Content uploaded by Charles Ungerleider
Author content
All content in this area was uploaded by Charles Ungerleider on Aug 04, 2015
Content may be subject to copyright.
Content uploaded by Olusola O. Adesope
Author content
All content in this area was uploaded by Olusola O. Adesope on Oct 30, 2014
Content may be subject to copyright.
http://rer.aera.net
Review of Educational Research
DOI: 10.3102/0034654310368803
2010; 80; 207 REVIEW OF EDUCATIONAL RESEARCH
Olusola O. Adesope, Tracy Lavin, Terri Thompson and Charles Ungerleider
Bilingualism
A Systematic Review and Meta-Analysis of the Cognitive Correlates of
http://rer.sagepub.com/cgi/content/abstract/80/2/207
The online version of this article can be found at:
Published on behalf of
http://www.aera.net
By
http://www.sagepublications.com
can be found at:Review of Educational Research Additional services and information for
http://rer.aera.net/cgi/alerts Email Alerts:
http://rer.aera.net/subscriptions Subscriptions:
http://www.aera.net/reprintsReprints:
http://www.aera.net/permissionsPermissions:
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
207
Review of Educational Research
June 2010, Vol. 80, No. 2, pp. 207–245
DOI: 10.3102/0034654310368803
© 2010 AERA. http://rer.aera.net
A Systematic Review and Meta-Analysis of the
Cognitive Correlates of Bilingualism
Olusola O. Adesope
Washington State University
Tracy Lavin and Terri Thompson
Directions Evidence and Policy Research Group
Charles Ungerleider
University of British Columbia
A number of studies have documented the cognitive outcomes associated with
bilingualism. To gain a clear understanding of the extent and diversity of
these cognitive outcomes, the authors conducted a meta-analysis of studies
that examined the cognitive correlates of bilingualism. Data from 63 studies
(involving 6,022 participants) were extracted and analyzed following estab-
lished protocols and procedures for conducting systematic reviews and
guidelines for meta-analysis. Results indicate that bilingualism is reliably
associated with several cognitive outcomes, including increased attentional
control, working memory, metalinguistic awareness, and abstract and sym-
bolic representation skills. Overall mean effect sizes varied from small to
large, depending on the cognitive outcomes measured, and were moderated
by methodological features of the studies.
Keywords: bilingual, cognitive correlates, biliteracy, meta-analysis, systematic
review.
Early research on bilingualism warned that bilingualism could be deleterious to
learning. These early studies concluded that monolingual students outperformed
bilingual students on a range of cognitive tasks (for reviews, see Bhatia & Ritchie,
2006; Hakuta, 1986; Macnamara, 1966).
In a seminal article, Peal and Lambert (1962) introduced the concept of “bal-
anced bilinguals” and demonstrated the methodological weaknesses of previous
bilingual studies, providing a new approach for research on bilingualism. Peal and
Lambert noted that early studies on the effects of bilingualism did not properly
match bilingual and monolingual participants along several dimensions, including
socioeconomic status (SES), second language proficiency (pseudobilingualism),
language of assessment, gender, age, and urban–rural contexts. They noted that
these and other factors may have confounded earlier results showing bilingual
disadvantages on cognitive measures. Controlling for these extraneous factors,
Peal and Lambert found that bilingual participants significantly outperformed
monolinguals on several measures of verbal and nonverbal intelligence.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
208208
Since Peal and Lambert’s (1962) original studies, a considerable body of evi-
dence has accumulated suggesting that bilingualism confers a number of cognitive
benefits. For example, researchers have observed that bilinguals may have greater
metalinguistic awareness (Bialystok, 1987, 1988, 2001b; Diaz, 1985; Diaz &
Klinger, 1991; Ferdman & Hakuta, 1985; Goetz, 2000; Hakuta, 1990; Huber &
Lasagabaster, 2000; Ricciardelli, 1993; Titone, 1997) and enhanced metacognitive
skills (Duncan, 2005). Bilinguals may have stronger symbolic representation and
abstract reasoning skills (Bamford & Mizokawa, 1990, 1992; Berguno & Bowler,
2004; Chan, 2005; Diaz, 1985; Goncz, 1988; Johnson, 1991; McLeay, 2003), as
well as better learning strategies (Bochner, 1996; Ponomarev, 1992). Bilinguals
may also have enhanced problem-solving skills because of their ability to selec-
tively attend to relevant information and disregard misleading information
(Bamford & Mizokawa, 1991; Bialystok, 1999, 2001a, 2005; Bialystok &
Majumber, 1998; Duncan, 2005; Stephens, 1997) and may be able to use this
selectivity to succeed at theory-of-mind tasks, which require the ability to attribute
the behavior of others to their own distinct beliefs, desires, and intentions (Chan,
2005; Goetz, 2000). Bilinguals may have enhanced creative and divergent thinking
skills (Braccini & Cianchi, 1993; Ho, 1987; Konaka, 1997; Ricciardelli, 1993;
Srivastava, 1991) and greater cognitive flexibility (Hakuta, 1990; Iannaccone,
Fraternali, & Vaccia, 1992; Kovacs & Teglas, 2002; Kozulin, 1999).
Although many studies have documented advantages for bilinguals on cogni-
tive tasks, other studies have reported negative, null, or mixed effects of bilingual-
ism (Macnamara, 1966; Rosenblum & Pinker, 1983). To make sense of these
conflicting findings, the current work synthesizes the available research on the
cognitive correlates of bilingualism. This review does not investigate the effective-
ness of bilingual education programs because previous reviews, meta-analyses,
and best-evidence syntheses have addressed the question of program effectiveness,
albeit with inconclusive results (Baker & de Kanter, 1981; Rossell & Baker, 1996;
Slavin & Cheung, 2005; Willig, 1985, 1987). Specifically, the current review
focuses on examining the cognitive correlates of bilingualism and the associated
effect sizes. The following section discusses potential relationships between bilin-
gualism and various cognitive skills.
Attentional Control
There is considerable evidence that bilingual speakers are more readily able
to control their attention while engaged in linguistic and nonverbal tasks com-
pared to monolingual learners (Bialystok, 2001a; Bialystok, Craik, Klein, &
Viswanathan, 2004; Bialystok, Craik, & Ryan, 2006; Emmorey, Luk, Pyers, &
Bialystok, 2008). Several explanations have been advanced for this cognitive
advantage. A dominant perspective suggests that the regular use of two lan-
guages requires that bilinguals control their attention and select the target lan-
guage. Some researchers have claimed that the ability to selectively attend to
different representations may be responsible for the greater attentional control
exhibited by bilingual participants in many studies (Bialystok, 2001a; Bialystok,
Martin, & Viswanathan, 2005; Yoshida, 2008). In other words, these researchers
speculate that bilingual learners’ ability to concurrently hold two languages in
the mind, resisting intrusions of words and grammar from one language into the
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
209209
other, might explain the greater control reflected by improved performance on
tasks with conflicting or distracting information.
More recently, researchers have also shown that the cognitive control of attention
found in studies with bilingual children appears to be sustained into adulthood. For
example, Bialystok et al. (2004) found that adults who have been bilingual since
childhood are more capable than comparable monolingual adults of managing
their attention when presented with tasks requiring cognitive control.
In addition, there is a growing body of evidence that bilingualism may help
offset some age-related cognitive declines by building cognitive reserves that slow
the aging process for adults (Bialystok, Craik, & Freedman, 2007; Bialystok et al.,
2004). In a recent study on the effect of lifelong bilingualism on age-related cogni-
tive decline, Bialystok et al. (2007) found that bilingual adults showed symptoms
of dementia 4 years later than comparable monolinguals, even when other factors
remained constant. The preliminary findings in the literature suggest that “the
lifelong experience of managing two languages attenuates the age-related decline
in the efficiency of inhibitory processing” (Bialystok et al., 2004, p. 301).
Working Memory
There are at least two contrasting hypotheses about the relationship between
bilingualism and working memory. First, the need to manage two languages con-
currently could place greater demands on working memory. This hypothesis sug-
gests that bilingualism may impede efficient processing of information in working
memory because of the cognitive load imposed on working memory (Lee, Plass,
& Homer, 2006; Sweller & Chandler, 1994; van Merrienboer & Sweller, 2005).
Conversely, bilinguals’ well-developed ability to inhibit one language while using
the other may increase the efficiency of their working memory capacity because
working memory resources are properly managed through such inhibitory process-
ing (Bialystok et al., 2004; Bialystok, Craik, & Luk 2008; Fernandes, Craik,
Bialystok, & Kreuger, 2007; Just & Carpenter, 1992; Michael & Gollan, 2005;
Rosen & Engle, 1997).
Research on these competing hypotheses has yielded inconclusive findings
with results depending on the nature of the task (Bialystok et al., 2008). On tasks
that require greater attentional control, bilinguals appear to have greater working
memory capacity than monolinguals (Engle, 2002; Kane, Bleckley, Conway, &
Engle, 2001). In attention-aided tasks, however, the bilingual advantage disappears
(Yang, Yang, Ceci, & Wang, 2005).
Metalinguistic Awareness
Metalinguistic awareness is the ability to think about language. It is the explicit
awareness of linguistic forms and structures and an understanding of how these
relate to and produce meaning (Cazden, 1974). It is hypothesized that the experi-
ence of acquiring and maintaining two different languages—with different forms
and structures—allows bilingual speakers to develop an explicit and articulated
understanding of how language works. For example, bilingual speakers have two
different words for most concepts. Reflecting on this can point to the insight that
words are only arbitrarily and symbolically related to their underlying concepts
(e.g., knowing that dog and chien are concepts for dog makes it obvious that the
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
210
word dog is only an arbitrary symbol). Similarly, when syntactic rules differ across
languages, bilingual speakers of those languages may notice the differences and
become explicitly aware of the syntactic rules—which most monolingual speakers
will know only implicitly. For example, English–Japanese bilinguals may note that
objects follow verbs in English sentences (e.g., “I like chocolate”) but objects
precede verbs in Japanese sentences (e.g., “I chocolate like”). Noting this distinc-
tion provides insight into the specific grammatical rules in each language as well
as into the universal properties of human language.
Over the past decades, researchers have investigated the effects of bilingualism
on children’s metalinguistic development. The majority of studies have found that
bilingual speakers, particularly those highly proficient in both languages, demon-
strate greater metalinguistic awareness than their monolingual counterparts
(Bialystok, Majumder & Martin, 2003; Campbell & Sais, 1995; Galambos &
Hakuta, 1988).
Metacognitive Awareness
Metacognitive awareness refers to knowledge about one’s own cognitive pro-
cesses. It is an awareness of one’s own learning strategies and the mental activities
required to self-regulate the learning process (Flavell, 1978). The process of learn-
ing the vocabulary, syntax, phonology, and morphology of more than one lan-
guage, as well as learning how to use this body of knowledge in contextually
appropriate fashion, may provide bilingual speakers special insight into their own
cognitive processes and learning strategies (Kemp, 2007). Research comparing the
metacognitive awareness of bilinguals and monolinguals is scant but has generally
found that bilinguals show greater metacognitive awareness than monolinguals
(Ransdell, Barbier, & Niit, 2006; Vorstman, De Swart, Ceginskas, & Van Den
Bergh, 2009).
Abstract or Symbolic Reasoning and Creative and Divergent Thinking
Across a number of studies, bilinguals have shown enhanced skills with respect
to creative and divergent thinking and to abstract and symbolic reasoning. In an
investigation on creativity and bilingualism, Ricciardelli (1992) found that bilin-
guals outperformed monolinguals in 20 of the 24 studies reviewed, showing a clear
positive relationship between bilingualism and creativity or divergent thinking.
Peal and Lambert (1962) suggested that bilingual children develop greater cogni-
tive flexibility and creativity as a result of switching between two languages and
two different perspectives. As well, Cummins (1976) has proposed that bilingual-
ism spurs the development of abstract and symbolic reasoning through the experi-
ence of having two different words for most concepts. This helps bilingual children
understand that the relationship between words and their referents is entirely arbi-
trary and represents an abstract symbolic relationship.
Problem Solving
Bilinguals also show evidence of enhanced problem-solving skills, particu-
larly on tasks requiring executive control (i.e., planning, cognitive flexibility,
abstract thinking, rule acquisition, initiating appropriate actions and inhibiting
inappropriate actions, and selecting relevant sensory information; Baddeley,
1996). A bilingual advantage has been demonstrated using the Simon task, the
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
211
dimensional change card sort task, and other similar tasks used to assess exec-
utive control for problem-solving tasks (Bialystok, 1999, 2006). Simon tasks
refer to a family of tasks typically used to investigate interference effects.
