Music training with Démos program positively influences cognitive functions in children from low socio-economic backgrounds
Music training with D?emos program positively influences cognitive functions in children from low socio-economic backgrounds
Myl?ne BarbarouxID 0 1
Eva Dittinger 0 1
Mireille Besson 0 1
0 CNRS & Aix-Marseille University, Laboratoire de Neurosciences Cognitives (LNC, UMR 7291) , Marseille , France , 2 CNRS & Aix-Marseille University, Laboratoire Parole et Langage (LPL, UMR 7309), Aix-en- Provence, France, 3 Brain and Language Research Institute (BLRI) , Aix-en-Provence , France
1 Editor: Linda Chao, University of California , San Francisco , UNITED STATES
This study aimed at evaluating the impact of a classic music training program (De? mos) on several aspects of the cognitive development of children from low socio-economic backgrounds. We were specifically interested in general intelligence, phonological awareness and reading abilities, and in other cognitive abilities that may be improved by music training such as auditory and visual attention, working and short-term memory and visuomotor precision. We used a longitudinal approach with children presented with standardized tests before the start and after 18 months of music training. To test for pre-to-post training improvements while discarding maturation and developmental effects, raw scores for each child and for each test were normalized relative to their age group. Results showed that De? mos music training improved musicality scores, total IQ and Symbol Search scores as well as concentration abilities and reading precision. In line with previous results, these findings demonstrate the positive impact of an ecologically-valid music training program on the cognitive development of children from low socio-economic backgrounds and strongly encourage the broader implementation of such programs in disadvantaged school-settings.
Funding: This research was conducted with the
support of the CNRS (MB), of the Philharmonie de
Paris who implemented the De?mos program in
Marseille (MB), of the Brain and Language
Research Institute (BLRI, MB) and of the Institute
or Communication and Language in the Brain
(ILCB; ANR-11-LABX-0036, MB and
ANR-11IDEX-0001-02 (A-MIDEX, MB). M. Barbaroux is
Music training programs for children and adolescents from low socio-economic status (SES),
defined based on parental income and education, have flourished in many countries since the
fifties, not only to promote music education but also to increase quality of life, general
education and the cohesion of social groups. For instance, the Yehudi Menuhin school was founded
in England by the well-known violinist in 1963; the El Sistema project was created in
Venezuela in 1975 by the musician and economist Jose? Antonio Abreu and was later extended to
Canada, Europe -England, France, Greece, Portugal- and more than 80 El Sistema-inspired
programs are active throughout the United States. More recently, the Harmony project was
launched in Los Angeles suburbs in the years 2000 to improve education for youth from
lowincome communities. Similarly, the De?mos project (?Dispositif d?e?ducation musicale et
orchestrale ? vocation sociale?, Musical and orchestral education with social vocation: http://
supported by a doctoral fellowship from the French
Ministry of Research and Education and ED by a
doctoral fellowship from the BLRI. The funders had
no role in study design, data collection and
analysis, decision to publish, or preparation of the
projetdemos.fr/qu-est-ce-que-demos.aspx) that is pivotal to the research presented here, was
initiated by the Paris Philharmonie in 2010, with the objective to promote the cognitive
development and the social integration of children from disadvantaged backgrounds through access
to culture and free classic music training. However, the societal impact of these music
programs has not often been measured using a scientific approach. Our main objective here was
to evaluate the impact of the De?mos program on the cognitive development of children from
disadvantaged backgrounds using a longitudinal approach. To this end, we tested the children
before the start of the music intervention and 18 months later (test?music training?retest
procedure) using standard tests of general intelligence (Intelligence Quotient, IQ), of phonological
awareness and reading abilities as well as of other cognitive functions, including auditory and
visual attention, working and short-term memory and sensorimotor precision.
The issue of whether music training ?makes you smarter? is highly controversial. There is
clear evidence that music training in children is positively correlated to higher cognitive
functioning. For instance, in several studies using the full or reduced (four tests) versions of the
Wechsler Intelligence Scale for Children (WISC), Schellenberg and collaborators showed
higher IQ scores in musically-trained children than in their untrained counterparts [
How to interpret this association between music training and IQ is less clear-cut. Schellenberg
] pointed out that while music training may be the cause of improved cognitive functioning,
children with higher cognitive functioning are possibly more likely to engage in (and to
pursue) the demanding task of learning to play music than children with lower cognitive abilities.
To directly address this issue, Schellenberg [
] conducted the first well-controlled longitudinal
study with a large group of 6-year old children (N = 132) trained either in music, drama or no
specific training. After one year of training, increase in IQ scores was significantly larger in the
music group than in the control groups (6.1 vs 3.9 IQ points), thereby providing causal
evidence for a positive impact of music training on general intelligence in children from middle
to high socio-economic backgrounds. More recently, Sala and Gobet [
] also concluded from
the results of a meta-analysis that music training had small but significant beneficial effects on
general intelligence. To our knowledge, however, no study has yet focused on socially
disadvantaged children. Thus, our first objective was to determine whether music training as
provided by the De?mos program would also improve measures of general intelligence in children
from low socio-economic background.
The second objective was to evaluate the impact of the De?mos music program on
phonological awareness and reading skills that are frequently impaired in children from low income
]. While there is evidence that music training improves phonological skills in
typically developing children [
] (see also results of meta-analysis  as well as studies in
children with dyslexia [
]), the evidence for an impact of music training on reading skills is
more controversial [
To evaluate the impact of the Harmony project mentioned above, Kraus and colleagues
] followed up a group of 6?9 year children from low SES, from the gang reduction zones
in Los Angeles, for two years. Results showed a positive correlation between reading fluency
(tested using word and non-word reading tests) and the amount of engagement in the
Harmony music program (measured as the percentage of attendance and the level of participation
in music classes ). Importantly, while age-normed reading scores were enhanced in
children that were more-engaged in music classes, they tended to deteriorate over time in children
that were less-engaged, possibly reflecting the negative consequences of living in low
socioeconomic backgrounds [
]. Moreover, children trained with music for one year
outperformed non-trained children in tasks requiring the silent discrimination of written words in a
continuous sequence of letters [
]. Again, musically-trained children maintained their
reading scores at the age-normed level while they decreased for children in the control group.
2 / 21
However, Slater and colleagues [
] reported no significant effects of music training on
phonological awareness after one year of music training in the Harmony project.
