficiency anemia, in which the body
has too few oxygen-carrying red blood cells (American Soci-
ety of Hematology, 2011). Frequently the children also suffer
from cognitive delays that result from the anemia. When the
brain lacks iron, it loses its ability to metabolize dopamine in
neuronal areas of the cortex responsible for attention (Beard,
1995; Pollitt, 1993) and working memory (Watanabe, Kodama,
& Hikosaka, 1997). Administering iron supplements to ane-
mic children can reverse these cognitive delays (Idjradinata &
Pollitt, 1993).
This reversibility of anemia’s detrimental effects is consis-
tent with the evidence that early iron supplementation is effec-
tive when intelligence is tested immediately after administration
of the supplements (Seshadri & Gopaldas, 1989) but ineffec-
tive when there is a delay between administration of the sup-
plements and intelligence testing (Zhou et al., 2006). Iron
supplementation allows the brain to temporarily reallocate
resources to dopamine transmission, which is responsible for
attention processes and working memory; once supplementa-
tion is terminated, its salutary effects fade.
This argument may be frustrated by the failure of the large-
scale Chilean iron supplementation study to improve scores on
the FTII (Lozoff et al., 2003). Although the iron-supplemented
children had longer looking times than did the control children
(which may represent increased ability in attention), novelty
preference—the FTII’s benchmark—was unaffected. The null
effect on the FTII, a test of information processing, paired with
the observed increases in overall looking times, is consistent
with reactive dopamine behavior. Therefore, although we can-
not state with full confidence that iron supplementation raises a
young child’s IQ, the hypothesis that iron supplementation can
produce salutary effects on IQ by influencing dopamine trans-
mission related to attention processes and working memory
capacity in the frontal cortex merits further investigation.
Thiamine, ascorbic acid, and B-complex vitamins. The
work on thiamine, ascorbic acid, and B-complex vitamins
comes primarily from one intervention conducted at two sites
(Harrell, Woodyard, & Gates, 1955). Hundreds of expecting
mothers at both sites (total N = 2,003) were randomly assigned
to receive one of four supplements: (a) 200 mg of ascorbic
acid; (b) 2 mg of thiamine (Vitamin B1); (c) a B-complex vita-
min consisting of 2 mg of thiamine, 4 mg of riboflavin (Vita-
min B2), 20 mg of niacin (Vitamin B3), and 15 mg of iron; or
(d) a placebo. Participants were instructed to take their
assigned supplements daily during the last trimester of preg-
nancy and the first 6 months of their children’s lives.
The results differed by location. None of the supplements
effectively increased IQ for children of participants at the first
site; all three supplements were effective at the second site,
with significant gains in IQ for children who received thia-
mine (g = 0.358), ascorbic acid (g = 0.255), and B-complex
vitamins (g = 0.507). The authors suggested that these site-
specific differences in effectiveness resulted from differences
in mothers’ diets and level of fidelity to the intervention.
Mothers participating at the site where the supplementation
failed to yield higher IQs had significantly better prenatal diets
than did mothers participating at the site where supplementa-
tion produced IQ gains. Therefore, it is only with extreme cau-
tion that we conclude that B-complex vitamins can effectively
raise young children’s IQ.
There is not enough research on the effects of riboflavin,
thiamine, and niacin on cognitive ability to draw confident
conclusions about the mechanisms by which they might ben-
eficially affect a young child’s intelligence. However, we can
speculate that these B-complex vitamins interact with other
nutrients in the body and that their salutary effects result from
these interactions. Specifically, each of the three B-complex
vitamins supplemented shares interaction effects with Vitamin
B6 (Sauberlich, 1980), an important vitamin for the use-depen-
dent growth of dendrites, specifically in the neocortex (Guil-
arte, 1993). Depriving rat neonates of Vitamin B6 alters the
behavior of the NMDA receptor-ion channel through a reduc-
tion in the number of glutamate- and glycerin-dependent 3[H]-
MK-801 binding sites in the cortex (Guilarte & Miceli, 1992).
This system is important for synaptic plasticity and dendritic
arborization (Cotman et al., 1989); blocking NMDA glutamate
receptors in rats prevents the growth of cells and dendrites in
the cortex in response to novel experiences (Rema, Armstrong-
James, & Ebner, 1998). In simpler terms, as the brain grows in
response to new information and activity, it needs B6 in order
to make new nerves.
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 29
Generalizing from the results of studies conducted with rats
to humans is not straightforward, especially when the human
trials are inconsistent. However, if researchers do replicate the
finding that B-complex supplementation can produce IQ gains
in future studies, they would be wise to consider the role the
interactions of these vitamins with B6 play in use-dependent
brain plasticity in their attempts to identify causal mechanisms.
Multivitamins. The majority of interventions involving mul-
tivitamin supplementation have been conducted with samples
of school-aged children, with only one study where authors
tested children before they entered school (Boggs, Scheaf,
Santoro, & Ritzman, 1985). Multivitamin supplementation did
yield some beneficial effects in an intervention conducted with
Head Start children. Researchers randomly assigned six pre-
schoolers to receive either daily multivitamins or placebos for
1 year. Before receiving the multivitamins, the children in the
experimental group had tested 20 IQ points lower than those in
the control group, a product of random sampling; after 1 year
of supplementation, they had gained 3.6 more IQ points than
the control group (Boggs et al., 1985). With such a small sam-
ple and such large differences in preintervention IQ, this study
does not provide enough evidence to make a firm statement on
the effects of multivitamin supplementation on young chil-
dren’s IQ.
Zinc. In a single study, women in their third trimester were
randomly assigned to receive 25 mg of zinc supplements or a
placebo daily through the remainder of their pregnancies.
When researchers tested these women’s children 5 years later,
zinc supplementation had no effect on their children’s IQ
(Tamura, Goldenberg, Ramey, Nelson, & Chapman, 2003).
Overall, we see that most attempts to raise the IQ of young
children through peri- and prenatal supplementation have not
been effective. We can be confident that supplementing with
LC-PUFA raises the IQ, and we can tentatively assert that sup-
plementing with B-complex vitamins also does so. In addition,
the work on iron supplementation is scattered enough that
more work needs to be done before we can understand iron’s
role in intelligence.
Environmental Changes
We now consider those interventions designed to increase
intelligence by enriching young children’s environments. In
this section, we include three meta-analyses, one on the effects
of early educational interventions and one on preschool. We
also review environmental interventions for which data are
insufficient for meta-analysis.
After searching the literature (see Appendix), we finally
found 16 RCTs where the authors provided intensive educa-
tional interventions to young children, yielding 43 effect sizes
over 19,238 participants (Table 2). These interventions involved
more than preschooling alone; rather, the interventions entailed
extensive alterations to the child’s environment.
We found that enrolling an economically disadvantaged
child into an early education intervention raised his or her IQ
by more than 4 points (g = 0.271, 95% CIWV = .114 to .429)
and that including a center-based education component raised
his or her IQ by more than 7 IQ points (g = 0.454, b = .183,
p < .001).
Because none of the studies involved children above lower-
middle socioeconomic status (SES), we did not examine SES
as a moderator. As a result, we must restrict the conclusions
we draw from this analysis to children from low SES back-
grounds and caution against broader generalization. In addi-
tion, contrary to theoretical assumptions (e.g., Ramey &
Ramey, 1998), we found that interventions that started earlier
in children’s lives were no more effective than those that
started later (see Appendix).
How Do Early Educational Interventions
Raise Intelligence?
The goal of early educational interventions is to raise young
children’s intelligence while also fostering other desirable out-
comes, such as improved social skills and self-regulation. The
results of our meta-analysis support the idea that early educa-
tional interventions raise a young child’s IQ by the time he or
she completes the intervention. Because these interventions
are multifaceted, we cannot identify any particular feature of it
as a causal mechanism. However, our findings are consistent
with the idea that environmental complexity promoting intel-
ligence and providing a more cognitively stimulating and
demanding environment raises the IQ of those who engage
with it.
To our surprise, we found no evidence to support the notion
that interventions conducted earlier in young childhood more
effectively boost IQ than those that begin later. Although this
finding contradicts the reasonable and widely held assumption
that earlier is always better when educating children, it is con-
sistent with the belief held by a prescient minority of research-
ers that the earliest few years of life are not the narrow
windows of opportunity they were once thought to be (e.g.,
Bruer, 1999). We also found evidence to suggest that educa-
tional interventions involving activities at a specially designed
center are more effective than simpler, home-based interven-
tions. Our results underscore the likelihood that environmental
complexity is the prime mechanism underlying gains in IQ;
however, which specific aspects of that complexity are most
effective or beneficial remain unknown.
Cognitive training
Researchers have made several attempts to raise young chil-
dren’s IQs by training early components of working memory
(WM), nonverbal reasoning, and effortful control. Despite
encouraging findings from studies conducted with adult par-
ticipants (e.g., Jaeggi, Buschkuehl, Jonides, & Perrig, 2008),
the work on WM training with samples of young children thus
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
30 Protzko et al.
Table 2. Studies Used in the Meta-Analysis on the Effects of Early Educational Interventions on IQ.