In the Simon task, stimuli are presented with different target features and in
different positions. For example, participants may be asked to indicate the
color of either a red or a green square presented on one side of the screen by
pressing a left or a right key. The general finding in the Simon task is that reac-
tion times are slowed when the spatial location of the target and its response
coding do not correspond (incongruent condition) versus when spatial location
and response coding correspond (congruent condition). An incongruent trial
occurs when a signal is presented to the right but its color requires a left-hand
button press. Conversely, signals that require a left-hand response and are also
presented on the left side are referred to as congruent trials. Typically, reaction
times are slower to incongruent compared to congruent trials, a finding referred
to as the congruency effect or interference effect.
The enhanced problem-solving ability may be because of the cognitive flexibil-
ity associated with bilingualism. Because bilinguals have the capacity to choose
between two languages, they may develop more flexibility with respect to thinking
that can be applied to solve problems.
Purpose of the Study
Since Peal and Lambert’s (1962) seminal work, a number of studies have docu-
mented the positive cognitive correlates of bilingualism; however, the magnitude
of these effects remains unclear. The majority of recent studies have shown posi-
tive effects of bilingualism, but some studies have shown that bilingual perfor-
mance is relatively impaired on some cognitive tasks. A few studies have also
demonstrated mixed effects of bilingualism on performance on cognitive tasks.
The current analysis is an attempt to synthesize these results. The goal of the pres-
ent study is to meta-analyze research on the cognitive outcomes of bilingualism by
estimating the effects of bilingualism on specific cognitive measures. Specifically,
the meta-analysis addresses the following research questions:
1. What are the cognitive correlates associated with bilingualism?
2. How do the effects of bilingualism vary when cognitive outcomes are mea-
sured in different geographical locations, in different settings, and at differ-
ent educational levels?
3. How are effect sizes influenced by different combinations of languages spo-
ken by bilinguals?
4. Are the effect sizes influenced by methodological features of the research?
Method
Selection Criteria
To capture all relevant studies on the cognitive benefits of bilingualism, specific
criteria for inclusion were developed. Studies were deemed eligible if:
a. Bilingual participants were reported to be equally (or almost equally) profi-
cient in two languages. Thus, participants who were learning second
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
212
languages were not regarded as bilinguals but rather as second language
learners. Studies with such second language learners were excluded from
this meta-analysis. Bilinguals with learning disabilities or other cognitive
disabilities were excluded.
b. They had an experimental group of bilingual participants and a control group
of monolingual participants.
c. Measured outcomes (cognitive benefits) were clearly reported. These
include attentional control, problem-solving skills, creative and divergent
thinking, cognitive flexibility, learning strategies, symbolic representation
and abstract reasoning skills, metalinguistic awareness, metacognitive skills,
and working memory. We excluded studies that measured only psychosocial
outcomes such as employability or social problem solving and other activi-
ties such as code switching, cross-language priming, and social identity.
d. They reported sufficient data to allow for effect size calculations. When
basic descriptive statistics were not included in a study, other statistics were
used (e.g., t and F statistics), but we coded for reviewer’s confidence in
effect size derivation. Studies with insufficient data for effect size calcula-
tions were excluded.
e. They were publicly available, either online or in library archives.
For multiple reports of the same study (e.g., dissertation, conference proceed-
ings, and journal article), the version published as a journal article was coded, but
in some cases other versions of the published article (e.g., conference proceedings)
were used to make the coding features more extensive and accurate.
Location and Selection of Studies
A comprehensive and systematic search was conducted in the following elec-
tronic databases to locate appropriate studies: Academic Search Premier, Education
Full Text, ERIC (including British and Australian ERIC), Linguistic and Language
Behavior Abstracts, PsycINFO, and Web of Science. A primary search was con-
ducted utilizing Boolean combinations of the controlled vocabulary within each
database for the terms immigrant students, bilingualism, and cognition. A manual
search of the reference lists of earlier reviews of the literature on bilingualism (e.g.,
Bialystok, 2002; Cenoz, 2003) was subsequently conducted.
A total of 5,185 articles were obtained from the search procedure. Two researchers
reviewed the titles, abstracts, and keywords of these articles for possible inclusion by
applying the selection criteria stated above. When abstracts did not contain sufficient
information to determine inclusion or exclusion, the full text of the article was obtained
and read. Duplicate studies were removed, and articles that did not meet the selection
criteria were excluded. Interrater agreement was computed to determine the reliability
of including or excluding articles based on reading only the abstracts. This yielded a
Cohen’s kappa of .88. Researchers discussed all disagreements until they were fully
resolved. A total of 157 articles were retained for secondary screening.
Two researchers independently read the full texts of each of the 157 articles
retained after first inclusion to further determine their suitability based on the
specified criteria for inclusion. Only 39 articles met the second inclusion criterion,
and data from these articles were extracted using EPPI-Reviewer, an online appli-
cation for managing and conducting systematic reviews (Thomas & Brunton,
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
213
2006). Coded variables were organized into 11 major categories in the database. These
include (a) study identification, (b) study characteristics and measured outcomes,
(c) research questions, (d) research design, (e) groups and randomization strategy, (f)
sampling strategy, (g) characteristics of samples in the study, (h) recruitment and
consent, (i) data collection, (j) data analysis, and (k) results and conclusion. In
cases where some variables were not explicitly stated in the study, reviewers made
reasonable inferences and noted the absence of explicit information. The appendix
shows the coding book containing a description of all variables coded under each
category.
Although there are many variants of bilingualism (early bilinguals, late bilin-
guals, balanced bilinguals, etc.), sufficient information was not provided in many
of the studies to code this variable. Nevertheless, as highlighted earlier in the arti-
cle, all of the studies that met our inclusion criteria had participants who were
proficient in two languages before the start of each study.
Some articles reported multiple studies. Hence, 63 studies with an overall sam-
ple size of 6,022 participants were reported in all 39 articles and were included for
meta-analysis. Another interrater reliability analysis was conducted to determine
agreement among researchers on inclusion or exclusion judgments based on full-
text review of all 63 studies, yielding a Cohen’s kappa of .92. Again, researchers
discussed all disagreements until they were fully resolved.
Throughout the design and implementation of this review, guidelines for meta-
analysis provided by Lipsey and Wilson (2001) were followed. The coding scheme
prevented inappropriately combining statistically dependent comparisons in cal-
culating mean effect sizes. To generate a single distribution of effect sizes, a mean
effect size was obtained for each set of statistically dependent effect sizes by aver-
aging over different cognitive outcomes and study characteristics.
Extraction and Calculation of Effect Sizes
Effect size is a standardized metric obtained by calculating the difference
between the means of the experimental (bilingual) and control (monolingual)
groups divided by the pooled standard deviation of the two groups. Hedges (1981)
observed that estimates may yield inflated effect sizes when samples are small. To
correct for such bias in effect size estimation, especially with small sample sizes
(Lipsey & Wilson, 2001, p. 48), the obtained Cohen’s d values were converted to
Hedges’s g, an unbiased estimate (Hedges & Olkin, 1985, p. 81) of the standard-
ized mean difference effect size. When other statistics such as F or t were provided,
these were also used to derive effect sizes or, in some cases, to verify the obtained
d (Cooper & Hedges, 1994).
Data Analysis
Standard meta-analytic guidelines and equations were followed in all data analyses
(Cooper & Hedges, 1994; Hedges & Olkin, 1985; Lipsey & Wilson, 2001). All data
analyses were conducted using Comprehensive Meta-Analysis Version 2.2.048
(Borenstein, Hedges, Higgins, & Rothstein, 2008) and SPSS Version 16.0 for Windows.
Aggregating Effect Sizes
To aggregate effect sizes, the inverse variance weight was computed for each
finding. An aggregate effect size was then obtained from the weighted effect sizes
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
214
to derive an overall weighted mean estimate of the effect of the treatment.
This allowed more weight to be assigned to studies with larger sample sizes. The
standard error of Hedges’s unbiased estimate of the mean effect size was then
computed.
A 95% confidence interval was computed around each weighted mean effect
size to determine statistical significance. Confidence intervals spanning a range
above zero were interpreted as indicating a statistically detectable result favoring
bilinguals with respect to the associated cognitive outcome.
An important aspect of meta-analysis involves the determination of whether the
various effect sizes that are averaged into a mean value all estimate the same pop-
ulation effect size. This assumption of homogeneity of variance was tested by the
Q statistic. When all findings share the same population value, Q has an approxi-
mate chi-square distribution with k – 1 degrees of freedom, where k is the number
of effect sizes or studies for a particular subset. When Q exceeded the critical value
of the chi-square distribution (i.e., p < .05), the mean effect size was reported to be
significantly heterogeneous, meaning that there was more variability in the effect
sizes than would be expected from sampling error and suggesting that each effect
size did not estimate a common population mean (Lipsey & Wilson, 2001). The I2
statistic is reported as a complement to interpret the result of the homogeneity test
(Higgins & Thompson, 2002; Huedo-Medina, Sánchez-Meca, Marín-Martínez, &
Botella, 2006).1
Results
After resolving statistical independence, 63 studies (from 39 articles) were
analyzed.
Table 1 shows a summary of the variables coded for each of the 63 studies. This
includes the study, grade level of participants, total number of participants involved
in each study, languages spoken by the bilingual participants, cognitive benefits
measured, and unbiased effect size, Hedges’s g. In this and subsequent tables,
positive effect sizes show bilingual advantages whereas negative effect sizes show
an advantage for monolinguals over bilinguals on cognitive measures.
For outlier analysis, we used the Comprehensive Meta-Analysis software to
determine the effect of removing a number of effect sizes from the distribution of
effect. The forest plot of the 63 standardized mean difference effect sizes for the
cognitive benefits of bilingualism was examined, and 3 potential outlying studies
were removed. The recalculated results did not increase the fit of the remaining
effect sizes to a simple model of homogeneity (g = 0.39; Qtotal(59) = 287.61, p < .001;
I2 = 79.50%). Because the removal of potential outliers did not produce a homoge-
neous model, a decision was made not to remove any effect sizes from the original
distribution.