In sum, results from the Kraus group provided evidence that the Harmony project helped
children from low SES to attain and maintain reading levels close to more privileged children
but the impact on phonological awareness was less clear-cut. By contrast, very recent results of
a longitudinal study by Linnavalli and collaborators [
] showed that music playschool, but
not dance lessons (both 45 minutes weekly, 30 times a year), enhanced phonological awareness
(e.g., choose the object whose name comprises the combination of phonemes pronounced by
the experimenter; suppress or replace a phoneme and say the resulting word) and vocabulary
(verbal knowledge) of 5-6-year-old preschool Finnish children. Similarly, Nan and
] showed that 6 months of piano training (45 minutes, 3 times per week) improved
word discrimination based on consonants compared to reading training in 4-5-year-old
Mandarin-speaking children. Importantly, Dege? & Schwartzer [
] also showed that music training
is possibly as efficient as phonological training and more efficient than sport training (in all
cases, 10 min of daily training for 20 weeks), to improve phonological awareness of large
phonological units (detecting rhymes and segmenting words into syllables). It was thus of interest
to compare these contrastive results to those of the children involved in the De?mos music
Finally, previous results also suggest that some aspects of auditory [
] and visual attention
] (but see [
] for negative results) are improved in adult musicians compared to
nonmusicians. Moreover, results of cross-sectional studies in children have shown that music
training is associated with better working and short-term memory as reflected by higher scores
at the backward and forward versions of the Digit Span test (i.e., repeat back series of orally
presented numbers) in musically-trained compared to untrained children [
recently, Guo and collaborators [
] provided evidence for a causal link between music
training and working memory in an interventional study: children quasi-randomly assigned to six
weeks of music training increased their backward digit span scores but the forward digit span
scores did not differ between the music and control groups. Moreover, Dege? and collaborators
] used a longitudinal approach and showed that children trained in music for two years
improved visual and auditory memory (recalling sequences of colors or sounds), while their
non-trained counterparts did not. Importantly, children and adolescents from low SES
typically show increased difficulties to focus attention [
] together with working memory deficits
that can, however, be counteracted by relevant interventions [
]. Based on these findings, our
third objective was to use a test-training-retest procedure with children from disadvantaged
socio-economic backgrounds to test for the relationship between music training and
improvements in these different cognitive abilities, auditory and visual attention, working and
shortterm memory, as well as in visuomotor precision (strongly needed to make the fine controlled
movements required to play a musical instrument). Evaluations of community-based musical
interventions are clearly too scarce and often incomplete, and it was of strong societal interest
to obtain a global screening of the impact of the De?mos music training program on different
cognitive and motor abilities.
To these aims, we tested children at two time points, before they started music training and
after 18 months of being involved in the De?mos program, using standardized tests of several
cognitive functions. Moreover, we computed normalized scores both before and after music
training so that significant pre-to-post improvements would reveal the influence of the De?mos
program on children? cognitive functions rather than the mere influence of maturation and
development (children were 18 months older post-than-pre training). Our general prediction
was that, rather than showing the typical deterioration of cognitive abilities across the course
of development [
] music training would increase the level of performance of children
3 / 21
from low SES to the age-normed level. Specifically and based on the literature reviewed above,
we hypothesized that classic music training would help children to improve their IQ scores [
] as well as their reading scores [
] and their backward digit span scores . However,
predictions were less clear-cut regarding the impact of music training on phonological
], auditory attention [
] and visual attention [
] because contrastive
results have been reported in the literature. Finally, based on previous results showing that
music training is more beneficial for children with lowest initial performance levels [
we also hypothesized that the beneficial effect of music training would be larger for the
children with initial lowest scores. In other words, we predicted that music training may, at least
to some extent, counteract the deleterious consequences of living in low socio-economic
backgrounds. Based on the relatively short duration of music training, we expected these effects to
be small but significant.
Participants and procedure
Fifty-four children (24 girls, 30 boys) from two primary schools, located in middle-class areas
in downtown Marseille, were involved in the De?mos music program. Even if the area is not
what would typically be called underprivileged, these schools were chosen by the organizers of
the De?mos program together with social institutions ?Apprentis d?Auteuil? because they are
specifically dedicated to the education of children from low socio-economic status. Since our
aim was to test the impact of the De?mos program implemented by the Paris Philharmonie in
two schools, the sample size was constrained by the number of children in each school.
Participation to the De?mos program was strongly encouraged (only a very few children did not
participate or withdraw from the program) and all children in the schools were tested. However,
nineteen children could not be re-tested in the second session 18 months later because some
children (7) left primary school to go to middle school and 11 children moved to another city/
school during the duration of the De?mos program. Only one child voluntary asked to stop
being involved in the De?mos program.
Analyses included the 35 children (18 girls, 17 boys), 7 to 12 years old (2nd to 5th grade in
the first session) that were tested both pre and post music training. All children were native
French speakers and had normal or corrected vision and normal audition. All children were
from low SES families, often single parent families, defined by French government criteria as
families with high unemployment levels, low income and social difficulties. This study was
verbally approved by the local ethics committee of Aix-Marseille University because it was
conducted under the responsibility of the directors and teachers of the two schools in which the
?De?mos project? was implemented. All parents gave their informed written consent for their
children to participate in the study. This study was conducted in agreement with guidelines for
the protection of human participants as defined in the declaration of Helsinki. The procedure
was carefully explained to the children to ensure that they agreed to participate in individual
testing sessions in a quiet classroom of the school. Children were informed that they could
stop performing the tests at any time and they were given presents at the end of the pre and
post training evaluations to thank them for their participation.
Music classes were taught by two professional music teachers specifically trained for
interventions in school settings. Children were trained over a 18 months period (except vacation
times), twice a week for two hours for a total of 4 hour/week. Small groups of 6?7 children
were formed based on the instrument (from the string or wind family) that they had freely
4 / 21
chosen. School teachers were involved in the music classes together with the children. The
teaching method was inspired from the Suzuki and Kodaly methods that are based on
listening, imitation and memorization. Each child learned to recognize the various instruments and
to reproduce the sounds and the music played by the teacher on their own instrument.
Children were also progressively trained to read musical scores. Finally, additional orchestra classes
were given once every 6 weeks for two hours to train the children for a public concert
performed with professional musicians from the Opera orchestra at the end of each school year.
The total number of training hours over the 18 months period was around 250 hours, without
considering the time children spent practicing their instrument at home. At the end of the
Post-training session, children were asked how many times and for how long they practiced
their instrument at home each week. Most children (around 90%) reported that they did
indeed practiced their instrument at home, 2 or 3 times a week for a total duration of 45 min
The experimental procedure and the different tests are presented on Fig 1.
Two musicality tests adapted from the MBEA (Montreal Battery of Evaluation of Amusia)
were used to evaluate musical abilities [
]. Children had to judge whether two
successivelypresented musical phrases were same or different, based either on melody or on rhythm.
An abbreviated version of the WISC-IV [
] was used to compute IQ scores that comprised
one subtest for each four indexes: Symbol Search (processing speed), Similarities (verbal
comprehension), Matrix Reasoning (perceptual reasoning), and Letter Number Sequencing
(auditory attention and working memory). This short-version has been shown to yield results
similar to the full-scale version [
] and provides a precise and reliable estimation of IQ in
approximately thirty minutes. The scores for each subtest were normalized and total IQ was
computed by adding the four normalized scores, and by converting this score into a standard
IQ score using a dedicated conversion table.
Symbol Search (WISC-IV): processing speed. On each row, two target symbols are
presented on the left side and five symbols are presented on the right side. Children are asked to
decide, as fast and as accurately as possible, whether one of the left target symbols is present or
not within the five symbols on the right. Each page comprises 15 rows.
Similarities (WISC-IV): verbal comprehension. Pairs of words were presented to the child
who described how they are alike (e.g. how are red and blue alike? Response: both are colors).
Matrix Reasoning (WISC-IV): non-verbal intelligence. A figure is missing from a series of
figures and children are asked to choose between five options, the picture that best completes
Letter Number Sequencing (WISC-IV): auditory attention and working memory. Series of
letters and numbers were read to the child who repeated back the letters in the alphabetical
order, and the numbers in numerical order (e.g. 4-A-3-C, response: 3-4-A-C).
The Digit Span test (WISC-IV, both forward and backward versions [
]) was administered
to further evaluate auditory short-term and working memory. Increasing series of numbers
were read to the child who repeated them back in the same order (Forward Digit Span:
shortterm memory) or in reverse order (Backward Digit Span: working memory).