Study g
Age
(years)
Duration
(years) Notes
No center-based component
Gutelius et al., 1972 0.82 −0.5 3.5
Goodson, Layzer, St. Pierre, Bernstein, &
Lopez, 2000
0.04 0 5
Wasik, Ramey, Bryant, & Sparling, 1990 −0.7 0.08 3 No center
Love et al., 2005 0.09 0.42 1.83 Home-based
Bridgeman, Blumenthal, & Andrews, 1981 0.25 1 2 Houston Wave 1
0.42 1 2 Houston Wave 2
0.26 1 2 Houston Wave 4
0.06 1 2 Houston Wave 5
0.46 1 2 Houston Wave 7
0.48 1 2 Detroit Wave 2
Jester & Guinagh, 1983 0.06 1 1 Experimental program in the 1st year, control
condition in the 2nd and 3rd years
−0.1 1 2 Experimental program in the 1st and 3rd years,
control condition in the 2nd year
0.40 1 2 Experimental program in the 1st and 2nd years,
control condition in the 3rd year
0.59 1 3 Experimental program in all 3 years
−0.1 2 1 Control condition in the 1st and 3rd years,
experimental program in the 2nd year
Scarr & McCartney, 1988 0.21 2 1.75
Jester & Guinagh, 1983 0.7 2 2 Control condition in the 1st year, experimental
program in the 2nd and 3rd years
Levenstein et al., 1983 0.11 2.17 2 1973 cohort
−0.47 2.17 2 1975 cohort
0.39 2.17 2 1976 cohort
0.02 2.63 .62 1970 cohort
Jester & Guinagh, 1983 0.33 3 1 Control condition in the 1st and 2nd years,
experimental program in the 3rd year
Average effect size (mixed) 0.27
Center-based component
Ramey et al., 1992 0.46 0 3
Ramey, Yeates, & Short, 1984 0.79 0 4
Wasik et al., 1990 1.1 0.08 3
Bridgeman et al., 1981 0.29 0.17 2.83 Houston Wave 6
1.13 0.17 2.83 New Orleans Wave 2
0.27 0.17 2.83 New Orleans Wave 4
−0.38 0.17 2.83 New Orleans Wave 5
1.38 0.17 2.83 New Orleans Wave 6
0.88 0.17 2.83 Detroit Wave 1
0.79 0.33 2.67 Birmingham Wave 1
0.32 0.33 2.67 Birmingham Wave 2
0.37 0.33 2.67 Birmingham Wave 3
Love et al., 2005 0.09 0.42 1.67 Center only
0.23 0.42 1.92 Home and center
Kagitcibasi, Sunar, & Bekman, 2001 −0.17 2 2 Home and center
0.3 2 2 Center only
Deutsch, 1971 0.63 3 .75 Wave 1
0.69 3 .75 Wave 2
0.73 3 .75 Wave 3
0.37 3 .75 Wave 4
Karnes, Hodgins, Stoneburner, Studley, &
Teska, 1968
1.43 3 .58
Average effect size (mixed) 0.45
Note. Age is the age at which the child began the intervention. Negative age indicates the intervention began before the child was born.
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 31
far has yielded disappointing results. In one intervention
(Rueda, Rothbart, McCandliss, Saccomanno, & Posner, 2005),
researchers trained 24 four-year-olds on WM tasks for 5 days.
These WM tasks consisted of a series of computer games
designed to exercise various aspects of attention and inhibi-
tion. After the training, these children scored only 3.4 points
higher than a control group of 25 four-year-olds who had not
received this training, a statistically insignificant difference
(g = 0.227, p < .432). In another study, children were assigned
to one of three experimental groups, one of which was a WM
training group (Nutley et al., 2011). The 24 children in this
study participated in 5 weeks of WM training but did not dem-
onstrate any IQ gains.
Although these efforts failed to produce IQ gains, Nutley
et al. (2011) did find that training young children to perform
nonverbal reasoning tasks (with items adapted from the Leiter
Battery, a test similar to the Ravens’ Colored Progressive
Matrices, or RCPM) led to their scoring significantly higher on
the RCPM (g = 0.614). Two other studies (Riding & Powell,
1985, 1986) serve as corroborating evidence for the beneficial
effects of training young children to complete reasoning tasks
on IQ. However, the high degree of similarity between inter-
vention stimuli and posttest items invites skepticism about
whether this improved IQ reflects a genuine increase in the
children’s intelligence or simple practice effects.
Effortful control is a critical executive function required for
attending to pertinent information while also manipulating it
to perform mental calculations (Blair & Ursache, 2011). Nine-
teen 5-year-old preschoolers who were trained on effortful
control tasks for over a month did not demonstrate IQ gains
immediately after completing the training program; however,
when they were tested again 2 months later, their improved
IQ revealed a positive effect (g = 0.178; Rueda, Checa, &
Cómbita, 2012). This delayed impact of the intervention is
unexpected and requires replication.
Listening to music
One RCT examined the effects of listening to music on young
children’s IQ. Forty-one children were randomly assigned to
either a phonological skills training program or a music listen-
ing program to evaluate potential effects on the phonological
ability of the preschoolers. Neither intervention produced IQ
gains (Dege & Schwarzer, 2011).
Training mothers to provide cognitively
complex environments
Maternal behavior and the environmental complexity of the
home vary significantly as a function of SES. Higher SES
mothers tend to provide their children with more complex and
enriching environments than do lower SES mothers (Neisser
et al., 1996). Can providing lower SES mothers with the means
and training to provide richer environments for their young
children increase these children’s IQs? The authors of one
study (Karnes, Studley, Wright, & Hodgins, 1968) randomly
assigned 26 mothers to either a control group or an 11-week
intervention. During this intervention, the experimenters
trained mothers to make educational materials for their chil-
dren, provided them with age-appropriate books and puzzles
for their children, and taught them how to help their children
learn to speak and identify objects in the home. By the end of
the 11 weeks, the children of mothers in the training group had
gained 7 IQ points (g = 0.342), whereas the IQs of the control
group mothers’ children remained unchanged.
Narrative talk and elaborate reminiscing
Young children’s ability to narrate their own experiences
shares strong associations with early school readiness (Cristo-
faro & Tamis-Lemonda, 2012). This relationship may not be
causal; both capacities may be outgrowths of intelligence, lin-
guistic skills, or other important cognitive abilities. Two stud-
ies shed light on these connections. In one, 20 three-year-olds
were randomly assigned to either a control group or a yearlong
intervention in which their mothers were trained to be as elab-
orative as possible with their children when asking them to
narrate past events. Specifically, these mothers were encour-
aged to talk to their children frequently and at length, to ask
many open-ended questions that required more than simple
one-word answers, to listen and encourage their children to
speak in multiple sentences, and to discuss topics of interest to
the children (Peterson, Jesso, & McCabe, 1999). This inter-
vention raised the children’s IQ more than 6 points (g = 0.438,
p < .01).
In a second study, researchers replicated the methods of
training mothers to elaborate and had mothers require their
children to be elaborative when recalling past events (Reese,
Leyva, Sparks, & Grolnick, 2010). Thirty-three mothers of
Head Start children were randomly assigned to one of three
groups: one in which mothers were trained to be elaborative
while reminiscing with their children, one in which mothers
were trained to be elaborative with their children while co-
reading a book during story time, or a control group in which
mothers received no training. Children whose mothers had
participated in the training groups did not have higher IQs than
children whose mothers had been assigned to the control group
(E. Reese & D. Leyva, personal communication, November 2,
2011).
Why did one study’s elaborative talk training program pro-
duce significant IQ gains in young children while another
study’s did not? These inconsistencies probably stem from the
different intensities of the two interventions. In the effective
intervention (Peterson et al., 1999), mothers received exten-
sive training, and experimenters followed up with them
throughout the yearlong intervention in order to aid and assist
them and to ensure fidelity in their implementation of the pro-
tocol. In the ineffective intervention (Reese et al., 2010),
experimenters provided mothers with a single 45-min training
session, then tested their children’s IQ 8 months later. Thus, it
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
32 Protzko et al.
is unlikely that the mothers in this unsuccessful intervention
received sufficient training, support, and monitoring to ensure
faithful adherence to the intervention.
How might elaborative narrative talk increase a young
child’s intelligence? Parent–child conversations and word use
are associated with later intelligence and achievement (Hart &
Risley, 1995). Future experiments examining such interven-
tions could help clarify whether such findings reflect the
effects of narrative talk in such interactions. However, with
only two studies on which to base our assumptions, we can
only speculate; but clearly such interventions are promising
and merit attempts to replicate.
Interactive reading: A meta-analysis
Children who grow up poor have lower IQs than children who
grow up wealthy (Herrnstein & Murray, 1994). One of the
mediating mechanisms that explain this gap is differences in
cognitive stimulation—specifically, differences in the number
of reading interactions a child had with his or her parents and
the number of books the child grows up with (Guo & Harris,
2000). Many interventions for altering the reading environ-
ment of children focus on remediating effects—for example,
treating children with preestablished speech and language dif-
ficulties (e.g., Crain-Thoreson & Dale, 1999; Danger & Lan-
dreth, 2005). Such studies are not included in this database. A
subset of these studies focuses on altering the reading environ-
ments of children in the normal range of ability.
The original meta-analysis included two experiments that
simply provided books to parents without instructions for
developing a more interactive reading style (Chow & McBride-
Chang, 2003; Lonigan, Anthony, Bloomfield, Dyer, & Sam-
wel, 1999). Neither of these interventions raised the children’s
IQ, so we can conclude that the mere presence of books in the
home does not raise IQ. In the remaining studies we consider,
experimenters provided parents both with books and with a
training program for effective reading with their children—
teaching them how to ask open-ended questions, encourage
their children to read, shadow their children’s interests, and so
on. This meta-analysis includes eight studies, providing 10
effect sizes across 499 participants (see Appendix).
Reading to a child in an interactive style raises his or her
IQ by over 6 points (g = 0.404, 95% CI = .153 to .654).
With one exception, interventions that begin after the child is
42 months old do not raise the IQ; however, in the random
effects model, age is not a significant moderator. In each of
these interventions, children and their parents engage with sto-
rybook reading in an interactive way. The child is an active
participant in the reading, with the adult encouraging the child
to be as elaborate as possible. With one exception, these inter-
ventions do not appear to raise the IQ if the child is more than
4 years old (see Table 3). Why might this be the case? It is
possible that interactive reading does not raise a young child’s
intelligence but instead merely accelerates language develop-
ment, which boosts IQ. If this is the case, once the child’s level
of language development is more advanced, added demands
may no longer act as accelerants.