Overall Relationship Between Cognitive Outcomes and Bilingualism
Table 2 shows the overall weighted mean and homogeneity statistics of all sta-
tistically independent effect sizes. Table 2 and subsequent tables include the num-
ber of participants (N) in each category, the number of findings (k), the weighted
mean effect size (g) and its standard error (SE), the 95% confidence interval around
the mean, a test of the null hypothesis (z), a test of heterogeneity (Q) with its
215
TABLE 1
Summary of coded studies and associated effect sizes
Study
Grade
range N
Languages
(bilingual)aDominant cognitive benefits measured
Effect
size (g)
Abu-Rabia and Siegel (2002) 4–7 63 Arabic–English Working memory for normally developed 0.20
Abu-Rabia and Siegel (2002) 4–7 31 Arabic–English Working memory for reading disabled –0.23
Bialystok (1997) PreK 81 French–English ASR: Symbolic representation of printb1.04*
Bialystok (1997) PreK 87 Chinese–English ASR: Symbolic representation of print 1.32*
Bialystok (1997) PreK 81 French–English ASR: Symbolic representationc0.07
Bialystok (1997) PreK 87 Chinese–English ASR: Symbolic representation –0.26
Bialystok (2006) PSec 97 Mixed–English Working memory using Simon tasks 0.51*
Bialystok (1999) PreK 30 Chinese–English Attentional Controld0.82*
Bialystok (1999) PreK 30 Chinese–English Attentional Controle0.69
Bialystok, Craik, et al. (2005) PSec 29 French–English Attentional Control 0.63
Bialystok, Craik, and Ruocco (2006)f, Exp. 1 PSec 48 Mixed–English Working memoryg0.48
Bialystok, Craik, and Ruocco (2006), Exp. 2 PSec 48 Mixed–English Working memoryh–0.30
Bialystok et al. (2004), Exp. 1 PSec 20 Tamil–English Attentional control: MRT for younger adultsi2.25*
Bialystok et al. (2004), Exp. 1 PSec 20 Tamil–English Attentional control: MRT for older adults 1.03*
Bialystok et al. (2004), Exp. 2 PSec 64 Tamil–English Attentional control: MRT for younger 2.63*
Bialystok et al. (2004), Exp. 2 PSec 30 Mixed–English Attentional control: MRT for older adults 1.35*
Bialystok et al. (2004), Exp. 2 PSec 20 French–English Attentional control: MRT for older adults 2.45*
Bialystok, Luk, et al. (2005b) 1 132 Mixed–English Metalinguistic awarenessj0.95*
Bialystok and Majumder (1998) K–3 71 Mixed–English Problem solvingk0.35
Bialystok et al. (2003), Exp. 1 K–3 72 French–English Metalinguistic awareness –0.02
Bialystok et al. (2003), Exp. 2 K–3 75 French–English Metalinguistic awareness –0.45
Bialystok et al. (2003), Exp. 3 K–3 64 Chinese–English Metalinguistic awarenessl–0.92*
Bialystok et al. (2003), Exp. 3 K–3 58 Spanish–English Metalinguistic awareness 0.51
Bialystok and Martin (2004), Exp. 1 PreK 67 Cantonese–English Attentional control using the DCCS taskm0.35
Bialystok and Martin (2004), Exp. 2 PreK 30 French–English Attentional control using manual DCCS task 0.83*
Bialystok and Martin (2004), Exp. 3 PreK 53 Chinese–English Attentional control: color shape and color object 0.63*
Bialystok and Martin (2004), Exp. 3 PreK 53 Chinese–English Attentional control: functional location and place 0.42
Bialystok, McBride-Chang, et al. (2005), Exp. 1 K–1 67 Chinese–English Metalinguistic awareness –0.04
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
216
Study
Grade
range N
Languages
(bilingual)aDominant cognitive benefits measured
Effect
size (g)
Bialystok, McBride-Chang, et al. (2005), Exp. 2 K–1 67 Chinese–English Metalinguistic awareness –0.30
Bialystok and Senman (2004), Exp. 2 PreK 95 Mixed–English Attentional controln1.59*
Bialystok and Shapero (2005), Exp. 1 K–1 48 Mixed–English ASR: Reversibility of ambiguous figureso1.14*
Bialystok and Shapero (2005), Exp. 2 K 53 Mixed–English Attentional control using different tasksp0.18
Bruck and Genesee (1995) K 137 French–English Metalinguistic awareness –0.44*
Campbell and Sais (1995) PreK 30 Italian–English Metalinguistic awarenessq0.64
Chen et al. (2004), Exp. 1 2 125 Canton-Mandarin Metalinguistic awareness: tone, onset and rime tasks 0.20
Chen et al. (2004), Exp. 1 4 150 Canton-Mandarin Metalinguistic awareness: tone, onset and rime tasks –0.30
Chen et al. (2004), Exp. 2 1 94 Canton-Mandarin Metalinguistic awareness tasksr0.36
Chen et al. (2004), Exp. 2 2 89 Canton-Mandarin Metalinguistic awareness: tone, onset and rime tasks 0.49*
Chen et al. (2004), Exp. 2 4 106 Canton-Mandarin Metalinguistic awareness: tone, onset and rime tasks 0.06
Clarkson and Galbraith (1992) 6 96 Pidgin–English Problem solving with mathematical tests 0.40
Cromdal (1999) K 54 Swedish–English Metalinguistic awareness taskss0.71*
Davidson and Tell (2005), Exp. 1 PreK 40 Urdu–English ASR: Mutual exclusivity in naming whole objectst0.18
Davidson and Tell (2005), Exp. 1 K–1 40 Urdu–English Mutual exclusivity in naming whole objects 1.22*
Davidson and Tell (2005), Exp. 2 K–1 20 Urdu–English Mutual exclusivity in naming whole objects 0.34
Demie and Strand (2006) 10 1026 Mixed–English Problem solving with a standardized test 0.23*
Demont (2001) 1 43 German-French Metalinguistic awareness: several tasksu0.56
Eviatar and Ibrahim (2000) K–1 76 Russian-Hebrew Metalinguistic awareness using several tasksv0.96*
Galambos and Goldin-Meadow (1990) PreK–1 64 Spanish–English Metalinguistic awareness—correcting errors 0.39
Galambos and Goldin-Meadow (1990) PreK–1 64 Spanish–English Metalinguistic awareness—explaining errors 0.09
Gutierrez-Clellen et al. (2004) 3 44 Spanish–English Working memory using two tasksw–0.09
Humphreys and Mumtaz (2001) 2–3 120 Urdu–English Working memory 1.52*
Jordà-Maria (2003) PSec 160 Catalan-Castilian Metalinguistic awareness: Metapragmatic 0.66*
Keshavarz and Astaneh (2004) PSec 60 Mixed–PersianxMetalinguistic awareness 0.56*
Loizou and Stuart (2003) 1 32 Greek–English Metalinguistic awareness using several tasksy0.96*
Loizou and Stuart (2003) 1 36 English-Greek Metalinguistic awareness using several tasks –0.34
Love et al. (2003) PSec 71 Mixed–English Metalinguistic awareness: syntactic processing 0.20
Merriman and Kutlesic (1993) K–2 78 Serbian–English Metalinguistic awareness 0.48*
Oller et al. (2007) 2 288 Spanish–English Metalinguistic awarenessz1.07*
TABLE 1 (continued)
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
217
Study
Grade
range N
Languages
(bilingual)aDominant cognitive benefits measured
Effect
size (g)
Oller et al. (2007) 4 296 Spanish–English Metalinguistic awareness with several measures 0.49*
Ransdell et al. (2006) PSec 137 Mixed–English Metacognitive awarenessaa 0.16
Reichard and Mokhtari (2004) PSec 350 Mixed–English Metacognitive awareness and strategies 0.39*
Sanz (2000) 8–12 201 Catalan-Spanish Metalinguistic awareness with several measures 0.68*
Sheng et al. (2006) K–3 24 Mandarin–English Metalinguistic awareness with several measuresbb 0.22
Note. PreK = prekindergarten; PSec = postsecondary; N = number of participants; ASR = abstract and symbolic representation; MRT = mean reaction time.
a. The language spoken by the monolingual group is presented last in the bilingual coupling (e.g., Arabic–English bilingual means that English was the language of the monolingual group).
b. Moving word problem was used to determine abstract and symbolic representation (ASR) for literacy.
c. Consistent and inconsistent word size problems were used to assess participants’ understanding of the principles of representing spoken language in printed text.
d. Postswitch items in the dimensional change card sort (DCCS) task were used for this sample.
e. Knowledge items in the DCCS task were used for this sample.
f. Results from 24 monolingual and 24 balanced full bilingual participants were used. We discarded results from the 24 unbalanced bilinguals because they reported minimal use of the second language. For example,
“a participant reporting moderate knowledge of Italian but the opportunity to use it only several times per month when visiting relatives was considered an unbalanced bilingual” (Bialystok et al., 2006, p. 1971).
Experiment 1 was conducted with younger participants and Experiment 2 with older participants.
g. A dual-task classification paradigm was used in both experiments.
h. Although the overall result here shows a bilingual disadvantage for adults, the breakdown of outcome measures actually shows bilingual superiority in letters or numbers (LN) conditions, but overall that advantage
was subdued by a much larger bilingual disadvantage in the animals or musical instruments (A-M) task.
i. Mean reaction times (MRT) for both congruent and incongruent trials on Simon task. Note that MRT was reversed to reflect the inverse relationship between RT and benefit. MRT was categorized as a cognitive
benefit closely associated with attentional control.
j. Chinese–English, Hebrew–English, and Spanish–English bilinguals were compared to English monolinguals on phoneme counting measure.
k. French–English and Bengali–English bilinguals were compared to English monolinguals on block design and water-level tasks to examine whether balanced bilinguals carry over their linguistic advantage into
nonlinguistic tasks.
l. This includes the phoneme segmentation task with first and second graders.
m. This was measured by a computerized DCCS task measuring inhibition of attention—categorized as a cognitive benefit closely associated with attentional control.
n. Appearance–reality tasks were used.
o. Outcome measures derived by using children’s embedded figures task and ambiguous figures task, categorized as abstract and symbolic representation.
p. Outcome measures for inhibition of attention include opposite worlds task and computerized DCCS.
q. MLA (metalinguistic awareness) outcomes include letter detection task.
r. Tone, onset, and rime awareness tasks were each assessed with two pholonogical tasks: the oddity task and the same–different task.
s. Outcome measures include symbol substitution, sentence judgment, and correction tasks.
t. Familiar and unfamiliar objects with or without spare parts were used as outcome measures in all experiments. These measures were categorized as abstract and symbolic representation.
u. This article was written in French. Outcomes used to measure linguistic awareness include the grammatical judgment and correction tasks.
v. Outcome measures include the arbitrariness test, initial and final phoneme tests, and phoneme–syllable deletion test.
w. Outcome measures are the competing language processing task and dual processing comprehension task.
x. Turkish–Persian and Armenian–Persian bilinguals were compared to Persian monolinguals using controlled productive ability test at 2,000- and 3,000-word levels.
y. Measures include the rhyme oddity, syllable completion, cluster onset oddity, single phoneme onset oddity, phoneme elision, and initial phoneme identification tasks.
z. Several metalinguistic awareness measures used including passage comprehension, proofreading, verbal analogies, and oral vocabulary.
aa. Metacognitive awareness was measured by self-ratings of reading, writing, speaking, and listening skills.
bb. Paradigmatic and syntagmatic Trials 1, 2, and 3.
*p < .05.
TABLE 1 (continued)
218
associated degrees of freedom (df), and the percentage of variability that is attribut-
able to true heterogeneity, that is, over and above the sampling error (I2).
Table 2 shows that the overall weighted mean effect size is moderate (g = 0.41)
but with substantial variability among studies (Qtotal = 362.62, p < .001).
Heterogeneity among the full set of studies was anticipated as different studies
measured different cognitive outcomes and there was no reason to expect similar
effect sizes for different outcomes. Following up on this heterogeneity, separate
analyses were conducted for each category of cognitive outcomes.
It was observed that 30 of the 63 independent effect sizes were obtained from
studies conducted by Bialystok and colleagues. A sensitivity analysis was per-
formed so as to investigate the potential bias of including about 47% of the entire
studies conducted by a single principal investigator. Table 2 shows statistically
detectable mean effect sizes irrespective of whether studies were authored by
Bialystok or not. There was an overlap in the confidence intervals across the two
categories. Hence, a decision was made to combine the data from both categories
in subsequent analyses.
Cognitive Correlates of Bilingualism
Table 3 shows the weighted mean effect sizes for different cognitive outcomes
associated with bilingualism. All the outcome measures produced statistically detect-
able mean effect sizes in favor of bilingualism. Attentional control produced the largest
effect with a weighted mean effect size of .96 across 14 studies. Although all the cogni-
tive outcomes in Table 3 had statistically detectable mean effect sizes, most of the effect
size distributions were highly heterogeneous, indicating that the variability among
effect sizes was greater than that expected from sampling error.
Moderator analyses were conducted to investigate this heterogeneity, but given
the small number of studies within some of the categories of cognitive outcomes,
a decision was made to collapse across categories with similar outcomes. We col-
lapsed studies that investigated metalinguistic and metacognitive awareness into
one group, and abstract and symbolic representation, attentional control, and prob-
lem solving were collapsed into another group. We did not include working mem-
ory studies with either of the recategorized groups because the dependent variables
used to measure working memory were markedly different from those used in
studies subsumed under the other two groups. Given the small number of studies
TABLE 2
Overall weighted mean effect size
Effect size
95% confidence
interval
Test
of null Test of heterogeneity
N k g SE Lower Upper z Q df p I2 (%)
All 6,022 63 0.41 0.03 0.36 0.46 15.00* 362.62 62 .00 82.90
Not conducted
by Bialystok
4,245 33 0.38 0.03 0.31 0.44 11.54* 141.26 32 .00 77.35
Conducted by
Bialystok
1,777 30 0.48 0.05 0.39 0.58 9.74* 218.17 29 .00 86.71
*p < .05.
219
TABLE 3
Weighted mean effect sizes for bilingual studies by outcome constructs
Effect size
95% confidence
interval
Test
of null Test of heterogeneity
N k g SE Lower Upper z Q df p I 2 (%)
Outcome constructs (cognitive
outcomes)
Metalinguistic awareness 2,813 29 0.33 0.04 0.26 0.41 8.44* 151.56 28 .00 81.53
Metacognitive awareness 487 2 0.32 0.09 0.14 0.50 3.47* 1.26 1 .26 20.80
Working memory 451 7 0.48 0.10 0.29 0.67 4.90* 41.36 6 .00 85.49
Abstract and symbolic representation 484 8 0.57 0.10 0.39 0.76 6.04* 42.10 7 .00 83.37
Attentional control 594 14 0.96 0.09 0.79 1.13 10.86* 73.30 13 .00 82.26
Problem solving 1,193 3 0.26 0.07 0.13 0.38 3.96* 0.65 2 .72 0.00
Collapsed outcome constructs
Metalinguistic and metacognitive
awareness
3,300 31 0.33 0.04 0.26 0.41 9.14* 152.61 30 .00 80.34
Abstract and symbolic representation,
attentional control, and problem
solving
2,271 25 0.52 0.05 0.43 0.61 11.33* 157.69 24 .000 84.78
*p < .05.