Auditory and visual attention abilities were evaluated by the Auditory Attention and
Response Set tests (from the NEPSY-II battery [
]) and by the d2-R test [
]. A paper with
four colored circles (yellow, red, blue, black) was presented in front of the child who listened
to a recorded list of words that included targets and distractors. In the Auditory Attention test,
the child pointed to the red circle when hearing the word ?red?. In the Response Set test, the
5 / 21
Fig 1. Experimental procedure with the different tests presented before and after 18 months of music training.
child pointed to the red circle when hearing the word ?yellow? and vice-versa, and also pointed
to the blue circle when hearing the word ?blue?. In the d2-R concentration ability test, the
child was presented with a sheet of paper where the letters ?d? and ?p? were surrounded by
one to four small strokes. The child was asked to cross out all target letters: the ?d? surrounded
by two strokes. For each line on the sheet, the child crossed out as many targets as possible in
20 seconds. The test lasted for 5 minutes. The Concentration Capacity index was computed by
taking the number of targets examined by the child (within the 20 seconds time-limit) minus
the number of missed targets minus the number of incorrectly crossed distractors.
Reading abilities and phonological awareness were assessed using the French standard
Alouette test (revised version [
]) and the First Phonemes Fusion and Syllabic Suppression
tests (from the BALE battery: Batterie Analytique du Langage Ecrit, analytic battery of written
], respectively. The Alouette test measures reading precision and reading speed.
The child was asked to read aloud a complex nonsense text within a 3 minutes time limit.
Based on the number of words read (M), the number of words correctly read (C), and the
reading time (TL), two indexes were computed: the reading precision index [CM = (C/M)
x100] and the reading speed index [CTL = (Cx180)/TL]. In the Phonemes Fusion test, two
words were spoken aloud by the experimenter and the child merged their first phonemes (e.g.
?good orange?, response: GO). In the Syllabic Suppression test, bi- or trisyllabic words were
spoken aloud by the experimenter and the child supressed either the first, the second or the
third syllabe (e.g ?suppress the first syllable of the word elephant?, response: LEFANT).
Finally, fine motor skills, speed and precicion of hand-eye visuomotor coordination were
measured using the Visuomotor Precision test (NEPSY-II). A sheet of paper picturing a
6 / 21
winding road with borders on each side was presented to the child who followed the road with
a pencil as fast and as accurately as possible (keeping inside the borders).
All tests were presented in a single session that lasted for about 75 min, with short breaks
regularly interspaced during the session and/or on children? demand.
Data processing and statistical analyses
All tests are standardized tests (i.e., a large sample of children, representative of the French
children population, have been presented with these tests thereby providing a mean reference
score for each age range). Moreover, raw scores for each child and for each test were
normalized relative to their age group to discard maturation and developmental effects (similar to a
passive control group) and these scores were compared between the pre and post training
sessions. No active control group was tested (see discussion). Finally, note that standard scores
for the musicality tests could not be computed for all children since, to our knowledge, the
MBEA only provides normative data for adults [
], 14 to 18 years old adolescents [
] and 6
to 8 years old children [
The variable number of children in the different tests (see Table 1) was mainly linked to
norms for some age-ranges not being available for some tests so that data from children in
these age-groups were not further considered. For instance, for the d2-R, no standardized
scores are available below 9 years old. By contrast, for the two BALE tests (Syllabic suppression
and Phoneme fusion), no norm is available above 5th grade. As a consequence, data from
children above 5th grade in the Post-training session were not included in the analyses (children
were included in our sample only when age-norms were available for children both in Pre and
Post training sessions). Only in very rare cases children did not want to perform some tests.
The normality of data distribution was tested using the Shapiro-Wilk test (W). Student
ttests were used to compare normalized results in the pre vs post training sessions for the tests
showing a normal distribution and Wilcoxon tests were used for non-normally distributed
dataset. The Statistica software was used for all statistical analyses (Version 12.0 StatSoft, Inc,
Tulsa, OK) including the cluster analyses below.
To further investigate whether the observed effects reflected a general trend or whether
different trends were present within the group of children, we conducted cluster analyses [
the tests showing significant improvements. This allowed us to separate children into three
groups: those who showed an improvement (cluster 1), those who showed no change (cluster
2), and those who showed a decrease in performance from pre to post music training (cluster
3). The differences in level of performance pre vs post training were analyzed by interactive
partitioning (K-means), minimizing the within-cluster variability and maximizing the
between-cluster variability. Then, t-test comparisons of a single value (0) to the mean
difference of each cluster were conducted with Bonferroni?s correction (p < .05 divided by 3 tests:
significant threshold at .02). Descriptive analyses were also conducted to determine the
percentage of children improving in the test(s) showing significant effects of music training.
Finally, simple and multiple linear regression analyses were computed for the tests showing
significant pre-post differences to determine which factor(s) contributed to explain the results.
Test for the normality of the data distribution
Data distribution was normal for rhythmic musicality, total IQ, Symbol Search, Similarities,
Matrix Reasoning, Letter-Number Sequencing, Digit Span, d2-R, and the speed index of the
Alouette reading test. Therefore, the Student t-test for normally-distributed dataset was used
to compare pre vs post results. By contrast, data were not normally distributed pre and/or post
7 / 21
Pre-post comparison (t-test or
t(29) = -2.45, p < .02
Data distribution Shapiro-Wilk:
Mean standard score
8 / 21
t(33) = -2.47, p < .05
t(31) = -0.07, p = .95
t(33) = -0.62, p = .54
t(31) = -0.98, p = .33
t(33) = -1.02, p = .31
t(26) = -7.11, p < .001
Wilcoxon: Z = 1.30, p = .19
Wilcoxon: Z = .30, p = .76
Wilcoxon: Z = .88, p = .38
Wilcoxon: Z = .22, p = .83
Wilcoxon: Z = .83, p = .41
Wilcoxon: Z = 1.96, p < .05
.92 (< .05)
.90 (< .01)
.91 (< .05)
.88 (< .01)
.85 (< .01)
.88 (< .01)
.92 (< .05)
.87 (< .001)
.84 (< .001)
z-score Cohen?s d
Wilcoxon: Z = 2.22, p < .05
Pre-post comparison (t-test or
t(33) = -0.64, p = .53
Data distribution Shapiro-Wilk:
training for melodic musicality, Auditory Attention, Response Set, Syllabic Suppression, First
Phonemes Fusion, the precision index of the Alouette reading test and for Visuomotor
Precision. In these cases, the Wilcoxon test for non-normally-distributed dataset was used to
compare pre vs post results (see Table 1).
Results at the different tests listed in the first column. N is the number of children who
performed the tests both in the pre and in the post-training sessions. Mean standard scores and
zscores are presented together with the effect size (Cohen?s d) and the results of statistical
analyses using t-tests or Wilcoxon tests depending upon the normality of the data distribution, as
reported in the last column.
Results at the different tests are presented in Table 1 and illustrated on Fig 2 (averaged data)
and on Fig 3 (individual data). They showed a significant improvement in musicality scores
with music training, as reflected by higher percentage of correct responses post- than
pretraining in both the melodic (Pre: 57.14, Post: 64.71; Z = 2.22, p < .05; Cohen?s d = 0.55) and
the rhythmic tests (Pre: 61.90, Post: 71.42; t(34) = -4.66, p < .001; Cohen?s d = 0.70).