Preschool and intelligence: A meta-analysis
We located 16 RCT studies, yielding 39 effect sizes based on
7,370 participants in which young children were enrolled in
preschool. We found that sending a child to preschool raises
the IQ by more than 4 points (g = 0.307, 95% CIWV = .135 to
.479). Preschools that include a specific language develop-
ment component boost IQ by more than 7 points (g = 0.512,
b = .205, p < .001).
As is the case in our earlier analyses, insufficient socioeco-
nomic variation prevented us from testing the reasonable pre-
diction that SES would moderate the effects of preschool. We
therefore must caution readers again that these findings cannot
be generalized beyond children from low-income homes. In
addition, we found that preschool interventions that last longer
are no more effective at raising the IQ than preschools that are
shorter (see Appendix). We explain this puzzling finding next.
Our meta-analysis indicates that sending a disadvantaged
child to preschool raises his or her IQ by as much as 7 points
Table 3. Studies Used in the Meta-Analysis on the Effects of Interactive Reading on IQ.
Study gAge (years) Notes
Whitehurst et al., 1988 0.58 2.33
Huebner, 2000 0.27 2.39
Arnold, Lonigan, Whitehurst, & Epstein, 1994 0.36 2.42 Trained directly
0.57 2.42 Trained by video
Valdez-Menchaca & Whitehurst, 1992 1.21 2.59
Whitehurst, Arnold, Epstein, & Angell, 1994 0.15 3.46 School reading
0.24 3.46 School and home reading
Lonigan, Anthony, Bloomfield, Dyer, & Samwel, 1999 0.03 3.76
Lamb, 1986 −0.01 4
Van Kleeck, Vander-Woude, & Hammett, 2006 1.41 4.17
Chow & McBride-Chang, 2003 −0.18 5.32
Effect size (random) 0.4
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 33
Table 4. Studies Used in the Meta-Analysis on the Effects of Preschool Attendance on IQ.
Study G
Duration
(years) Notes
No language development component
Rainey, 1968 0.33 0.5
Di Lorenzo, Salter, & Brady, 1969 0.52 0.75 Cortland, no language
–.05 0.75 Greenburgh
–0.54 0.75 Hempstead
–0.02 0.75 Long Beach
0.16 0.75 Spring Valley
.47 0.75 Yonkers
Karnes, Zehrbach, & Teske, 1974 0.48 0.75
Erickson, McMillan, Bonnell, Hofmann, & Callahan, 1969 0.67 0.75 Enrichment program
Average effect size (mixed) 0.31
Language development component
Abbott-Shim, Lambert, & McCarty, 2003 0.08 0.17 Summer program
Skeels, Ruth, Wellman, & Williams, 1938 0.32 0.32 1–199 days
Blank & Solomon, 1968 1.50 0.33 5 days per week
0.28 0.33 3 days per week
Peta, 1973 0.52 0.34
Edwards & Stern, 1970 0.20 0.46 Cohort 1
0.07 0.46 Cohort 2
Ametjian, 1965 0.87 0.5
Klaus & Gray, 1968 0.73 0.5 2-year program
0.59 0.5 3-year program
Dawe, 1942 0.96 0.6
Di Lorenzo & Salter, 1969 0.52 0.75 Cortland, language
0.36 0.75 Mt. Vernon
0.41 0.75 Schenectady
Abbott-Shim et al., 2003 0.24 0.75 School year
0.32 0.75 School year with posttest delay
Erickson et al., 1969 0.77 0.75 Bereiter-Englemann program
Skeels et al., 1938 0.81 0.77 200–399 days
Weikart, 1966 0.56 1 Wave 0
0.83 1 Wave 3
Puma, Bell, Cook, & Heid, 2010 0.11 1 4-year-olds
Deutsch, 1971 0.81 1.5 Wave 1
0.01 1.5 Wave 2
0.77 1.5 Wave 3
0.35 1.5 Wave 4
Skeels et al., 1938 1.10 1.65 400+ days
Weikart, 1966 0.55 2 Wave 1
0.96 2 Wave 2
Herzog, Newcomb, & Cisin, 1974 0.66 2
Puma et al., 2010 0.07 2 3-year-olds
Average effect size (mixed) .51
if it includes a specific language-development component.
Schooling is known to both increase and maintain intelligence
(Ceci, 1991), though the specific mechanisms remain
unknown. Many intelligence tests tap knowledge of informa-
tion and vocabulary. Although there are specific mechanisms
responsible for the recall, comprehension, and retrieval of
information and vocabulary words, the most important factor
in a young child’s success or failure with these test items may
be exposure. Young children cannot define a word that they
have not encountered, nor can they identify a picture of an
item that they have never seen. Therefore, preschool may raise
intelligence test performances merely by exposing young chil-
dren to the information and vocabulary words included on
these tests. If exposure is the mechanism by which preschool
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
34 Protzko et al.
boosts IQ test performance, then these improved scores do not
indicate genuine gains in intelligence.
A more optimistic explanation of preschool’s causal role in
boosting young children’s intelligence is that the cognitive
complexity inherent to the preschool environment and experi-
ence leads to increases in underlying intelligence. Attending
preschool provides lower SES children with the opportunity to
engage with novel stimuli, to practice complex problem solv-
ing, to navigate social interactions, and to confront other cog-
nitive challenges they do not face in their home environments.
This explanation is consistent with the positive results of
extensive maternal training and early educational interven-
tions and provides additional support for the argument that
increasing the cognitive demands of a young child’s environ-
ment causes him or her to adapt and become more intelligent.
Yet why do we fail to find a positive relationship between
duration of preschooling and IQ gains? Recall that preschool
programs of longer durations were no more effective than
shorter ones. We suspect that the reason for this seeming con-
tradiction is that as the program wears on, preschool may
become less challenging and complex.
For example, consider a physically weak young boy who
cannot do more than one or two push-ups. Our goal is to
increase the number of push-ups he can do, so we enroll him
in a push-up intervention. This intervention involves daily
push-up exercises of up to 15 push-ups per day. After partici-
pating in this intervention for 6 weeks, our formerly weak
young boy can reliably perform 15 push-ups on demand. If the
intervention continues for another 12 weeks and the boy still
has to do only 15 push-ups a day, we would not see another
increase in the young boy’s strength. Why? Because the daily
demands of the intervention did not change. Preschoolers may
have similar experiences. If a young child gains intelligence
after attending preschool for 1 year, for example, and the pre-
school curriculum’s demands do not improve, the child may
not be adequately challenged to further increase his or her
intelligence if the preschool’s demands on intelligence remain
constant. As a result, his or her intellectual progress may
stagnate.
Conclusion
The present work has been a systematization and synthesis of
as much of the available knowledge on raising young chil-
dren’s intelligence as we could locate. Our analysis allows us
to draw the following conclusions:
Supplementing the diets of pregnant women and
neonates with LC-PUFA raises the children’s IQ in
young childhood. Providing preschool-aged children
with iron supplements may boost their IQ, but giving
these supplements to infants does not.
Enrolling a lower SES infant in an intense early edu-
cational intervention will raise his or her IQ in young
childhood. Enrolling him or her in such a program at
a younger age has no additional benefits for his or her
IQ. The more complex the intervention is, the greater
these gains will be.
Reading interactively with young children raises their
IQ. The earlier the interactive reading takes place, the
larger the benefits.
Attending preschool increases a young child’s IQ. If
the preschool program includes a specific language-
development component, these gains are even larger.
We currently lack sufficient data to determine whether any
of the other interventions we examined are effective. Using the
most scientifically rigorous and conservative standards avail-
able, we will continue to compile and synthesize all of the
available scientific evidence on raising intelligence. Our cur-
rent findings strengthen earlier conclusions (e.g., that complex
environments build intelligence), cast doubt on others (e.g.,
that earlier interventions are always most effective), and give
rise to tantalizing new questions for future research (e.g., can
essential fatty acids increase intelligence?).
Database Activity
We made every effort to include every intervention in our
study. However, we acknowledge that additional studies may
exist; some may even be under way at the time of writing and
publication. It is for this reason that DORI will remain an
active database, updated continually as new studies are pub-
lished and file drawers are opened. We invite scientists in all
fields who have completed or are in the process of conducting
studies that meet DORI’s inclusion criteria to send their data.
Please contact the corresponding author with any research you
may have to contribute.
Appendix
What follows is the statistical methods and procedures we
used to come to our conclusions, including all models run and
the results from those models. Across all four meta-analyses,
we included multiple experiments from the same study.
Though a more conservative approach is to remove studies
with multiple interventions and only keep one, we find this
method troublesome as the choice of which experiment to
keep can be too subjective and leave open the possibility for
bias. To examine whether experiments that came from the
same study could bias our results, we tested whether effect
sizes were dependent on groups of studies. In each of the
meta-analyses reported here, no such grouping was found at
any significant level (Card, 2011). All confidence intervals are
given using the Weighted-Variance method (Sidik & Jonkman,
2003).
Statistical issues on the meta-analysis of
LC-PUFA supplementation and intelligence
There was no evidence of publication bias in our data (pBegg <
.158, where Begg is the metric of publication bias). We
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 35
originally ran the data in STATA (StataCorp, 2011) using a
weighted regression and clustering the covariance matrix to be
able to take into account longitudinal data. Two effect sizes
were outliers in the data set, both from studies with the same
authors (Carlson & Werkman, 1996; Werkman & Carlson,
1996; both greater than three standard deviations above). In
both studies, the intervention included giving infants formula
supplemented with 0.2% docosahexaenoic acid (DHA, the
most common form of supplemented LC-PUFA), one for 2
months (Carlson & Werkman, 1996) and the other for 9 months
(Werkman & Carlson, 1996). Two other studies (Auestad
et al., 2003; Birch et al., 2007) included a third intervention
group (both included DHA plus arachidonic acid). We removed
these studies from the analysis and present them separately.