Adesope et al.
220
in the working memory category, moderator analyses were not conducted on this
category. However, we note that bilingualism was associated with greater working
memory, resulting in a moderate effect size of .48. Henceforth, the results section
deals with the two new categories (metalinguistic or metacognitive awareness and
attention and representation).
Table 3 shows weighted mean effect sizes for the recategorized outcome con-
structs. The new categories produced statistically detectable mean effect sizes, and
significant heterogeneity was observed in both categories. In subsequent analyses,
we examined the different moderators that may help explain the variability within
each of these two categories.
Moderator Variable Analyses
A mixed-effects model was used for all moderator variable analyses. A mixed-
effects model uses a random-effects model to combine studies within subgroups and a
fixed-effect model to combine studies across subgroups to yield an overall effect
(Borenstein, Hedges, Higgins, & Rothstein, 2009). By using the random-effects model
to combine studies within subgroups in moderator analyses, a mixed-effects model
typically allows for population parameters to vary across studies, reducing the proba-
bility of committing a Type I error, and is usually regarded as a more rigorous meta-
analytical model than a fixed-effects model only (Borenstein et al., 2009; Denson,
2009; Hedges & Vevea, 1998; National Research Council, 1992).
Relationship Between Bilingualism and
Metalinguistic or Metacognitive Awareness Across
Different Locations, Educational Levels, Settings, Language Groups,
SES, and Region
Table 4 shows weighted mean effect sizes for metalinguistic or metacognitive
awareness outcomes under various conditions. When disaggregated by the geographi-
cal location of the research, the effect of bilingualism was statistically detectable across
studies conducted in the United States, Europe, and the Middle East.
Because the total between-levels variance was statistically detectable, QB(5) =
14.47, p = .01, further analyses showed that studies conducted in Europe were sig-
nificantly different from those conducted in other geographical locations. Although
studies conducted in the United States and the Middle East produced a statistically
detectable effect, they were not significantly different from those conducted in China
or Canada and/or those that are part of the various or mixed category.2
In Table 4, a majority of the studies were conducted with primary school stu-
dents in kindergarten through third grade. A mean effect size was statistically
detectable only for these early primary-level students, possibly because of the
small number of studies with students at other levels.
Classroom studies in which learning activities contributed toward performance
assessment in a program produced statistically detectable effect sizes along with
studies that did not specify the setting. The classroom studies were significantly
different from studies conducted in the laboratory. However, the certainty of this
interpretation is limited by the high number of studies that did not specify the set-
ting under which those studies were undertaken.
Bilingualism was reliably associated with greater metalinguistic or metacogni-
tive awareness when bilinguals spoke Spanish and English or a mix of language
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
221
TABLE 4
Weighted mean effect sizes for studies investigating bilinguals’ metalinguistic and metacognitive awareness under various conditions
Effect size
95% confidence
interval
Test of
null
Test of
heterogeneity
N k g SE Lower Upper z QBdf p
Geographical locations 14.47 5 .01
Canada 672 8 –0.08 0.15 –0.37 0.20 –0.57
United States 757 5 0.55 0.18 0.19 0.90 2.99*
Europe 556 7 0.57 0.17 0.24 0.89 3.41*
Middle East 136 2 0.77 0.30 0.18 1.35 2.56*
China 564 5 0.15 0.18 –0.19 0.49 0.87
Mixed 615 4 0.26 0.20 –0.12 0.65 1.34
Educational level 1.99 4 .74
Preschool 30 1 0.64
Primary (K–3) 1,739 21 0.26 0.11 0.04 0.48 2.33*
Intermediate (4–7) 552 3 0.09 0.27 –0.45 0.63 0.33
Secondary (8–12) 261 2 0.63 0.35 –0.05 1.31 1.80
Postsecondary 718 4 0.36 0.24 –0.11 0.83 1.50
Setting 10.54 2 .01
Classroom 1,232 6 0.65 0.16 0.34 0.96 4.11*
Laboratory 1,359 16 0.07 0.10 –0.13 0.27 0.72
Not specified 709 9 0.42 0.14 0.15 0.70 2.99*
Language of the monolingual group 2.99 3 .56
English 2,042 18 0.22 0.11 0.01 0.44 1.97*
Romance (Castilian, French, and Spanish) 468 4 0.58 0.23 0.13 1.04 2.51*
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
222
TABLE 4 (continued)
Effect size
95% confidence
interval
Test of
null
Test of
heterogeneity
N k g SE Lower Upper z QBdf p
Chinese 564 5 0.15 0.20 –0.24 0.55 0.76
Othera226 4 0.50 0.25 0.02 0.99 2.03*
Languages of the bilingual group 16.52 3 .00
French–English 284 3 –0.32 0.22 –0.74 0.11 –1.44
Spanish–English 770 5 0.54 0.17 0.21 0.87 3.24*
Chinese–English 918 10 0.08 0.12 –0.16 0.32 0.69
Mixedb1,328 13 0.51 0.11 0.29 0.72 4.57*
Socioeconomic status 5.99 3 .11
Medium 485 4 –0.04 0.22 –0.47 0.39 –0.19
High 313 4 0.28 0.23 –0.17 0.74 1.23
Mixed 681 4 0.72 0.22 0.28 1.16 3.21*
Not reported 1,821 19 0.27 0.11 0.07 0.48 2.60*
Region 3.05 3 .39
Urban or inner city 552 8 0.28 0.18 –0.07 0.63 1.59
Suburban 225 3 0.00 0.27 –0.54 0.54 0.01
Rural 76 1 0.96
Not reported 2,447 19 0.30 0.11 0.09 0.51 2.83*
Note. QB = QBetween.
a. The “other” subset refers to studies in which participants spoke languages other than English, Spanish, French, and Mandarin or Chinese. Specifically, monolingual control
participants in the “other” subset spoke Swedish (Cromdal, 1999), Hebrew (Eviatar & Ibrahim, 2000), Persian (Keshavarz & Astaneh, 2004), or Greek (Loizou & Stuart, 2003).
b. The “mixed” subset refers to studies in which participants spoke various language couplings apart from French–English, Spanish–English, and Chinese–English.
*p < .05.
Cognitive Benefits of Bilingualism
223
not including French–English or Chinese–English. Similarly, bilinguals showed a
significant metalinguistic or metacognitive advantage over monolinguals for most
monolingual groups except Chinese-speaking monolinguals.
Many of the studies did not report the SES of participants involved in the studies in
Table 4. Statistically detectable mean effect sizes were found among participants
with mixed SES and those studies that did not report the SES of participants.
However, mean effect sizes did not differ statistically among the four SES catego-
ries (i.e., middle, high, mixed, and not reported).
Relationship Between Bilingualism and Attention and
Representation Across Different Locations, Educational Levels,
Settings, Language Groups, SES, and Region
Table 5 shows weighted mean effect sizes for studies investigating bilinguals’
abstract and symbolic representation, attentional control, and problem-solving
outcomes under various conditions. Studies conducted in Canada and various geo-
graphical locations (mixed) produced statistically detectable mean effect sizes.
The mean effect size for studies conducted in different locations was much higher
than those conducted specifically in the United States, Canada, and Europe.
Studies conducted with preschool, primary, and postsecondary students pro-
duced statistically detectable mean effect sizes, with bilingual postsecondary stu-
dents showing the largest mean effect size (g = 1.76). Post hoc comparisons were
conducted because the total between-levels variance for educational level was sig-
nificant, QB(4) = 12.92, p = .01. Results showed that studies conducted with par-
ticipants from preschool up to 12th grade were not significantly different from one
another but that they were all significantly different and produced lower mean
effect sizes than studies conducted with postsecondary students.
There were no classroom studies in which participants engaged in learning
activities that contributed toward performance assessment. Most of the studies did
not report the setting in which they were conducted. Nevertheless, studies con-
ducted in the laboratory and those that did not report the setting both produced
statistically detectable effect sizes.
Except Urdu–English bilinguals (who were included in only three studies), all
categories of language pairs spoken by bilinguals showed statistically detectable
mean effect sizes, indicating that bilingualism is reliably associated with better atten-
tion and representation skills. Because the total between-levels variance was statisti-
cally detectable, QB(4) = 11.95, p = .02, further analyses showed that studies conducted
with bilinguals who spoke Tamil–English were significantly different from those
conducted with Chinese–English bilinguals and those in the “mixed” category.
In Table 5, statistically detectable mean effect sizes were found among partici-
pants with medium SES and those studies that did not report the SES of partici-
pants. Mean effect sizes did not differ statistically among the four SES categories
(i.e., low, middle, mixed, and not reported). These findings indicate that bilingual-
ism contributes to cognitive benefits irrespective of the SES of participants. Across
all regional categories, bilinguals showed more positive and statistically detectable
mean effect sizes than monolinguals.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
224
TABLE 5
Weighted mean effect sizes for studies investigating bilinguals’ abstract and symbolic representation, attentional control, and
problem solving under various conditions
Effect size
95% confidence
interval
Test of
null Test of heterogeneity
N k g SE Lower Upper z QBdf p
Geographical locations 12.06 4 .02
Canada 915 16 0.72 0.16 0.40 1.03 4.52*
United States 100 3 0.58 0.38 –0.17 1.33 1.52
Europe 1,026 1 0.23
Papua New Guinea 96 1 0.40
Mixed 134 4 1.92 0.36 1.23 2.62 5.42*
Educational level 12.92 4 .01
Preschool 734 12 0.63 0.18 0.28 0.99 3.47*
Primary (K–3) 232 5 0.64 0.29 0.07 1.21 2.20*
Intermediate (4–7) 96 1 0.40
Secondary (8–12) 1,026 1 0.23
Postsecondary 183 6 1.76 0.30 1.18 2.34 5.93*
Setting 0.30 1 .59
Laboratory 421 8 0.72 0.24 0.25 1.19 2.98*
Not specified 1,850 17 0.88 0.17 0.55 1.20 5.28*
Language of the monolingual group 0.44 1 .51
English 2,175 24 0.85 0.14 0.57 1.12 6.01*
Pidgin English 96 1 0.40
Languages of the bilingual group 11.95 4 .02
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
225
TABLE 5 (continued)
Effect size
95% confidence
interval
Test of
null Test of heterogeneity
N k g SE Lower Upper z QBdf p
French–English 342 7 0.80 0.24 0.33 1.28 3.30*
Chinese–English 407 7 0.55 0.23 0.09 1.01 2.36*
Tamil–English 134 4 1.92 0.35 1.23 2.62 5.47*
Urdu–English 100 3 0.58 0.38 –0.16 1.33 1.53
Mixeda1,288 4 0.63 0.29 0.05 1.20 2.14*
Socioeconomic status 4.40 3 .22
Low 1,026 1 0.23
Medium 716 13 1.12 0.20 0.73 1.52 5.52*
Mixed 100 3 0.58 0.43 –0.26 1.42 1.36
Not reported 429 8 0.56 0.25 0.07 1.06 2.23*
Region 4.99 2 .08
Urban or inner city 356 8 0.57 0.27 0.05 1.09 2.16*
Suburban 472 7 0.57 0.27 0.04 1.10 2.09*
Not reported 1,443 10 1.25 0.24 0.78 1.73 5.23*
a. The “mixed” subset refers to studies in which bilinguals spoke languages other than French–English, Chinese–English, Tamil–English, and Urdu–English. Specifically, bilin-
gual participants spoke Bengali/French–English (Bialystok & Majumder, 1998); Armenia, Hebrew, Russian, Spanish, and a host of other languages (Bialystok & Senman, 2004,
p. 573); Pidgin–English (Clarkson & Galbraith, 1992); or various languages (Demie & Strand, 2006).
*p < .05.
226
Effects of Different Methodological Features
Tables 6 and 7 show how effect sizes varied with research design and imple-
mentation features. The studies were categorized according to the researchers’
confidence in the calculated effect size (rated as medium or high depending on
whether sufficient data were provided to calculate an effect size), the reliability and
validity of instruments used in the studies, and a measure of the level of control for
bias in the studies. An additional source of variance often reported in meta-analyses is
the research designs of the studies under consideration (Abrami & Bernard, 2006).