Results also showed a significant improvement of total IQ from pre (81.17) to post music
training (85.47; t(29) = -2.45, p = .02; Cohen?s d = 0.24) when considering the four subtests of
the WISC-IV abbreviated version. Considering each subtest separately, the pre to post
improvement was only significant in the Symbol Search test (pre: 7.56, Post: 8.53; t(33) = -2.47,
p = .05; Cohen?s d = 0.33). The ability to focus attention was measured by the concentration
ability index of the d2-R. Results showed significant improvements from pre (84.63) to post
training (93.48; t(26) = -7.11, p < .001; Cohen?s d = 0.85). Finally, the improvement of reading
abilities, as measured by the Alouette test, was significant on reading precision (z-score, Pre:
-1.74, Post: -1.37; Wilcoxon: Z = 1.96, p = .05; Cohen?s d = 0.19) but not on reading speed). No
pre to post improvements were found for the tests of Auditory Attention, Response Set, Digit
Span and Visuomotor Precision.
In order to better understand the impact of the De?mos music program at the individual level,
each child was assigned to one of three groups, ?increase?, ?equal? or ?decrease?, according to
her/his pre-post-test evolution at the three main tests showing significant effects (total IQ,
d2-R and Alouette reading precision). Results showed that 37% of children improved in all
three tests, 50% improved in two tests, and 13% improved in one test.
9 / 21
Fig 2. Pre- and post-training results for all tests. A. Results for total Intelligence Quotient (IQ) score, for the four subtests of the reduced version of WISC-IV
and for the digit span test are compared in the pre- and post-music training sessions. B. Results at the d2-R and NEPSY-II subtests are compared in the
preand post-music training sessions. For both A and B, standard scores are indicated on the left ordinate and z-scores on the right ordinate. C. Results at the
Alouette reading test and at the phonological tests are compared in the pre- and post-music training sessions. Z-scores are indicated on the left ordinate.
Significant pre vs post differences with : p < .05; : p < .001; ns: not significant. Error bars are standard errors of mean (SEM).
Statistical results of cluster analyses are presented in Table 2 and illustrated on Fig 4. Except
for the melodic and rhythmic tests (for which norms did not exist for our age groups), cluster
analyses were performed on z-scores for the tests showing significant pre to post improvement
(Rhythmic, Melodic, IQ, d2-R, Alouette). Results of the ANOVAs including Cluster as a
within-subject factors confirmed the presence of three clusters that significantly differed from
one another. To simplify results presentation, results are summarized in the text, presented in
detail in Table 2 and illustrated on Fig 3. For the melodic test, the first cluster included 13
children (37%) showing a large significant improvement in melodic scores, the second included
16 children (46%) showing no significant improvement, and the third cluster included 6
children (17%) with a significant decrease in performance. In the rhythmic test, the first cluster
included 6 children (17%) showing a large significant improvement in rhythmic scores, the
second included 13 children (37%) with medium significant improvement, and the third
cluster included 16 children (46%) with no significant improvement. Regarding IQ scores, the first
cluster included 7 children (23%) with a large significant improvement in IQ scores, the
second included 10 children (33%) with a medium significant improvement, and the third cluster
10 / 21
Fig 3. Violin plots for each test showing significant improvements from pre to post-music training. Violin plots (R Development Core Team, 2018)
show the shape of the distribution. Each black dot represents one child, the red dot corresponds to the mean and the red line to the standard deviation (SD).
Z-scores are indicated on the ordinate.
included 13 children (43%) with a medium significant decrease in IQ scores. In the d2-R test,
the first cluster included 8 children (30%) showing a large significant improvement in d2-R
scores, the second included 9 children (33%) with medium significant improvement, and the
third cluster included 10 children (37%) with no significant improvement. Finally, for the
reading precision scores, the first cluster included 3 children (9%) showing a large significant
increase in reading precision, the second included a larger group of 23 children (70%) with a
small significant increase in reading precision and the third cluster included seven children
(21%) with a small significant decrease in reading precision.
Results of simple regression analyses to model the significant effects reported above and in
particular, to determine whether improvements in IQ or in d2-R accounted for improvement in
reading precision, as well as results of multiple linear regression analyses computed to model
reading precision as a function of both IQ and d2-R showed no significant effects (effect of IQ
on reading precision: t(27) = 0.22, Beta = .04, p = .82; effect of d2-R on reading precision : t
(25) = -1.38, Beta = -0.27, p = .18). Interestingly, however, the improvement in IQ was
significantly larger for children with initial lower IQ scores (t(28) = -3.58, Beta = -0.56, p < .001; see
Finally, results of multiple linear regression analysis, with IQ improvement (post minus
pre) as the dependent variable and improvements in all four subtests as independent variables
showed that all four subtests contributed to the improvement in IQ (Symbol Search: t(25) =
51.20, Beta = 0.44, p < .001; Similarities: t(25) = 56.19, Beta = 0.49, p < .001; Matrix Reasoning:
t(25) = 59.20, Beta = 0.50, p < .001; Letter Number Sequencing: t(25) = 64.03, Beta = 0.55, p <
Evaluation of the impact of the De?mos program, initiated by the Paris Philharmonie and
conducted in ecologically valid school-settings, on the cognitive development of children from
11 / 21
low socio-economic backgrounds revealed several findings of interest after 18 months of
music training. Results showed significant improvements of musicality scores, general
intelligence (total IQ), processing speed (Symbol search) and concentration ability (d2-R) as well as
increased reading precision (Alouette test). Results at the other tests were not significant
although a trend in the same direction was observed (see Table 1).
The De?mos music intervention was successful in improving the children?s level of
performance in both the melodic and the rhythmic tests. In both cases, the effect sizes were relatively
large (Cohen?s d = 0.55 and d = 0.70 respectively) and results of cluster analyses showed that
more than half of children (54%) improved their abilities to judge whether two short musical
phrases had identical or different rhythm. By contrast, only 37% of the children improved
their ability to decide whether two musical phrases had identical or different melodic contour
and 63% showed no improvement or a small decrease in melodic scores. Thus, while most
children developed a better sense of rhythm, 18 months of music training were possibly not
long enough to develop a ?musical ear? for melody in most children. These results are in line
with results from children of the middle class [
] showing that the melodic task is typically
more difficult than the rhythmic task.
General intelligence, processing speed and concentration abilities
In line with previous findings with typically developing children [
], one of the main
finding of this study was a significant improvement of general intelligence (IQ scores) in children
from low SES after 18 month of music training. This finding is of primary importance when
12 / 21
Fig 4. Results of cluster analyses for the five tests showing significant post minus pre- music training differences. For
each test, each box plot corresponds to one cluster and the dots represent the children within each cluster. For each box plot,
the upper quartile, the median and the lower quartile are represented together with the upper and lower whiskers. Significant
increases or decreases in post minus pre-differences compared to 0 (no change) are indicated with : p < .05; : p < .01 and
: p < .001.
considering that these children typically show lower IQ scores than children from higher
socio-economic backgrounds ([
] for a review) and that this gap tends to increase over the
course of development, the so-called ?Matthew effect? [
]. For instance, Shaywitz &
Shaywitz  followed-up a cohort of children for 7 years, from kindergarten to 6th grade. IQ
scores were evaluated every two years and, as in the present study, normalized-age scores were
computed to rule out maturation and developmental effects. While children with higher IQ
scores (e.g. average = 110) showed an increase in IQ, children with lower IQ scores (e.g.
average = 80) showed a decrease in IQ over the course of elementary school. The children involved
in the De?mos program showed relatively low initial IQ scores (i.e., average = 80) and a small
(effect size, Cohen?s d = 0.24) but significant increase in IQ of 4.3 points on average, that was
close to the increase found in the control group of the Schellenberg? study ([
]; mean IQ
increase = 3.9). These results strongly suggest that the De?mos music intervention program
helped children from low SES to counteract the potential decrease in IQ over the school years
] to get closer to the level of more privileged children. Future studies will aim at testing
whether other types of training (e.g., dance, theater, painting . . .) can also counteract the
decrease in IQ scores often encountered in children from low socio-economic backgrounds
13 / 21
Fig 5. Results of simple regression analysis showing a significant correlation (p < .001) between initial IQ score
and improvement in IQ.