All of the studies involved supplementing either pregnant
or breast-feeding mothers or neonates, except one study (Ryan
& Nelson, 2008) in which 4-year-old children received 400
mg of LC-PUFA supplements daily for 4 months. To deter-
mine whether LC-PUFA supplementation in the first years of
life raises intelligence while trying to reduce the heterogeneity
in the sample, we ran the analysis both with and without this
study; removing this study from the analysis did not substan-
tively change the results.
We found nine other studies on LC-PUFA supplementation
and young children’s intelligence, yielding 13 effect sizes. There
was significant heterogeneity in the sample, Q(12) = 184.748,
p < .001, so it was necessary to look for moderators, remove
outliers, or run a random- or mixed-effects analysis on the data.
We thought it best to remove the two outliers from the data first;
doing so reestablished normality to the data set’s distribution
(pK-S < .2; K-S = Kolmogorov–Smirnov normality test).
To determine the role of LC-PUFA supplementation on sub-
sequent intelligence, we restricted longitudinal investigations
to the last time point at which participants were still taking
supplements and had completed an intelligence test. In doing
so, we removed five of our effect sizes, reducing the final sam-
ple to 10 effect sizes across 844 participants. Although in some
(n = 4) of the studies researchers tested children’s intelligence
in toddlerhood, in others (n = 6) researchers tested children’s
intelligence in infancy by using the Fagan Test of Infant Intel-
ligence (FTII; Fagan & Shepard, 1987). Test type (FTII vs.
tests for older children, such as the Stanford-Binet or Wechsler
Intelligence Scale for Children) did moderate the results. As
there was still significant heterogeneity, we used a mixed-
effects analysis4 (Overton, 1998) with whether the FTII was
used as our moderator. We reached convergence (defined as
Δτ2 < 10-10; Erez, Bloom, & Wells, 1996) in seven iterations.
Statistical issues on the meta-analysis of
intense early educational interventions and
intelligence
There were no outliers in the early educational intervention
data. Effect sizes were normally distributed (pK-S < .2), but
there was a grouping problem with effect sizes coming from
the same study being more similar than others. We found no
evidence for publication bias in the data (pBegg < .832).
In order to determine the maximum effect of each interven-
tion while minimizing the amount of delay between the inter-
vention and the posttest, we restricted the data to time points
immediately following the interventions, reducing the sample
to 43 effect sizes (see Table 2). By doing so, we eliminated the
aforementioned grouping issue, F(1, 12) = .397. However,
there was significant heterogeneity in the sample, Q(42) =
558.713, p < .001, so further analyses were in order.
To provide the most complete list of theoretically appropri-
ate moderators, we coded the following variables in the stud-
ies: socioeconomic status (SES) of the family, age of the child
when the intervention started, duration (years) of the interven-
tion, whether the child’s diet was supplemented, and whether
the intervention occurred in the children’s home, in a special
development center, or both. We removed studies including a
specific nutritional component from the analysis because of
insufficient variability.
Initial analyses indicated significant heterogeneity in the
sample, Q(42) = 558.713, p < .001, so we identified character-
istics of the interventions that might moderate the effects.
With 43 effect sizes we had enough statistical power to test
only the two most important moderators: child’s age at the
intervention’s inception and the inclusion of a center-based
education component (e.g., nursery school or program at
another specially designed development center that delivered
developmentally appropriate and cognitively demanding
activities). Both of these moderators were statistically
significant.
Because the effect sizes were still grouped within studies,
we tested for heteroskedasticity in the errors (Dickens, 1990),
as possible effect size errors are correlated with group level
errors; heteroskedasticity existed, Breusch–Pagan χ2(1) =
1203.87, p < .001. Combining the statistical issues of hetero-
skedastic errors from data grouping and heterogeneity across
the samples, we conducted the analysis again, this time using
a mixed-effects model (Overton, 1998) in R. We first tested
the moderators under a nonweighted situation and found that
interventions that included a center-based education compo-
nent were more effective at boosting young children’s IQ, but
interventions that began earlier in a child’s life were no more
effective than those that started later. Therefore, in the subse-
quent mixed-effects analysis, we included the presence of a
center-based education component as the sole moderator. We
reached convergence (defined as Δτ2 < 10-10; Erez et al., 1996)
in 20 iterations.
Statistical issues on the meta-analysis of
interactive reading and intelligence
Data were normally distributed (pK-S < .114). There was no
indication of publication bias (pBegg < .335). Under a fixed-
effects analysis, reading to a child raises his or her IQ by over
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
36 Protzko et al.
2 points (g = 0.159, 95% CI = .129 to .189); however there was
significant heterogeneity in this model, Q(15) = 100.586, p <
.001, so we looked further into the data. We first removed lon-
gitudinal data points so that results would reflect only the
immediate result of the intervention. After doing so, we found
that reading to a young child raises his or her IQ by over 3
points (g = 0.204, 95% CI = .169 to .239), but we also found
heterogeneity in the data, Q(11) = 105.846, p < .001. Because
we had enough power to run only one moderator in our attempt
to remove this heterogeneity, we examined our three most
informed moderators: age when the intervention began (cen-
tered on age 2), SES of the parents, and duration of the inter-
vention. The younger the child was when the intervention
began, the better the results (b = −.168, p < .001); however,
the interventions did not differ in effectiveness as a function
of SES (p < .963). Because multiple effect sizes came from
the same studies, errors may not be independent; indeed, we
found that to be the case under both models (all Breusch–
Pagan ps < .001). Two studies included two experimental
conditions, one a dialogic reading condition (see Whitehurst
et al., 1988) and another a regular reading condition in which
parents received books without specific instructions or train-
ing regarding how to read with their children. Neither of these
interventions raised the IQ of children, and they were removed
from the analysis. Continuing with this analysis changes it
from a meta-analysis on reading to your child to a meta-anal-
ysis on interactive reading. Clustering the standard errors for
nonindependence in the fixed effects model did not substan-
tively change the results; further models were therefore
run without clustering. Because of significant heterogeneity,
Q(9) = 69.345, p < .001, we ran the final model as a random-
effects model with no moderators. We then explored whether
the age at which the intervention began moderated the results.
Statistical issues in the meta-analysis of
preschool and intelligence
There were no outliers in the preschool data, effect sizes were
normally distributed (pK-S < .2), and there was no evidence of
publication bias (pBegg < .25).
The first problem with the preschool analysis was group-
ing. We coded effect sizes by the study from which we obtained
them. Studies with multiple sites or multiple waves of an inter-
vention presented in a single article were coded as such, and
we investigated to ensure that effect sizes from the same stud-
ies did not differ significantly from others, which would repre-
sent a problem of grouping. Effect sizes were dependent on
groups of studies, F(1, 15) = 6.942, p < .001.
Removing the longitudinal data points and keeping only
those that represented the longest duration of preschool while
minimizing the delay between the end of preschool and the
posttest removed the grouping problem, F(1, 15) = 1.87, p <
.086, and reduced the number of effect sizes to 39 across 7,370
participants. Because doing so did not eliminate the amount of
heterogeneity in the sample, Q(38) = 510.64, p < .001, we then
examined characteristics that moderate a preschool’s effec-
tiveness at raising a young child’s IQ.
Although a host of predictors could help identify the most
effective components of a preschool program, we coded the
following possible moderators as part of our meta-analysis:
children’s families’ SES, age at which children started pre-
school, duration of the preschool (in years), number of days
per week preschool was attended, number of hours per day
preschool was attended, presence of a nutritional component,
maternal involvement (either through home visits or presence
at the preschool), presence of a specific language-develop-
ment component, presence of a specific cognitive-skills com-
ponent, and measure of intelligence used as posttest.
Our initial analyses indicated significant heterogeneity
existed in the sample so we looked for characteristics of the
interventions that could moderate the effects of preschool on
intelligence.
Limited statistical power permitted the testing of only two
critical moderators: duration of preschool and inclusion of a
specific language-development component. Although both
moderators were significant in the analysis, there was still sig-
nificant heterogeneity in the sample, Q(36) = 465.795, p <
.001; and heteroskedasticity of the errors, Breusch-Pagan
χ2(1) = 38.09, p < .001. With correlated error components
within groups, significant heterogeneity, and moderators of
interest, we conducted a mixed-effects model on the data
(Overton, 1998) in R to obtain a more accurate estimate of the
effects. To ensure the acceptability of both moderators, we ran
the model again, this time without the weighting; in this analy-
sis, preschools that included a specific language-development
component were found to be somewhat more effective at rais-
ing IQ (p < .099), but program duration did not moderate the
effectiveness of a preschool program (p < .651). As a result, in
the final analysis, we included only the presence of a specific
language-development component. We reached convergence
(defined as Δτ2 < 10-10; Erez et al., 1996) in 19 iterations.
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with
respect to their authorship or the publication of this article.
Notes
1. LC-PUFA are frequently referred to as omega-3 fatty acids.
2. All point values are the effect sizes multiplied by a standard
deviation of 15. These represent conservative values as many of the
studies have smaller distributions, so the actual point values would
be higher. These values are given for ease of interpretation only.
3. ARA—or omega-6—is a nonessential fatty acid.
4. A mixed-effects analysis is a random-effects meta-analytic model
with a fixed moderator. See Overton (1998) for a full explanation.
References
Abbott-Shim, M., Lambert, R., & McCarty, F. (2003). A comparison
of school readiness outcomes for children randomly assigned to a
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 37
Head Start program and the program’s wait list. Journal of Edu-
cation for Students Placed at Risk, 8, 191–214.
American Society of Hematology. (2011). Iron-deficiency anemia.
Retrieved from http://www.hematology.org/Patients/Blood-
Disorders/Anemia/5263.aspx
Ametjian, A. (1965). The effects of a preschool program upon the intel-
lectual development and social competency of lower class children
(Doctoral dissertation). Stanford University, Palo Alto, CA.