However, because none of the studies in the current meta-analysis used random
assignment—almost all the studies included were matched quasi-experimental
designs except two studies that did not clearly report the design—we could not
meta-analyze the variations based on research designs. As well, all the analyzed
studies were published, hence precluding further analyses based on whether stud-
ies were published or not but heightening the potential for publication bias.
Table 6 shows the weighted mean effect sizes for metalinguistic or metacogni-
tive awareness studies by different methodological features. High coder confi-
dence in the effect size calculation was associated with statistically detectable
mean effect size, but medium coder confidence was not. Studies in which the reli-
ability of measures was not reported produced a statistically detectable mean effect
size, but studies in which reliability was reported did not. Studies produced a sta-
tistically detectable mean effect size whether they reported validity measures or
not. Studies in which bias was tightly controlled or not controlled produced a sta-
tistically detectable mean effect size, but studies in which biases were loosely
controlled did not.
Table 7 shows the weighted mean effect sizes for representation and attention
studies by different methodological features. High coder confidence in the effect
size calculation was associated with a statistically detectable mean effect size, but
medium coder confidence was not. Mean effect sizes were statistically detectable
whether reliability of the measures used were reported or not. Similarly, mean
effect sizes were statistically significant whether bias was tightly or loosely con-
trolled.
Publication Bias
Researchers have observed that published studies are a biased sample of studies
in a particular domain because research reports are more likely to be published
when significant results are reported (Orwin, 1983; Rosenthal, 1979). Hence, stud-
ies with nonsignificant findings are often either tucked away in file drawers or
reported in the less accessible gray literature. This problem, referred to as the “file-
drawer problem,” becomes apparent in meta-analyses, where there is a tendency to
exclude unpublished and gray literature studies, thereby potentially skewing meta-
analytical findings toward a positive mean effect size. This poses a threat to the
validity of results obtained from any meta-analyses. Researchers have proposed
different methods to examine the validity of results obtained from meta-analyses.
In the current work, publication bias seems to be a potential threat to the validity
of results obtained because all the studies analyzed were published in peer-
reviewed journals. Two statistical tests were performed using Comprehensive
Meta-Analysis software to assess the potential for publication bias in this meta-
analysis. First, a “classic fail-safe N” test was performed to determine the number
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
227
TABLE 6
Weighted mean effect sizes for studies investigating bilinguals’ metalinguistic and metacognitive awareness by different
methodological features
Effect size
95% confidence
interval
Test of
null
Test of
heterogeneity
N k g SE Lower Upper z QBdf p
Confidence in effect size derivation 0.84 1 .36
Medium 238 2 0.58 0.33 –0.06 1.21 1.77
High 3,062 29 0.27 0.09 0.09 0.44 3.02*
Reliability 0.44 1 .51
Not reported 2,028 19 0.33 0.11 0.12 0.55 3.06*
Reported 1,272 12 0.22 0.14 –0.05 0.49 1.57
Validity 0.54 1 .46
Not reported 2,644 26 0.26 0.09 0.08 0.45 2.76*
Reported 656 5 0.44 0.22 0.01 0.86 2.00*
Control for bias in studies 3.11 2 .21
Tightly controlled 1,556 17 0.34 0.12 0.12 0.57 2.99*
Loosely controlled 1,023 11 0.12 0.14 –0.15 0.39 0.86
Not controlled 721 3 0.58 0.25 0.10 1.06 2.35*
*p < .05.
228
of null effect studies needed to raise the p value associated with the average effect
above an arbitrary alpha level (set at a = .05). This test revealed that 3,791 addi-
tional studies would be required to invalidate the overall effect found in this meta-
analysis.
The second test, Orwin’s fail-safe N, was used to estimate the number of
file-drawer studies with null results required to nullify the effects found in this
meta-analysis. Using a criterion trivial level of .05, the fail-safe N was found to be
453 studies, meaning that 453 missing null studies would be needed to bring the
current mean effect size found in this meta-analysis to .05. Researchers have
claimed that meta-analytical results could be interpreted as valid and thus resistant
to the “file-drawer problem” if the fail-safe N reaches the 5k+10 limit (Carson,
Schriesheim, & Kinicki, 1990; Rosenthal, 1979). The results of the two computed
statistical tests suggest that it is unlikely that publication bias poses a significant
threat to the validity of findings reported in the current work because both fail-safe
N values are larger than the 5k+10 limit.
Discussion
The meta-analysis presented here synthesized data from 63 studies to examine
the cognitive correlates of bilingualism and the magnitude of such benefits. We
found a moderate positive overall effect of bilingualism on different cognitive
measures. Nevertheless, significant variability existed among studies, with some
yielding a positive cognitive effect of bilingualism and others a negative cognitive
effect. To more appropriately explain the variability among findings, we examined
the study features that may account for the variable effects. Specifically, this meta-
analysis provided answers to the following research questions.
TABLE 7
Weighted mean effect sizes for studies investigating bilinguals’ abstract and
symbolic representation, attentional control, and problem solving by different
methodological features
Effect
size
95% confidence
interval
Test of
null
Test of
heterogeneity
N k g SE Lower Upper z QBdf p
Confidence in effect
size derivation
0.65 1 .42
Medium 1,180 4 0.59 0.34 –0.07 1.24 1.75
High 1,091 21 0.88 0.15 0.58 1.18 5.69*
Reliability 1.43 1 .23
Not reported 1,779 18 0.94 0.17 0.61 1.27 5.58*
Reported 492 7 0.57 0.26 0.07 1.08 2.21*
Control for bias in
studies
3.11 2 .21
Tightly controlled 1,862 18 0.89 0.16 0.57 1.21 5.43*
Loosely controlled 409 7 0.67 0.26 0.17 1.17 2.61*
*p < .05.
229
What Cognitive Correlates Is Bilingualism Associated With?
Results of this meta-analysis show that bilingualism is positively associated
with a range of cognitive benefits. Specifically, bilinguals were found to outper-
form monolinguals on the combined measures of metalinguistic and metacognitive
awareness (g = 0.33) and on measures of abstract and symbolic representation,
attentional control, and problem solving (g = 0.52). There was, however, signifi-
cant variability in these effect sizes.
These results indicate that the process of acquiring two languages and of simul-
taneously managing those languages—of inhibiting one so the second can be used
without interference—allows bilinguals to develop skills that extend into other
domains. These skills appear to give bilingual speakers insight into the abstract
features of language and into their own learning processes. They also appear to
give bilingual speakers an enhanced capacity to appropriately control and distrib-
ute their attentional resources, to develop abstract and symbolic representations,
and to solve problems.
How Do the Effects of Bilingualism Vary When
Cognitive Outcomes Are Measured in Different
Geographical Locations and Settings, for Different Language Groups,
at Different SESs, and at Different Educational Levels?
Sociocultural attitudes toward bilingualism and the use of particular languages
vary across different countries and communities (Hamers & Blanc, 2000).
Consistent effects associated with geography were not observed in the current
work. For example, Canada and the United States provide rather different contexts
for bilingualism: Canada maintains two official languages, and a number of poli-
cies supporting bilingualism in both official languages are in place (Government
of Canada, 2009). As well, Canada’s multicultural attitude toward immigration
encourages immigrants to maintain their native language while acquiring at least
one of Canada’s official languages (Esses & Gardner, 1996). In contrast, the United
States is officially unilingual and has adopted a “melting pot” rather than multicul-
tural approach to immigration (Ravitch, 1990). Despite these differences, we
observed no consistent differences regarding cognitive correlates of bilingualism
in Canada and the United States. It may be that limiting our inclusion criteria to
cover only studies in which participants were balanced bilinguals eliminated any
potential differences across bilingual speakers in different countries.
Bilinguals who acquire their second language at an early age often master that
second language more fully than those who acquire their second language later in
life (Johnson & Newport, 1989). The evidence reviewed in the current analysis
suggests that earlier, rather than later, acquisition of a second language is also more
likely to be associated with greater metalinguistic and metacognitive awareness.
Although bilingual speakers of all ages demonstrated significant advantages with
respect to representation and attention, only the youngest bilinguals (who, by def-
inition, must have acquired their second language early in life) showed significant
advantages with respect to metalinguistic or metacognitive awareness.
Studies conducted in the classroom yielded a moderately high effect of bilin-
gualism over monolingualism on measures of metalinguistic and metacognitive
awareness, but this effect was not observed in laboratory studies conducted outside
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
230
of the classroom. Because of the high stakes associated with classroom studies in
which testing contributes to students’ performance assessments, it is possible that
bilinguals in such studies exhibited their cognitive skills to the fullest relative to
bilinguals in the laboratory studies in which outcome measures typically do not
contribute to performance assessments.
How Are Effect Sizes Influenced by Different Combinations
of Languages Spoken by Bilinguals?
For most language pairings—including French–English and Chinese–English
pairings—bilinguals outperformed monolinguals on the set of representation and
attention measures that includes abstract and symbolic representations, attentional
control, and problem solving. However, when measuring metalinguistic or meta-
cognitive outcomes, French–English and Chinese–English pairings did not show
a significant advantage over monolinguals. This may be a result of the specific
manners in which metalinguistic or metacognitive outcomes were measured in the
French and Chinese speaker studies. Alternatively, this pattern of results may indi-
cate that any language pairing can yield general cognitive benefits for bilinguals
but specific pairs of languages are necessary for bilinguals to develop metalinguis-
tic or metacognitive advantages. For example, it may be necessary for both lan-
guages to contain certain features for bilinguals to become metalinguistically
aware of those features. This would be in line with the findings from research on
cross-linguistic transfer of reading skills, where it has been observed that the
amount of cross-linguistic transfer is maximized when both languages have alpha-
betic writing systems. As a result, cross-linguistic transfer occurs easily for
Spanish–English bilinguals but not for Mandarin–English bilinguals (Bialystok,
Luk, & Kwan, 2005; Koda & Zehler, 2008).
Are the Effect Sizes of Cognitive Benefits Influenced by
Methodological Features of the Research?
Across different methodological features, the meta-analysis found cognitive advan-
tages of bilingualism. Specifically, on measures of metalinguistic or metacognitive
awareness, statistically detectable benefits of bilingualism were obtained when
researchers’ confidence in calculating effect sizes was high and when biases in studies
were tightly controlled. However, the large number of studies that did not report any
reliability or validity measures somewhat undermines the methodological quality of
the studies we analyzed. On measures of representation and attention, a similar result
was found with researchers’ confidence in calculating effect sizes. A large, statistically
detectable effect size was obtained showing that a high coder confidence in the effect
size was associated with a high mean effect size. Also, statistically detectable effect
sizes were obtained whether reliability of instruments used was reported or not and
irrespective of how biases in studies were controlled.
Conclusion
Although monolingualism is often depicted as normative, the best available
evidence indicates that, around the world, bilingual and multilingual speakers out-
number monolingual speakers (Tucker, 1998). The current work suggests that
bilingualism (and, presumably, multilingualism) is associated with a number of
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
231
cognitive benefits. These findings point to the need for further work investigating
the utility of these benefits in a variety of contexts. For example, cognitive benefits
documented in the current work may be of use to bilingual speakers in classrooms
where the language of instruction is not their native language. As the pace of
immigration to developed countries increases, the incidence of bilingualism and
multilingualism in these countries will also increase—as will the number of sec-
ond language learners in public school classrooms. Although second language
learners often present challenges within the classroom, the current analysis sug-
gests they may also bring a number of advantages. It remains unclear how, in
practice, second language learners and their instructors may capitalize on these
advantages. Further work investigating the cognitive correlates of bilingualism
within educational contexts is required to clarify this issue.
APPENDIX
Codebook on the cognitive benefits of bilingualism: Coded variables
Section A: Identification of studies/
reviewer
A.1 Name of the reviewer A.1.1 Details
A.2 Date of the review A.2.1 Details
A.3 Please enter the details of the paper A.3.1 Author (last name, first name)
A.3.2 Date of publication
A.3.3 Title of the article
A.3.4 Journal, issue number, pages
A.3.5 Credentials—institutional affiliation of
the authors
Section B: General information
B.1 What is the focus of the study? B.1.1 ESL instructional practices
B.1.2 Cognitive benefits of bilingualism
B.2 In what country was the study
conducted?
B.2.1 USA
B.2.2 Canada
B.2.3 United Kingdom
B.2.4 Australia/New Zealand
B.2.5 Other (please specify)
B.3 What is the language of the
monolingual group?