Results of cluster analyses are important in showing moderate (33% of the children) to
strong (23% of the children) improvements in IQ scores after music training in more than half
of the children (56%) involved in the De?mos program compared to 43% showing a moderate
decrease in IQ scores. While these increase and decrease in IQ may seem surprising, we need
to keep in mind that they are computed on normalized scores that is, in reference to the results
of a large children population within an age group. This population is, by definition,
representative of the entire population and does not only include children from low SES. Thus, the
decrease in IQ for children with low initial IQ scores, for instance, is relative to the entire
population and may not be found if compared to specific norms for children from low SES.
Moreover, the IQ increase was unlikely to be driven only by the significant improvement in the
Symbols Search task since results of multiple regression analyses showed that the
improvements in all four subtests contributed to the increase in IQ scores. Final but not least, results
were very encouraging in revealing that children with lower initial IQ scores showed larger
increases in IQ than children with higher initial scores. In other words, the impact of music
training is potentially stronger when there is more space for improvement. These results are in
line with recent results by Linnavalli et al. [
] showing that 5 to 6-year-old children with low
scores in linguistic tasks benefitted more from music playschool activities than children who
started with higher scores. They are also reminiscent of findings by Swaminathan and
] showing that the impact of music training on musical competence (defined as the
ability to perceive and remember sequences of tones or beats) was only significant in
undergraduate (19-year old on average) who scored below the mean on tests of working memory
and non-verbal intelligence. In sum, a tentative conclusion that needs to be tested in future
studies is that individuals who initially scored relatively low on different tests may benefit
14 / 21
more (as seen in the improvements of musical competence, linguistic tasks or IQ scores) from
the positive influence of music training.
Based on previous results from cross-sectional studies in adults showing that musicians
outperformed non-musicians in tasks similar to the Symbol Search task and the d2-R, taken to
measure processing speed and concentration abilities [
], we expected children to improve at
these tests after 18 months of music training. In line with these predictions, results showed
significant improvements, with medium to large effect sizes in the Symbol Search (Cohen?s
d = 0.33) and d2-R tests (Cohen?s d = 0.85). These findings provide further support to the
proposal that playing music improves the ability to rapidly process information and to focus on
the stimuli and task at hand. In this respect, it is remarkable that music training had a
particularly strong impact (Cohen?s d = 0.85) on the concentration ability of a large percentage (63%)
of children (as mentioned above, this increase is relative to the general child population in this
age group and not compared to specific norms for children from low SES), possibly because
they learned to focus attention on the teacher and on their instrument during the music
classes. This finding is particularly interesting when considering that children from low SES often
encounter difficulties to focus attention [
]. Along these lines, Neville et al. [
] showed that a
family-based training program, targeting improvements of family stress regulation, parental
responsiveness. . ., and of child attention, succeeded in improving selective attention in
children from low SES. Thus, different types of training seem to exert a positive influence on
attentional abilities. Moreover, by showing rapid improvement of selective attention (within 8
weeks in the Neville et al.? intervention study) and of concentration abilities (within 18 months
in the present study), these findings open new perspectives for research and education. While
a few studies have tested the influence of musicianship on the variability of auditory cortical
activity in adults [
] and in children [
], the impact of music training (longitudinal studies)
on the different components of attention (e.g., selective vs divided attention, focused vs
switching attention, distractors etc.) at the behavioral and electrophysiological levels has not yet been
thoroughly examined. This is of primary importance for future research since the ability to
focus attention on the task at hand is a prerequisite for successful learning in many, if not all,
domains of knowledge. Future studies could test whether improvements in concentration
ability associated to music training positively influence learning in other domains such as language
or mathematics and whether these effects are specific to music training or could be found with
other training activities.
Phonological awareness, reading abilities and working memory
In line with results from the Kraus group with children from low SES musically-trained for 24
months within the Harmony program [
], results with children from the De?mos program
musically-trained for 18 months showed improvements in reading precision (Alouette test).
Overall, these reading improvements were relatively small (Cohen?s d = 0.19) in most children
(70%) but 9% of the children showed a large improvement and 21% showed a small decrease
in reading precision, as shown by results of cluster analyses. Surprisingly, but in line with
results from the Kraus group, we found no influence of music training on phonological
awareness (First Phonemes Fusion and Syllabic Suppression). While it is very encouraging that
similar results were found in two independent and culturally-different samples of children from
low SES involved in music training, these results stand in contrast with the findings of
Linnavalli et al.  and Nan et al. [
] showing a positive impact of playschool music and piano
training on phonemic awareness and word discrimination based on consonants. However, in
these two studies, children were younger (4-6-year old), they were not only issued from
disadvantaged backgrounds and the phonological tests were different from the standardized tests
15 / 21
used here, which possibly account for these differences. Nevertheless, our results were also
unexpected based on results of meta-analyses pointing to stronger effects of music training on
phonological awareness than on reading abilities [
] and in view of several findings
showing that phonological abilities critically influence the development of reading skills .
Which factors contribute most to reading success is still a hotly debated issue [
instance, recent results from a large sample of French children (N = 703) from low SES families
] highlighted the primary importance of listening comprehension on the earliest phases of
reading acquisition. Thus, lexical and semantic knowledge (e.g., what the words mean) seem
to strongly influence reading abilities, while other factors, such as vocabulary, morphological
and phonemic awareness may only indirectly contribute to reading skills (via listening
comprehension). More generally, working and short-term memory [
], attention and executive
], as well as articulatory rehearsal strategies [
] likely contribute to
reading abilities. For instance, children with reading disabilities (dyslexics or poor readers)
often show poor short-term phonological memory as revealed by scores at the forward digit
span test and at the nonword repetition task (see meta-analysis [
]). Directly related to this
issue, results of a longitudinal study by Perez and collaborators [
] showed that short term
memory capacity predicted reading abilities one year later. These authors proposed that the
ability to read new words is causally impacted by the ability to recall sequences of phonemes in
the right order (taken to reflect core short-term memory [
]), rather than by phonological
These views open the intriguing possibility that the positive influence of music training
on reading precision reported here was mediated by improvements in general intelligence,
concentration abilities, working and short-term memory rather than only by phonological
skills. For instance, musical practice may improve concentration abilities on visual stimuli
that, in turn, may facilitate grapheme-phoneme decoding and thereby improve reading
performance. However, results of simple and multiple regression analyses showed that
improvement in IQ scores and/or in concentration ability did not account for the
improvement in reading precision. It may be that the effects size or the participant sample size were
too small, with not enough children showing a large effect. Clearly more studies are needed
to determine the relative weight of the different factors that contribute to the development
of reading skills.