Anderson, J. W., Johnstone, B. M., & Remley, D. T. (1999). Breast-
feeding and cognitive development: A meta-analysis. American
Journal of Clinical Nutrition, 70, 525–535.
Arnold, D. S., Lonigan, C. J., Whitehurst, G. J., & Epstein, J. N.
(1994). Accelerating language development through picture book
reading: Replication and extension to a videotape training format.
Journal of Educational Psychology, 86, 235–243.
Auestad, N., Halter, R., Hall, R. T., Blatter, M., Bogle, M. L., Burks,
W., . . . Bornstein, M. H. (2001). Growth and development in
term infants fed long-chain polyunsaturated fatty acids: A double-
masked, randomized, parallel, prospective, multivariate study.
Pediatrics, 108, 372–381.
Auestad, N., Scott, D. T., Janowsky, J. S., Jacobsen, C., Carroll, R. E.,
Montalto, M. B., . . . Hall, R. T. (2003). Visual, cognitive, and lan-
guage assessments at 39 months: A follow-up study of children
fed formulas containing long-chain polyunsaturated fatty acids to
1 year of age. Pediatrics, 112 , e177–e183.
Barkley, R. A. (1997). Behavioral inhibition, sustained attention, and
executive functions: Constructing a unifying theory of ADHD.
Psychological Bulletin, 121, 65–94.
Bayley, N. (1969). Bayley Scales of Infant Development. New York,
NY: Psychological Corporation.
Beard, J. (1995). One person’s view of iron deficiency, development,
and cognitive function. American Journal of Clinical Nutrition,
62, 709–710.
Birch, E. E., Garfield, S., Castaneda, Y., Hughbanks-Wheaton, D.,
Uauy, R., & Hoffman, D. (2007). Visual acuity and cognitive
outcomes at 4 years of age in a double-blind, randomized trial
of long-chain polyunsaturated fatty acid-supplemented infant for-
mula. Early Human Development, 83, 279–284.
Blair, C., & Ursache, A. (2011). A bidirectional model of executive
functions and self-regulation. In K. D. Vohs & R. F. Baumeister
(Eds.), Handbook of self-regulation: Research, theory and appli-
cations (2nd ed., pp. 300–320). New York, NY: Guilford Press.
Blank, M., & Solomon, F. (1968). A tutorial language program to
develop abstract thinking in socially disadvantaged preschool
children. Child Development, 39, 379–389.
Boggs, U. R., Scheaf, A., Santoro, D., & Ritzman, R. (1985). The
effect of nutrient supplements on the biological and psycho-
logical characteristics of low IQ preschool children. Journal of
Orthomolecular Psychiatry, 14, 97–127.
Bridgeman, B., Blumenthal, J. B., & Andrews, S. R. (1981). Parent-
child development centers: Final evaluation report. Unpublished
manuscript. Princeton, NJ: Educational Testing Service.
Bruer, J. T. (1999). The Myth of the first three years a new under-
standing of early brain development and lifelong learning. New
York, NY: The Free Press.
Card, N. (2011). Applied meta-analysis for social science research.
New York, NY: Guilford Press.
Carlson, S. E., & Werkman, S. H. (1996). A randomized trial of visual
attention of preterm infants fed docosahexaenoic acid until two
months. Lipids, 31, 85–90.
Ceci, S. J. (1991). How much does schooling influence general intel-
ligence and its cognitive components? A reassessment of the evi-
dence. Developmental Psychology, 27, 703–722.
Chow, B. W.-Y., & McBride-Chang, C. (2003). Promoting language
and literacy development through parent-child reading in Hong
Kong preschoolers. Early Education and Development, 14, 233–
248.
Columbo, J. (1993). Infant cognition: Predicting later intellectual
functioning. Newbury Park, CA: SAGE.
Cotman, C. W., Bridges, R. J., Taube, J. S., Clark, A. S., Geddes,
J. W., & Monaghan, D. T. (1989). The role of the NMDA recep-
tor in central nervous system plasticity and pathology. Journal of
NIH Research, 1, 65–74.
Crain-Thoreson, C., & Dale, P. S. (1999). Enhancing linguistic per-
formance: Parents and teachers as book reading partners for chil-
dren with language delays. Topics in Early Childhood Special
Education, 19, 28–39.
Cristofaro, T. N., & Tamis-Lemonda, C. S. (2012). Mother-child
conversations at 36 months and at pre-kindergarten: Relations to
children’s school readiness. Journal of Early Childhood Literacy,
12, 68–97.
Danger, S. & Landreth, G. (2005). Child-centered group play therapy
with children with speech difficulties. International Journal of
Play Therapy, 14, 81–102.
Dawe, H. C. (1942). A study of the effect of an educational program
upon language development and related mental functions in
young children. Journal of Experimental Education, 11, 200–209.
Dege, F., & Schwarzer, G. (2011). The effect of a music program
on phonological awareness in preschoolers. Frontiers in Psychol-
ogy, 2, 1–7.
Deutsch, M. (1971). Regional research and resource center in early
childhood (Final Report). Washington, DC: Office of Economic
Opportunity.
Dickens, W. T. (1990). Error components in grouped data: Is it ever
worth weighting? Review of Economics and Statistics, 72, 328–
333.
Di Lorenzo, L. T., Salter, R., & Brady, J. J. (1969). Prekindergar-
ten programs for educationally disadvantaged children (Final
Report). Albany, NY: University of the State of New York, State
Education Department, Office of Research and Evaluation.
Dunstan, J. A., Simmer, K., Dixon, G., & Prescott, S. L. (2008). Cog-
nitive assessment of children at age 2(1/2) years after maternal
fish oil supplementation in pregnancy: A randomised controlled
trial. Archives of Diseases in Childhood: Fetal & Neonatal Edi-
tion, 93, F45–F50.
Edwards, J., & Stern, C. (1970). A comparison of three intervention
programs with disadvantaged preschool children. Journal of Spe-
cial Education, 4, 205–214.
Erez, A., Bloom, M. C., & Wells, M. T. (1996). Using random
rather than fixed effects models in meta-analysis: Implications
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
38 Protzko et al.
for situational specificity and validity generalization. Personnel
Psychology, 49, 276–306.
Erickson, E. L., McMillan, J., Bonnell, J., Hofmann, L., & Callahan,
O. D. (1969). Experiments in head start and early education: Cur-
riculum structures and teacher Attitudes (Final Report). Washing-
ton, DC: Office of Economic Opportunity.
Fagan, J., Holland, C. R., & Wheeler, W. (2007). The prediction, from
infancy, of adult IQ and achievement. Intelligence, 35, 225–232.
Fagan, J., & Shepard, P. (1987). The Fagan test of infant intelligence.
Cleveland, OH: InfanTest Corporation.
Goodson, B. D., Layzer, J. I., St. Pierre, R. G., Bernstein, L. S., &
Lopez, M. (2000). Effectiveness of a comprehensive five-year
family support program on low-income children and their fami-
lies: Findings from the comprehensive child development pro-
gram. Early Childhood Research Quarterly, 15, 5–39.
Guilarte, T. R. (1993). Vitamin B6 and cognitive development:
Recent research findings from human and animal studies. Nutri-
tion Reviews, 51, 193–198.
Guilarte, T. R., & Miceli, R. C. (1992). Maternal vitamin B6 defi-
ciency alters postnatal development and zinc regulation of 3[H]-
MK-801 binding to cortical NMDA receptor-ion channels. Social
Neuroscience Abstracts, 18, 1157.
Guo, G., & Harris, K. M. (2000). The mechanisms mediating the
effects of poverty on children’s intellectual development. Demog-
raphy, 37, 431–447.
Gutelius, M. F., Kirsch, A. D., MacDonald, S., Brooks, M. R.,
McErlean, T., & Newcomb, C. (1972). Promising results from a
cognitive stimulation program in infancy. A preliminary report.
Clinical Pediatrics, 11, 585–593.
Harrell, R. F., Woodyard, E., & Gates, A. I. (1955). The effect of
mothers’ diets on the intelligence of offspring. New York, NY:
Teachers College.
Hart, B., & Risley, T. R. (1995). Meaningful differences in the every-
day experience of young American children. Baltimore, MD:
Brookes.
Helland, I. B., Smith, L., Saarem, K., Saugstad, O. D., & Drevon, C.
A. (2003). Maternal supplementation with very long-chain n-3
fatty acids during pregnancy and lactation augments children’s
IQ at 4 years of age. Pediatrics, 111, e39–e44.
Herrnstein, R. A., & Murray, C. (1994). The bell curve. New York,
NY: Grove Press.
Herzog, E., Newcomb, C. H., & Cisin, I. H. (1974). Double depri-
vation: The less they have the less they learn. In S. Ryan (Ed.),
A report on longitudinal evaluations of pre-school programs
(Vol. 1, pp. 62–76). Washington, DC: Department of Health, Edu-
cation, and Welfare.
Huebner, C. E. (2000). Promoting toddlers’ language development
through community-based intervention. Journal of Applied
Developmental Psychology, 21, 513–535.
Idjradinata, P., & Pollitt, E. (1993). Reversal of developmental delays
in iron deficient anemic infants treated with iron. Lancet, 341,
1–4.
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008).
Improving fluid intelligence with training on working memory.
Proceedings of the National Academy of Sciences, USA, 105,
6829–6833.
Jester, R. E., & Guinagh, B. J. (1983). The Gordon parent education
infant and toddler program. In The Consortium for Longitudinal
Studies (Ed.), As the twig is bent—Lasting effects of preschool
programs (pp. 103–132). Hillsdale, NJ: Erlbaum.
Judge, M. P., Harel, O., & Lammi-Keefe, C. J. (2007). Maternal con-
sumption of a docosahexaenoic acid-containing functional food
during pregnancy: Benefit for infant performance on problem-
solving but not on recognition memory tasks at 9 mo. American
Journal of Clinical Nutrition, 85, 1572–1577.