B.3.1 English
B.3.2 Other (please specify)
B.4 What are the languages of the
bilingual group?
B.4.1 (Please list)
B.4.2 (Please list)
B.5 What is/are the cognitive variable(s)
being measured?
B.5.1 Problem-solving skills
B.5.2 Theory-of-mind tasks
B.5.3 Creative and divergent thinking
B.5.4 Cognitive flexibility
B.5.5 Abstract representation and reasoning
skills
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
232
B.5.6 Learning strategies
B.5.7 Metalinguistic awareness
B.5.8 Metacognitive skills
B.5.9 Working memory
B.5.10 Others (Please provide details)
B.6 What is the age range of the
participants?
B.6.1 Grades K–3 (5–8 years old)
B.6.2 Grades 4–7 (9–12 years old)
B.6.3 Grades 8–12 (13–18 years old)
B.6.4 Longitudinal range (Please provide details)
B.7 What is the location of the study? B.7.1 Classroom
B.7.2 Pull-out room (for instance, a resource
room or another room within the school)
B.7.3 Laboratory
B.7.4 Other (please specify)
Section C: Study research questions
C.1 What is the overarching question the
researcher is trying to address? Please
write in authors’ description if there is
one. Elaborate if necessary, but
indicate which aspects are reviewers’
interpretations. Other, more specific
questions about the research questions
and hypotheses are asked later.
C.1.1 Explicitly stated (please specify)
C.1.2 Implicit (please specify)
C.1.3 Not stated/unclear (please specify)
C.2 What is the author’s specific
research question? Research questions
operationalize the overarching
question. Please write in author’s
description if there is one. Elaborate if
necessary, but indicate which aspects
are reviewer’s interpretations.
C.2.1 Explicitly stated (please specify)
C.2.2 Implicit (please specify)
C.2.3 Not stated/unclear (please specify)
C.3 What is the author’s hypothesis?
Research questions or hypotheses
operationalize the aims of the study.
Please write in authors’ description if
there is one. Elaborate if necessary,
but indicate which aspects are
reviewers’ interpretations.
C.3.1 Explicitly stated (please specify)
C.3.2 Implicit (please specify)
C.3.3 Not stated/unclear (please specify)
C.4 What is the theoretical/empirical basis
for this study? Please write in author’s
description if there is one. Elaborate if
necessary, but indicate which aspects
are reviewers’ interpretations.
C.4.1 Explicitly stated (please specify)
C.4.2 Implicit (please specify)
C.4.3 Not stated/unclear (please specify)
APPENDIX (continued)
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
233
Section D: Methods–designs
D.3 What variables are included? D.3.1 Independent variables
List independent moderator variables
D.3.2 Dependent (outcome) variables
List outcome variables
D.4 What measurement tool(s) is/are
used?
D.4.1 Standardized test
Please provide the name of the test if listed
D.4.2 Classroom or teacher developed test
Please describe and give page number
D.4.3 Observation
Please give a description and the page
number
D.4.4 Other
Please describe and give page number
Section E: Methods–groups
E.1 What is the design of the study E.1.1 Nonrandomized with treatment and
control groups. How were the groups
assigned/created?
E.1.2 Repeated measures design (Where the
same sample of individuals is measured in
all of the conditions)
E.1.3 Other (please specify)
E.2 How do the groups differ? E.2.1 Explicitly stated (please specify)
E.2.2 Implicitly stated (please specify)
E.2.3 Not applicable (not more than one group)
E.2.4 Not stated/unclear (please specify)
E.3 Number of groups. For instance, in
studies in which comparisons are made
between groups, this may be the number
of groups into which the dataset is
divided for analysis (e.g., social class,
or form size).
E.3.1 Not applicable
E.3.2 One
E.3.3 Two
E.3.4 Three
E.3.5 Four or more (please specify)
E.3.6 Other/unclear (please specify)
Section F: Method–sampling strategy
F.1 Are the authors trying to produce
findings that are representative of a
given population? Please write in
authors’ description. If authors do not
specify, please indicate reviewers’
interpretation.
F.1.1 Explicitly stated (please specify)
F.1.2 Implicit (please specify)
F.1.3 Not stated/unclear (please specify)
F.2 What is the sampling frame (if any)
from which the participants are chosen?
e.g., telephone directory, electoral
register, postcode, school listing, etc.
There may be two stages—e.g., first
sampling schools and then classes or
pupils within them.
F.2.1 Not applicable (please specify)
F.2.2 Explicitly stated (please specify)
F.2.3 Implicit (please specify)
F.2.4 Not stated/unclear (please specify)
APPENDIX (continued)
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
234
F.3 Which method does the study use to
select people, or groups of people (from
the sampling frame)? e.g., selecting
people at random, systematically—
selecting for example every 5th person,
etc.
F.3.1 Not applicable (no sampling frame)
F.3.2 Explicitly stated (please specify)
F.3.3 Implicit (please specify)
F.3.4 Not stated/unclear (please specify)
Section G: Sample information
G.1 What was the total number of
participants in the study (the actual
sample)? If more than one group is
being compared, please give numbers
for each group.
G.1.1 Not applicable (e.g., study of policies,
documents, etc.)
G.1.2 Explicitly stated (please specify)
G.1.3 Implicit (please specify)
G.1.4 Not stated/unclear (please specify)
G.2 What is the sex of the individuals in
the actual sample? Please give the
numbers of the sample that fall within
each of the given categories. If
necessary refer to a page number in the
report (e.g., for a useful table). If more
than one group is being compared,
please describe for each group.
G.2.1 Not applicable (e.g., study of policies,
documents, etc.)
G.2.2 Single sex (please specify)
G.2.3 Mixed sex (please specify)
G.2.4 Not stated/unclear (please specify)
G.2.5 Coding is based on: Authors’
description
G.2.6 Coding is based on: Reviewers’ inference
G.3 What is the socioeconomic status of
the individuals within the actual
sample? If more than one group is
being compared, please describe for
each group.
G.3.1 Not applicable (e.g., study of policies,
documents, etc.)
G.3.2 Explicitly stated (please specify)
G.3.3 Implicit (please specify)
G.3.4 Not stated/unclear (please specify)
G.4 What is the ethnicity of the individuals
within the actual sample? If more than
one group is being compared, please
describe for each group.
G.4.1 Not applicable (e.g., study of policies,
documents, etc.)
G.4.2 Explicitly stated (please specify)
G.4.3 Implicit (please specify)
G.4.4 Not stated/unclear (please specify)
G.5 What is known about the special
educational needs of individuals
within the actual sample? (choose all
that apply) Please note whether it was
explicitly stated or implicit.
G.5.1 Normally developing children
G.5.2 Language impaired
G.5.3 Learning disabled
G.5.4 Reading disabled
G.5.5 Late talkers
G.5.6 Intellectual difficulties
G.5.7 Other (please specify)
G.5.8 Not applicable (e.g., study of policies,
documents, etc.)
G.5.9 Not stated/unclear (please specify)
G.6 What are the regional characteristics
of individuals/groups in sample?
G.6.1 Not applicable (please specify)
G.6.2 Urban/inner city
G.6.3 Suburban
G.6.4 Rural
G.6.5 Not stated/unclear (please specify)
APPENDIX (continued)
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
235
G.6.6 Other
G.7 Do the authors describe strategies used
to control for bias from confounding
variables and groups? Please include
information on: Were the groups similar
at the start of the study?
G.7.1 Age (please specify)
G.7.2 Gender (please specify)
G.7.3 Social class (please specify)
G.7.4 Not stated/Unclear
G.7.5 Other (please specify)
G.8 What are additional sample
information/characteristics if any?
G.8.1 Not applicable
G.8.2 Sample was obtained from another
study (specify study)
G.8.3 Other
G.8.4 Not stated/unclear (please specify)
G.9 How many participants left before the
end of the study? If more than one group,
please give numbers for each group.
G.10 If the study involves following
samples prospectively over time, do
authors provide baseline values of key
variables such as those being used as
outcomes and relevant socio-
demographic variables?
G.10.1 Not applicable (e.g., study of policies,
documents, etc.)
G.10.2 Not applicable (not following samples
prospectively over time)
G.10.3 Yes (please specify)
G.10.4 No
Section H: Recruitment and consent
H.1 Which methods are used to recruit
people into the study? e.g., letters of
invitation, telephone contact, face-to-
face contact.
H.1.1 Letter of invitation
H.1.2 Telephone contact
H.1.3 Other (please specify)
H.1.4 Not applicable (please specify)
H.1.5 Not stated/unclear (please specify)
H.2 Were any incentives provided to
recruit people into the study?
H.2.1 Not applicable (please specify)
H.2.2 Explicitly stated (please specify)
H.2.3 Not stated/unclear (please specify)
H.3 Whose consent was sought? Please
comment on the quality of consent if
relevant.
H.3.1 Participants
H.3.2 Parental consent
H.3.3 Other (please specify)
H.3.4 No consent was sought
H.3.5 Not stated/unclear
Section I: Data collection
I.1 Which methods were used to collect
the data? Please indicate all that
apply and give further detail where
possible.
I.1.1 Experimental
I.1.2 Curriculum-based assessment
I.1.3 Focus group
I.1.4 Group interview
I.1.5 One to one interview (face to face or by
phone)
I.1.6 Observation
I.1.7 Self-completion questionnaire
APPENDIX (continued)
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
236
I.1.8 Self-completion report or diary
I.1.9 Exams
I.1.10 Clinical test
I.1.11 Practical test
I.1.12 Psychological test
I.1.13 Hypothetical scenario including
vignettes
I.1.14 School/college records (e.g.,
attendance records, etc.)
I.1.15 Secondary data such as publicly
available statistics
I.1.16 Other documentation
I.1.17 Not stated/unclear (please specify)
I.1.18 Coding is based on: authors’ description
I.1.19 Coding is based on: reviewers’
interpretation
I.2 Who collected the data? Please
indicate all that apply and give further
detail where possible.
I.2.1 Researcher
I.2.2 Head teacher/senior management
I.2.3 Teaching or other staff
I.2.4 Parents
I.2.5 Pupils/students
I.2.6 Governors
I.2.7 LEA/government officials
I.2.8 Other educational practitioner
I.2.9 Other (please specify)
I.2.10 Not stated/unclear
I.2.11 Coding is based on: Authors’ description
I.2.12 Coding is based on: Reviewers’ inference
I.3 Do the authors describe any ways
they addressed the reliability of their
data collection tools/methods? e.g.,
test–retest methods (Where more than
one tool was employed, please provide
details for each.)
I.3.1 Details
I.4 Do the authors describe any ways
they have addressed the validity of
their data collection tools/methods?
e.g., mention previous validation of
tools, published version of tools,
involvement of target population in
development of tools. (Where more
than one tool was employed, please
provide details for each.)
I.4.1 Details
APPENDIX (continued)
(continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
237
Section J: Data analysis
J.1 Which statistical methods, if any,
were used in the analysis? (check all
that apply)
J.1.1 Descriptive
J.1.2 Correlation
J.1.3 Group differences (e.g., t test, ANOVA)
(please specify)
J.1.4 Growth Curve analysis/multilevel
modeling (HLM)
J.1.5 Structural equation modeling (SEM)
J.1.6 Path analysis
J.1.7 Regression
J.1.8 Latent growth curve
J.1.9 Other (please specify)
Section K: Results and conclusion
K.1 What are the results of the study as
reported by authors? Please give
details and refer to page numbers in
the report(s) of the study, where
necessary (e.g., for key tables).
Group means:
SD:
N:
Estimated effect size:
Appropriate SD:
F, t statistic:
Significance:
Inverse variance weight:
K.2 Are there any shortcomings in the
reporting of the data? Please list all
implicit and explicit shortcomings of
the study.
K.2.1 Yes (please specify)
K.2.2 No
K.3 Do the authors report on all
variables they aimed to study as
specified in their aims/research
questions? This excludes variables
just used to describe the sample.
K.3.1 Yes
K.3.2 No (please specify)
K.4 What do the author(s) conclude
about the findings of the study?
Please give details and refer to page
numbers in the report of the study,
where necessary.
K.4.1 Details
Note. ESL = English as a second language; LEA = local educational agencies; HLM = hierarchical linear
modeling. Thanks to the EPPI-Reviewer database system team (Thomas & Brunton, 2006).
APPENDIX (continued)
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
238
Notes
This research was supported by a grant from the Canadian Language & Literacy
Research Network (CLLRNet) and funds from the Canadian Council on Learning
(CCL).