One could consider that the main limitation of the present study is the lack of an active control
group which is needed to ascertain that the reported effects are specific to music training and
would not obtain with another activity, such as dance, painting or cooking, that would be as
interesting and as motivating for the children [
]. This, however, was not our aim and
does not reduce the societal impact of the present findings. Whether effects similar to those
obtained with the De?mos music program were to be found with another type of training
would also be very positive for the children. Moreover, since the pre-to-post comparisons
reported here are based on normalized scores for each age group, we can ascertain that the
results were not age-related, with children performing better because they were 18 months
older post-than-pre training (i.e., we can rule out the influence of developmental and
maturation effects). In fact, computing normalized scores is equivalent to comparing results of
children trained with the De?mos program to a passive control group that would include a large
number of children representative of the population (e.g. 1103 children were tested for the
French version of WISC-IV [
]) who are not involved in any specific training or who are
waiting to be involved in music training, as in some previous studies [
16 / 21
Test?Music intervention (Keyboard)?Retest
A second potential limitation is linked to the relative high number of children who could
not be tested post-music training because they left primary school to go to middle school or to
another school. This is often the case in longitudinal studies and particularly when testing
children from low SES. However, since only one child decided to stop music training, we are
confident that the effects that we observed do not only come from children who were highly
motivated by the De?mos program compared to children that were less motivated.
A third limitation is linked to the repetition of the different tests with potentially higher
level of performance on second than on first presentation. We partly controlled for this aspect
by using tests that are typically chosen in the literature to evaluate the effects of an
intervention. For instance, in a recent study [
] (see Table 3), children performed the French version
of the WISC-IV twice, with an average of 20 months between test repetition. The improvement
in IQ scores (Full Scale IQ) was significant but smaller (2.5) than in the present study (4.3). In
the study by Ryan and collaborators [
], pre-post differences in IQ scores were not significant
and smaller (1.6) than in our study, although the between-test repetition interval was shorter
(11 months; see also [
]; Table 3). Finally, in the Shaywitz & Shaywitz? study [
discussed above, children with low initial IQ scores showed decreases, rather than increases, in
IQ scores across the three repetitions of the same tests. It is thus unlikely that test repetition
was driving the present results.
Results of the present longitudinal experiment are important in showing that music training
may counter-balance the negative influence of living in low socio-economic backgrounds [
by enhancing several core cognitive functions: general intelligence, processing speed,
concentration abilities and reading precision. Moreover, because these improvements were computed
from normalized scores, it is possible to rule out the influence of maturation and
developmental effects. However, as discussed above, we cannot ascertain that similar effects would not be
found with another training program, as interesting for the children as the De?mos music
training program. But this is not problematic, on the contrary, as it would be important to use any
training program that can counteract the negative effects of living in disadvantaged
backgrounds. It is also important to note that several cognitive functions tested in the present study
did not benefit from music training (auditory attention, visuomotor precision). For instance,
in contrast to the results of Guo and colleagues [
] showing significant improvement in
backward Digit Span scores after only 6 weeks of musical training, we did not find similar
improvements of working memory. While the reasons for such variability in the results are difficult to
determine, the question of whether music training has a general impact on cognitive abilities
17 / 21
or whether its influence is specific to cognitive functions that share common processes with
music, such as language for instance, is still an open issue [
]. In our view, the
outcome most likely depends upon the duration of music training and upon the tasks used to test
the effect of music training.
To conclude, and in agreement with the proposals made by several authors [
believe that ecological studies in school-settings, developed in partnership with existing
programs and even if they are typically not as well controlled as laboratory-based experiments, are
of crucial importance to better understand the impact of music training and of other types of
training on the cognitive development of children from diverse cultural and economic
backgrounds. Importantly, all children in the schools were from low SES and all were involved in
the De?mos music program: this reduced potential bias linked to children from high SES being
more likely to choose music training. Moreover, children were strongly encouraged to practice
their musical instruments. This is of crucial importance since motivation to pursue a
demanding music training is likely to be lower in families facing social difficulties than in families with
higher social status [
]. In sum, our results are encouraging in showing that, after only
18 months of music training, 37% of the children improved on three tests (general intelligence,
concentration and reading abilities), and that 100% of the children improved at least in one of
these tests. Thus, even if the effects were overall of small to medium size, as expected based
upon the literature and the relatively short duration of music training, these positive results
may help promote a wider use of music training in school-settings.
We would like to thank the ?Apprentis d?Auteuil? and two primary schools in Marseille,
Vitagliano and Ozanam, for their collaboration. We are grateful to the children who participated
in the experiment and to their parents, as well as to Benjamin Furnari, Pauline Maurel, Laure
Tosatto, Audrey Foro and Julien de Maria for help in data analysis and to Indiana Wollman
for interesting comments on this study. This research was conducted with the support of the
CNRS (MB), of the Philharmonie de Paris (MB) who implemented the De?mos program in
Marseille, of the Brain and Language Research Institute (BLRI, ANR-11-LABX-0036, MB) and
of the Institute of Communication and Language in the Brain (ILCB, ANR-11-IDEX-0001-02,
A-MIDEX, MB). M. Barbaroux is supported by a doctoral fellowship from the French Ministry
of Research and Education and ED by a doctoral fellowship from the BLRI.
Conceptualization: Myl?ne Barbaroux, Eva Dittinger, Mireille Besson.
Data curation: Myl?ne Barbaroux.
Formal analysis: Myl?ne Barbaroux.
Funding acquisition: Mireille Besson.
Investigation: Myl?ne Barbaroux.
Methodology: Myl?ne Barbaroux, Mireille Besson.
Project administration: Myl?ne Barbaroux, Eva Dittinger, Mireille Besson.
Resources: Mireille Besson.
Supervision: Mireille Besson.
Validation: Myl?ne Barbaroux, Mireille Besson.
18 / 21
Visualization: Myl?ne Barbaroux, Mireille Besson.
Writing ? original draft: Myl?ne Barbaroux, Mireille Besson. Writing ? review & editing: Myl?ne Barbaroux, Eva Dittinger, Mireille Besson.
19 / 21
20 / 21
1. Schellenberg EG . Long-term positive associations between music lessons and IQ . Journal of Educational Psychology . 2006 ; 98 ( 2 ): 457 - 68 .
2. Schellenberg EG , Mankarious M. Music training and emotion comprehension in childhood . Emotion . 2012 ; 12 ( 5 ): 887 - 91 . https://doi.org/10.1037/a0027971 PMID: 22642351
3. Schellenberg EG . Examining the association between music lessons and intelligence . British journal of psychology . 2011 ; 102 ( 3 ): 283 - 302 . https://doi.org/10.1111/j.2044- 8295 . 2010 . 02000 .x PMID: 21751987
4. Schellenberg EG . Music Lessons Enhance IQ. Psychological Science . 2004 ; 15 ( 8 ): 511 - 4 . https://doi. org/10.1111/j.0956- 7976 . 2004 . 00711 . x PMID : 15270994
5. Sala G , Gobet F . When the music's over. Does music skill transfer to children's and young adolescents' cognitive and academic skills? A meta-analysis . Educational Research Review . 2017 ; 20 : 55 - 67 .
6. Cole? P, Cavalli E , Duncan LG , Theurel A , Gentaz E , Sprenger-Charolles L , et al. What is the influence of morphological knowledge in the early stages of reading acquisition among low SES children? A graphical modeling approach . Frontiers in Psychology. 2018 Apr;9.