Kagitcibasi, C., Sunar, D., & Bekman, S. (2001). Long-term effects
of early intervention: Turkish low-income mothers and children.
Journal of Applied Developmental Psychology, 22, 333–361.
Karnes, M. B., Hodgins, A. S., Stoneburner, R. L., Studley, W. M.,
& Teska, J. A. (1968). Effects of a highly structured program
of language development on intellectual functioning and psy-
cholinguistic development. Journal of Special Education, 2,
405–412.
Karnes, M. B., Studley, W. M., Wright, W. R., & Hodgins, A. S.
(1968). An approach for working with mothers of disadvantaged
preschool children. Merrill-Palmer Quarterly, 14, 174–184.
Karnes, M. B., Zehrbach, R. R., & Teske, J. A. (1974). The Karnes’
preschool program: Rationale, curricula offerings and follow-up
data. In S. Ryan (Ed.), A report on longitudinal evaluations of
pre-school programs, Vol. 1. Washington, DC: Department of
Health, Education, and Welfare.
Klaus, R. A., & Gray, S. W. (1968). The early training project for
disadvantaged children: A report after five years. Monographs of
the Society for Research in Child Development, 33, 1–66.
Klingberg, T., Fernell, E., Olesen, P., Johnson, M., Gustafsson, P.,
Dahlstrom, K., . . . Westerberg, H. (2005). Computerized train-
ing of working memory in children with ADHD: A randomized,
controlled trial. Journal of the American Academy of Child &
Adolescent Psychiatry, 44, 177–186.
Kurlak, L. O., & Stephenson, T. J. (1999). Plausible explanations for
effects of long chain polyunsaturated fatty acids (LC-PUFA) on
neonates. Archives of Diseases in Childhood Fetal & Neonatal
Edition, 80, F148–F154.
Lamb, H. A. (1986). The effects of a read-aloud program with lan-
guage interaction (Unpublished doctoral dissertation). Tallahas-
see, FL: Florida State University.
Levenstein, P., O’Hara, J., & Madden, J. (1983). The mother–child
home program of the verbal interaction project. In Consortium
for Longitudinal Studies (Ed.), As the twig is bent—lasting effects
of preschool programs (pp. 237-263). Hillsdale, NJ: Erlbaum.
Lonigan, C. J., Anthony, J. L., Bloomfield, B. G., Dyer, S. M., &
Samwel, C. S. (1999). Effects of two preschool shared reading
interventions on the emergent literacy skills of children from low
income families. Journal of Early Intervention, 22, 306–322.
Love, J. M., Kisker, E. E., Ross, C., Raikes, H., Constantine, J.,
Boller, K., . . . Vogel, C. (2005). The effectiveness of Early Head
Start for 3-year-old children and their parents: Lessons for policy
and programs. Developmental Psychology, 41, 885–901.
Lozoff, B., De Andraca, I., Castillo, M., Smith, J. B., Walter, T.,
& Pino, P. (2003). Behavioral and developmental effects of
preventing iron-deficiency anemia in healthy full-term infants.
Pediatrics, 112 , 846–854.
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
How to Make a Young Child Smarter 39
McNamara, R. K., Able, J., Jandacek, M. P. R., Rider, T., Tso, P., Elias-
sen, J. C., . . . Adler, C. M. (2009). Docosahexaenoic acid supple-
mentation augments prefrontal cortex activation during sustained
attention in healthy boys: A placebo-controlled dose ranging fMRI
study. American Journal of Clinical Nutrition, 91, 1060–1067.
Neisser, U., Boodoo, G., Bouchard, T. J., Jr., Boykin, A. W., Brody,
N., Ceci, S. J., . . . Urbina, S. (1996). Intelligence: Knowns and
unknowns. American Psychologist, 51, 77–101.
Nisbett, R. E. (2009). Intelligence and how to get it: Why schools and
cultures count. New York, NY: W. W. Norton.
Nisbett, R. E., Aronson, J., Blair, C., Dickens, W., Flynn, J., Halpern,
D. F., & Turkheimer, E. (2012). Intelligence: New findings and
theoretical developments. American Psychologist, 67, 130–159.
Nutley, S. B., Soderqvist, S., Bryde, S., Thorell, L. B., Humphreys,
K., & Klingberg, T. (2011). Gains in fluid intelligence after train-
ing non-verbal reasoning in 4-year-old children: A controlled,
randomized study. Developmental Science, 14, 591–601.
O’Connor, D. L., Jacobs, J., Hall, R., Adamkin, D., Auestad, N.,
Castillo, M., . . . Zemon, V. (2003). Growth and development of
premature infants fed predominantly human milk, predominantly
premature infant formula, or a combination of human milk and
premature formula. Journal of Pediatric Gastroenterology and
Nutrition, 37, 437–446.
Overton, R. C. (1998). A comparison of fixed-effects and mixed (ran-
dom-effects) models for meta-analysis tests of moderator vari-
able effects. Psychological Methods, 3, 354–379.
Peta, E. J. (1973). The effectiveness of a total environment room on an
early reading program for culturally different pre-school children
(Unpublished doctoral dissertation). Lehigh University, Bethle-
hem, PA.
Peterson, C., Jesso, B., & McCabe, A. (1999). Encouraging narra-
tives in preschoolers: An intervention study. Journal of Child
Language, 26, 49–67.
Pollitt, E. (1993). Iron deficiency and cognitive function. Annual
Review of Nutrition, 13, 521–537.
Puma, M., Bell, S., Cook, R., & Heid, C. (2010). Head start impact
study. Washington, DC: Administration for Children and Families.
Rainey, E. W. (1968). The development and evaluation of a language
development program for culturally deprived preschool children
(Unpublished doctoral dissertation). University of Mississippi,
Oxford.
Ramey, C. T., Bryant, D. M., Wasik, B. H., Sparling, J. J., Fendt, K.
H., & La Vange, L. M. (1992). The infant health and development
program. Program elements, family participation, and child intel-
ligence. Pediatrics, 89, 454–465.
Ramey, C. T., & Ramey, S. L. (1998). Early intervention and early
experience. American Psychologist, 53, 109–120.
Ramey, C. T., Yeates, K., & Short, E. (1984). The plasticity of intel-
lectual development: Insights from preventive intervention. Child
Development, 55, 1913–1925.
Reese, E., Leyva, D., Sparks, A., & Grolnick, W. (2010). Maternal
elaborative reminiscing increases low-income children’s narrative
skills relative to dialogic reading. Early Education and Develop-
ment, 21, 318–342.
Rema, V., Armstrong-James, M., & Ebner, F. F. (1998). Experience-
dependent plasticity of adult rat S1 cortex requires local NMDA
receptor activation. Journal of Neuroscience, 18, 10196–10206.
Riding, R. J., & Powell, S. D. (1985). The facilitation of thinking
skills in pre-school children using computer presented activities.
Educational Psychology, 5, 171–178.
Riding, R. J., & Powell, S. D. (1986). The improvement of thinking
skills in young children using computer activities: A replication
and extension. Educational Psychology, 6, 179–183.
Rueda, M. R., Checa, P., & Cómbita, L. M. (2012). Enhanced effi-
ciency of the executive attention network after training in
preschool children: Immediate and after two months effects.
Developmental Cognitive Neuroscience, 2S, S192–S204.
Rueda, M. R., Rothbart, M., McCandliss, B., Saccomanno, L., &
Posner, M. (2005). Training, maturation, and genetic influences
on the development of executive attention. Proceedings of the
National Academy of Sciences, USA, 102, 14931–14936.
Ryan, A. S., & Nelson, E. B. (2008). Assessing the effect of doco-
sahexaenoic acid on cognitive functions in healthy, preschool
children: A randomized, placebo-controlled, double-blind study.
Clinical Pediatrics, 47, 355–362.
Salem, N., Jr., Kim, H. Y., & Yergey, J. A. (1986). Docosahexaenoic
acid: Membrane function and metabolism. In A. P. Simopoulos,
R. R. Kiefer, & R. E. Martin (Eds.), Health effects of polyun-
saturated fatty acids in seafoods (pp. 263–317). New York, NY:
Academic Press.
Sauberlich, H. E. (1980). Interactions of thiamin, riboflavin, and
other B-vitamins. Annals of the New York Academy of Science,
355, 80–97.
Scarr, S., & McCartney, K. (1988). Far from home: An experimental
evaluation of the mother-child home program in Bermuda. Child
Development, 59, 531–543.
Seshadri, S., & Gopaldes, T. (1989). Impact of iron supplementation
on cognitive functions in pre-school and school-aged children:
The Indian experience. American Journal of Clinical Nutrition,
50, 675–686.
Sidik, K., & Jonkman, J. N. (2003). On constructing the confidence
intervals for a standardized mean difference in meta-analysis.
Communications in Statistics–Simulation and Computation, 32,
1191–2003.
Skeels, H. M., Ruth, U., Wellman, B. L., & Williams, H. M. (1938).
A study of environmental stimulation: an orphanage pre-
school project. University of Iowa Studies in Child Welfare, 15,
37–74.
Spitz, J. (1986). The raising of intelligence. Englewood, NJ: Erlbaum.
StataCorp. (2011). Stata statistical software: Release 12. College Sta-
tion, TX: Author.
Stevens, L., Zhang, W., Peck, L., Kuczeck, T., Grevstad, N., Mahon,
A., . . . Burgess, J. R. (2003). EFA supplementation in children
with inattention, hyperactivity, and other disruptive behaviors.
Lipids, 38, 1007–1021.
Tamura, T., Goldenberg, R. L., Ramey, S. L., Nelson, K. G., &
Chapman, V. R. (2003). Effect of zinc supplementation of preg-
nant women on the mental and psychomotor development of their
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from
40 Protzko et al.
children at 5 y of age. American Journal of Clinical Nutrition, 77,
1512–1516.