1When the value of Q is less than or equal to the degree of freedom associated with
a subset, I2 is assigned a value of zero. Similarly, negative values of I2 are assigned a
value of zero so that I2 lies between 0% and 100%. A value of 0% indicates no observed
heterogeneity, and larger values show increasing heterogeneity. Higgins and Thompson
(2002) recommend that percentages of around 25% (I2 = .25), 50% (I2 = .50), and 75%
(I2 = .75) should be interpreted to mean low, medium, and high heterogeneity, respec-
tively.
2The “mixed” category consists of studies conducted in multiple locations (e.g.,
Ransdell, Barbier, & Niit, 2006, reported a study conducted in three disparate geo-
graphical locations—Estonia, France, and the United States).
References
References marked with an asterisk indicate articles included in the meta-analysis.
Abrami, P. C., & Bernard, R. M. (2006). Research on distance education: In defense of
field experiments. Distance Education, 27, 5–26.
*Abu-Rabia, S., & Siegel, L. S. (2002). Reading, syntactic, orthographic, and working
memory skills of bilingual Arabic-English speaking Canadian children. Journal of
Psycholinguistic Research, 31, 661–678.
Baddeley, A. D. (1996). Exploring the central executive. Quarterly Journal of
Experimental Psychology, 49A, 5–28.
Baker, K. A., & de Kanter, A. A. (1981). Effectiveness of bilingual education: A review
of the literature. Washington, DC: U.S. Department of Education, Office of Planning,
Budget and Evaluation.
Bamford, K. W., & Mizokawa, D. T. (1990, April). Cognitive development of children
in an additive-bilingual program: The third report. Paper presented at the meeting
of the American Educational Research Association, Boston, MA.
Bamford, K. W., & Mizokawa, D. T. (1991, April). Cognitive and language development in
an additive-bilingual program: Report after four observations. Paper presented at the
meeting of the American Educational Research Association, Chicago, IL.
Bamford, K. W., & Mizokawa, D. T. (1992, April). Spanish-immersion children in
Washington State: Fourth year of a longitudinal study. Paper presented at the meet-
ing of the American Educational Research Association, Seattle, WA.
Berguno, G., & Bowler, D. M. (2004). Communicative interactions, knowledge of a
second language, and theory of mind in young children. Journal of Genetic
Psychology, 165, 293–309.
Bhatia, T., & Ritchie, W. (Eds.). (2006). Handbook of bilingualism. Cambridge, MA:
Blackwell.
Bialystok, E. (1987). Words as things: Development of word concept by bilingual
children. Studies in Second Language Acquisition, 9, 133–140.
Bialystok, E. (1988). Levels of bilingualism and levels of linguistic awareness.
Developmental Psychology, 24, 560–567.
*Bialystok, E. (1997). Effects of bilingualism and biliteracy on children’s emerging
concepts of print. Developmental Psychology, 33, 429–440.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
239
*Bialystok, E. (1999). Cognitive complexity and attentional control in the bilingual
mind. Child Development, 70, 636–644.
Bialystok, E. (2001a). Against isolationism: Cognitive perspectives on second language
research. Selected Proceedings of the Second Language Research Forum, 97–103.
Bialystok, E. (2001b). Bilingualism in development: Language, literacy, and cognition.
Cambridge, UK: Cambridge University Press.
Bialystok, E. (2002). Acquisition of literacy in bilingual children: A framework for
research. Language Learning, 52, 159–199.
Bialystok, E. (2005). Consequences of bilingualism for cognitive development. In J. F.
Kroll & A. M. B. de Groot (Eds.), Handbook of bilingualism: Psycholinguistic
approaches (pp. 417–432). New York, NY: Oxford University Press.
*Bialystok, E. (2006). Effect of bilingualism and computer video game experience on
the Simon task. Canadian Journal of Experimental Psychology, 60, 68–79.
Bialystok, E., Craik, F. I. M., & Freedman, M. (2007). Bilingualism as a protection
against the onset of symptoms of dementia. Neuropsychologia, 45, 459–464.
*Bialystok, E., Craik, F. I. M., Grady, C., Chau, W., Ishii, R., Gunji, A., & Pantev, C.
(2005). Effect of bilingualism on cognitive control in the Simon task: Evidence from
MEG. NeuroImage, 24, 40–49.
*Bialystok, E., Craik, F. I. M., Klein, R., & Viswanathan, M. (2004). Bilingualism, aging,
and cognitive control: Evidence from the Simon task. Psychology & Aging, 19, 290–303.
Bialystok, E., Craik, F. I. M., & Luk, G. (2008). Cognitive control and lexical access
in younger and older bilinguals. Journal of Experimental Psychology: Learning,
Memory and Cognition, 34, 859–873.
*Bialystok, E., Craik, F. I. M., & Ruocco, A. C. (2006). Dual-modality monitoring in
a classification task: The effects of bilingualism and aging. Quarterly Journal of
Experimental Psychology, 59, 1968–1983.
Bialystok, E., Craik, F. I. M., & Ryan, J. (2006). Executive control in a modified anti-
saccade task: Effects of aging and bilingualism. Journal of Experimental Psychology:
Learning, Memory, and Cognition, 32, 1341–1354.
*Bialystok, E., Luk, G., & Kwan, E. (2005). Bilingualism, biliteracy, and learning to
read: Interactions among languages and writing systems. Scientific Studies of
Reading, 9, 43–61.
*Bialystok, E., & Majumder, S. (1998). The relationship between bilingualism and the
development of cognitive processes in problem solving. Applied Psycholinguistics,
19, 69–85.
*Bialystok, E., Majumder, S., & Martin, M. M. (2003). Developing phonological
awareness: Is there a bilingual advantage? Applied Psycholinguistics, 24, 27–44.
*Bialystok, E., & Martin, M. M. (2004). Attention and inhibition in bilingual children:
Evidence from the dimensional change card sort task. Developmental Science, 7,
325–339.
Bialystok, E., Martin, M. M., & Viswanathan, M. (2005). Bilingualism across the lifespan:
The rise and fall of inhibitory control. International Journal of Bilingualism, 9, 103–119.
*Bialystok, E., McBride-Chang, C., & Luk, G. (2005). Bilingualism, language profi-
ciency, and learning to read in two writing systems. Journal of Educational
Psychology, 97, 580–590.
*Bialystok, E., & Senman, L. (2004). Executive processes in Appearance-reality tasks:
The role of inhibition of attention and symbolic representation. Child Development,
75, 562–579.
*Bialystok, E., & Shapero, D. (2005). Ambiguous benefits: The effect of bilingualism
on reversing ambiguous figures. Developmental Science, 8, 595–604.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
240
Bochner, S. (1996). The learning strategies of bilingual versus monolingual students.
British Journal of Educational Psychology, 66, 83–93.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2008).
Comprehensive meta-analysis (Version 2.2.048). Englewood, NJ: Biostat.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction
to meta analysis. Chichester, UK: Wiley.
Braccini, F., & Cianchi, R. (1993). The influence on some linguistic and cognitive
skills of the early learning of a foreign language. Rassegna Italiana Di Linguistica
Applicata, 25, 53–66.
*Bruck, M., & Genesee, F. (1995). Phonological awareness in young second language
learners. Journal of Child Language, 22, 307–324.
*Campbell, R., & Sais, E. (1995). Accelerated metalinguistic (phonological) aware-
ness in bilingual children. British Journal of Developmental Psychology, 13, 61–68.
Carson, K. P., Schriesheim, C. A., & Kinicki, A. J. (1990). The usefulness of the “fail-
safe” statistic (N) in meta-analysis. Educational and Psychological Measurement,
50, 233–243.
Cazden, C. R. (1974). Play with language and metalinguistic awareness: One dimen-
sion of language experience. Urban Review, 7, 28–29.
Cenoz, J. 2003. The additive effect of bilingualism on third language acquisition: A
review. International Journal of Bilingualism, 7, 71–87.
Chan, K. T. (2005). Chinese-English bilinguals’ theory-of-mind development. Dissertation
Abstracts International, A: The Humanities and Social Sciences, 65 (10-A).
*Chen, X., Anderson, R. C., Li, W., Hao, M., Wu, X., & Shu, H. (2004). Phonological
awareness of bilingual and monolingual Chinese children. Journal of Educational
Psychology, 96, 142–151.
*Clarkson, P. C., & Galbraith, P. (1992). Bilingualism and mathematics learning—
another perspective. Journal for Research in Mathematics Education, 23, 34–44.
Cooper, H. M., & Hedges, L. V. (1994). The handbook of research synthesis. New York,
NY: Russell Sage.
*Cromdal, J. (1999). Childhood bilingualism and metalinguistic skills: Analysis and
control in young Swedish English bilinguals. Applied Psycholinguistics, 20, 1–20.
Cummins, J. (1976). The influence of bilingualism on cognitive growth: A synthesis of
research findings and explanatory hypotheses. Working Papers on Bilingualism, 9, 1–43.
*Davidson, D., & Tell, D. (2005). Monolingual and bilingual children’s use of mutual
exclusivity in the naming of whole objects. Journal of Experimental Child
Psychology, 92, 25–45.
*Demie, F., & Strand, S. (2006). English language acquisition and educational attain-
ment at the end of secondary school. Educational Studies, 32, 215–231.
*Demont, E. (2001). Contribution of early second language learning in the develop-
ment of language awareness and learning to read. International Journal of
Psychology, 36, 274–285.
Denson, N. (2009). Do curricular and co-curricular diversity activities influence racial
bias? A meta-analysis. Review of Educational Research, 79, 805–838.
Diaz, R. M. (1985). The intellectual power of bilingualism. Quarterly Newsletter of the
Laboratory of Comparative Human Cognition, 7, 16–22.
Diaz, R. M., & Klinger, C. (1991). Towards an explanatory model of the interaction between
bilingualism and cognitive development. In E. Bialystok (Ed.), Language processing in
bilingual children (pp. 167–192). Cambridge, UK: Cambridge University Press.
Duncan, S. E. (2005). Child bilingualism and cognitive functioning: A study of four
Hispanic groups. Dissertation Abstracts International, A: The Humanities and
Social Sciences, 65(8), 2895–A.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
241
Emmorey, K., Luk, G., Pyers, J. E., & Bialystok, E. (2008). The source of enhanced
cognitive control in bilinguals. Psychological Science, 19, 1201–1206.
Engle, R. W. (2002). Working memory capacity as executive attention. Current
Directions in Psychological Science, 11, 19–23.
Esses, V. M., & Gardner, R. C. (1996). Multiculturalism in Canada: Context and cur-
rent status. Canadian Journal of Behavioural Science, 28, 145–152.
*Eviatar, Z., & Ibrahim, R. (2000). Bilingual is as bilingual does: Metalinguistic abili-
ties of Arabic-speaking children. Applied Psycholinguistics, 21, 451–471.
Ferdman, B. M., & Hakuta, K. (1985, April). A population perspective on bilingualism
in Puerto Rican children. Paper presented at the meeting of the Society for Research
in Child Development, Toronto, Canada.
Fernandes, M. A., Craik, F. I. M., Bialystok, E., & Kreuger, S. (2007). Effects of bilin-
gualism, aging, and semantic relatedness on memory under divided attention.
Canadian Journal of Experimental Psychology, 61, 128–141.
Flavell, J. H. (1978). Metacognitive development. In J. M. Scandura & C. J. Brainerd
(Eds.), Structural process theories of complex human behavior (pp. 213–245).
Alphen aan den Rijn, Netherlands: Sijthoff & Noordhoff.
*Galambos, S. J., & Goldin-Meadow, S. (1990). The effects of learning two languages
on levels of metalinguistic awareness. Cognition, 34, 1–56.
Galambos, S., & Hakuta, K. (1988). Subject-specific and task-specific characteristics
of metalinguistic awareness in bilingual children. Applied Psycholinguistics, 9,
141–162.
Goetz, P. J. (2000). A comparison of 3- and 4-year-old English monolinguals, man-
darin monolinguals, and mandarin-English bilinguals on a series of theory of
mind-related tasks. Dissertation Abstracts International, B: Sciences and
Engineering, 61, 1110–B.
Goncz, L. (1988). A research study on the relation between early bilingualism and
cognitive development. Psychologische Beitrage, 30, 75–91.
Government of Canada. (2009). Roadmap for Canada’s linguistic duality, 2008–2013:
Acting for the future. Retrieved from http://www.pch.gc.ca/pgm/slo-ols/strat-eng.
cfm
*Gutierrez-Clellen, V. F., Calderon, J., & Weismer, S. E. (2004). Verbal working mem-
ory in bilingual children. Journal of Speech, Language, and Hearing Research, 47,
863–876.