7. Gentaz E , Sprenger-Charolles L , Theurel A . Differences in the predictors of reading comprehension in first graders from low socio-economic status families with either good or poor decoding skills . PloS one . 2015 ; 10 ( 3 ):e0119581. https://doi.org/10.1371/journal.pone. 0119581 PMID: 25793519
8. Hemmerechts K , Agirdag O , Kavadias D. The relationship between parental literacy involvement, socio-economic status and reading literacy . Educational Review . 2017 Jan; 69 ( 1 ): 85 - 101 .
9. Anvari SH , Trainor LJ , Woodside J , Levy BA . Relations among musical skills, phonological processing, and early reading ability in preschool children . Journal of experimental child psychology . 2002 ; 83 ( 2 ): 111 - 130 . PMID: 12408958
10. Forgeard M , Schlaug G , Norton A , Rosam C , Iyengar U , Winner E. The relation between music and phonological processing in normal-reading children and children with dyslexia . Music Perception: An Interdisciplinary Journal . 2008 ; 25 ( 4 ): 383 - 90 .
11. Gordon RL , Fehd HM , McCandliss BD . Does music training enhance literacy skills? A meta-analysis . Frontiers in psychology. 2015 ; 6 : 1777 . https://doi.org/10.3389/fpsyg. 2015 .01777 PMID: 26648880
12. Habib M , Lardy C , Desiles T , Commeiras C , Chobert J , Besson M. Music and dyslexia: a new musical training method to improve reading and related disorders . Frontiers in Psychology. 2016 Jan;7.
13. Butzlaff R . Can music be used to teach reading? Journal of Aesthetic Education . 2000 ; 34 ( 3 - 4 ).
14. Kraus N , Hornickel J , Strait DL , Slater J , Thompson E. Engagement in community music classes sparks neuroplasticity and language development in children from disadvantaged backgrounds . Frontiers in psychology. 2014 ; 5 : 1403 . https://doi.org/10.3389/fpsyg. 2014 .01403 PMID: 25566109
15. Slater J , Strait DL , Skoe E , O'Connell S , Thompson E , Kraus N. Longitudinal effects of group music instruction on literacy skills in low-income children . PLoS ONE . 2014 Nov; 9 ( 11 ):e113383. https://doi. org/10.1371/journal.pone. 0113383 PMID: 25409300
16. Hackman DA , Farah MJ , Meaney MJ . Socioeconomic status and the brain: mechanistic insights from human and animal research . Nature reviews neuroscience . 2010 Sep; 11 ( 9 ): 651 - 9 . https://doi.org/10. 1038/nrn2897 PMID: 20725096
17. Shaywitz BA , Shaywitz SE . A Matthew effect for IQ but not for reading: Results from a longitudinal study . Reading Research Quarterly. 1995 ; 30 ( 4 ).
18. Sirin SR. Socioeconomic status and academic achievement: a meta-analytic review of research . Review of Educational Research . 2005 ; 75 ( 3 ): 417 - 453 .
19. Linnavalli T , Putkinen V , Lipsanen J , Huotilainen M , Tervaniemi M. Music playschool enhances children's linguistic skills . Scientific Reports . 2018 Jun; 8 ( 1 ): 8767 . https://doi.org/10.1038/s41598-018 - 27126-5 PMID: 29884891
20. Nan Y , Liu L , Geiser E , Shu H , Gong CC , Dong Q , et al. Piano training enhances the neural processing of pitch and improves speech perception in Mandarin-speaking children . Proceedings of the National Academy of Sciences . 2018 Jul; 115 ( 28 ): E6630 - 9 .
21. Dege ? F, Schwarzer G. The effect of a music program on phonological awareness in preschoolers . Front Psychol . 2011 Jun; 2 : 124 . https://doi.org/10.3389/fpsyg. 2011 .00124 PMID: 21734895
22. Strait DL , Kraus N , Parbery-Clark A , Ashley R . Musical experience shapes top-down auditory mechanisms: evidence from masking and auditory attention performance . Hearing research . 2010 ; 261 ( 1 ): 22 - 29 .
23. Rodrigues AC , Loureiro MA , Caramelli P. Long-term musical training may improve different forms of visual attention ability . Brain and cognition . 2013 ; 82 ( 3 ): 229 - 235 . https://doi.org/10.1016/j.bandc. 2013 . 04 .009 PMID: 23694752
24. Clayton KK , Swaminathan J , Yazdanbakhsh A , Zuk J , Patel AD , Kidd G Jr. Executive function, visual attention and the cocktail party problem in musicians and non-musicians . PLoS ONE . 2016 Jul; 11 ( 7 ): e0157638. https://doi.org/10.1371/journal.pone. 0157638 PMID: 27384330
25. Lee Y , Lu M , Ko H . Effects of skill training on working memory capacity . Learning and Instruction . 2007 Jun; 17 ( 3 ): 336 - 44 .
26. Guo X , Ohsawa C , Suzuki A , Sekiyama K. Improved digit span in children after a six-week intervention of playing a musical instrument: an exploratory randomized controlled trial . Frontiers in psychology. 2017 ; 8 : 2303 . https://doi.org/10.3389/fpsyg. 2017 .02303 PMID: 29358927
27. Dege ? F, Wehrum S , Stark R , Schwarzer G. The influence of two years of school music training in secondary school on visual and auditory memory . European Journal of Developmental Psychology . 2011 Sep; 8 ( 5 ): 608 - 23 .
28. Stevens C , Lauinger B , Neville H. Differences in the neural mechanisms of selective attention in children from different socioeconomic backgrounds: an event-related brain potential study . Developmental Science . 2009 ; 12 ( 4 ): 634 - 646 . https://doi.org/10.1111/j.1467- 7687 . 2009 . 00807 . x PMID : 19635089
29. Neville HJ , Stevens C , Pakulak E , Bell TA , Fanning J , Klein S , et al. Family-based training program improves brain function, cognition, and behavior in lower socioeconomic status preschoolers . Proceedings of the National Academy of Sciences. 2013 Jul; 110 : 12138 - 43 .
30. Strait DL , Slater J , O'Connell S , Kraus N. Music training relates to the development of neural mechanisms of selective auditory attention . Developmental Cognitive Neuroscience . 2015 Apr; 12 : 94 - 104 . https://doi.org/10.1016/j.dcn. 2015 . 01 .001 PMID: 25660985
31. Bradley RH , Corwyn RF . Socioeconomic status and child development . Annual Review of Psychology . 2002 Feb; 53 ( 1 ): 371 - 99 .
32. Peretz I , Champod AS , Hyde KL . Varieties of Musical Disorders . Ann NY Acad Sci . 2003 ; 999 : 58 - 75 . PMID: 14681118
33. Wechsler D. Manuel de l'Echelle d'Intelligence de Wechsler pour Enfants- 4e e? dition [Manual for the Wechsler Intelligence Scale for Children-fourth edition] . Paris: Editions du Centre de Psychologie Applique?e. 2005 ;
34. Gre?goire J. L'examen clinique de l'intelligence de l 'enfant: fondements et pratique du WISC-IV [Clinical examination of children intelligence: foundations and practice of WISC-IV] . Editions Mardaga; 2006 .
35. Korkman M , Kirk U , Kemp S. NEPSY-Second Edition (NEPSY-II) . San Antonio: The Psychological Corporation. 2007 ;
36. Brickenkamp R , Schmidt-Atzert L , Liepmann D. d2-R: test d'attention concentre?e [d2-R: test of attention]. E?ditions Hogrefe France. 2015 . 108 p.