Valdez-Menchaca, M. C, & Whitehurst, G. J. (1992). Accelerating
language development through picture book reading: A system-
atic extension to Mexican day care. Developmental Psychology,
28, 1106–1114.
van Kleeck, A. G. J., Vander-Woude, J., & Hammett, L. A. (2006).
Fostering literal and inferential language skills in head start pre-
schoolers with language impairment using scripted book-sharing
discussions. American Journal of Speech-Language Pathology,
15, 85–95.
Wasik, B. H., Ramey, C. T., Bryant, D. M., & Sparling, J. J. (1990).
A longitudinal study of two early intervention strategies: Project
CARE. Child Development, 61, 1682–1696.
Watanabe, M., Kodama, T., & Hikosaka, K. (1997). Increase of extra-
cellular dopamine in primate prefrontal cortex during a working
memory task. Journal of Neurophysiology, 78, 2795–2798.
Weikart, D. P. (1966). Preschool programs: Preliminary findings.
Journal of Special Education, 1, 153–181.
Werkman, S. H., & Carlson, S. E. (1996). A randomized trial of visual
attention of preterm infants fed docosahexaenoic acid until nine
months. Lipids, 31, 91–97.
Whitehurst, G. J., Arnold, D. S., Epstein, J. N., & Angell, A. L.
(1994). A picture book reading intervention in day care and home
for children from low-income families. Developmental Psychol-
ogy, 30, 679–689.
Whitehurst, G. J., Falco, F., Lonigan, C. J., Fischel, J. E., DeBaryshe,
B. D., Valdez-Menchaca, M. C., & Caulfield, M. (1988). Accel-
erating language development through picture-book reading.
Developmental Psychology, 24, 552–558.
Zhou, S. J., Gibson, R. A., Crowther, C. A., Baghurst, P., & Makrides,
M. (2006). Effect of iron supplementation during pregnancy on
the intelligence quotient and behavior of children at 4 y of age:
Long-term follow-up of a randomized controlled trial. American
Journal of Clinical Nutrition, 83, 1112–1117.
Zigler, E., & Butterfield, E. C. (1968). Motivational aspects of
changes in IQ test performance of culturally deprived nursery
school children. Child Development, 39, 1–14.
at Bobst Library, New York University on January 18, 2013pps.sagepub.comDownloaded from

Supplementary resource (1)

... As noted by Price and Kalil (2019), these are lacking in this area of inquiry. A fairly recent meta-analytical review identifies only eight of such trials (Protzko et al., 2013) most of which, however, tend to have passive control conditions in which children were simply not read to or were just left to play. Furthermore, the focus of these trials has been on preschool children and, with a notable, recent exception , to which we shall return in the discussion, they have focused on the effects of reading on language development. ...
... The VCI comprises of three subscales: (1), abstract verbal reasoning (Similarities); (2) the definition of words without context (Vocabulary); and (3) the ability to understand abstract or idiomatic expressions (Comprehension). This measure was selected because among the WISC-IV subscales, the VCI is the most strongly correlated with achievement tests (Kaufman, 2018) and because previous work on the effects of reading-aloud has used either this measure or similar language-focused measures (Protzko et al., 2013). This allows for direct comparison between the results of the present trial and the results of some of the studies reported above on the effects of (self-report) parental reading on younger children. ...
... Although rare, previous randomized-control trials exist on the effects of reading aloud, but have typically focused on single measures of language development as opposed to standardized measures of intelligence, and have been conducted with relatively small samples (for a review see Protzko et al., 2013). The most comparable study to the three studies reported here was conducted amid low-income families in Brazil in the context of The Universidade do Bebê program . ...
Article
Storytelling played a crucial role in human evolution. To this day, through stories humans gain declarative and procedural knowledge, and learn the skills that support learning itself. Research shows that reading stories to children enhances their reading and language skills. Does it also enhance their intelligence? To answer this question , we conducted three (N = 626, 254, 195) longitudinal, cluster-randomized control trials in Italian elementary and middle schools. Over a 4-month period, for half of the participants 1 h/day of standard, active language instructional activities were substituted with reading-aloud of stories by a teacher. Compared to those who kept doing language instructional activities, read-aloud condition children showed a stronger increase on two measures of intelligence focusing on knowing things and thinking skills. This result, which emerged in three independent trials conducted in different regions of Italy, suggests avenues for easily scalable interventions to improve children's intelligence.
... Furthermore, adoption studies show that disadvantaged children adopted into better circumstances have higher IQs . Some of the environmental factors responsible for socioeconomic differences in IQ may be breastfeeding or diet (Protzko et al., 2013), differential exposure to complex spoken language, differential education, and differential exposure to traumatic stress (Nisbett et al., 2012). Additional evidence for environmental contributions to intelligence comes from the Flynn effect (Flynn, 1987) (Figure 9-2). ...
... High quality teachers in kindergarten and first grade also can boost IQ . Intervention programs increase disadvantaged students' IQs by 4-7 points depending on the quality of the intervention (Protzko et al., 2013). Imaging studies indicate that IQ changes correlate with typical changes in the cerebral cortex (Schnack et al., 2015). ...
Book
Full-text available
We wrote this book for undergraduate and graduate students of the helping professions. It is a complete view of development from an ethological and cognitive behavioral perspective. It is unique in its synthesis of biological, anthropological and psychological accounts of development.
... Four meta-analyses yielded significant results; supplementing infants with longchain polyunsaturated fatty acids, enrolling children in early educational interventions, reading to children in an interactive manner, and sending children to preschool all raise the intelligence of young children. (Protzko, Aronson, & Blair, 2013). ...
Preprint
Full-text available
This multidisciplinary study addresses pressing societal issues such as terrorism, crime, divorce , homicides, theft, bribery, corruptions, war, and drug abuse by exploring their underlying psychological roots. Identifying over 3,000 "mind viruses" that distort human intelligence and decision-making, this research proposes a transformative approach to mental health through the introduction of 3,000 positive "healthy mind viruses." By examining the evolution of intelligence from a biological perspective and considering the implications of consciousness in the context of an afterlife, the study suggests that a psychological framework based on the 'eight-fold path' may serve as an effective method for identifying and mitigating these detrimental influences. Additionally, it explores concepts like nature and nurture, the distinction between clock time and psychological time, and the pursuit of true happiness, answering over forty innovative big questions. A 3D graph illustrating the mechanisms of individual and social intelligence evolution underscores the potential for fostering a more harmonious world.
... In one of their meta-analyses using a subset of intervention studies from the Database of Raising Intelligence (DORI), Protzko et al. (2013) found that regularly reading to a child in an interactive fashion raised his or her IQ by over six points. However, this only occurred if the parents started this habit before the child was four years old, and benefits were larger for early starters. ...
Article
Full-text available
Objective: This study investigates whether greater identification of mothers and fathers in different-sex couples with the stimulation dimension of intensive parenting promoted both parents' involvement in reading and benefited children's language development between ages 3 and 5 in the UK. In addition, we explore social class variations. Background: Our study tests the frequently assumed relationship of parenting beliefs about stimulation with language development, and contributes to our understanding of parenting practices and their implications for child development. Method: We draw on a large representative sample of young children from the UK Millennium Cohort Study (N = 8,071) and apply path analyses in the framework of structural equation models. Results: For mothers and fathers, stimulation beliefs partially mediated the relation between parental education and reading frequency. Mothers' and fathers' stimulation beliefs had positive effects on children’s language abilities, but their impact was small compared to the direct associations with parental education. Conclusion: While parental education emerges as a key determinant of children's language development, our study reveals the nuanced role of beliefs about stimulation within intensive parenting, prompting further investigation into the multifaceted nature of parental involvement.
... Die Annahme, dass die Intelligenz einen vermittelnden Mechanismus zwischen sozialer Herkunft und Lernerfolg darstellt, wird weitestgehend konsistent bestätigt (Kriegbaum & Spinath, 2016, S. 52;Steinmayr et al., 2010, S. 542;Steinmayr et al., 2012, S. 343). Dieser Befund kann und sollte allerdings nicht zur Legitimation sozialer Ungleichheit genutzt werden (Brake & Büchner, 2012, S. 86 ff.), da die Intelligenz eine, bis zu einem gewissen Grad, veränderbare Größe darstellt (Protzko et al., 2013). ...
Book
Full-text available
Das Buch verbindet den Forschungsbereich der Wirksamkeit von Unterricht und Schule mit dem Forschungsbereich der Bildungsungleichheiten. Es befasst sich mit den Ursachen von Kompetenzunterschieden nach sozialer Herkunft und möglichen schulischen Maßnahmen zur Verringerung dieser Kompetenzunterschiede. Im theoretischen Teil wird ein Modell zu vermittelnden Mechanismen zwischen sozialer Herkunft und Lernerfolg entwickelt. Weiterhin wird erörtert, auf welche Art schulische Einflussfaktoren Kompetenzunterschiede nach sozialer Herkunft beeinflussen. Im empirischen Teil werden Analysen auf der Grundlage des Nationalen Bildungspanels präsentiert. Es wird untersucht, durch welche Faktoren Kompetenzunterschiede nach sozialer Herkunft entstehen (z.B. elterliche Lernbegleitung, Freizeitaktivitäten). Weiterhin werden die Effekte von Unterrichtsmerkmalen, Ganztagsangeboten und Klassenmerkmalen auf den Lernerfolg sowie auf Kompetenzunterschiede nach sozialer Herkunft untersucht.
... Four meta-analyses yielded significant results; supplementing infants with longchain polyunsaturated fatty acids, enrolling children in early educational interventions, reading to children in an interactive manner, and sending children to preschool all raise the intelligence of young children. (Protzko, Aronson, & Blair, 2013). ...