Hakuta, K. (1986). Mirror of language: The debate on bilingualism. New York, NY:
Basic Books.
Hakuta, K. (1990). Bilingualism and bilingual education: A research perspective
(Occasional Papers in Bilingual Education No. 1). Washington, DC: National
Clearinghouse for Bilingual Education.
Hamers, J. F., & Blanc, M. H. A. (2000). Bilinguality and bilingualism (2nd ed.).
Cambridge, UK: Cambridge University Press.
Hedges, L. V. (1981). Distribution theory for Glass’s estimator of effect size and related
estimators. Journal of Educational Statistics, 6, 107–128.
Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL:
Academic Press.
Hedges, L. V., & Vevea, J. L. (1998). Fixed and random-effects models in meta-analysis.
Psychological Methods, 3, 486–504.
Higgins, J. P. T., & Thompson, S. G. (2002). Quantifying heterogeneity in a meta-
analysis. Statistics in Medicine, 21, 1539–1558.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
242
Ho, D. Y. (1987). Bilingual effects on language and cognitive development: With spe-
cial reference to Chinese-English bilinguals. Bulletin of the Hong Kong Psychological
Society, 18, 61–69.
Huber, E., & Lasagabaster, D. (2000). The cognitive effects of bilingualism. ITL,
Review of Applied Linguistics, 129–130, 191–224.
Huedo-Medina, T. B., Sánchez-Meca, F., Marín-Martínez, F., & Botella, J. (2006).
Assessing heterogeneity in meta-analysis: I2 or Q statistic? Psychological Methods,
11, 193–206.
*Humphreys, G. W., & Mumtaz, S. (2001). The effects of bilingualism on learning to
read English: Evidence from the contrast between Urdu-English bilingual and
English monolingual children. Journal of Research in Reading, 24, 113–134.
Iannaccone, A., Fraternali, O. T., & Vaccia, R. (1992). Cognitive flexibility and early
bilingual education. Rassegna Italiana Di Linguistica Applicata, 24, 149–165.
Johnson, J. (1991). Constructive processes in bilingualism and their cognitive growth
effects. In E. Bialystok (Ed.), Language processing in bilingual children (pp. 193–
221). Cambridge, UK: Cambridge University Press.
Johnson, J. S., & Newport, E. L. (1989). Critical period effects in second language
learning: The influence of maturational state on the acquisition of English as a sec-
ond language. Cognitive Psychology, 21, 60–99.
*Jordà-Maria, P. S. (2003). Metapragmatic awareness and pragmatic production of
third language learners of English: A focus on request acts realizations. International
Journal of Bilingualism, 7, 43–68.
Just, M. A., & Carpenter, P. A. (1992). A capacity theory of comprehension: Individual
differences in working memory. Psychological Review, 98, 122–149.
Kane, M. J., Bleckley, M. K., Conway, A. R. A., & Engle, R. W. (2001). A controlled-
attention view of working-memory capacity. Journal of Experimental Psychology:
General, 130, 169–183.
Kemp, C. (2007). Strategic processing in grammar learning: Do multilinguals use more
strategies? International Journal of Bilingualism, 4, 241–261.
*Keshavarz, M. H., & Astaneh, H. (2004). The impact of bilinguality on the learning
of English vocabulary as a foreign language (L3). International Journal of Bilingual
Education and Bilingualism, 7, 295–302.
Koda, K., & Zehler, A. (Eds.). (2008). Learning to read across languages: Cross-
linguistic relationships in first- and second-language literacy development. New
York, NY: Routledge.
Konaka, K. (1997). The relationship between degree of bilingualism and gender to
divergent thinking ability among native Japanese-speaking children in the New York
Area. Dissertation Abstracts International, AAT 9718715.
Kovacs, A. M., & Teglas, E. (2002). Integrating two languages, theories of minds, and
executive functions. Odense Working Papers in Language and Communication, 3, 1.
Kozulin, A. (1999). Reality monitoring, psychological tools, and cognitive flexibility
in bilinguals: Theoretical synthesis and pilot experimental investigation. In P. Lloyd
& C. Fernyhough (Eds.), Lev Vygotsky: Critical assessments: Future directions (Vol.
4, pp. 187–198). New York, NY: Routledge.
Lee, H., Plass, J. L., & Homer, B. D. (2006). Optimizing cognitive load for learning
from computer-based science simulations. Journal of Educational Psychology, 98,
902–913.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA:
Sage.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
243
*Loizou, M., & Stuart, M. (2003). Phonological awareness in monolingual and bilin-
gual English and Greek five-year-olds. Journal of Research in Reading, 26, 3–18.
*Love, T., Maas, E., & Swinney, D. (2003). The influence of language exposure on
lexical and syntactic language processing. Experimental Psychology, 50, 204–216.
Macnamara, J. (1966). Bilingualism and primary education. Edinburgh, UK: Edinburgh
University Press.
McLeay, H. (2003). The relationship between bilingualism and the performance of
spatial tasks. International Journal of Bilingual Education and Bilingualism, 6,
423–438.
*Merriman, W. E., & Kutlesic, V. (1993). Bilingual and monolingual children’s use of
two lexical acquisition heuristics. Applied Psycholinguistics, 14, 229–249.
Michael, E., & Gollan, T. H. (2005). Being and becoming bilingual: Individual differ-
ences and consequences for language production. In J. F. Kroll & A. M. B. de Groot
(Eds.), The handbook of bilingualism: Psycholinguistic approaches (pp. 389–407).
New York, NY: Oxford University Press.
National Research Council. (1992). Combining Information: Statistical issues and
opportunities for research. Washington, DC: National Academy Press.
*Oller, D. K., Pearson, B. Z., & Cobo-Lewis, A. B. (2007). Profile effects in early
bilingual language and literacy. Applied Psycholinguistics, 28, 191–230.
Orwin, R. (1983). A fail-safe N for effect size in meta-analysis. Journal of Educational
Statistics, 8, 157–159.
Peal, E., & Lambert, W. E. (1962). The relation of bilingualism to intelligence.
Psychological Monographs, 76, 1–23.
Ponomarev, S. Y. (1992). Bilingualism and its influence on personality development.
Vestnik Sankt-Peterburgskogo Universiteta, Seriya 2: Istoriya, Yazykoznanie,
Literaturovedenie, 3, 46–49.
*Ransdell, S., Barbier, M., & Niit, T. (2006). Metacognitions about language skill and
working memory among monolingual and bilingual college students: When does
multilingualism matter? International Journal of Bilingual Education and
Bilingualism, 9, 728–741.
Ravitch, D. (1990). Multiculturalism: E pluribus plures. American Scholar, 59, 339–340.
*Reichard, C., & Mokhtari, K. (2004). Investigating the strategic reading processes of first
and second language readers in two different cultural contexts. System, 32, 379–394.
Ricciardelli, L. A. (1992). Creativity and bilingualism. Journal of Creative Behavior,
26, 242–254.
Ricciardelli, L. A. (1993). An investigation of the cognitive development of Italian-
English bilinguals and Italian monolinguals from Rome. Journal of Multilingual and
Multicultural Development, 14, 345–346.
Rosen, V. M., & Engle, R. W. (1997). The role of working memory capacity in retrieval.
Journal of Experimental Psychology: General, 126, 211–227.
Rosenblum, T., & Pinker, S. A. (1983). Word magic revisited: Monolingual and bilin-
gual children’s understanding of the word-object relationship. Child Development,
54, 773–780.
Rosenthal, R. (1979). The “file drawer problem” and tolerance for null results.
Psychological Bulletin, 86, 638–641.
Rossell, C. H., & Baker, K. (1996). The educational effectiveness of bilingual educa-
tion. Research in the Teaching of English, 30, 7–69.
*Sanz, C. (2000). Bilingual education enhances third language acquisition: Evidence
from Catalonia. Applied Psycholinguistics, 21, 23–44.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Adesope et al.
244
*Sheng, L., McGregor, K. K., & Marian, V. (2006). Lexical-semantic organization in
bilingual children: Evidence from a repeated word association task. Journal of
Speech, Language, and Hearing Research, 49, 572–587.
Slavin, R. E., & Cheung, A. (2005). A synthesis of research on language of reading instruc-
tion for English language learners. Review of Educational Research, 75, 247–284.
Srivastava, B. (1991). Creativity and linguistic proficiency. Psycho-Lingua, 21,
105–109.
Stephens, M. A. (1997). Bilingualism, creativity, and social problem-solving. Dissertation
Abstracts International, AAT 9729615.
Sweller, J., & Chandler, P. (1994). Why some material is difficult to learn. Cognition
& Instruction, 12, 185–233.
Thomas, J., & Brunton, J. (2006). EPPI-Reviewer 3.0: Analysis and management of
data for research synthesis. London, UK: Social Science Research Unit, Institute of
Education.
Titone, R. (1997). Early second-language learning: Bilingualism and metalinguistic
development. Europe Plurilingue, 12–13, 138–149.
Tucker, G. R. (1998). A global perspective on multilingualism and multilingual educa-
tion. In J. Cenoz & F. Genesee (Eds.), Beyond bilingualism: Multilingualism and
multilingual education (pp. 3–15). Clevedon, UK: Multilingual Matters.
van Merrienboer, J., & Sweller, J. (2005). Cognitive load theory and complex learning:
Recent developments and future directions. Educational Psychology Review, 17,
147–177.
Vorstman, E., De Swart, H., Ceginskas, V., & Van Den Bergh, H. (2009). Language
learning experience in school context and metacognitive awareness of multilingual
children. International Journal of Multilingualism, 6, 258–280.
Willig, A. (1985). A meta-analysis of selected studies on the effectiveness of bilingual
education. Review of Educational Research, 55, 269–317.
Willig, A. (1987). Examining bilingual education research through meta-analysis
and narrative review: A response to Baker. Review of Educational Research, 57,
363–376.
Yang, H., Yang, S., Ceci, S., & Wang, Q. (2005). Effects of bilinguals’ controlled-
attention on working memory and recognition. In J. Cohen, K. McAlister, K.
Rolstad, & J. MacSwan (Eds.), Proceedings of the 4th international symposium on
bilingualism. Somerville, MA: Cascadilla Press.
Yoshida, H. (2008). The cognitive consequences of early bilingualism. Journal of Zero
to Three, 29, 26–30.
Authors
OLUSOLA ADESOPE holds a Ph.D. in Educational Psychology from Simon Fraser
University, Canada. He recently accepted a position as Assistant Professor of Educational
Psychology in the Department of Educational Leadership & Counseling Psychology at
Washington State University, Pullman, WA 99164-2136; email: adesope@yahoo.com.
His research interests include the cognitive and psychological processes underlying mul-
timedia learning, teaching and learning with interactive concept maps and diagrams,
language pedagogy, technology-mediated instruction and research synthesis, especially
meta-analysis.
TRACY LAVIN is a Researcher with the Directions Evidence and Policy Research Group,
1055 Dunsmuir–Suite 1254, Four Bentall Centre, Vancouver, BC V7X 1A2; email:
lavin.t@gmail.com. She received a Ph.D. from the University of British Columbia and
pursued postdoctoral studies at Northwestern University. She has studied and written
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from
Cognitive Benefits of Bilingualism
245
about a number of learning and educational issues from both academic and public policy
perspectives. Her research interests include studying how children and adults learn lan-
guage and acquire its underpinning cognitive concepts. She has written several research
papers for the National Literacy Strategy initiative and also leads the team of analysts
responsible for developing, analyzing, and reporting on Canadian Council on Learning’s
annual Survey of Canadian Attitudes toward Learning.
TERRI THOMPSON is a Researcher with the Directions Evidence and Policy Research
Group, 1055 Dunsmuir–Suite1254, Four Bentall Centre, Vancouver, BC V7X 1A2;
email: territho@gmail.com. Over the past five years, she has participated in, and man-
aged the completion of, over 60 systematic reviews and question scans devoted to issues
about education and learning. She holds a Master’s degree in Education from the
University of British Columbia where she studied issues of social justice and education
policy.
CHARLES UNGERLEIDER is a professor of the sociology of education at the University
of British Columbia, 2125 Main Mall, Vancouver, BC V6T 1Z4, Canada; e-mail: charles.
ungerleider@ubc.ca. His research interests include educational policy, finance and gov-
ernance, educational change, and the educational trajectories of students in elementary
and secondary schools.
at SIMON FRASER LIBRARY on June 28, 2010 http://rer.aera.netDownloaded from