37. Lefavrais P. Alouette R : test d'analyse de la vitesse de lecture ? partir d'un texte [Alouette R: analysis test of text-reading speed] . Paris, France: Les Editions du Centre de Psychologie Applique?e. 2006 ;
38. Jacquier-Roux M , Lequette C , Pouget G , Valdois S , Zorman M. BALE: Batterie Analytique du Langage Ecrit [Analytic Battery of Written Language] [Internet] . Cogni-Sciences . 2010 . Available from: http:// www.cognisciences.com/accueil/outils/article/bale
39. Nunes-Silva M , Haase VG . Montreal Battery of Evaluation of Amusia: Validity evidence and norms for adolescents in Belo Horizonte , Minas Gerais, Brazil. Dementia & Neuropsychologia. 2012 Dec; 6 ( 4 ): 244 - 52 .
40. Peretz I , Gosselin N , Nan Y , Caron-Caplette E , Trehub SE , Be?land R. A novel tool for evaluating children's musical abilities across age and culture . Frontiers in systems neuroscience . 2013 ; 7 : 30 . https:// doi.org/10.3389/fnsys. 2013 .00030 PMID: 23847479
41. Danna J , Paz-Villagra?n V , Gondre C , Aramaki M , Kronland-Martinet R , Ystad S , et al. ? Let me hear your handwriting!? Evaluating the movement fluency from its sonification . PloS one . 2015 ; 10 ( 6 ):e0128388. https://doi.org/10.1371/journal.pone. 0128388 PMID: 26083384
42. Dittinger E , Chobert J , Ziegler JC , Besson M. Fast brain plasticity during word learning in musicallytrained children . Frontiers in Human Neuroscience . 2017 May; 11 : 233 . https://doi.org/10.3389/fnhum. 2017 .00233 PMID: 28553213
43. Stanovich K. Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy . Reading Research Quarterly. 1986 ; 48 .
44. Swaminathan S , Schellenberg EG . Musical competence is predicted by music training, cognitive abilities, and personality . Scientific Reports . 2018 Jun; 8 ( 1 ): 9223 . https://doi.org/10.1038/s41598-018 - 27571-2 PMID: 29907812
45. Strait DL , Kraus N. Can you hear me now? Musical training shapes functional brain networks for selective auditory attention and hearing speech in noise . Frontiers in Psychology. 2011 ; 2 : 113 . https://doi. org/10.3389/fpsyg. 2011 .00113 PMID: 21716636
46. Baddeley A . Working memory and language: an overview . Journal of Communication Disorders . 2003 May; 36 ( 3 ): 189 - 208 . PMID: 12742667
47. Dege ? F, Kubicek C , Schwarzer G. Music lessons and intelligence: A relation mediated by executive functions . Music Perception: An Interdisciplinary Journal . 2011 ; 29 ( 2 ): 195 - 201 .
48. Moreno S , Bialystok E , Barac R , Schellenberg EG , Cepeda NJ , Chau T . Short-term music training enhances verbal intelligence and executive function . 2011 Oct; 22 ( 11 ): 1425 - 33 . https://doi.org/10. 1177/0956797611416999 PMID: 21969312
49. Roden I , Grube D , Bongard S , Kreutz G . Does music training enhance working memory performance? Findings from a quasi-experimental longitudinal study . Psychology of Music . 2014 Oct; 42 ( 2 ): 284 - 298 .
50. Swanson HL , Zheng X , Jerman O . Working memory, short-term memory, and reading disabilities: A selective meta-analysis of the literature . Journal of Learning Disabilities . 2009 ; 42 ( 3 ): 260 - 287 . https:// doi.org/10.1177/0022219409331958 PMID: 19255286
51. Perez TM , Majerus S , Poncelet M. The contribution of short-term memory for serial order to early reading acquisition: Evidence from a longitudinal study . Journal of Experimental Child Psychology . 2012 ; 111 ( 4 ): 708 - 723 . https://doi.org/10.1016/j.jecp. 2011 . 11 .007 PMID: 22209402
52. Chobert J , Francois C , Velay J-L , Besson M. Twelve months of active musical training in 8- to 10-yearold children enhances the preattentive processing of syllabic duration and voice onset time . Cerebral Cortex . 2012 ; 24 ( 4 ): 956 - 67 . https://doi.org/10.1093/cercor/bhs377 PMID: 23236208
53. Moreno S , Marques C , Santos A , Santos M , Castro SL , Besson M. Musical training influences linguistic abilities in 8-year-old children: more evidence for brain plasticity . Cerebral Cortex . 2009 Mar; 19 ( 3 ): 712 - 723 . https://doi.org/10.1093/cercor/bhn120 PMID: 18832336
54. Moreno S , Besson M. Musical training and language-related brain electrical activity in children . Psychophysiology . 2006 May; 43 ( 3 ): 287 - 91 . https://doi.org/10.1111/j.1469- 8986 . 2006 . 00401 . x PMID : 16805867
55. Kraus N , Slater J , Thompson EC , Hornickel J , Strait DL , Nicol T , et al. Music enrichment programs improve the neural encoding of speech in at-risk children . The Journal of Neuroscience . 2014 Sep; 34 ( 36 ): 11913 - 11918 . https://doi.org/10.1523/JNEUROSCI.1881- 14 . 2014 PMID: 25186739
56. Kieng S , Rossier J , Favez N , Lecerf T. Long-term stability of the French WISC-IV: Standard and CHC index scores . Revue Europe?enne de Psychologie Applique?e. 2017 Jan; 67 ( 1 ): 51 - 60 .
57. Ryan JJ , Glass LA , Bartels JM . Stability of the WISC-IV in a sample of elementary and middle school children . Applied Neuropsychology . 2010 Feb; 17 ( 1 ): 68 - 72 . https://doi.org/10.1080/ 09084280903297933 PMID: 20146124
58. Bartoi MG , Issner JB , Hetterscheidt L , January AM , Kuentzel JG , Barnett D . Attention problems and stability of WISC-IV scores among clinically referred children . Applied Neuropsychology: Child. 2015 Jul; 4 ( 3 ): 133 - 40 .
59. Besson M , Chobert J , Marie C . Transfer of training between music and speech: common processing, attention, and memory . Frontiers in Psychology. 2011 May; 2 : 94 . https://doi.org/10.3389/fpsyg. 2011 . 00094 PMID: 21738519
60. Kraus N , Chandrasekaran B . Music training for the development of auditory skills . Nature Reviews Neuroscience . 2010 Aug; 11 ( 8 ): 599 - 605 . https://doi.org/10.1038/nrn2882 PMID: 20648064
61. Tervaniemi M. Musical sounds in the human brain . In: Neuroaesthetics. Routledge; 2018 . p. 221 - 231 .
62. Habibi A . Neural correlates of accelerated auditory processing in children engaged in music training . Developmental Cognitive Neuroscience . 2016 ; 14 .
63. Habibi A , Damasio A , Ilari B , Veiga R , Joshi AA , Leahy RM , et al. Childhood music training induces change in micro and macroscopic brain structure: results from a longitudinal study . Cerebral Cortex . 2017 Nov; 1 - 12 . https://doi.org/10.1093/cercor/bhw362
64. Habibi A , Damasio A , Ilari B , Elliott Sachs M , Damasio H . Music training and child development: a review of recent findings from a longitudinal study: Music training and child development: a review . Annals of the New York Academy of Sciences. 2018 Jul; 1423 ( 1 ): 73 - 81 .