Preprint
Full-text available
Humankind faces many problems, such as terrorism, crime, divorce, drug abuse, war, theft, etc. I performed this multidisciplinary study to determine whether there are primary and common psychological reasons and solutions for significant issues. I have discovered more than 3000 mind viruses that make grand delusions that significantly regress human intelligence and decisions in social and individual well-being. How can those mind viruses be destroyed by ('3000') healthy mind viruses? Here I suggest the fundamental nature of intelligence and why/how it resulted from biological evolution and (if) the afterlife(occurs) how does universally evolve the consciousness. The psychological(scientific) course on the 'eight-fold path' might be a better/best way to learn and train to scan mind viruses to prevent/ cure issues. Furthermore, discuss better nurture, clock time vs. psychological time, true happiness, and more than forty new ideas. A 3D graph showing individual and social intelligence evolution mechanisms make a happier world.
... To speculate further, the ultimate (as opposed to proximate or immediate) effect of maternal supportiveness or other environmental influences, such as interventions designed to increase cognitive abilities (Protzko, Aronson, & Blair, 2013), on general intelligence (Protzko, 2016) could be due to a slowing life history strategy. Maternal supportiveness is thought to push a child's development toward a slower life history trajectory (e.g., Szepsenwol & Simpson, 2019) freeing bioenergetic resources to be allocated to cognitive growth and, hence, greater general intelligence (Dunkel et al., 2021). ...
Article
Data from the Early Head Start Research and Evaluation Project (N = 1075) were used to test the hypothesis that maternal supportiveness (measured at three waves from 14 to 36 months) is positively and prospectively associated with a child's general intelligence (measured at five waves from 14 months to 10 years). Bivariate correlations showed that maternal supportiveness was consistently and positively associated with a child's general intelligence. For example, maternal supportiveness as measured at 14 months was correlated with a child's general intelligence at age 10; r = 0.35. Results of autoregressive cross-lagged panel models showed maternal supportiveness directly predicted future general intelligence through age four and indirectly, via age four general intelligence, up to age 10. Additional analyses verified that the effect of maternal supportiveness was on general intelligence and not specific abilities. The results point to the importance of maternal supportiveness on general intelligence in the first decade of life.
... 24 GAME is provided in the natural home environment where training is personalised to the infant's enjoyment and family customs-translating to more intense, specific and relevant practice. GAME intervention enriches the: (1) physical environment by setting up spaces within the home with activities and materials selected to entice infant-generated motor practice at the appropriate level of challenge; (2) cognitive environment by encouraging infant problem-solving and self-correction of errors plus reading to children in an interactive manner which has been proven to advance IQ 37 ; (3) sensory environment by providing evidencebased interventions that improve backdrop capacity for learning, including: pain management, feeding interventions that ensure adequate nutrition for attention 37 and sleep management that produces a wakeful state for learning; and (4) social environment by coaching parents to be sensitive, responsive and communicative to infant cues, 36 and promote their child's involvement in family events and routines. How well Only the GAME therapists will be educated in GAME. ...
Article
Full-text available
Introduction Cerebral palsy (CP) is the most common physical disability of childhood worldwide. Historically the diagnosis was made between 12 and 24 months, meaning data about effective early interventions to improve motor outcomes are scant. In high-income countries, two in three children will walk. This evaluator-blinded randomised controlled trial will investigate the efficacy of an early and sustained Goals–Activity–Motor Enrichment approach to improve motor and cognitive skills in infants with suspected or confirmed CP. Methods and analysis Participants will be recruited from neonatal intensive care units and the community in Australia across four states. To be eligible for inclusion infants will be aged 3–6.5 months corrected for prematurity and have a diagnosis of CP or ‘high risk of CP’ according to the International Clinical Practice Guideline criteria. Eligible participants whose caregivers consent will be randomly allocated to receive usual care or weekly sessions at home from a GAME-trained study physiotherapist or occupational therapist, paired with a daily home programme, until age 2. The study requires 150 participants per group to detect a 0.5 SD difference in motor skills at 2 years of age, measured by the Peabody Developmental Motor Scales-2. Secondary outcomes include gross motor function, cognition, functional independence, social–emotional development and quality of life. A within-trial economic evaluation is also planned. Ethics and dissemination Ethical approval was obtained from the Sydney Children’s Hospital Network Human Ethics Committee in April 2017 (ref number HREC/17/SCHN/37). Outcomes will be disseminated through peer-reviewed journal publications, presentations at international conferences and consumer websites. Trial registration number ACTRN12617000006347.
Article
L’alimentation maternelle durant la grossesse et en période postnatale est cruciale pour le développement du fœtus, la santé du nouveau-né et celle de l’adulte en devenir. Un apport nutritionnel insuffisant sur les plans quantitatif et qualitatif favorise, à plus ou moins long terme, certaines pathologies métaboliques, chroniques mais aussi mentales. Un apport excessif en graisses et en sucres peut quant à lui faire le lit d’une obésité tardive. Plus l’alimentation de la mère est diversifiée, plus l’enfant acceptera de manger varié et équilibré, ce qui préservera sa santé. © 2024 Publié par Elsevier Masson SAS Mots clés – apport nutritionnel ; carence ; grossesse ; maladie chronique ; maladie métabolique ; trouble cognitif The importance of maternal nutrition for child health. Maternal nutrition during pregnancy and in the postnatal period is crucial for the development of the fetus, the health of the newborn and that of the future adult. An inadequate nutritional intake, in terms of both quantity and quality, is conducive to certain metabolic, chronic and mental pathologies over the long term. Excessive fat and sugar intake, on the other hand, can lead to obesity later in life. The more diversified the mother’s diet, the more varied and balanced the child’s diet will be, and the healthier it will be. © 2024 Published by Elsevier Masson SAS Keywords – chronic disease; cognitive disorder; deficiencie; metabolic disease; nutritional intake; pregnancy
Article
Full-text available
For 4 decades, vigorous efforts have been based on the premise that early intervention for children of poverty and, more recently, for children with developmental disabilities can yield significant improvements in cognitive, academic, and social outcomes. The history of these efforts is briefly summarized and a conceptual framework presented to understand the design, research, and policy relevance of these early, interventions. This framework, biosocial developmental contextualism, derives from social ecology, developmental systems theory, developmental epidemiology, and developmental neurobiology. This integrative perspective predicts that fragmented, weak efforts in early intervention are not likely to succeed, whereas intensive high-quality, ecologically pervasive interventions can and do. Relevant evidence is summarized in 6 principles about efficacy of early intervention. The public policy challenge in early intervention is to contain costs by more precisely targeting early interventions to those who most need and benefit from these interventions. The empirical evidence on biobehavioral effects of early experience and early intervention has direct relevance to federal and state policy development and resource allocation.
Article
The Infant Health and Development Program was an eight-site randomized controlled trial testing the efficacy of early intervention to enhance the cognitive, behavioral, and health status of low birth weight, premature infants. The 377 intervention families received for the first 3 years of life: (1) pediatric follow-up, (2) home visits, (3) parent support groups, and (4) a systematic educational program provided in specialized child development centers. The control group (n = 608) received the same pediatric follow-up and referral services only. This paper describes the delivery of the intervention and its outcomes. A Family Participation Index that was the sum of participation frequencies in each of the program modalities unique to the intervention revealed that program implementation was not different across the eight sites. Index scores did not vary systematically with mother's ethnicity, age, or education or with child's birth weight, gender, or neonatal health status; but they were positively related to children's IQ scores at age 3. Only 1.9% of children of families in the highest tercile of participation scored in the mentally retarded range (IQ ≤70), whereas 3.5% and 13% of children in the middle and lowest participation terciles, respectively, scored in the retarded range. Similar findings were obtained for borderline intellectual functioning. These findings are consistent with previous research linking intensity of intervention services with degree of positive cognitive outcomes for high-risk infants. The determinants of variations in individual family participation remain unknown.
Article
Objective. To determine the behavioral and developmental effects of preventing iron-deficiency anemia in infancy. Methods. Healthy full-term Chilean infants who were free of iron-deficiency anemia at 6 months were assigned to high- or low-iron groups or to high- or no-added-iron groups. Behavioral/developmental outcomes at 12 months of age included overall mental and motor test scores and specific measures of motor functioning, cognitive processing, and behavior. There were no differences between high- and low-iron groups in the prevalence of iron-deficiency anemia or behavioral/developmental outcome, and they were combined to form an iron-supplemented group (n = 1123) for comparison with the no-added-iron group (n = 534). Results. At 12 months, iron-deficiency anemia was present in 3.1% and 22.6% of the supplemented and unsupplemented groups, respectively. The groups differed in specific behavioral/developmental outcomes but not on global test scores. Infants who did not receive supplemental iron processed information slower. They were less likely to show positive affect, interact socially, or check their caregivers’ reactions. A smaller proportion of them resisted giving up toys and test materials, and more could not be soothed by words or objects when upset. They crawled somewhat later and were more likely to be tremulous. Conclusions. The results suggest that unsupplemented infants responded less positively to the physical and social environment. The observed differences seem to be congruent with current understanding of the effects of iron deficiency on the developing brain. The study shows that healthy full-term infants may receive developmental and behavioral benefits from iron supplementation in the first year of life.
Book
With a careful study of the recent research literature on individual differences in the preverbal infant, this book presents the historical and procedural contexts of four measures of infant attention and learning that have proved the most promising in predicting later childhood intellectual performance. The author examines the psychometric properties for each measure (within-age and cross-age test-retest coefficients), the concurrent relations of the measures with other measures of early cognition and development, and the available evidence on how well the measures predict cognitive and intellectual development in later childhood.
Article
This is a review of the relationship between schooling, IQ, and the cognitive processes presumed to underpin IQ. The data suggest that much of the causal pathway between IQ and schooling points in the direction of the importance of the quantity of schooling one attains (highest grade successfully completed). Schooling fosters the development of cognitive processes that underpin performance on most IQ tests. In western nations, schooling conveys this influence on IQ and cognition through practices that appear unrelated to systematic variation in quality of schools. If correct, this could have implications for the meaning one attaches to IQ in screening and prediction as well as for efforts to influence the development of IQ through changes in schooling practices.